VOL. VII JANUARY, 1953 NO. PACIFIC SCIENCE A QUARTERLY DEVOTED TO THE BIOLOGICAL AND PHYSICAL SCIENCES OF THE PACIFIC REGION

IN THIS ISSUE: Banner—The Crangonidae, or Snapping Shrimp,

of Hawaii • NEWS NOTES • INDEX

Published by THE UNIVERSITY OF HAWAII HONOLULU, HAWAII BOARD OF EDITORS LEONARD D. TUTHILL, Editor-in-Chief Department of Zoology and Entomology, University of Hawaii O. A. BUSHNELL, Associate Editor Department of Bacteriology, University of Hawaii MARJORIE MILNES, Assistant to the Editors Office of Publications and Information, University of Hawaii

EARL J. ANDERSON WENDELL A. MORDY Pineapple Research Institute Pathologist Department of Meteorology Pineapple Research Institute Pineapple Research Institute Honolulu 14, Hawaii Honolulu 14, Hawaii

TOWNSEND CROMWELL NORMAN S. NOBLE Pacific Oceanic Fishery Investigations Commonwealth Scientific and Industrial Oceanographer Research Organization Pacific Oceanic Fishery Investigations 314 Albert Street Honolulu 14, Hawaii East Melbourne, C2, Victoria, Australia HAROLD ST. JOHN MAXWELL S. DOTY Department of Botany, University of Hawaii Department of Botany, University of Hawaii G. DONALD SHERMAN R. A. FALLA Chemist, University of Hawaii Dominion Museum, Wellington, New Zealand Agricultural Experiment Station

CHRISTOPHER GREGORY ALBERT L. TESTER Department of Mathematics Department of Zoology and Entomology University of Hawaii University of Hawaii D. ELMO HARDY CLAUDE E. ZOBELL Associate Entomologist, University of Hawaii Scripps Institution of Oceanography Agricultural Experiment Station La Jolla, California THOMAS NICKERSON, Managing Editor Office of Publications and Information, University of Hawaii SUGGESTIONS TO AUTHORS Contributions to Pacific biological and physi­ 30 printed pages before sending their manu­ cal science will be welcomed from authors in all script. Authors should not overlook the need for parts of the world. (The fields of anthropology, good brief papers, presenting results of studies, agriculture, engineering, and medicine are not notes and queries, communications to the editor, included.) Manuscripts may be addressed to the or other commentary. Editor-in-Chief, PACIFIC SCIENCE, University PREPARATION OF MANUSCRIPT of Hawaii, Honolulu 14, Hawaii, or to indi­ Although no manuscript will be rejected vidual members of the Board of Editors. Use of merely because it does not conform to the style air mail is recommended for all communications. of PACIFIC SCIENCE, it is suggested that authors Manuscripts will be acknowledged when re­ follow the style recommended below and exem­ ceived and will be read promptly by members of plified in the journal. the Board of Editors or other competent critics. Manuscript form. Manuscripts should be typed Authors will be notified of the decision reached on one side of standard-size, white bond paper as soon as possible. and double-spaced throughout. Pages should be Manuscripts of any length may be submitted, consecutively numbered in upper right-hand but it is suggested that authors inquire concern­ corner. Sheets should not be fastened together ing possibilities of publication of papers of over (Continued on inside back cover) PACIFIC SCIENCE A QUARTERLY DEVOTED TO THE BIOLOGICAL AND PHYSICAL SCIENCES OF THE PACIFIC REGION

VOL. VII JANUARY, 1953 NO. 1

Previous issue published October 3, 1952

CONTENTS

PAGE

The Crangonidae, or Snapping Shrimp, of Hawaii. ALBERT H. BANNER . 3

NEWS NOTES . 145

INDEX . 147

PACIFIC SCIENCE, a quarterly publication of the University of Hawaii, appears in Janu­ ary, April, July, and October. Subscription price is three dollars a year; single copies are one dollar. Check or money order payable to University of Hawaii should be sent to PACIFIC SCIENCE, Office of Publications, University of Hawaii, Honolulu 14, Hawaii.

i i. The Crangonidae, or Snapping Shrimp, of Hawaii1

ALBERT H. BANNER2

INTRODUCTION Island, a species previously known from the THE DECAPOD FAMILY CRANGONIDAE, or the Hawaiian Islands. Coutiere in The Fauna and family Alpheidae of the earlier workers and of Geography of the Maldive and Laccadive Archi­ the present European workers, is a family of pelagoes (1906) reported five species from the shrimp known popularly as snapping or pistol Hawaiian Islands—three on the basis of re­ shrimp. The members of the family seldom ports of previous workers, one on a specimen exceed 30 millimeters in length, are usually in the Museum of Paris, and one without fur­ of stout shape, are noticeably laterally com­ ther reference (the last species is discussed on pressed, and, though their color is variable, page 142). In 1909 Coutiere, in his treatment are most often reddish- to grayish-white in of the synalpheids of the United States Na­ color. The outstanding characteristics of the tional Museum, listed four species of the most common genera of the family are the genus Synalpheus collected in Hawaiian waters large size and asymmetrical development of by the steamer "Albatross." By far the most the chela of the first pereiopod. It is the vio­ extensive studies on the Hawaiian crangonids lent closure of the enlarged chela that pro­ were those of Edmondson. In 1925 he re­ duces the clicking or snapping sound which ported on 17 species collected by the "Tana- has given the members of the family their ger" in the western islands and shoals of the common name. Hawaiian Archipelago; in 1930 he reported Considering how common the snapping two new species of a genus previously un­ shrimp are in Hawaii, surprisingly little work known to Hawaii; and in 1933 in his book, has been done on them. The earliest record of Reef and Shore Fauna of Hawaii, he gives a brief members of this family from the Hawaiian description of 13 species. Islands was made by Randall in 1839 when he The geographical limits of the present study reported Crangon ventrosa (Milne-Edwards) as coincide with the geographical limits of the Alpheus laevis Randall. Dana in 1852 [this vol­ Hawaiian Archipelago (Fig. 1). Most of the ume was distributed on February 4, 1853, ac­ collecting has been done around the larger or cording to Haskell (1942: 79) and therefore windward islands of the archipelago, which for questions of nomenclatorial priority the are, from east to west, Hawaii, Maui, Kahoo- later date should be used] reported upon the lawe, Lanai, Molokai, Oahu, Kauai, and Nii- collections made by the United States Explor­ hau. Extending over a thousand miles to the ing Expedition and listed three species, all west and north of these large islands is a chain new, from the Hawaiian Islands. Stimpson in of shoals and normally uninhabited small is­ I860 reported four species from Hawaii. Only lands known as the leeward islands; these are, two species were reported from Hawaii by again in their order from east to west, Nihoa, Bate in the Challenger Reports in 1888. Lenz Necker, French Frigate Shoal, Gardner Pin­ in 1901 reported a single species from Laysan nacles, Raita Bank, Maro Reef, Laysan, Lisi- anski, Pearl and Hermes Reef, Midway, and 1 Contribution No. 27, Hawaii Marine Laboratory. Kure, or Ocean, Island. 2 Department of Zoology and Entomology, Univer­ sity of Hawaii, Honolulu, Hawaii. Manuscript received The present study was begun in the year July 5, 1951. 1937-38, when quite extensive collections

C 3 ] 4 PACIFIC SCIENCE, Vol. VII, January, 1953

were made of the species that could be col­ a single or several specimens indicate that lected in the very shallow water of the reef other rare species not collected may exist in flats (in water up to about 5 feet deep). The the islands. However, it is unlikely that spe­ specimens in the collections of the Bernice P. cies not discussed here will be found in the Bishop Museum also were examined at that course of routine collecting. time. My departure from the Islands in the fall of 1938 prevented the completion of the Measurements study, and the work was put aside until my In the numerous measurements used in this return to Hawaii in 1946. At that time a new paper, the following standards are adopted. series of collections was made, the new col­ For total length, the body is measured from the lection in large part being made in the waters tip of the rostrum to the end of the telson- beyond the reef flats, the specimens being following the middorsal line of the body. If taken by "skin-diving" in water up to 20 or the abdomen is flexed, a slightly greater meas­ 25 feet deep. The specimens taken by the urement is obtained than if it were straight, United States Fish Commission Steamer "Al­ but the difference is not great. For the append­ batross" in the Hawaiian Archipelago in 1902 ages, the breadth is taken at the broadest point were obtained on loan from the United States unless otherwise specified; the length is meas­ National Museum; these were a valuable con­ ured from articular surface to articular sur­ tribution to the study as they represented to a face (this is especially important in the carpal large extent a deeper fauna that was impossi­ articles, as the articular surface is toward the ble to collect without dredging. Additional inferior side of the meral-carpal joint, and, if specimens in the Bishop Museum were ex­ the maximum length to the edge of the su­ amined. Some deep-water specimens dredged perior portion of the article were taken, it by the "Makua" of the Territorial Board of would be found to vary greatly with the de­ Agriculture and Forestry and by the "Salpa" gree of flexion of the joint). The lengths of the of the University of Hawaii were available. antennular articles are taken in the mid-line of Finally, I was loaned a small but very impor­ the atticle in dorsal view, with the exception tant collection of crangonids taken off the of the first article, where the length is taken island of Hawaii by R. W. Hiatt. Several from the most proximal portion visible dor- thousand specimens from these various sally to the middle of the distal end. The sources were examined. height or breadth of the large chela is taken to be I believe that this study is quite complete the diameter from superior to inferior surface, for the crangonid fauna of the Hawaiian Is­ the thickness as the diameter at right angles to lands. It is true that collections were not made this height, the length of the chela as including from all the reefs on all the islands, but enough the fingers; unless otherwise specified the collections were made on Oahu, Maui, Ha­ length of the fingers is given as a proportion of waii, and the leeward islands of the archi­ the length of the entire chela. The breadth of pelago to show that the fauna does not seem the merus of the chelipeds, in those cases where to vary to any major extent from island to is­ the cheliped was definitely triangular in cross land. On Oahu all possible types of habitats section, is taken as the width of the outer face have been examined, from muddy bays to the of the triangle; the length as from the middle clean coral growing on the outer sides of of the ischial-meral articulation to the meral- wave-swept reefs. The deep-water dredgings carpal articulation. Finally, throughout the of the "Albatross" and other ships have sam­ paper the relative lengths of the carpal articles of pled extensively the deeper fauna from the the second legs are expressed as ratios based islands. On the other hand, the several spe­ on the length of the first article, to which the cies in the collections that are represented by value 10 is assigned; it is possible to carry Hawaiian Crangonidae •— BANNER 5

Kure Mid* Pearl and Hermes Reef

Lisianski , . Lav/ear Rai-ta Bank Maro Reef Gardner Pinnacles Necker French "Frigate Shoal Nihoa

Kauai Niihau^ 4$ Oahu Kaula ^ Molokai

300 Nautical Mile Lanai»^Mai Kahoolawe

•Hawai

FIG. 1. Map of the Hawaiian Archipelago, showing principal islands, reefs, and shoals.

these ratios to two or more significant places, his cooperation and advice on several points; but in most cases the figures are rounded to and Dr. L. D. Tuthill, Editor of Pacific Science, the first significant place (for example, 10 : for advice on several taxonomic points. 7:5:5:7) because so much variation was I also wish to acknowledge the help of the found in the relative lengths of the articles following artists employed by the University within one species that it is felt that any fur­ of Hawaii. Miss Inger Achton drew Figures ther refinement of figures would be scientifi­ 6a-f, j-l. Mrs. Florence Lambeth drew Fig­ cally inaccurate. The measurements of the tel- ures \Aa-b, lld-e, 18a-d, f, /',/', m, and 24 son are self-evident. Throughout the study, a-b. Evan I. Gillespie drew the frontispiece, fractions are expressed as their decimal equiv­ redrew Figures 1, 12a-c, 42a-h from their alents and rounded with the exception of one original sources, and completed the following fourth and three fourths. drawings after I had delineated them: Fig­ ures 16 a-j; lla-c, f-j; 18?, gy h, k, I; 19a-g; Acknowledgments 2Qa-m; 34a-i; 37; 41g-b; 43a-k; 44a-l; 45 I wish to thank the following institutions a-g; 46a-i; 47a-k; 48a-h; 50a-i. and individuals for their assistance in this work: The Bernice P. Bishop Museum; the DISTRIBUTION United States National Museum; Dr. C. H. The members of the family Crangonidae Edmondson, Professor Emeritus of the Uni­ are commonly circumtropical in distribution versity of Hawaii and Curator of Zoology at although in the Pacific Ocean they have been the Bernice P. Bishop Museum, for his help reported from as far north as Sitka, Alaska and cooperation in all phases of the work, es­ (Betaeus harrimani Rathbun, 1910), and as far pecially during the initial year's study; Dr. R. south as Hermite Island, Tierra del Fuego (Be­ W. Hiatt, Professor of Zoology at the Uni­ taeus truncatus Dana, 1852). The center of dis­ versity of Hawaii, for the use of collections of tribution for the family in the Pacific appears the shrimp made by him; Dr. Fenner A. Chace, to be the East Indian Archipelago, from which Jr., of the United States National Museum for de Man (1911) reported 113 species and 20 6 PACIFIC SCIENCE, Vol. VII, January, 1953 varieties. Though he noted that the members ance of the currents in this portion of the of the family were largely limited to the upper Pacific flowing from Hawaii toward the west­ littoral zone of the sea, four species of Syn- ern Pacific, it is difficult to understand how alphem and eight species of Crangon were these larvae, without any marked powers of dredged from depths greater than 100 meters locomotion, were able to reach Hawaii from during the Siboga Expedition. Two species the closest island groups, over a thousand have been reported from tropical fresh-water miles away. It is likely that they were carried lakes. here by fortuitous combinations of eddies and Like most of the marine fauna ol the Ha­ temporary shifts in the superficial currents in waiian Archipelago, the crangonids belong to changing weather conditions. the Indo-Pacific faunal group and not to the It is likewise fruitless to speculate on the Western American. Those species that are not number of endemic species. According to the endemic range southward and westward, in present records 19 of the 44 species recorded many instances as far as the Red Sea. Crangon from these islands, or some 45 per cent, are ventrosa (Milne-Edwards) is the only excep­ known only from the Hawaiian Archipelago. tion to this statement: its range extends from However, no one knows how many of these the Red Sea through Hawaii to the Gulf of species may occur at Johnston Island or Can­ California. ton Island or in Tahiti or Samoa; no one In the present state of knowledge of Crus­ knows how many of these species may reach tacea in the Pacific it is useless to speculate on to the northern Marianas where conditions the distributional patterns of the nonendemic somewhat similar to those in Hawaii exist. It species. There are too many islands and archi­ is to be expected that the isolated position of pelagoes where no collecting of marine inver­ Hawaii would produce some speciation, but tebrates has been done at all; on the few major how much cannot yet be determined. islands where some collecting has been done, it was done usually in a random fashion, and ECOLOGY the lists of species cannot be regarded as even In Hawaii the members of the family seem approaching completion. The few exceptions to be confined largely, if not exclusively, to to this in the tropical Indo-Pacific region are the eulittoral zone, which is deeper in tropical the area around the mouth of the Red Sea, than in temperate waters. Many of the species those portions of the Indian Ocean that were are found exclusively in the shallow water of visited by J. Stanley Gardiner, and the portion reef flats and in waters less than 50 feet deep of the southwestern Pacific visited by the Si­ at the outer side of the reef; a few have never boga Expedition. It is likely that the report on been collected except on the reef flat itself, the Bikini collections, now in preparation by where at low tide the water varies from only a Fenner A. Chace, Jr., of the United States Na­ few inches to several feet in depth. The spe­ tional Museum, will add another area. But cies inhabiting the more shallow waters are whether the range of a species extends from found in five types of habitats: in dead corals, Hawaii to the Marianas or to the Marquesas where they live in cracks, folds, old worm or to New Caledonia is not known. holes, or even in tubes of algae which they Without the knowledge of the geographic construct for themselves; in living coral heads, range of the species, speculation on paths of among the branches of the coral; in sandy or distribution would be without basis. As the muddy portions of the inner reef and of bays, adults are bottom dwellers, and as the larvae where they hide about the bases of rocks and are planktonic, it is safe to assume that most partially buried objects; in the dense growth of the species that reached Hawaii were carried of algae found on exposed wave-beaten here as larvae. However, with the preponder­ coasts; and commensally with sponges and Hawaiian Crangonidae — BANNER 7 tube worms. Lack of data prevents the tabu­ are occasional species that appear to be rather lation of the habitats of the deeper-dwelling constant and do not show as much variation species, but the few I have collected in dredge as would be expected from observation of re­ hauls from deeper water were either living in lated species collected in similar numbers, dead coral boulders or in sponges. e. g., C. pacifica (Dana), C. nanus Banner, and Although the general habitats of the spe­ C. collumiana (Stimpson). cies are known, very little is known of their This apparently unusual degree of variation, actual ecology. For example, what are their reported in detail in the text, may be due to food habits? They are reported to be carnivor­ three factors: ous, yet the digestive tracts of specimens kept The first possibility is the lack of discrimi­ in aquaria with Ulva turn a bright green. In nation on my part, resulting in my confusing the case of a dozen species living in one dead several discrete but narrowly separated spe­ coral head, are there zones typical of each spe­ cies and considering them as a single species. cies, and, if so, what factors determine this However, whenever during this study any selection? Does a pair found living together considerable variation of a species was noted, in a single tube represent a pair mated for life? that species was examined critically, and, in What are the food and defensive relationships instances of complicated variations, tables of of this pair in which the smaller male has the measurements were prepared to facilitate the larger chela? What is the biological signifi­ drawing of conclusions. Characters which ex­ cance of the sexual dimorphism of the small hibited a more or less continuous spectrum of chela of some species? Why are the two close­ variation were not considered as of specific ly related species, Crangon crassimanus (Heller) worth. Although other workers have some­ and C.pacifica (Dana), which seem to inhabit times used a combination of variable charac­ the same type of coral in the same zone, al­ ters as a basis for distinguishing species, I do most never found together in the same coral not believe that, even if the procedure is valid, head? What are the limitations of the distri­ the knowledge of the family is sufficiently ad­ butional range of C. diadema (Dana)? To my vanced to warrant the use of such criteria. knowledge, none of these problems has ever A second possible cause of the variation is been investigated. a species radiation in Hawaii. This would be caused by the immigration of one or a few VARIATION WITHIN SPECIES species to an isolated location and then the A number of species reported upon in this rapid evolution of the species to fit into the study are noted to show great variation, a var­ unoccupied ecological niches. When suffi­ iation which transcends the limits previously cient time has elapsed, this process should accepted to be of subspecific and, in some produce many new but closely related spe­ cases, of specific worth. These variations are cies; when the process is just starting, it should encountered particularly in the armature and be noticeable as variations within a species. It proportions of the appendages, characteristics is true that the Hawaiian Archipelago has that have been accepted as stable within quite fewer species than other parts of the Pacific; narrow limits for a single species. Some of the for example, de Man (1911) reported 38 spe­ species from Hawaii, especially those that cies of the genus Synalphens from the area vis­ have been collected in large numbers, like ited by the Siboga Expedition, whereas in Ha­ Synalphens paraneomeris (Coutiere), Crangon waii the genus is represented by only six brevipes (Stimpson), and C. clypeata (Coutiere), species, three of which are common. There­ show great variation; many of those that are fore, those three species in Hawaii may be in less plentiful, such as S. charon (Heller) and C. the process of evolution to fit the ecological edmondsoni Banner, show less variation. There niches filled by the 38 species of the Nether- 8 PACIFIC SCIENCE, Vol. VII, January, 1953

lands Indies. In Hawaii this type of species ra­ variation often renders it necessary to erect diation has been observed in groups as diverse new species for what may later be found to be as the birds, the drosophilids, and the land divergent members of the same species. This snails. However, possibly because the marine has been done in the past, and I fear that in forms have not been as completely isolated, this paper it may have been done again. For or possibly because the marine animals have example, Crangon pseudopugnax Banner is sep­ not been as thoroughly studied, the phenom­ arated from C. percyi (Coutiere) chiefly by the enon has not been especially noted in the differences in the base of the rostrum, and the animals inhabiting the reefs and shores. character seems to be quite clear-cut and de­ The third possibility is that the variation is cisive. However, of the two species only nine not unusual or unique to Hawaii and occurs specimens are known from Hawaii, and per­ in all areas but has been overlooked to a large haps, if a large number of specimens were col­ extent by other workers. They may not have lected, it might be found that the two species had a large enough series of specimens of one described are merely the extremes of a vari­ species to give a picture of the variation, or. able single species, with specimens of inter­ having the specimens, they may have been too mediate form of the rostrum more plentiful. imbued with the type-specimen concept to Only by extensive studies can it be deter­ discuss the variations in their papers. It is true mined which of these postulated causes for that in many cases the earlier workers did dis­ the observed variation is the correct one. To cuss these variations; de Man, for example, arrive at a solution of the problem it will be presented long numerical tables on the varia­ necessary to study not only a large series of tions in some of his species. Yet in his keys, specimens from Hawaii but also many speci­ de Man (1911: 192) used characteristics that mens of these species from other portions of could easily overlap, e.g., their ranges. Specimens from the southwest i. Rostrum usually 3-times (2.75-3-25) as long as Pacific especially should be studied, as that wide. . . . Dactylus of third legs 3-times as long area appears to be the center of distribution as thick at the base. . . . for the Indo-Pacific members ol the family. ii. Rostrum 2 2-times as long as wide. . . . Dactylus of Through such a study some of the criteria third legs 2.51 times as long as thick at its base. . .. based on small differences in the relative pro­ Coutiere has erected species upon differ­ portions, like those of the antennular or carpal ences with even less allowance for variation. articles, will perhaps have to be discarded. DeMan states in his introduction (1911: 133): Coutiere also first called attention to the great impor­ With any such change in the criteria, some of tance of relative measurements . . . and it was just by the species now recognized will fall into sy­ means of these new characters that often specimens of nonymy. It may also be necessary to find new small size of Alpheus and Synalpheus proved to be species and more stable characteristics upon which to that were still unknown, while formerly such specimens base specific descriptions. usually would be regarded as juvenile forms or at most varieties. In any case, the work done in this present Coutiere stated (1909: 36) that some of these study is sufiicient to cast doubts on the valid­ species were even based largely on "geo­ ity of some of the species that have been de­ graphic characteristics." For example, "S. scribed on the basis of slight differences in paulsoni and the other Indo-Pacific forms are what appear to be variable characteristics. not represented [in America] by identical forms, but the differences are at times so Family CRANGONIDAE Weber slight that, without indication of locality, the Carapace smooth, provided with cardiac identification would be very difficult." grooves; rostrum reduced, antennal and bran- The lack of a sufficient number of speci­ chiostegal spines always absent; in known Ha­ mens of a given species to show the extent of waiian species carapace more or less project- Hawaiian Crangonidae •— BANNER 9

ing over eyes. Antennular base cylindrical, he used the name Crangon for species called with basal article not longer than sum of other Alpheus by Fabricius in 1798. Her interpreta­ two; scaphocerite rarely longer than pedun­ tion of this, according to the International cles. Mandible always bipartite, with palpus Rules of Zoological Nomenclature, made it of 2 articles. Distal article of first maxilla bi­ necessary to change the genus previously furcate at its extremity. Epipodites of first and known as Alpheus Fabricius to Crangon Weber second maxillipeds always undivided. Chela and the genus known as Crangon Fabricius to of first thoracic leg predominant, always large. Crago Lamarck. She presented the case to the (In the Hawaiian species the chelae are usual­ International Commission on Zoological No­ ly asymmetrically developed.) Carpus of che- menclature, and they confirmed her views in liped short, cup-shaped or hemispherical. Sec­ Opinion 17: "Weber's Nomenclator Entomologi- ond legs chelate, weakly developed. (The cus 1793 complies with the requirements of multiarticulate carpus in Hawaiian species is Article 25, hence the genera in question are to always composed of four to five articles.) Fol­ be accepted. . . ." Apstein in 1915 proposed lowing legs short, compressed, with spinous that the name Alpheus be a nomen conservandum, propodi (in Hawaiian species the dactyli are but the proposal was rejected (although not either simple or biunguiculate). Propodus of specifically for Alpheus versus Crangon) by the the fifth leg with more or less well-developed International Commission's Opinion 74. The "brush" of bristles placed in transverse- whole controversy is reviewed by Hult (1938). oblique rows. Abdomen usually with grad­ As a result of these rulings most American ual curve, without any pronounced bending and Australian workers are using the name at third segment; sixth segment broad and Crangon instead of Alpheus, whereas most Eu­ short; size and shape of pleura showing sexual ropean workers are still using the name Al­ dimorphism. Pleopods of usual form, show­ pheus. As this paper goes to press there is an ing only slight sexual dimorphism. The bran­ appeal to the International Commission for a chial formula always includes five pleuro- suspension of the rules to permit the re-estab­ branchs. (Description adapted from Coutiere, lishment of the name Alpheus. 1899.) The position of the family Crangonidae dif­ fers but slightly in the various modern sys­ Of the 18 (?) previously known genera of tems of classification of the lower decapods. the family Crangonidae, only 4 are represented However, because the names and definitions in the fauna of the Hawaiian Archipelago. In applied to the subsidiary classifications differ addition, one new genus has been described; with the various schemes, that of Balss (1927) this genus, however, is not endemic to the has been selected arbitrarily for use in this pa­ Hawaiian Islands, as it includes one species per. Balss followed, with only minor changes, from the southwest Pacific which de Man the scheme of classification earlier devised by tentatively placed in the genus Alpheopsis Cou­ Borradaile. In it the Crangonidae are placed tiere. The great majority of the species in Ha­ in the suborder Natantia, the division Eucyph- waii are members of the genus Crangon Weber idae (which seems to be the equivalent of the as they are in all other investigated regions of tribe Caridea of others), and the tribe Palae- the Indo-Pacific. moniida (equivalent to the superfamily of the The generic name Crangon and the family same name used by other workers). Except for name Crangonidae replaced the widely used differences in names as pointed out above, all names Alpheus and Alpheidae as the result of present workers appear to be in accord on the the work of Rathbun in 1904. At that time she classification of the Crangonidae to this point. published the results of an examination of a Within the tribe Palaemoniida there is a paper by Weber, published in 1795, in which difference of opinion as to the relative position 10 PACIFIC SCIENCE, Vol. VII, January, 1953 of the family with respect to the three or fout not angular; with anal tubercles; with other families. Primarily on the basis of the 7 to 8 epipodites Crangon (p. 46) characteristics of adults of the lower genera of JOUSSEAUMEA Weber the family, Coutiere (1899) placed the cran- Carapace extending forward as rostrum of gonids below but adjacent to the Hippolyti- variable size and as orbital teeth forming dae. Gurney (1938), on the basis of charac­ broad eave-like projection that conceals eyes teristics of the larvae, concluded that the in dorsal view. family was most closely related to the Palae- Antennular peduncle with strong stylocer- monidae and of higher phylogenetic position. ite. Scaphocerite broad, oval, with squamous KEY TO THE HAWAIIAN GENERA portion equal to or longer than lateral spine. Carpocerite of antennular peduncle short and 1. Eyes at most only partially concealed by broad, sometimes shorter than merocerite. projection of margin of carapace, and First chela very asymmetrical, carried flexed visible at least anteriorly; large chela at carpus so that distal end approaches body. without a plunger on the dactylus fit­ Fingers of large chela serrate with large teeth ting into a cavity on the finger oppo­ joining exactly. Merus long. site, but instead usually bearing a se­ Carpus of second legs composed of 5 ar­ ries of strong teeth 2 ticles. Following legs robust with merus un­ Eyes completely concealed by projec­ armed, ischium bearing 2 spines, dactylus tion of anterior carapace and visible simple. Sixth abdominal segment without ar­ only in anteroventral aspect; large ticulated pleura. No anal tubercles on telson. chela almost always with a plunger Posterior margin of telson emarginate. that fits into a cavity on the finger op­ Branchial formula: 5 pleurobranchs, 1 ar- posite, never serrate 4 throbranch, 8 epipodites. 2. Sixth abdominal segment without ar­ DISCUSSION: The nine species recorded for ticulated pleura; large chela carried the world by de Man (1911, 1922) have been flexed toward body, with the palm increased to eleven by the addition of the two excavate to accommodate merus; tel- Hawaiian species described by Edmondson son with tip emarginate (1930). Of these, three species occur in the Jousseaumea (p. 10) Red Sea and the Gulf of Aden, two species in Sixth abdominal segment usually with the Banda Sea south of the Celebes Islands, pleura articulated; if large chela is car­ one "from the coasts of America," one from ried flexed, then merus is flattened to Puerto Rico, and the other two species from accommodate palm; telson with tip the Hawaiian Islands. There are no records in­ either straight, slightly convex, or ex­ dicating the occurrence of any of the species tended to form a triangle 3 at other than the type locality. 3. Large chela carried extended; posterior margin of telson straight to slightly HAWAIIAN SPECIES OF JOUSSEAUMEA arcuate Alpheopsis (p. 14) 1. Rostrum long and broad, extending Large chela carried flexed; posterior mar­ from basal angles anterior to middle gin of telson projecting as a strong of eyes to apex beyond end of second triangle Neoalpheopsis (p. 20) article of antennular peduncle 4. With pterygostomial margin pro­ J. mauiensis (p. 12) duced into a definite angle; without Rostrum short, extending from base be­ anal tubercles; with only 2 epipod- tween eyes to or slightly beyond end ites Synalpheus (p. 26) of first antennular article With pterygostomial margin rounded, J. brevirostris (p. 12) Hawaiian Crangonidae — BANNER 11

jlmm A _i 2mrr B j0.5mm

FIG. 2. Jousseaumea mauiensis Edmondson. a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, ventral aspect; e, small cheliped;/, second legs; g, third legs; h, telson. (a, b, scale A; c-g, scale B; h, scale C.) 12 PACIFIC SCIENCE, Vol. VII, January, 1953

Jousseaumea mauiensis Edmondson dactylus moderately heavy, curved, simple, Fig. 2 a-h with acute tip. Telson 1.5 times as long as broad at base; Jousseaumea mauiensis Edmondson, Bernice P. tip 0.4 as broad as base. Terminal emargina- Bishop Mus., Occas. Papers 9(10): 5-7, fig. tion 0.1 total length of telson, rounded and 2a-g, 1930. bearing 3 to 4 pairs of plumose setae. Both DESCRIPTION: Rostrum nearly equilaterally pairs of terminal spines almost as long as triangular, with base arising anterior to cor­ width of tip. nea of eye and only slightly shorter than con­ DISCUSSION: (See discussion under/, brevi- cave lateral margins; tip reaching to middle of rostris). third antennular article. Orbital teeth acute DISTRIBUTION: The type specimen came and about 0.2 to 0.25 as long as rostrum. Ros­ from the island of Maui; additional specimens trum with a rounded keel extending back to have been collected about Oahu at Hanauma its base. Pterygostomial angle rounded. Bay, Kahala Beach, off Waikiki Reef, off Antennular peduncle short and heavy, with Nanakuli, and at Kawela Bay, at depths from third article longer than visible portion of the reef flat to about 20 feet. Edmondson either first or second article; second article al­ (1946: 256) states that this species is commen- most twice as broad as long. Stylocerite acute, sally associated with a large polychaete of the reaching slightly beyond end of rostrum. Bi­ genus Eurythoe. furcation of outer flagellum occurring near base; secondary flagellum short. Jousseaumea brevirostris Edmondson Antennal peduncle with basicerite bearing Fig. 3 r-g a poorly developed lateral tooth. Scaphocer- Jousseaumea brevirostris Edmondson, Bernice ite slightly longer than antennular peduncle, P. Bishop Mus., Occas. Papers 9(10): 7-9, with squamous portion exceeding poorly de­ figs. 2a-e, 1930. veloped lateral spine in length. Carpocerite short and heavy, reaching almost to end of DESCRIPTION: Carapace with dorsal keel for antennular peduncle. Flagellum large in diam­ over 0.5 its length, continuous anteriorly with eter basally, only as long as carapace. rostral carina. Rostrum triangular, reaching Large chela 0.3 length of body, with basal slightly beyond end of first antennular article. articles almost as long as chela. Chela large, Eyes normally concealed by orbital hoods in subcylindrical, with palm 1.3 to 1.5 times dorsal and lateral view; orbital hoods contin­ length of fingers. Fingers curved and com­ ued as small acute orbital teeth, teeth about pressed, with faces in opposition armed with 0.3 length of rostrum. Area between base of 4 large and 2 smaller exactly fitting teeth; orbital teeth and dorsal carina slightly con­ tips curved, acute, and crossing each other. cave. Merus flattened on side toward palm. Small Antennular peduncle short and heavy, with chela about 0.25 length of larger chela, not visible portion of first and second articles sub- swollen, and without teeth. equal in length, about 1.5 times as long as Carpal articles of second legs with ratio broad. Heavy stylocerite with acute tip, reach­ 10 : 2.6 : 1.6 : 1.6 : 3.3. ing almost to end of second antennular arti­ Merus of third legs 3.5 times length of cle. Flagella short, with shorter branch of bi­ ischium, 1.3 times length of carpus, 0.75 times furcated upper flagellum not as long as last length of propodus, 2.7 times length of dac- 2 antennular articles together. tylus. Merus over 4 times as long as broad, Basicerite heavy, bearing acute lateral tooth unarmed; propodus bearing about 8 movable and short spiniform process (in lateral view) spines and about 6 shorter tufts of bristles; above articulation of scaphocerite. Scaphocer- Hawaiian Crangonidae •— BANNER

d

FIG. 3- Jousseaumea brevirostris Edmondson. a, b, Anterior region, dorsal and lateral aspects; c, d, large chela, lareral and ventral aspects; e. second legs;/, third legs; g, telson. 14 PACIFIC SCIENCE, Vol. VII, January, 1953 ite short and broad, with weak lateral spine gles), whereas in J. mauiensis it is long, reach­ just reaching end of antennular peduncle, ing beyond the end of the second antennular with squamous portion noticeably longer. article, and broad at its base. Furthermore, in Carpocerite short, heavy, reaching slightly be­ /. brevirostris the rostrum and anterior cara­ yond end of second antennular article. pace carry a carina that is lacking in J. maui­ Large chela rounded, subcylindrical, length ensis. Second is the nature of the terminal cleft about 3.3 times breadth at maximum diam­ of the telson, which in J. brevirostris is narrow eter; surface of palm toward merus excavate and deep, about 1.5 times as deep as broad at permitting flexure of appendage. Fingers over the outer edge, whereas in J. mauiensis it is 0.4 length of chela, distally compressed and broad, less than half as deep as broad at the somewhat curved; dactylus and fixed finger outer edge. Other slight differences can be with 8 large triangular teeth in opposition, found in the length of the stylocerite, the ratio tips acute and crossing. Merus slightly longer of the lengths of the articles of the antennular than palm, slender, rounded on outer side, peduncles, the merus of the third legs, and flattened on side toward carpus and chela. other characteristics. Ischium also slender, 0.5 as long as merus. DISTRIBUTION: The two specimens known Small chela lacking. were collected from Waikiki Reef (the type Carpal articles of second legs with ratio locality) and from the nearby reef at Kahala, 10 : 3 : 2 : 2 : 4. Oahu. Third legs with ischium 0.5 length of me­ rus; merus 4.1 times as long as broad, un­ ALPHEOPSIS Coutiere armed; carpus over 0.6 length of merus, with Frontal border of carapace produced into superior margin projecting, acute; propodus rostrum and projection that screens eyes from 0.75 length of merus, with 3 weak spines and above; orbital teeth present or absent. Cornea several bristles; dactylus 0.25 length of merus, of eyes visible from front and maybe from somewhat curved, slender, simple. sides. Telson almost twice as long as broad at Antennular peduncle short and stout; sty­ base; tip almost 0.4 breadth of base; anterior locerite variable. Scaphocerite usually broad surface strongly convex, posterior surface without heavy lateral spine. Carpocerite long. more flattened. Terminal cleft of telson about Large chela carried extended, without 1.5 times as deep as wide, only 0.1 of total grooves or sheaths. Carpus globular, merus length of telson. Inner terminal spines equal roughly triangular. Chelae show slight asym­ in length to breadth of tip. metry; either entire and subcylindrical or with DISCUSSION: The description given above lines and depressions. Fingers compressed, and the figures are those of the type specimen either without teeth or with simple arrange­ previously described and figured by Edmond- ment of exactly fitting teeth. son. The slight differences to be noted on a Second thoracic legs with carpus of three point by point comparison of the two descrip­ articles (A. idiocarpus Coutiere) or five. tions are merely differences in interpretation. Following legs robust, without teeth on The two species of this genus from Hawaii, merus; propodus weakly spinose; dactylus bi- J. brevirostris and J. mauiensis, can best be sep­ unguiculate or simple. Propodus of fifth legs arated by two characteristics. First is the na­ with or without "brush" of bristles. ture of the rostrum, which in J. brevirostris is Sixth abdominal segment without articu­ short, reaching only to near the end of the lated pleura only in A. biunguiculatus Banner. first antennular article and correspondingly Posterior border of telson rounded. narrow at its base (rostrums of both species Branchial formula: 5 pleurobranchs, 1 ar- approach the dimensions of equilateral trian­ throbranch, 8, 7, or 6 epipodites. Hawaiian Crangonidae — BANNER 15

DISCUSSION: The description of this genus either first or third. Stylocerite with tip ex­ as given by Coutiere (1899: 330) is modified tending almost to end of second article of to accommodate those species described sub­ antennular peduncle. sequently and further to permit the assign­ Scaphocerite short and broad, as long as ment of a new species of doubtful relation­ antennular peduncle, slightly shorter than pe­ ship, A. biunguiculatus Banner, thereto. duncle of antenna. Lateral spine of scaphoce­ The genus is confined largely to the west rite not pronounced and equal in length to Pacific and the Indian Oceans; aside from A. squamous portion. Lateral spine of basicerite chilensis Coutiere, found off South America in acute but not well developed. the Pacific, the two species listed below are . Basal articles of third maxilliped with strong the only members of the genus that have been spiniform bristles. recorded as far east as the central Pacific Chelae almost symmetrical in size and Ocean. shape, not showing more than 10 per cent difference in size; rounded and without HAWAIIAN SPECIES OF ALPHEOPSIS grooves. Fingers compressed, with slight ir­ 1. Dactyli of third to fifth legs simple. . . regularities on opposing faces; usually with A. equalis (p. 15,) stiff recurving hairs. Carpus elongate with Dactyli of third to fifth legs biunguicu- outstanding flange into which proximal end late (?) A. biunguiculatus (p. 18) of propodus fits, with slight constriction be­ fore this flange. Merus with upstanding collar Alpheopsis equalis Coutiere on outer distal side into which proximal end Fig. 4 a-o of carpus fits, with rounded edges, with vari­ Alpheopsis equalis Coutiere, Paris Mus. d'Hist. able set of long stiff bristles on inner lateral Nat., Bui. 2(8): 380, 1896. and superior margins. Ischium also with bris­ Alpheopsis equalis var. truncatus Coutiere, Soc. tles on superior distal face. Philomath. Paris, Bui. IX, 5(2): 89, 1903. Carpal articles of second thoracic leg with Alpheopsis equalis and Alpheopsis equalis var. ratio 10 : 6 : 4 : 4 : 5. truncatus Coutiere, Fauna and Geog. Maid, Third legs relatively long and slender, with and Laccad. 2(4): 868, 1905. [A fuller de­ ischium 0.5 as long as merus; merus 6.5 times scription and discussion, with figures, of as long as wide; carpus slightly less than 0.5 both forms.] as long as merus; propodus slightly longer Alpheopsis aequalis de Man, Siboga Exped., than merus; dactylus about 0.3 length of pro­ 1 39a (2): 177, 1911. podus. Merus without spines or teeth; carpus DESCRIPTION: Frontal border of carapace with long slender spine on distal portion of produced into acute rostrum extending about inferior margin; propodus bearing 4 or 5 0.7 length of visible portion of first antennu- spines on inner margin. Appendage glabrous lar article. Termination of rostrum, in lateral except several setae on distal end of merus, view, varying from acute to obliquely trun­ distal end of propodus, and middle of dac­ cate. Frontal border projecting beyond cor­ tylus. "Brush" of fifth thoracic legs of 10 neas of eyes, masking them entirely in dorsal rows of weak bristles. view and almost or entirely in lateral view; Species small, maximum length about 12 marked by 2 pairs of slight convexities, one mm.; color in life not noted. pair dorsad, one pair laterad of eyes. Ptery- DISCUSSION: In the collections available gostomial angle rounded or produced into from Hawaii there are 13 specimens of this acute tooth. species, some broken and some entire. These Antennular peduncle short and heavy with specimens vary in a series of points, yet plain­ second article shorter than visible portion of ly all are members of the same species for they 16 PACIFIC SCIENCE, Vol. VII, January, 1953 show essentially the same structure. The both chelae attached, the larger is only about points of variation are as follows: 10 per cent longer than its mate, but in sev­ 1. The nature of the tip of the rostrum (Fig. eral the larger chela was at least twice the size Ab, c, d). In his three specimens Coutiere was of the smaller. It is presumed that in these able to distinguish between the species A. asymmetrical specimens the smaller chela was equalis, which had a rostrum with a uniform in the process of regeneration. Some of the taper, and the variety, A. equalis var. truncatus, specimens, moreover, had their fingers armed which had an abruptly truncate rostrum. with a dense covering of recurving bristles; in Specimens approaching both conditions are others the bristles, while present, were neither represented in the Hawaiian collections; how­ dense nor conspicuous. Finally, all of the meri ever, in those with the pointed rostrum it is a of the chelipeds had spine-like bristles, but in little less pointed than in those described by some they were conspicuous and strong and Coutiere, and in those in which the rostrum is in others they were inconspicuous and weak. truncate it is a little less abruptly so than in Between the specimens showing the extremes his variety. In addition to the two extremes, of variation of the chelipeds were intergrading there are a series of intergrading individuals individuals. which show various degrees of truncation. A If these variations were constant, so the number of these specimens have a very slight truncated rostrum was always found with a tooth at the lower side of the terminal trun­ pterygostomial spine and a spiny merus, etc., cation. I conclude, therefore, that the variety and if there were no intergrading individuals, described by Coutiere is merely an individual the differences would be sufficient for the erec­ variation. tion of a new species; as it is, the species must 2. The presence of a tooth on the pterygo- be considered merely as one in which rather stomial angle of the carapace (Fig. Ac, d). In great variation occurs. most specimens in the Hawaiian collections DISTRIBUTION: The species has been re­ the carapace below the base of the antennae is ported from the Hawaiian Islands by Ed- rounded, as Coutiere shows for A. equalis; in a mondson (1946). Specimens in the present few it has a tooth somewhat similar to that collection have been collected on Oahu as fol­ shown for A. equalis var. truncatus. However, lows: on algal holdfasts in the intertidal zone among the specimens possessing the tooth, it near Koko Head and Kawela Bay; in water is not always at the same stage of develop­ from 10 to 20 feet deep at Hanauma Bay, ment, and in one specimen there is a tooth on Waikiki, and Nanakuli. On Maui it was col­ one side but none on the other. Therefore this lected at three localities between Lahaina and too must be considered an individual varia­ Kalama Park in 3-8 feet of water. On Hawaii tion. it was collected by Hiatt from a tidal pool on 3. The coverage of the eyes by the carapace Keaoi Island, Kau Coast. (Fig. Ab, d). In lateral view, the eyes of most Three specimens were collected by the "Al­ specimens are completely obscured by the batross" in Auau Channel, in water from 32 carapace which may project beyond the cor­ to 43 fathoms deep (U. S. N. M. 63454, 63455, neas for a varying distance; in several, how­ 63456, "Albatross" Stations 3873, 3876, 3872). ever, the anterior portion of the corneas of the Either the species proper or its variety has eyes projects beyond the margin of the cara­ been reported from Djibouti and Permin at pace. the mouth of the Red Sea, from the Maldives 4. The size and the armature of the chelae in the Indian Ocean, and from New Caledonia (Fig. Ae-k). In most of the specimens with and Samoa. Hawaiian Crangonidae — BANNER 17

FIG. 4. Alpheopsis equalis Coutiere. a, b, Anterior region, dorsal and lateral aspects; c, d, variations in anterior carapace; e, f, large chela, lateral and medial aspects; g, h, merus, large cheliped, lateral and medial aspects; /,_/', variations in large chela; k, small chela of same specimen as chela /'; /, second leg; m, third leg; n, fifth leg; a, telson and uropods of specimen shown in c. {a, b, /, m, scale A; c, d, i-k, scale B; e—h, scale C; n, o, scale D.) 18 PACIFIC SCIENCE, Vol. VII, January, 1953

(?) Alpheopsis biunguiculatus sp. nov. Second thoracic legs 3.2 mm. long. Carpal Fig. 5 a-h articles with ratio 10 : 8 : 6 : 6 : 10. Length of first secondary article 5.8 times breadth. TYPE SPECIMEN: A male 8.4 mm. long col­ Third legs moderately heavy, 3.1 mm. long. lected from a head of Pocillopora meandrina Ischium 0.34 mm. long, with strong movable Verrill from the intertidal zone at Halape, Kau spine. Merus 0.95 mm. long, 0.23 mm. broad Coast, Hawaii, by R. W. Hiatt. This is the (4.1 times as broad as long), with 2 movable only specimen of this species in the collection spines. Carpus 0.45 mm. long, 0.16 mm. (U. S. N. M. 93451). broad, unarmed. Propodus 1.04 mm. long, DESCRIPTION: Front of carapace projecting 0.14 mm. broad, with 6 small movable spines, over eyes but abruptly truncate and flattened 2 on distal end. Dactylus 0.29 mm. long, 0.09 mm. broad, with strong secondary un­ dorsally; without keels, ridges, or grooves. guis. Rostrum, in dorsal view, very small, narrow and acute, reaching only 0.7 of length of first Fifth legs similar to third with slight differ­ article of antennular peduncle; ventral keel ences in proportions. Propodus with no sem­ and abrupt truncation in lateral view. Hoods blance of "brush," but with 4 feeble spines. over eyes slightly inflated, without teeth or Secondary unguis of dactylus poorly devel­ other projections. Eyes extending slightly be­ oped. yond margins of hoods in lateral view. Ptery- Basiopodites of first 4 pairs of pereiopods gostomial angle rounded. with rudimentary exopodites(?) (Fig. 5h). Antennular peduncle short and heavy, with Pleura of sixth abdominal segment not ar­ visible portion of first and second article each ticulated. 0.2 mm. long, with third article 0.25 mm. Telson 1.18 mm. long, 0.55 mm. broad at long; second article 0.23 mm. wide. Styloce- base (2.1 times as long as broad), with tip rite short, rounded, with short acute spine 0.25 mm. broad. Lateral margins slightly con­ arising abruptly from lower margin, tip not vex; terminal margin more markedly convex. reaching to end of first antennular article. Sec­ Four pairs of long setiferous bristles between ondary bifurcation of outer flagehum well de­ 2 terminal spines. Dorsal spines 0.5 and 0.8 veloped. All flagella broken before tip. mm. from anterior end. Anal tubercles absent. Lateral spine of basicerite acute, short, not Branchial formula: 5 pleurobranchs, 1 ar- reaching to end of broad portion of styloce- throbranch (rudimentary), 7 epipodites (mas- rire. Scaphocerite broad, with lateral spine tigobranch on seventh leg poorly formed; poorly developed, about as long as squamous setobranch lacking on eighth). portion. Scaphocerite and antennular pedun­ DISCUSSION: Unfortunately, this most per­ cle equal in length, both considerably longer plexing species is represented by a single than carpocerite. Antennular flagella lost. broken specimen, so any conclusions about it Third maxillipeds with basal article (ischio- must be tentative. It was placed in this genus merus) 0.68 mm. long, 0.17 mm. broad; car­ because in most characteristics it is very much pus 0.30 mm. long, 0.12 mm. broad; third like A. equalis Coutiere. However, the lack of article (propo-dactylus) 0.30 mm. long, 0.10 articulation of the pleura of the sixth abdom­ mm. broad. Appendages with scattered hairs inal segment and the presence of exopodites along basal joints, 8 rows of short stiff bristles on the periopods raise serious doubts as to its on terminal article, 2 spiniform bristles on end relationship to this genus. The second char­ of first article, and 4 on second article. Exopod acteristic may not be of any worth, possibly reaching to middle of carpus. being merely a persistent larval characteristic Large chelipeds lost. or an atavism. Because the specimen is incom- Hawaiian Crangonidae •— BANNER 19

FIG. 5. (J)Alpheopsis bimiguicnlatus sp. nov. a, h, Anterior region, dorsal and lateral aspects; c, third maxilliped; ./, second leg; e, third leg; j, fifth leg; g, telson; h, lateral view of the branchiaJ cavity, the two anterior pleuro- branchs shown only in outline, (a-g, scale A; h, scale B.) 20 PACIFIC SCIENCE, Vol. VII, January, 1953 plete and because of the dubious worth of the portion. Carpocerite normal. Flagellum al­ exopodites as a diagnostic character, it has most as long as body. seemed advisable not to create a new genus Last article of third maxilliped with trans­ for this specimen. When another specimen, verse rows of short setae on inner side, similar complete with the large chelipeds, is found it to that of Crangon. may be necessary to remove this species from First pair of chelae well developed, large, this genus. approaching bilateral symmetry, without sex­ RELATIONSHIP: If A. biunguiculatus belongs ual dimorphism. Chela can be folded back to this genus, it can be separated from all against merus, merus triangular, somewhat other species by the lack of articulated pleura excavate and twisted to accommodate chela; on the sixth abdominal segment; it can be merus without spines or teeth. Carpus cyan- separated from all species save A. fissipes Cou­ thiform, more elongate than in Crangon. Chela tiere by the presence of the biunguiculate dac- proper rounded, smooth, tapering, without tyli on the third to fifth legs, and from A. fis­ crests or lobes, rotated so dactylus rests sipes (Coutiere, 1908, and more completely, against merus when flexed. Cutting surface of 1921), by the characteristics presented in the dactylus with serrations that meet exactly following tabulation: with similar serrations on fixed finger, or de­ veloped as knife-like ridge. Palmar and digital A. biunguiculatus A. fissipes Rostral tip without hairs Rostral tip with hairs adhesive plaques poorly developed. Stylocerite shorter than Stylocerite almost as long Second legs with carpus of 5 articles, first first antennular article as second antennular article secondary article slightly shorter than remain­ Carpocerite shorter than Carpocerite longer than ing 4 together. antennular peduncle antennular peduncle Third legs with proximal articles with mov­ Terminal article of third Terminal article of third maxillipeds with bristles maxillipeds without able spinules; propodus with few weak spines; bristles dactylus simple. Fifth legs with usual brush Ratio of carpal articles of Ratio of carpal articles of on inner side of propodus poorly developed. second legs 10 : 8 : 6 : second legs 10 : 5 : 5 : 6 : 10 4 : 6 Abdomen of usual form. Pleura of sixth ab­ Ratio of articles of third Ratio of articles of third dominal segment articulated. Uropods of legs: ischium, 0.75; me- legs: ischium, 0.8; me- usual form. Telson tapering, convex dorsally. rus,'2.1; carpus, 1.0; pro­ rus, 1.4; carpus, 1.0; podus, 2.3 propodus, 1.6 Tip of telson with 2 pairs of spines, medial­ ly to them margin projects markedly as acute NEOALPHEOPSIS gen. nov. triangle; triangle with 3 or 4 pairs of setiferous Carapace projecting to form rostrum and bristles. No anal tubercles. orbital teeth which obscure eyes, in dorsal See Table 1 for branchial formula. view; corneas may or may not extend slightly beyond edge of carapace in lateral view. No This genus is distinguished from almost pterygostomial tooth or even acute angle. all other genera of crangonids by a series of Antennule with heavy peduncle, basal ar­ characteristics, the more conspicuous of ticle nearly as long as both distal articles. Sty­ which are given here. locerite broad, acute tip reaching distal por­ It is distinguished from Athanas Leach, tion of second antennular article. Inner flagel- Athanopsis Coutiere, Arete Stimpson, Areteopsis lum as long as carapace, outer somewhat de Man, Automate de Man, Pterocaris Heller, shorter and with setiferous bifurcation on ba­ and Ogyris Stimpson by the covering of the sal portion. eyes, in dorsal view, by the projecting cara­ Antenna with spine of basicerite broad, but pace; from Betaeus Dana and Parabetaeus Cou­ apically acute. Scaphocerite broad, rounded tiere by the presence of a projecting rostrum; with lateral spine not longer than squamous from Synalpheus Bate and Crangon Weber by Hawaiian Crangonidae — BANNER 21

TABLE 1 the merus, the fingers are serrate or toothed, BRANCHIAL FORMULA OF THE GENUS Neoalpheopsis and the chelipeds are almost symmetrical; whereas in Alpheopsis the chelae are carried ex­ PLEURO- ARTHRO- EPI- tended, the fingers are usually without serra­ BRANCHS BRANCHS PODITES* tions, and the chelipeds usually have marked Maxillipeds asymmetry. The telson in Neoalpheopsis pro­ 1 1 jects terminally to form a large and acute tri­ 2 1 3 a angle, whereas in Alpheopsis the terminal mar­ gin is either straight or slightly arcuate. Thoracic legs 1 1 1 a, /3 TYPE SPECIES: Neoalpheopsis hiatti Banner. 2 . 1 a, (8 3 1 a, /3 SPECIES OF NEOALPHEOPSIS 4 1 «t. 0 5 1 1. Rostrum reaching to near middle of sec­ ond antennular article; merus of third * The designations a and (3 are used here as they legs 4.5 times as long as broad are applied by Coutiere (1899: 268-286), where a re­ fers to the arm-like mastigobranch, and /3 refers to the N. hiatti (p. 21) small setiferous lobe-like setobranch. Rostrum reaching only to distal third of | This epipodite reduced to a small protuberance. first antennular article; merus of third legs 7.5 times as long as broad the lack of the well-developed plunger and N. euryone (p. 25) adhesive plaques on the dactylus of the large chela. Neoalpheopsis hiatti sp. nov. Neoalpheopsis bears a superficial resemblance Fig. 6 a-l to Jousseaumea Coutiere as both the frontal border of the carapace and the teeth on large TYPE SPECIMEN: A male 13.5 mm. long chela are similar and there is a general simi­ collected from a head of coral in 20 feet of larity of form. However, they can be distin­ water, Hanauma Bay, Oahu. (Sizes and places guished easily by the modification for the of collection of paratypes listed below.) (U. S. flexure of the large chela, as the merus is N. M. 93452.) grooved in Neoalpheopsis and the chela is DESCRIPTION: Rostrum broadly triangular grooved in Jousseaumea; by the pleura of the with tip reaching to initial third of second ar­ sixth abdominal segment which are articu­ ticle of antennular peduncle; rostrum slightly lated in Neoalpheopsis and not in Jousseaumea; convex dorsally, with ventral keel. Orbital by the telson, the posterior margin of which teeth broadly triangular with rounded tips; projects as a triangle in Neoalpheopsis but which length 0.3 of rostrum. Carapace slightly in­ is posteriorly cleft in Jousseaumea. The two flated over eyes. No pterygostomial angle. can be distinguished further by the branchial Eyes concealed in dorsal view; in lateral formula which is 5-1-7 in Neoalpheopsis and view slightly visible ventrad and anterior to 5-1-8 in Jousseaumea. margin of orbital hood. In all ways Neoalpheopsis shows a close rela­ First antennular article 0.85 mm. long from tionship to Alpheopsis Coutiere. The form of base, with visible portion 0.30 mm. long; sec­ the carapace is similar in the two genera; the ond article 0.30 mm. long; third article 0.40 cephalic appendages are similar; the walking mm. long; all about 0.35 mm. in diameter. legs of the thorax are similar; the pleura of the Stylocerite broadly acute, not spiniform, ex­ sixth abdominal segment are similar. The che­ tending to end of second antennular article. lae and the telson differ, however. The chelae Internal crest of first article well developed, in Neoalpheopsis are carried folded back against with acute tooth extending to end of article. 22 PACIFIC SCIENCE, Vol. VII, January, 1953

Inner flagellum 4.1 mm. long; outer flagellum both chelae with ridge, armed with fine serra­ 3.7 mm. long, with base thickened for 0.30 tions, apposable to bare spot on dactylus mm. and produced as secondary lobe 0.35 opposite; both fingers taper to unarmed, mm. long. curved, acute tip that crosses opposite mem­ Lateral margin of basicerite of antennal pe­ ber. Palmar and digital adhesive plaques poor­ duncle with broad, acute tooth extending to ly developed. middle of first article of antennular peduncle. Second legs 4.3 mm. long. Articles of car­ Scaphocerite reaching to end of antennular pus with ratio 10 : 2 : 2 : 2 : 4. Propodus al­ peduncle, almost 0.5 as broad as long, with most twice as long as last article of carpus, setae around rounded distal and inner mar­ fingers 0.5 of length. gins; lateral spine poorly developed, shorter Third legs with ischium 0.8 mm. long; than squamous portion. Carpocerite not reach­ merus, 1.6 mm. long, 0.35 mm. broad; carpus, ing to end of the scaphocerite. Antennal fla­ 0.7 mm. long, 0.2 mm. broad; propodus, 1.3 gellum 8.0 mm. long. mm. long, 0.2 mm. broad; dactylus, 0.35 mm. Basal article of endopod of third maxilliped long, 0.06 mm. broad. Ischium unarmed; me­ (ischio-merus) 1.5 mm. long, with basal 0.7 rus with strong movable spine on middle of flattened and curved (not shown in aspect inferior margin, usually pressed against the drawn), straight and subcylindrical distally. article; carpus with single small movable spine Carpus 0.46 mm. long, cylindrical. Third arti­ at articulation of propodus; propodus with 4 cle (propodo-dactylus) 0.7 mm. long, circular spines, 3 proximal less than 0.5 width of pro­ in cross section, tapering; with 8 transverse podus, fourth 0.7 width of adjacent propodus. rows of spinules. Dactylus with uniform taper to acute tip, Chelipeds well developed, 7.5 mm. long, slightly curved. quite symmetrical, usually carried partially or Fourth leg similar to third. Fifth leg with completely folded back on merus. Ischium armature of propodus reduced to 2 short short, delimited from merus by slight groove. spines, usual "brush" on distal portion of ar­ Merus 2.9 mm. long, 0.45 mm. broad; con­ ticle reduced to 4 small bunches, not rows, of vex-concave distally in cross section, concav­ bristles, each composed of at most 6 or 7 in­ ity accommodating chela; edges of concavity dividual bristles. produced as low ridges distally, outer ridge Anterior 5 abdominal pleura rounded. obsolete proximally, inner ridge distinct to Pleura of sixth abdominal segment acute, tri­ base. Merus without spines or setae, inferior angular, with margins slightly convex, 0.42 internal margin produced as small tooth. Car­ mm. long, 0.32 mm. broad at base. Uropods pus conical, 0.7 mm. in length and distal of normal form, broad, rounded. breadth, unarmed. Upon flexion of both car­ Telson convex dorsally, concave ventrally; pal articulations chela rests against merus. sides straight with marked taper. Proportions Propodus 3.6 mm. long, fingers 1.3 mm. Palm as follows: Total length, 1.80 mm.; width at somewhat flattened, curved, without keels or base, 0.74 mm.; width at tip, 0.32 mm.; thick­ sharp angles, 1.0 mm. high. Maximum width ness at base, 0.42 mm.; width of base of ter­ of fixed finger 0.3 mm.; dactylus slightly thin­ minal triangle, 0.14 mm.; length (or height) ner (0.25 mm.). Proximal 0.75 of fixed finger of terminal triangle, 0.14 mm.; anterior pair of of right chela with 11 large flattened triangu­ dorsal spines 0.98 mm. from base; posterior lar teeth, larger distally. These alternate with pair of dorsal spines 1.40 mm. from base; an­ corresponding smaller teeth on dactylus. Dac­ terior pair of dorsal spines 0.10 mm. long; tylus of left chela with knife-like crest show­ posterior pair of dorsal spines 0.14 mm. long; ing minor irregularities homologous to teeth lateral terminal spines 0.09 mm. long; medial of right dactylus. Distal 0.25 of fixed finger of pair of terminal spines 0.50 mm. long. Tip of

i Hawaiian Crangonidae — BANNER 23

FIG. 6. Neoalpheopsis hiatti gen. et sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, anterior region of specimen 4; d, third maxilliped; e, large cheliped; /, merus; g, left chela; h, right chela; i, large chela of speci­ men 3;j, third leg; k, fifth leg; /, telson, uropods, and articulated pleura of the sixth abdominal segment (drawing foreshortened), (a, b, d-f,j-l, type specimen; c, g, h, specimen 4; /', specimen 3.) {a, b, d-f, scale A; c,j-l, scale B; g-i, scale C.) 24 PACIFIC SCIENCE, Vol. VII, January, 1953

telson with 3 additional pairs of long setae. ous teeth on both fingers instead of fine teeth Anus at anterior end of ventral side of telson; on the fixed finger in opposition to the blade no anal tubercles. of the movable finger (Fig. 6i). Some of the Color variable, with red and brown chroma- teeth appear to be blunted or broken. tophores. It may well be that this last specimen repre­ DISCUSSION: In addition to the type speci­ sents another species. However, inasmuch as men, this series is represented in the collec­ all other characteristics are well within the tions by four other specimens: range of expected variation, it must be as­ 1. An ovigerous female, 27 mm. long, col­ sumed, until other specimens are studied, that lected off Waikiki Reef, Oahu, from a coral this is an extreme variation from the normal. head in water about 20 feet deep. Another reason for this conclusion is the fact 2. A male, 21 mm. long, collected in Hana- that this specimen and specimen 4 were col­ uma Bay, Oahu, from a coral head in about lected from the same habitat and are of differ­ 10-15 feet of water (chelae lacking). ent sexes. It is possible that these two speci­ mens were mates living in the same burrow, 3. An ovigerous female, 20 mm. long, col­ as is common among the crangonids. lected offshore at Halape, Kau Coast, Hawaii, from a head of Pontes in 8-9 feet of water by The species is named in honor of Robert H. R. W. Hiatt. The eggs borne by this specimen Hiatt, the collector of some of the specimens were near hatching and contained larvae in the of this species and of other specimens valu­ protozoal stage of development; the eggs able to the study. were 0.25 mm. by 0.72 mm. in diameter. RELATIONSHIP: There is only one other spe­ 4. A male, 17.2 mm. long, collected at the cies in this genus, N. euryone (de Man). If the same place as specimen 3 by R. W. Hiatt. specimens discussed above all belong to N. The four specimens agree well with the type hiatti, then apparent differences between this in almost all characteristics except for minor species and N. euryone in the chelae, second variations in proportions, e.g., the chela varies legs, etc., are within the range of variation of from a little over 3 to a little less than 4 times the species. However, there are four sets of as long as broad, and the length-breadth rela­ characteristics that exceed the range of vari­ tionship of the telson varies by 10-15 per cent. ations noted and which will serve for the These differences are well within the range of separation of the two species. In N. hiatti normal and expected variation. the rostrum reaches beyond the end of the Specimen 4 shows greater differences from first antennular article; the stylocerite reaches the type than do 1 and 2. In 4 the anterior re­ to the end of the second antennular article; gion of the carapace, while bearing the same the merus of the third legs is 4.5 times as proportions as that of the type in respect to long as broad, 2.3 times as long as the car­ the length tf the rostrum, etc., is more closely pus, and 4.5 to 5.5 times as long as the dac- pressed to the bases of the antennae; in con­ tylus; the telson is 13 times as long as the sequence, the eyes are almost entirely con­ terminal triangle. On the other hand, in N. cealed in lateral view, so that at first glance the euryone the rostrum does not reach past the rostrum appears to be longer than that of the middle third of the first antennular article; the type (Fig. 6c). The chelae also show a series stylocerite reaches only to the middle of the of finer teeth and generally longer and thinner second antennular article; the merus of the proportions (Fig. 6g-h). third legs is 7.5 times as long as broad, 1.6 However, it is specimen 3 that shows the times as long as the carpus,, and only 3.0 times greatest differences. In this specimen the chela as long as the dactylus; the telson is 10 times (only one is present) bears few and conspicu­ as long as the terminal triangle. Hawaiian Crangonidae — BANNER 25

Neoalpheopsis euryone (de Man) pendage is flexed; inner and outer inferior Fig. 7 a-g distal corners of merus project slightly as rounded teeth. Chela proper 5.4 mm. long, Alpheopsis ? euryone de Man, Ned. Dierk. Ver., 1.3 mm. broad (3 times as long as broad), 1.1 Tijdschr. (2), 11(4): 308, 1910. mm. thick, oval in cross section. Palm smooth Alpheopsis ? euryone de Man, Siboga Exped. without ridges or depressions. Fingers 2.0 mm. 39aJ(2): 184, pi. 5, fig. 19, 1911. [This de­ long; fixed finger with narrow crest along in­ scription is more available, more complete, ner face, with 11 teeth; ridge depressed distad and is provided with plates.] of teeth, then produced as small crest with Rostrum triangular, acute, tip extending poorly defined teeth; apex curved, strong, over 0.7 of length of visible portion of first acute. Dactylus similarly with narrow crest, article of antennular peduncle. Rostral base but knife-like except for small portion in mid­ and rostrum without dorsal keel, not distinct dle with feeble teeth, bidentate crest opposite from gradual curvature of carapace; rostrum depressed area on fixed finger, knife edge with ventral keel. Orbital teeth strongly re­ diminishing to acute, curved tip. Cheliped duced, hence frontal border of carapace lateral glabrous except for several setae near end and to rostrum almost straight in dorsal view; apex sparse row of short stiff bristles along knife of slight teeth rounded. No pterygostomial ridge of dactylus. Palmar and digital adhesive angle. plaques poorly developed. Eyes concealed in dorsal view, partially vis­ Second legs 6.8 mm. in total length. Carpal ible in lateral view. articles with ratio 10 : 1.9 : 1.7 : 1.7 : 3.8. First article of antennular peduncle 0.74 Third legs slender and elongate, 7.8 mm. mm. long; second 0.44 mm. long and as broad long. Ischium 1.5 mm. long, with 2 movable as long; the third 0.37 mm. long. Stylocerite spines on inferior surface. Merus 2.2 mm. broad, tip acute, reaching more than 0.7 length long, 0.29 mm. broad (7.6 times as long as of second article. Inner flagellum 8.5 mm. broad), with 2 movable spines on inferior long; outer 7.3 mm. long, with secondary fla­ edge. Carpus 1.4 mm. long, without spines or gellum reaching over 1 mm. from base of fla­ teeth. Propodus 2.0 mm. long, with 4 mov­ gellum. able spines. Dactylus long and slightly curved, Lateral angle of basicerite produced and 0.7 mm. long, 0.09 mm. broad, with 2 tufts truncate [however, as in the specimen avail­ of bristles on convex surface. able, it ends in such abrupt truncation on both Articulated pleura of sixth abdominal seg­ sides that it may have had a strong and acute ment well developed, triangular, with free cor­ tip broken off.] Scaphocerite broad, flat, al­ ner acute. most as long as antennular peduncle; lateral Telson 2.95 mm. long, 1.17 mm. broad at spine shorter than rounded portion. Carpoc- base, 0.48 mm. broad at tip. Anterior portion erite of antenna extending only to middle of strongly convex dorsally, but posteriorly al­ last article of antennular peduncle; flagellum most flat; lateral margins almost straight. Pos­ 15.3 mm. long, extending to fifth abdominal terior margin produced as large triangle, 0.25 segment. mm. broad at base and 0.29 mm. long, with Third maxilliped lacking. straight sides and acute tip. Two pairs of ter­ Chelipeds quite symmetrical in develop­ minal bristles, inner slightly shorter than tri­ ment, over 11 mm. long from base, capable of angle (0.28 mm. long), outer about 0.3 as long being folded back against merus. Ischium as inner. Triangle with 4 pairs of setose bris­ similar to N. hiatti. Merus 3.4 mm. long, 0.6 tles. First pair of dorsal spinules of telson 0.4 mm. broad, with diagonal longitudinal exca­ distance from base (1.1 mm.); second pair vation accommodating propodus when ap­ about 0.7 of distance from base (1.9 mm.). 26 PACIFIC SCIENCE, Vol. VII, January, 1953

DISCUSSION: De Man's specimen came from SYNALPHEUS Bate the Netherlands Indies; because it was lacking Synalpheus Bate, Challenger Rpts. 24: 572, in chelipeds, he referred it to the genus Alphe- 1888. opsis "with some doubt." Synalpheus Coutiere, Les Alpheidae, p. 334, The specimen here described agrees very 1899 [redefined]. well with de Man's description and plates ex­ Anterior portion of carapace with rostrum cept in a few small details. In his specimen and orbital hoods which completely envelop the rostrum was shorter, only one third the eyes except on anteroventral side. Anterolat­ length of the visible portion of the first anten- eral margin of carapace produced as distinct nular article instead of two thirds. He also states that the carapace on his specimen had pterygostomial angle or tooth. its upper surface finely punctate, whereas Antennules with basal articles preponder­ that on this specimen was smooth. The inner ant, stylocerites large and well formed. Anten- pair of spines on the tip of the telson are nular flagellum with weak bifurcation. longer than the projecting triangle in de Squamous portion of scaphocerite shorter Man's specimen, whereas in this they are than lateral spine. Basicerite bearing one or shorter; however, the ones in this specimen more teeth. appear to be broken. The lateral tooth of the Large chela carried extended, entire and basicerite in his specimen is described as smooth, ovaloid, much larger than small strong and acute, whereas in this it is short chela; dactylus very short, with cylindrical and truncate, but it probably has been broken. process that penetrates into corresponding Finally, the merus of the third leg in the type cavity on fixed finger. Small chela simple, with is 6 times as long as broad instead of 7.5 fingers joining exactly, palm entire, carpus times. frequently elongate. Second pereiopods with carpus composed Most, if not all, of these differences could of 4 or 5 articles. Following legs short and easily be individual variation, or they could compressed. In all Hawaiian species merus be subspecific differences due to geographic unarmed, propodus armed with weak spines, isolation. The only difference that could be of dactylus biunguiculate. No anal tubercles. specific importance is the sculpturing of the Branchial formula: 5 pleurobranchs, 1 ar- carapace in the original specimen, and this throbranch, 2 epipodites. difference, also, is too slight to warrant spe­ cific separation without any other strong The generic description has been altered to differences in nonvarying characteristics, es­ accommodate Synalpheus redactocarpus Banner pecially as only two specimens of the species described below, which has only four instead are known. of the usual five articles in the carpus of the second legs. For a comparison of this species and N. This genus, represented by over 60 species hiatti Banner see the discussion under the in the Indo-Pacific and more than 25 from the latter. coasts of the Americas, is represented in the DISTRIBUTION: De Man's species came Hawaiian Archipelago by only seven species. from the Karkaralong group, south of Min­ As most of the Indo-Pacific forms have been danao (about 5°N, 125°30'E); the specimen collected from off-shore waters up to about described here is a male 20 mm. long collected 100 meters deep but not from the shallow by R. W. Hiatt at Halape, Kau Coast, Ha­ reef flats, it is possible that more extensive waii, from a head of Pontes in the intertidal dredging around the Hawaiian Islands may zone, west end of Keaoi Island. produce more species. However, considering Hawaiian Crangonidae — BANNER 27

FIG. 7. Neoalpheopsis euryone (de Man), a, b, Anterior region, dorsal and lateral aspects; c, large cheliped; d, distal end of large chela; e, second leg;/, third leg; g, telson. (a, b, d, g, scale A; c, scale B; e,f, scale C.) 28 PACIFIC SCIENCE, Vol. VII, January, 1953 that the collections made by the "Albatross" spines of most other species of the group, the with its extensive dredging and that the col­ spines are not even subequal. lections made by diving have produced only "... first segment of the carpus of the sec­ two new species, it is likely that the present ond pair of feet very long . . ."—the first arti­ list comprises almost all the species found in cle of the carpus is about 1.2 times as long as these waters. the sum of the following, when compared to At least one species of this genus, 5. para- the usual 1.0 times as long found in both S. neomeris Coutiere, manifests a remarkable vari­ paraneomeris and S. charon and 1.7 times as ation. If this condition has been correctly long as in S. prolificus, all of which belong to interpreted, it casts doubts upon many species other groups which are defined as having a of this genus that are separated by relatively short carpus. fine distinctions in the proportions of the "... following feet cylindrical . . ."—I do various parts. In fact, the perusal of the de­ not know what is meant here, but at least the scriptions of Coutiere in particular (as for articles of the legs of S. albatrossi are as flat­ example in his 1909 publication) would make tened as were the corresponding articles of one wonder about the validity of some of the other synalpheids examined. described species and varieties, even without such an example as that of S. paraneomeris in "... ventral hook of the dactyl obsolete mind. Coutiere, too, was doubtful of the va­ . . ."—the ventral hook of the dactylus is no lidity of some of the species he described and more obsolete in S. albatrossi than is the sup­ suggested that some of them be raised in posedly well-developed hook of 5. towsendi "'pure cultures" to determine the amount of brevispinis Coutiere of the Paulsoni Group (op. variation found in a homogenous group. Cer­ cit., p. 35). tainly many of the described species have been "... telson with an oval median lobe ..." established on criteria so fine that almost no —the median lobe of the telson is no more variation could exist without destroying the oval than any other Hawaiian species of this validity of the species. genus. Coutiere (1909) divided the genus into six A similar critique could be applied, I be­ groups (not subgenera, but groups which are lieve, to most of the other characteristics listed without standing under the International in the key to the groups given by Coutiere. Rules of Zoological Nomenclature). With Even Coutiere did not seem to distinguish only the small group of species studied for between the groups, let alone the species, for this paper it is impossible to decide how ade­ a specimen from Hawaii (fragmentary, it is quate and helpful these groups are. However, true) in the United States National Museum it appears that there are so many exceptions which he determined as S. townsendi in the to the definitions as they were drawn up that Paulsoni Group appears to me to be well the groups are of scant value. For example, S. within the range of variation found in S. para­ albatrossi Coutiere, which was placed by that neomeris of the Neomeris Group. In other author in the Comatularum Group, violates words, the groups that were established to the following parts of his definition of the show relationships and to be of aid to the Group: classification of the genus appear to show "Supraorbital spines insignificant compared only dubious relationships and are of almost to the rostrum . . ."—the supraorbital spines no aid in classification. As a consequence, in and the rostrum are of equal size. this paper the groups are not accepted. "... spines of the basicerite almost equal If later workers wish to use these groups, . . ."—even in S. albatrossi, where the spines the Hawaiian species may be divided as fol­ approach equality more closely than do the lows: Hawaiian Crangonidae — BANNER 29

Comatularum Group Palm of large chela inflated, with diam­ S. albatross/ Coutiere eter at fingers 0.5 maximum diameter; S. biunguiculatus (Stimpson) fingers less than 0.25 length of palm Neomeris Group S. macromanus (p. 45) S. charon (Heller) S. prolificus (Bate) Synalpheus redactocarpus sp. nov. S. paraneomeris Coutiere Fig. 8 a-i S. macromanus Edmondson TYPE SPECIMEN: A male 5.1 mm. long, col­ For the other species described in this paper, lected at Halape, Kau Coast, Hawaii, from S. redactocarpus Banner, a new group should coral heads in the intertidal zone on Keaoi be established.With the shifting of S. biungui­ Island by R. W. Hiatt (U. S. N. M. 93454). culatus on the basis of its new description DESCRIPTION: Rostrum narrow and acute, below from the Biunguiculatus Group to the reaching to near middle of second antennular Comatularum Group, it would be advisable article; in lateral view tip slightly depressed to change the name of the former group. but higher than orbital teeth. Orbital hoods slightly inflated; orbital teeth acute, 0.7 as HAWAIIAN SPECIES OF SYNALPHEUS long as rostrum. 1. Carpus of second legs with 4 articles . . Antennular peduncle short and heavy with S. redactocarpus (p. 29) visible portion of first article subequal to both Carpus of second legs with 5 articles. . 2 second and third. Second article slightly 2. Inferior hook of dactylus of third legs broader than long. Flagella also short and reduced in size, shorter and smaller in heavy, outer slightly longer than peduncle, diameter than superior hook 3 inner twice length of outer. Stylocerite reach­ Inferior hook of dactylus of third legs ing nearly to middle of second antennular larger than superior hook in basal di­ article. ameter and usually longer 4 Lateral spine of basicerite feebly developed, 3. Stylocerite reaching only to end of first reaching only to base of spine of stylocerite; antennular article. . .S. albatrossi (p. 30) basicerite otherwise inermous. Scaphocerite Stylocerite reaching to end of second with lateral spine heavy and broad at base, tip antennular article exceeding antennular peduncle and carpocer- S. biunguiculatus (p. 32) ite; squamous portion broad, about 0.9 length 4. Inferior hook of dactylus of third legs 3 of lateral spine. Carpocerite as long as anten­ to 5 times diameter of distal hook, nular peduncle. Antennal flagella lost. broad and blunt at tip 5 Inferior hook of dactylus of third legs Distal article of third maxilliped rounded twice diameter of distal hook, acute in section, heavily spinose. at tip 6 Large chela showing compression and some 5. First article of carpus of second legs ap­ torsion, without grooves, ridges, or other proximately as long as rest combined sculpture. Chela almost 3 times as long as S. charon (p. 37) broad, with arcuate finger occupying distal First article of carpus of second legs 1.7 0.23. Carpus short, broad, expanded. Merus times length of rest combined without spines but with superior distal angle S. prolificus (p. 38) produced into small tooth; article 2.2 times as 6. Palm of large chela tapering only slight­ long as broad, less than 0.3 length of chela. ly to fingers; fingers equal to 0.3 Small chela less than 2.5 times as long as length of palm. .S. paraneomeris (p. 40) broad with fingers subequal to palm. Fingers 30 PACIFIC SCIENCE, Vol. VII, January, 1953 with strong tufts of setae. Carpus as large as found in Synalpheus. The important branchial propodus, expanded, on some sides almost formula is the same as that of other members forming collar for base of palm; article twice of the genus. as broad as long. Merus without spines, twice The unique four-jointed carpus will separ­ as long as broad, and equal in length to chela. ate this species from all other members of the Second legs very heavy and short, 2.3 mm. genus, however. Other genera of crangonids long, merus 0.16 mm. in maximum width. have fewer articles than the usual five, e.g., Ischium 0.48 mm. long, 0.13 mm. broad. Pterocaris Heller, Ogyris Stimpson, Arete Stimp- Merus 0.71 mm. long, 0.16 mm. broad. Only son, and Alpheopsis Coutiere. Moreover, in 4 secondary carpal articles with ratio 10 : 3.6 Alpheopsis there are species with both three and : 2.8 : 7.6; fourth secondary article 1.5 times five articles to the carpus. However, the other as long as broad. Chela 0.6 as long as carpus, characteristics of this species do not fit these slightly broader than carpus; fingers 0.5 genera. length of chela. To my knowledge the lack of dorsal spi­ Third legs also relatively heavy, 2.1 mm. nules on the telson is a characteristic that is long. Ischium 0.33 mm. long, inermous. Me­ unique among the Pacific synalpheids, al­ rus 0.76 mm. long, 0.22 mm. broad, unarmed. though Coutiere states (1899: 316) "Quelques Carpus 0.29 mm. long, with superior distal especes [of the genus Synalpheus] . . . montrent margin produced as rounded tooth. Propodus une reduction tres grande des epines, ou slightly curved, 0.77 mm. long, 0.14 mm. meme leur disparition totale." broad, bearing 4 feeble spines on inferior This species does not fit the subgeneric margin. Dactylus 0.20 mm. long, 0.07 mm. groups created by Coutiere very satisfactorily. broad at base, biunguiculate, with superior or It apparently belongs to either the Neomeris distal claw not as broad at base as inferior, or Paulsoni group. From the former it differs about 1.5 times as long as inferior; superior in that the two hooks of the third and fourth claw more acute than inferior. Fourth legs dactyli are subequal; from the latter in that similar. the rostrum has no ventral prolongation that Fifth legs with "brush" feebly developed, "embraces the ocellary beak" (Coutiere 1909: consisting of but 3 tufts of 3 bristles each. 4). Should other species with similar develop­ Dactylus with superior unguis thinner than ment of the carpus be described, they should on preceding legs. be included with this species in a special Telson 1.2 times as long as broad at base, group; however, the creation of a new group with tip 0.5 as broad as base. Lateral margins does not seem practical for this single species. convex distally; distal margin strongly arcuate Of the Hawaiian specimens studied, S. re- and bearing 4 plumose setae. Large pair of dactocarpus appears to be most closely related terminal spines approximately 0.5 as long as to S. biunguiculatus (Stimpson), but it differs breadth of tip of telson. Dorsal surface slight­ from this species in the length of the spines of ly convex, without any trace of usual 2 pairs the stylocerite and basicerite, in the large and of dorsal spinules. small chela, and in the dactyli of the third to DISCUSSION: This species, in spite of the fifth legs. four-jointed carpus of the second legs, plainly Synalpheus albatrossi Coutiere belongs to the genus Synalpheus. The shape Fig. 9 a-i and development of the anterior region of the carapace, including the pterygostomial angle, Synalpheus albatrossiCoutiere, U. S. Natl. Mus., are similar to other species of the genus; ex­ Proc. 36: 89, fig. 54a-k, 1909- cept for the second legs, the appendages in DESCRIPTION: Rostrum acute, narrow, pro­ their development parallel the development jecting to 0.3 of second article of antennular Hawaiian Crangonidae — BANNER 31

FIG. 8. Synalpheus redactocarpus sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, dactylus flat; e, small cheliped;/, second leg; g, fourth leg; h, fourth leg, dactylus; i, telson. {a, b, e-g, i, scale A; c, d, scale B; h, scale C.) 32 PACIFIC SCIENCE, Vol. VII, January, 1953 peduncles. Orbital teeth acute, elongate, be­ ous bristles; inner pair of lateral spines almost tween 0.4 and 0.5 as long as rostrum. Margin as long as width of tip. Dorsal spinules lo­ between orbital teeth and rostrum U-shaped. cated near margins, at 0.5 and 0.75 length of Antennule with visible portion of basal ar­ telson posterior to articulation. ticle 1.2 times length of second and 1.3 times Colors in life not observed. length of third. Second article about as long DISCUSSION: The two specimens studied as broad. Distal ends of articles without con­ were both males 8 mm. long. These speci­ spicuous setae. Stylocerite reaching end of mens agree well with Coutiere's description first antennular article. and plates except for two characteristics: First, Basicerite with both upper and lower outer the rostrum is described as equal in length to spines developed and acute, lower consider­ the orbital spines and shorter than the first ar­ ably longer than upper, reaching beyond base ticle of the antennular peduncle, whereas in of spine of stylocerite. Outer spine of scaph- these specimens it is longer than both. Sec­ ocerite acute, reaching to middle or near end ond, the figures of the type specimen show of last article of antennular peduncle; squa­ the articles of the antennular peduncle to be mous portion relatively broad and well devel­ relatively more broad than the articles in these oped. Carpocerite reaching past base of an­ specimens. However, as these characteristics tennular peduncle. were found to be very variable in S. paraneo- Large chela only slightly compressed, mar­ meris Coutiere, and as the specimens agree on gins rounded. Palm twice as long as wide, other points, they are referred to this species. with strong tooth above articulation of dac- For a discussion of the points of difference tylus. Dactylus 0.3 as long as palm, strongly between this species and the related S. biungui- arcuate. Carpus of cheliped short. Merus culatus (Stimpson) see the discussion under slightly more than 0.4 as long as chela, twice the latter. as long as broad; superior distal margin ex­ tended as slight rounded projection. Small DISTRIBUTION: This species has been col­ chela with fingers 0.5 as long as palm, narrow, lected only twice, once by the "Albatross" acute; palm slightly over twice as long as off Laysan in 10-18 fathoms (the type local­ broad. Carpus somewhat smaller than chela; ity), and by myself (the two specimens de­ merus almost as long as chela, 3 times as long scribed here) from off Waikiki Reef, Oahu, as broad. from coral in 20 feet of water. Carpal articles of second legs with ratio 10 : 1.3 : 1.3 : 1.5 : 3.0. Synalpheus biunguiculatus (Stimpson) Third legs with ischium unarmed, slightly Fig. 10 a-j over 0.3 as long as merus. Merus also un­ Alpheus biunguiculatus Stimpson, Acad. Nat. armed, 4.3 times as long as broad. Carpus Sci. Phila., Proc. 1860: 31. slightly over 0.5 as long as merus, with tooth (Nee) Alpheus biunguiculatus de Man, Arch. f. on superior margin and 2 movable spines on inferior margin. Propodus 1.1 times as Naturgesch. 53: 502, figs. 6, 6a, 1887. long as merus, with 9 single and double spines (Nee) Alpheus biunguiculatus Bate, Challenger along inferior margin. Dactylus 0.2 as long as Rpt. 24: 562, pi. 51, figs. 4, 4o, 4z, 1888. propodus, 3 times as long as wide, inferior (Nee) Synalpheus biunguiculatus Coutiere, Soc. claw much smaller in diameter at base and Ent. de France, Bui. 11: 232, figs. 1-4, much shorter than superior or distal claw. 1898b. Telson less than 1.1 times as long as broad (Nee) Synalpheus biunguiculatus Coutiere, Fau­ at base, tip 0.3 breadth of base. Tip strongly na and Geog. Maid, and Laccad. 2(4): 873, arcuate between lateral spines, with 8 setifer- pi. 71, fig. 8, 1905. Hawaiian Crangonidae — BANNER 33

FIG. 9- Synalpheus albatrossi Coutiere. a, Anterior region, dorsal aspect (note: specimen lacking bilateral sym­ metry in the scaphocerite); b, anterior region, lateral aspect; c, large cheliped, lateral aspect; d, large chela, dactylus flat; e, small chela; /, second leg; g, third leg; h, third leg, dactylus; /', telson. {a, b, i, scale A; c, d, scale B; e-g, scale C; h, scale D.) 34 PACIFIC SCIENCE, Vol. VII, January, 1953

NEOTYPE: A female 10.5 mm. long, col­ as long as merus, with terminal rounded tooth lected from coral in about 10 feet of water in on superior margin, with terminal movable Hanauma Bay, Oahu (U. S. N. M. 93453). spine on inferior margin. Propodus as long as DESCRIPTION: Rostrum acute, narrow, merus, 7 times as long as broad, with 6 mov­ about 1.5 times as long as broad at base, tip able spines. Dactylus relatively long and thin reaching slightly beyond end of first anten- for genus, almost 4 times as long as wide at nular article; lateral margins concave. Orbital point of maximum breadth, not strongly teeth reaching almost to end of rostrum, con­ curved. Lower unguis only slightly divergent siderably broader than rostrum at base; with from axis of dactylus, conical, acute, marked­ concave lateral margins. Dorsal surface of ly smaller than superior unguis, length only carapace and orbital hoods uniformly rounded; 0.3 and basal diameter 0.7 that of superior both rostrum and orbital teeth somewhat de­ unguis. Superior unguis acute and lying on pressed toward tip, in lateral view. axis of basal portion of dactylus; about 0.3 as Antennular peduncle relatively broad and long as entire dactylus. Fourth legs similar to heavy, with second and third articles of ap­ third. Fifth legs similar to but thinner than proximately equal size, slightly broader than anterior legs, with no trace of "brush." long; visible portion of first article slightly Telson about 1.3 times as long as broad at longer than others. Outer nagellum with base, tip 0.5 as broad as base. Posterior mar­ thickened basal portion about as long as an­ gins with usual 2 pairs of spines, length of tennular base. Stylocerite strong, lateral spine longer over 0.5 width of posterior margin. reaching end of second antennular article. Margin strongly arcuate between spines. An­ Scaphocerite with lateral margin slightly terior pair of dorsal spinules over 0.3 distance concave, lateral spine strong, tip reaching be­ from articulation, posterior pair slightly over yond ends of antennular peduncle and car- 0.7. pocerite. Squamous portion well developed, DISCUSSION: The name Alpheus (now Synal- reaching end of antennular peduncle. Upper pheus) biunguiculatus was applied by Stimpson teeth of basicerite entirely lacking, lower to a specimen from an unspecified locality in tooth acute but not reaching beyond end of the Hawaiian Islands. He described it as fol­ first antennular article. lows (I860—translated from the Latin): Terminal article of third maxillipeds round­ Closely related to Alpheus neptuna, yet having front ed, with usual strong spines. short-toothed; spine at the base of the dactylus of large Large chela compressed but with rounded chela missing; posterior foot small, dactylus two- clawed, secondary claw ventral. The merus of the third and entire margins. Palm slightly less than and fourth legs without spines on the lower side. twice as long as high, 0.7 as thick as high. Length .5 inch. Habitat: Hawaiian Islands; in coral. Fingers slightly less than 0.3 length of palm; This description was without plates. Evi­ dactylus strongly arcuate and closing at 90° dently Stimpson planned to publish more to plane of chela. Merus of cheliped 3 times than this preliminary account, as he did for as long as broad, with small terminal tooth on the other groups (the descriptions of the superior distal margin. Small chela less than Brachyura and Anomura of the expedition 0.5 as long as large chela, with palm about were published posthumously in 1907 by the 1.3 times length of fingers, twice as long as Smithsonian Institution). However, nothing broad. further on the Macrura was ever published. Carpal articles of second legs with ratio The type specimen is now lost—whether or 10 : 2.0 : 2.0 : 1.7 : 4.2. not it was destroyed in the fire of the Phila­ Merus of third legs 4.3 times as long as delphia Academy of Science could not be broad, twice length of ischium; both ischium ascertained—so there is' nothing left but the and merus unarmed. Carpus slightly over 0.5 description, which will fit many synalpheids Hawaiian Crangonidae — BANNER 35

FIG. 10. Synalpheus biunguiculatus (Stimpson). a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, dactylus flat; e, small cheliped, lateral aspect;/, second leg; g, third leg; h, third leg, dactylus; /, fifth leg, distal end;/', telson. {a, b, j, scale A; c-e, scale B;/ g, i, scale C; h, scale D.) 36 PACIFIC SCIENCE, Vol. VII, January, 1953 with equal ease. In 1898 and 1905 Coutiere ing only to the end of the first antennular ar­ redescribed the species on the basis of speci­ ticle was short in Stimpson's opinion. More­ mens from the Indian Ocean. Coutiere evi­ over, when more specimens are examined, dently did not establish a neotype. this characteristic will perhaps be found to be Unfortunately the species Coutiere de­ variable, as it is in S. paraneomeris. scribes has not been found in the Hawaiian This species is most closely related to S. al­ Islands. This confirms de Man's doubts about batross! Coutiere of those known from the Ha­ the validity of redescribing a species from waiian Islands. They can be separated by the such a distant locality (1911: 202-203). Cou­ characteristics shown in the following tabu­ tiere's specimens do not, then, meet the one lation: definite characteristic given in the original de­ S. biunguiculatus S. albatrossi scription: "Habitat: Hawaiian Islands.". Rostral base narrower than Rostral base broader Moreover, the original description states that base of orbital teeth than base of orbital the tooth at the base of the dactylus of the teeth Third antennular article as Third antennular article large chela is missing, but Coutiere's plates long as second antennu­ shorter than second clearly show his specimens to have this tooth lar article antennular article or projection. Finally, the statement in the Stylocerite reaching to end Stylocerite reaching of second antennular ar­ slightly beyond end of original description that the dactyli of the ticle first antennular article third and fourth legs had their "secondary Basicerite without upper Basicerite with acute up­ claws ventral" would imply that the ventral spine, lower spine reach­ per spine, lower spine ing end of first antennu­ 0.5 length of first an­ claw is not as heavy as the distal or superior lar article tennular article claw, a condition contrary to that found in Scaphocerite longer than Scaphocerite shorter than Coutiere's specimens. For these reasons Cou­ antennular peduncle antennular peduncle Large chela without tooth Large chela with tooth tiere's description of Synalpheus biunguiculatus over dactylar articulation over dactylar articula­ must be discarded and a neotype selected or on superior margin of tion and small tooth from the fauna of the Hawaiian Islands. merus on distal end of su­ perior margin of merus The name biunguiculatus has priority over Dactylus of third legs with Dactylus of third legs all names of synalpheids from the Hawaiian lower unguis almost par­ with upper and lower area, and Stimpson's description can be inter­ allel to upper unguis ungui divergent Telson with posterior mar­ Telson with posterior preted to apply to any local species with gin strongly convex margin "very little greater or less precision. For example, it could convex" (Coutiere, be applied to S. paraneomeris Coutiere with 1909) ease. However, if it were applied to that spe­ In addition, there are differences in propor­ cies, the entire literature would be thrown into tions of several of the appendages, especially an even greater snarl. To avoid such further in the merus of the large cheliped and of the confusion, the name has been applied to this, small chela. Of these characteristics, probably the third most common synalpheid in these the most important are the relative lengths of waters, a species that has not been named pre­ the stylocerites and the presence or absence of viously. It agrees with the few characteristics the upper spine of the basicerite. It will be given in the original description, especially noted that the characteristics of S. albatrossi in with the portions discussed above. The only the tabulation are derived from Coutiere's difference is in the indefinite statement of the original description, and that the first, second, original description: "... having front short- and eighth points do not apply well to the toothed . . .". It would appear that this de­ specimens placed in S. albatrossi in this paper. scription possibly was of a different species, It is necessary to assign a new name to Cou­ but one cannot tell what Stimpson meant by tiere's species. I propose, therefore, Synalpheus "short." Perhaps, having the rostrum reach­ coutierei nom. nov. to replace Synalpheus biun- Hawaiian Crangonidae — BANNER 37 guiculatusCoutiere (1898^: 232, figs. 1-4), non ued into strong spine, almost equal in length S. biunguiculatus (Stimpson) (I860: 31). All to stylocerite. previous records under the names Alpheus, or Large chela slightly compressed and twist­ Synalpheus, biunguiculatus other than the orig­ ed, with margins rounded and entire; shape inal description should be attributed to S. cou- somewhat inflated, 2.5 times as long as broad. tierei, with the possible exception of the record Superolateral edge of face of chela behind ar­ made without description by Lenz in 1901 ticulation of dactylus projecting over articu­ wherein he reported the species from Laysan. lation as small rounded tubercle. Dactylus However, it is impossible to know exactly strongly arcuate, between 0.25 and 0.3 length upon which species he was reporting. of entire chela. Merus with superior distal margin projecting as acute tooth. Small chela DISTRIBUTION: As stated above, the type likewise rounded and entire, dactylus slightly locality was not designated. The neotype over 0.3 length of entire chela. Greatest diam­ came from the deeper water along the west eter of carpus slightly larger than that of chela side of Hanauma Bay, Oahu; other specimens proper. Merus similar in form to large chela. have been collected on Oahu off Waikiki in 18 feet of water and off Nanakuli in 15 feet of Carpal articles of second legs with ratio water; at Maui off Kalama Park in 8 feet of 10 : 3 : 3 : 3 : 5. water; and by R. W. Hiatt off Hawaii at Ha- Third and fourth legs with merus unarmed lape, Kau Coast, near Keaoi Island, from Poc- and 3 times as long as broad. Carpus pro­ illopora meandrina Verrill in 25-30 feet and longed as rounded lobe on superior distal 40-45 feet of water. margin, armed with spine on inferior distal margin. Propodus curved and armed with 5 Synalpheus charon (Heller) or 6 slender spines. Inferior hook of dactylus Fig. 11 a-k thick and obtuse, with broad spoon-like ex­ cavation on inferior margin. Distal hook, Alpheus charon Heller, K. Akad. Wiss. Wien, when seen in profile, longer than inferior Sitzungsber. 44: 272, 1861. hook, less than 0.3 as broad at base as ventral hook, tapering to acute tooth. All aspects of DESCRIPTION: Rostrum acute, narrow tri­ distal hook except straight profile showing angle with straight margins, tip reaching near­ basal half as broad and flat, paddle-like, with ly to end of first antennular article. Acute tip sudden constriction to narrow acute tip (Figs. of orbital hoods reaching to near tip of ros­ 11^, /, /; the last is in such rotation that the trum; outer margins of orbital hoods convex, acute tip is somewhat obscured). inner margins slightly concave and separated from base of rostrum by shallow depression. Telson about as long as broad at base, with Antennular peduncle with visible portion tip 0.5 width of base. Lateral margins anteri­ of first article longest, third slightly shorter orly concave, then posteriorly convex; tip than first, second correspondingly shorter arcuate. than third; each article about as broad as long. The largest specimen in the collection is a Stylocerite well developed, with tip of lateral female 22 mm. long; most specimens are spine reaching almost to end of second an­ much smaller. The color is often a brilliant tennular article. orange red. Lateral spine of scaphocerite slightly longer DISCUSSION: Only slight variation was than squamous portion, approximately as noted in the specimens of this species in the long as antennular peduncle, and somewhat collections at hand, possibly in part because shorter than carpocerite. Superior angle of of the small number of specimens available. basicerite rounded, but lateral angle contin­ Only one specimen showed a notable varia- 38 PACIFIC SCIENCE, Vol. VII, January, 1953 tion from the specimens described; in this of Maui in 13 fathoms (Stations 3962, 3955, specimen the merus of the third legs was only 4073). 2.5 times as long as broad and the basal por­ S. charon has been reported from many tion of the superor unguis was only slightly places between the Red Sea and Hawaii, in­ broadened. This specimen had all the other cluding several localities in the Indian Ocean, characteristics of the species, including the South China and southern Japan, Australia, excavated ventral hook of the dactylus, so it and Thursday Island in the Pacific. was assigned to this species. Synalpheus prolificus (Bate) These Hawaiian specimens agree well with Heller's original description and with de Fig. 12 a-c Man's redescription of the same species from Alpheus prolificus Bate, Challenger Rpt. 24: the East Indies (de Man, 1911). Heller did 556-557, pi. 99, figs. 4, 4c, 4m, 1888. not remark upon the peculiar superior unguis As no specimen of this species was avail­ of the dactylus of the third legs, but it was able the original description is given: described fully by de Man. All differences Carapace two-thirds the length of the pleon. The noted were minor ones of proportions, such rostrum narrow. The orbital lobes pointed, broad and as the length of the stylocerite, which was nearly equal in length to the rostrum. found to be quite variable in the closely re­ First pair of antennae have the second joint of the peduncle shorter than the first, the third subequal with lated S. paraneomeris. the second. Stylocerite a little longer than the first joint, This species is very similar to S. paraneo­ nagella subequal and about as long as the carapace. Second pair of antennae as long as the animal, pe­ meris Coutiere in general body form and pro­ duncle longer than that of the first pair, basal tooth as portion, especially when the variation of S. long as the stylocerite and nearly half as long as the paraneomeris is considered. However, the na­ scaphocerite, which is subequal with the peduncle. First pair of pereiopoda lost. Second slender, having ture of the dactyli of the third and fourth legs the carpos five-articulate, the first articlus being longer affords a rapid and certain method of distin­ than the four following ones. Third and fourth pairs guishing the two. To separate this species robust, meros broad, compressed, without distal tooth; carpos anteriorly produced on the upper distal margin; from S. prolificus (Bate) see the discussion propods curved, furnished with spines on the posterior under that species. Ramadan (1938) considers margin; dactylos stout, bluntly pointed, with a small S. helleri de Man (1911: 194, 245) to be prob­ secondary unguis on the outer surface (fig. 4m). The pleopods are broad and foliaceous. ably a synonym of S. charon, but reaches no Telson broad, smooth, furnished with two spinules definite conclusion. on the dorsal lateral surface, tapering and rounded at the extremity. DISTRIBUTION: This species appears to oc­ Length, entire 20 mm. (0.8 in.) of carapace 8 cur almost exclusively in living heads of Pocil- Depth of carapace 5.5" lopora meandrina Verrill, which it inhabits Length of pleon 12 with Crangon ventrosa (Milne-Edwards) and Habitat—Off Honolulu, Sandwich Island; depth 18 fathoms. One specimen; female. crabs of the genus Trapezia. It therefore could Observations—This species is one of those that ap­ be expected from all reefs and other areas proach Alpheus neptunus Dana (PI. CI, fig. 2) and Al­ where this coral grows. It has been collected pheus biunguiculatus Stimpson. It differs from the former at Mokulua, Waimanalo, Hanauma Bay, in having the carpos of the third and fourth pairs of legs anteriorly produced at the upper distal extremity Black Point, Waikiki, Nanakuli, Mokuleia, and in having the dactylos biunguiculate, and from the and Kawela Bay on Oahu, and at Lahaina on latter (PI. CI, fig. 4) in being generally more robust, in Maui, in water ranging from a few feet to having the rostrum and supraorbital teeth less prom­ inent, the carpos of the third pair of pereiopoda pro­ about 20 feet deep. As reported by Coutiere duced on the upper margin, and the dactylos short and (1909), the "Albatross" collected the species thick, the second unguis being rudimentary and situ­ at two stations offLaysan in water from 10 to ated behind the larger on the cfuter margin. Our specimen is a female and carries a very large 30 fathoms deep and off the northeast coast mass of ova; hence the name. Hawaiian Crangonidae — BANNER 39

FIG. 11. Synalpheus charon (Heller), a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, dactylus flat; e, small chela;/, second leg; g, third leg; h, i,j, dactylus of third leg, anterior, posterior, and posteroventral sides; k, telson. (a, b, scale A; c, d, scale B; e-g, scale C; h-j, scale D; k, scale E.) 40 PACIFIC SCIENCE, Vol. VII, January, 1953

DISCUSSION: This species appears to be most closely related to S. charon (Heller) of all the Hawaiian species. However, to judge from Bate's description and figures, the two spe­ cies can be separated by a series of character­ istics. Probably the most important are in the carpus of the second leg, which has the first article 1.7 times as long as the following four together in S. prolificus (according to the fig­ ure) and only equal to the sum of the follow­ ing articles in S. charon, and in the biunguicu- late dactylus of the third legs, where the base of the superior claw is one sixth and the base of the inferior claw five sixths the diameter of the dactylus in Bate's figures instead of one fourth and three fourths as in S. charon. Final­ ly, in the antennules the visible portion of the first antennular article is longer than the second and third, which are subequal, and the tip of the stylocerite reaches only slightly be­ yond the end of the first article in S. prolificus, whereas in S. charon the first and second an­ tennular articles are subequal, the third is FIG. 12. Synalpheusprolificus (Bate), a, Lateral aspect shorter, and the tip of the stylocerite reaches (slightly over 3X); b, anterior region, dorsal aspect; c, third leg, propodus and dactylus. (Redrawn from Bate, to the middle of the second antennular article. Challenger Rpts., Vol. 24; Plates, 1888, pi. 99, fig. 4.) Bate makes no mention of, nor shows any detail of, the dactylus of the third leg. If it species is probably restricted to the deeper does show the concavity of the inferior ungis zones and does not reach the reefs; it may also and the basal expansion of the superior unguis indicate that the species is endemic to the Ha­ characteristic of S. charon (and such a small waiian Islands, but insufficient sampling of character could easily have been overlooked), deeper water in other areas would make such the two species are certainly most closely re­ a conclusion questionable. lated. However, the nature of the second legs Synalpheus paraneomeris Coutiere and of the stylocerite would be valid charac­ Figs. 13 a-l, 14 a-b teristics for the separation of the two species even if the dactyli were similar. Synalpheus paraneomeris Coutiere, Fauna and Geog. Maid, and Laccad. 2: 872, 1905. De Man (1911) suggests that S. prolificus Synalpheus townsendi Coutiere (partim), U. S. may be a synonym of S. biunguiculatus (Stimp- Natl. Mus., Proc. 36: 35, 1909- son); the redescription of S. biunguiculatus in this paper removes that possibility. DESCRIPTION: Rostrum narrow and acute, There are no specimens of S. prolificus in the not quite reaching to end of first antennular present collection; it has not been reported in article. Orbital hoods produced into teeth the literature since its original description ex­ that reach almost to end of rostrum. Depres­ cept for two specimens recorded by Ortmann sion between rostrum and orbital hoods not which were later re-examined and placed in extensive, but lateral margins of rostral base S. gravieri Coutiere (de Man, 1911). This lack so abrupt that rostrum appears separated from of distributional records indicates that the anterior carapace in dorsal view. Hawaiian Crangonidae — BANNER 41

Antennular peduncles broad and heavy; 1. The length of the rostrum and of the orbital second article slightly broader than long. Sty- teeth. The rostrum itself varied from two to locerite reaching almost to end °f second an­ three times as long as broad, and its tip reached tennular article. Scaphocerite slightly longer variously from two thirds the length of the than antennular peduncle, not reaching to end visible portion of the first antennular article to of carpocerite. Squamous portion of scapho­ beyond the end of the same article. In some cerite rounded anteriorly, comparatively cases the rostral tip was depressed when seen broad. External tooth of basicerite only slight­ in side view, with the curve paralleling that of ly shorter than stylocerite; upper angle not the orbital teeth; in others the rostrum con­ produced into tooth. tinued straight in the dorsal line of the cara­ Large chela rounded in outline, only slight­ pace while the orbital teeth alone were de­ ly compressed, without tubercle at base of pressed anteriorly. In most specimens the or­ articulation of dactylus. Length of fingers bital teeth were of the same length as the ros­ somewhat more than 0.3 length of palm; chela trum; in some they were somewhat shorter, more than 3 times as long as broad. Merus and in a few they were only two thirds the curved, without spines, with superior margin length of the rostrum, a condition that was obtuse distally. approached by the specimen Coutiere de­ Small chela 0.3 length of large chela. Car­ scribed. pus greater in diameter than propodus. Merus 2. The antennules. These were depicted by without spines but produced on upper distal Coutiere as being variable in the proportions margin as broad dentate process. of the articles. The local specimens were found Carpal articles of second legs with ratio to parallel this described variability. In addi­ 10 • 2.5 : 2.5 : 2.5 : 5.5. tion, the stylocerite also varied in length, but Merus of third legs broad, over 3 times as its tip always reached beyond the end of the long as wide, inermous. Carpus with movable first antennular article. spine on distal inferior margin, with superior 3. The spine of the basicerite. When the tip of margin produced distally as sharp tooth. Pro­ this spine was compared to the tip of the sty­ podus with 6 spines along inferior margin. locerite, the two were usually found to reach Dactylus biunguiculate, with inferior tooth to the same level, but in a few cases the tooth solid, conical, and with slightly protuberant of the basicerite was much shorter. tip; distal or superior tooth longer than in­ 4. The breadth of the squamous portion of the ferior tooth, 0.5 as broad at base as inferior scaphocerite. As shown by Coutiere, the squa­ tooth. mous portion was two thirds the length of the Length of largest specimens 15 mm.; usual­ lateral spine. In these specimens it varied from ly 10 mm. or less. Color typically olive brown a narrow blade two thirds the length of the to gray, but those living in heads of Pocillopora lateral spine to a broad squamous portion five meandrina reddish. sixths the length of the spine. DISCUSSION: According to Coutiere's de­ 5. The large cheliped. The variation in the scription, this species manifests a rather wide proportions of the large chela were like those variation in the size and shape of the rostral noted by Coutiere, but, in addition, the shape front, in the antennules and antennae with of the superior distal end of the merus was also their spines, and in the proportions of the found to be variable. In about half the speci­ chelae. An examination of almost 100 speci­ mens it was rounded, as specified and shown mens available showed all of these and other by Coutiere; in the rest of the specimens this variations. The more important variations ob­ corner was produced into a somewhat variable served were in the following characteristics: but small, acute, triangular tooth. 42 PACIFIC SCIENCE, Vol. VII, January, 1953

6. The carpal articles of the second legs. These An attempt was made to determine if any were described as having the first article slight­ of these variable characteristics would be con­ ly longer than the sum of the remaining four; stant and discrete enough to separate the however, this ratio was found to vary so that population into either distinct species or va­ in some the first article was shorter than the rieties. However, all the characteristics seemed sum of the distal articles (as is the case for the to vary within the limits given above. For ex­ specimen described above); on the other ample, although most of the specimens seemed hand, the longest article measured was 1.3 to be either the form with a long rostrum or times the length of the distal articles. the form with the short depressed rostrum, some individuals were intermediate between 7. The dactylus of the third and fourth legs. the extremes. Again, careful measurements This was almost always as described and de­ were made on a series of specimens from the picted by Coutiere, but some variation was same location to determine if the length of noted. Instead of having the lower unguis the squamous portion of the scaphocerite markedly larger than the distal unguis, oen would be a good criterion for separation; in specimen with three normal dactyli had a dac­ the specimens examined the shortest squa­ tylus on the fourth legs which had the inferior mous portion was slightly less than two thirds hook much shorter and only slightly larger the length of the lateral spine, the longest was in basal diameter than the distal tooth (Fig. more than five sixths the length, but many 15k). Others showed conditions approaching were intermediate between the two. It was this extreme on one or more legs. The dactyli concluded that neither these nor the other also varied in the angle between the two characteristics discussed above were constant hooks; in most the spread was 45 degrees or enough to be the basis of separation of any less (when the main axis of the hooks was subspecific forms. projected from the point of meeting); in one specimen the spread was about 60 degrees. However, the species has been separated into five varieties by Coutiere (1905, 1909) The series of measurements necessary to and de Man (1911): S. paraneomeris var. pro- study the variation in proportions of the ap­ latus de Man, S. paraneomeris var. halmaherensis pendages and telson were not made, but a de Man, S. paraneomeris var. praedabundus de simple visual comparison showed these pro­ Man, S. paraneomeris var. prasalini Coutiere, portions to exhibit a range comparable to that and S. paraneomeris var. seychellensis Coutiere. discussed above. There is confusion about the varietal name This variation is a mere extension of the S. paraneomeris var. prolatus, which first ap­ variability already noted by Coutiere, and pears as a name in Coutiere (1909: 9), where these Hawaiian forms agree rather well with he states "S. paraneomeris prolatus, new name the characteristics noted by Coutiere, except ( = S. paraneomeris oxyceros Coutiere)," yet on one point. In his description of the dactyli there are no references in the Zoological Rec­ of the third and fourth legs Coutiere states ord or in Coutiere's bibliographies to an S. that the ventral hooks are, "... courbe, ob­ paraneomeris oxyceros. A possible clue to the tuse, suivie d'une concavite du bord inferieur name appears on page 12 of the 1909 work: parfois assez marquee pour indiquer une S. lockingtoni differs from S. paulsoni almost solely by ebauche de triunguiculation." Yet his draw­ the spine of the scaphocerite being longer in the latter ing of the dactylus shows the ventral hook and surpassing the carpocerite; with the exception of the place of origin, the second species would corre­ far from obtuse and gives no indication of the spond to the "oxyceros" form so often met with that location of a concavity. The Hawaiian speci­ it appears to be almost a constant variation among the mens agree with the figure, not the descrip­ sub-species of a given species." ' [Footnote]* ... in deference to the accepted rule of tion. nomenclature which forbids duplication of names with- Hawaiian Crangonidae — BANNER 43

FIG. 13. Synalpheus paraneomeris Coutiere. a, b, Anterior region, dorsal and lateral aspects; c, d, anterior region, showing variations; e, large cheliped, lateral aspect; /, large chela, dactylus flat; g, small chela; h, second leg; /', third leg;/', third leg, dactylus; k, third leg, aberrant dactylus; /, telson. {a, d, /, scale A; c, scale B; b, e, f, scale C; g-i, scale D;/', k, scale E.) 44 PACIFIC SCIENCE, Vol. VII, January, 1953 in a single genus, I have in this paper used different names for oxyceros having a similar meaning, as . . . pro- latus. . . . Later, de Man describes and depicts two specimens similar to what Coutiere meant to be S.paraneomerisprolatus (or oxyceros). There­ fore, it appears to me that if the International Rules were to be applied to these unrecog­ nized names (which are varietal names, not names of subspecies), Coutiere merely created a nomen nudum and de Man is the true author of the variety. Only a small number of specimens of each of these varieties was examined: Coutiere's original description was based on 14 speci­ mens—one variety upon four specimens and the other four varieties upon only two speci­ mens each. It is evident that such a small number of specimens is not an adequate basis for the separation of varieties. For example, if, of the 100 specimens examined for this study, 10 or 20 were selected and studied carefully, it is likely that at least three seemingly separate and distinct subspecies or even species could be distinguished; it is only the large number of specimens that permits the range of varia­ tion to be understood. Consequently in my opinion the varieties described merely fall in the line of normal variation of a large popula­ tion and are not of taxonomic worth; for this reason the previously named varieties have not been accepted nor have new varieties been proposed in this paper. FIG. 14. Synalpheus paraneomeris Coutiere. Specimen Coutiere (1909) listed S. townsendi as occur­ from "Albatross" Station 3969, identified by Coutiere ring in the Hawaiian Islands; the record was as S. townsendi Coutiere. a, Anterior region, dorsal as­ based on a single specimen taken by the "Al­ pect; b, telson. batross" on French Frigate Shoal. I have ex­ amined the specimen which is in the United that S. townsendi is known only from the States National Museum. It has none of the Americas, and that it has never been reported thoracic legs, and these evidently (from Cou­ from these islands nor is it represented in the tiere's notes in the vial) were lacking when it present large collection of synalpheids, the was examined by him. Without the important specimen has been assigned to S. paraneomeris. dactyli and chelipeds it is impossible to identi­ DISTRIBUTION: This species is one of the fy the specimen positively. However, in view most common in the Hawaiian Islands. In of the facts that the anterior body regions and addition to the usual habitat in dead coral its appendages and the telson were within the heads, it is found in heads of living Pocillopora range of variation for the local S. paraneomeris, meandrina Verrill. On Oahu it has been col- Hawaiian Crangonidae — BANNER 45 lected from the reef flats and in water to about and the Netherlands Indies, and in the central 20 feet deep at the following localities: Off Pacific from Wake and Palmyra Islands. Mokulua, Waimanalo, Hanauma Bay, Black Point, Waikiki, Nanakuli, Kawela Bay, and Synalpheus macromanus Edmondson Kaaawa. On Maui it has been collected at Fig. 15 a-g Makena and at five places along the coast north and south of Lahaina. R. W. Hiatt took Synalpheus macromanus Edmondson, Bernice some specimens from the intertidal zone at P. Bishop Mus., Bui. 27: 9-11, figs, la-j, Halape, Kau Coast, Hawaii. C. H. Edmond- 1925. son (1925) reported the species from Laysan DESCRIPTION: Rostrum apparently reaching Island, Lisianski Island, and Pearl and Hermes to near end of first antennular article, orbital Reef. Collections by the "Albatross" reported teeth acute, about 0.5 length of rostrum. by Coutiere included specimens from off Antennular peduncle short and heavy with southwest Oahu in 18-69 fathoms, off Laysan second and third articles about 1.7 times as in 19-60 fathoms, and the specimen which he broad as long. Stylocerite reaching to near end referred to as S. townsendi from French Frigate of second article. Lateral spine of basicerite Shoal in 15-16 fathoms. strong, almost as long as stylocerite; dorso­ The species and its varieties have been re­ lateral angle rounded. Lateral spine of scapho- ported from the Red Sea, the Indian Ocean, cerite 1.2 times as long as squamous portion;

FIG. 15. Synalpheus macromanus Edmondson. a, Anterior region, dorsal aspect, left side not shown; b, anten- nules and antennae, lateral aspect; c, large cheliped, lateral aspect; d, large chela, dactylus flat; e, second leg; /, third leg; g, third leg, dactylus. (a, b, g, scale A; c, d, scale B; e, f, scale C.) 46 PACIFIC SCIENCE, Vol. VII, January, 1953

tip reaching beyond end of antennular pe­ of that species would easily encompass these duncle. Carpocerite dorsoventrally com­ differences. However, no specimen of S.para­ pressed, as long as scaphocerite. neomeris has been seen with a large chela ap­ Large chela very large, inflated, almost cir­ proaching the inflated condition found in that cular in section, only twice as long as broad of S. macromanus, and it appears that this is a at maximum diameter, with fingers much valid criterion for the recognition of two spe­ thinner and lower than palm, occupying dis­ cies. Of course, it may be found that the huge tal 0.25 of chela. Carpus of usual form. Merus chela of S. macromanus is merely a monstrous 0.25 length of chela, twice as long as broad, variation of S. paraneomeris, but no specimens with subacute tooth terminating superior dis­ in the collection studied would indicate it; it tal margin. may also be that the gross chela is merely a Small chela of usual configuration, with growth anomaly, for no other specimens have chela proper compressed and fingers occupy­ been collected with a similar chela. At the ing distal 0.3- present, in any case, the species appears to be Carpal articles of second legs with ratio distinct and valid. 10 : 1.7 : 1.7 : 1.9 : 6.5. DISTRIBUTION: The type specimen, a female Ischium of third legs 0.3 length of merus; 13.5 mm. long, is from Lisianski Island. merus about 3.5 times as long as broad, un­ armed; carpus 0.45 length of merus, superior CRANGON Weber distal margin projecting as rounded tooth, in­ Crangon Weber, Nomen. Ent. Sec. Ent. Syst., ferior distal margin not projecting but with 1795. slender spine; propodus 0.8 length of merus, somewhat curved, with 5 movable spines on Alpheus Fabricius, Sup. Ent. Syst., 1798. inferior margin; dactylus 0.25 as long as me­ Carapace developed as orbital hoods which rus, biunguiculate, conical inferior unguis 1.3 completely enclose eyes except on ventral times as broad at base and 0.75 as long as side. Orbital hoods frequently with spiniform curved superior or distal unguis. apex and usually demarked from rostral base DISCUSSION: The description and figures are by more or less pronounced depression. Ros­ of the type specimen previously described and trum often present and carapace usually with figured by Edmondson. In this specimen, the at least slight rostral carina. Pterygostomial only one known, the frontal regions of the margin of carapace rounded. carapace are so distorted as to render accurate Antennules usually short, frequently with determination of the characteristics impos­ basal peduncular article and stylocerite re­ sible. duced. Scaphocerite frequently reduced; ba- This species approaches S. paraneomeris sicerite either armed with inferolateral spioe Coutiere in almost all characteristics. It is pos­ or rounded; carpocerite usually reaching tn sible that the rostrum and orbital hoods are or beyond end of scaphocerite. quite different in the two species, but, as men­ Chela of first legs very asymmetrically de­ tioned, the exact condition of the frontal re­ veloped. Large chela of variable form, from gion of the carapace of S. macromanus cannot subcorneal to compressed and twisted, sur­ be determined. The articles of the antennules faces from smooth and entire to deeply sculp­ may be broader in this species; the stylocerite tured with grooves and pronounced "alphe- may be longer; the relative sizes of the claws opsidean" lobes. Dactylus usually with pis­ of the dactylus of the third legs may be differ­ ton-like process that fits into cavity on fixed ent. As these characteristics are among the finger and always with digital and palmar ad­ most variable in what was interpreted to be hesive plaques. Chela always carried extended S. paraneomeris, I believe the range of variation from body. Large cheliped with hemispherical Hawaiian Crangonidae — BANNER 47 carpus, with merus at least somewhat triangu­ groups were defined, for example, Coutiere lar in section. Small chela of simple form, at (1905) described 23 new species from the times showing marked sexual dimorphism. Maldives and Laccadives, and de Man (1911) Carpus of second legs of five articles of described some 20 new species from the col­ variable proportions. lections of the Siboga Expedition. Since these Third and following legs robust, com­ did not exactly conform to the original de­ pressed; armature and proportions of merus, scriptions of the groups and subgroups, mod­ carpus, and propodus various, dactylus sim­ ifications of the definitions were made. ple or biunguiculate. Fifth legs with "brush" On the basis of the Hawaiian species it ap­ on propodus. pears that these groups, unlike those of Syn­ Abdomen usually without lateral compres­ alpheus, probably are still quite valid in deter­ sion, with pleura in females larger than those mining the relationships of the species. How­ of males and usually without acute projections ever, should some future worker have avail­ on margins. Pleura of sixth abdominal seg­ able to him a collection of a large number of ment not jointed. Telson usually with pos­ species, these groups should certainly be re­ terior margin convex and lateral angles dis­ evaluated and some of the species be assigned tinct but not acute. Anal tubercles well de­ either to others of the present groups or to veloped. new groups; by doing this some of the am­ Branchial formula: 5 pleurobranchs, 1 ar- biguity of the definitions of the groups could throbranch, 8 epipodites, and sometimes a be avoided. If that were done it would be de­ supplementary arthrobranch on first thoracic sirable to raise the groups to subgenera. legs (except in C. paragracilis Coutiere and C. While the groups in general show close re­ nanus Banner). lationships of the species within them, the This genus, previously known to American utility of the groups as redefined by de Man systematists and still known to European sys- is definitely questionable. Some species could tematists as Alpheus, is by far the largest in the fall into two or three groups, as defined, with family, not only in Hawaiian waters, but also equal ease, as the revised definitions often are in the entire Pacific. In 1911 de Man listed equivocal and ambiguous. For example, de 128 species of Crangon in comparison with 62 Man's definition of the Macrochirus Group species of Synalpheus, the next largest genus. (1911: 307; italics mine): In this paper 31 species from Hawaii are dis­ cussed; in addition there are two species re­ Lateral (extracorneal) spines usually present. Inferior spine of basicerite sometimes large. Larger chela usually ported from Hawaii, the records of which are longitudinally grooved or emarginate anteriorly. Merus doubtful. of third legs usually unarmed, dactyli of three posterior legs biunguiculate or simple. External spine of exopod of To facilitate the classification of the genus, caudal fan often black. Coutiere (1899) divided it into five groups, principally on the basis of evolutionary devel­ Because the groups do seem to follow the opment of the large chela and to a lesser natural relationships, they have been used in extent on other characters, such as the charac­ this paper. However, because the definitions ter of the frontal margins of the carapace, the of the groups are not clear and definitive, it character of the dactylus of the third legs, etc. has been impossible for the major divisions Later he divided the third group, Crinitus, into of the key to the species to follow the group­ three subgroups (1905). These groups as orig­ ings. Instead, the key keeps all members of inally defined were sharply distinguished one one group in a single series and in the se­ from another and constituted a valuable aid quence in which they are discussed in the to the separation of the species. text. For ready reference the Hawaiian species Many new species were described after the are listed here by groups. 48 PACIFIC SCIENCE, Vol. VII, January, 1953

Megacheles Group Edwardsi Group | 1. C. hailstonei (Coutiere) var. paucispi- 29. C. leptochirus (Coutiere) nata Banner 30. C. crassimanus (Heller) 2. C. brachymerus Banner 31. C. pacifica (Dana) 3. C. albatrossae Banner With the change of the generic name from 4. C. tuthilli Banner Alpheus, a masculine noun, to Crangon (from 5. C. oahuensis Banner Gr. Kpayyov) a feminine noun, it has been 6. C. collumiana (Stimpson) necessary to change the endings of some of 7. C. deuteropus (Hilgendorf) the specific names. 8. C. paradentipes (Coutiere) It was hoped that in the key to the species Macrochirus Group it would be possible to avoid the use of all 9. C. gracilis (Heller) var. simplex Banner sexually dimorphic characteristics and the 10. C. edmondsoni Banner large and small chelae which are so often lost. 11. C. latipes Banner However, it was soon found that such a key 12. C. ventrosa (Milne-Edwards) would be difficult to use and would throw the 13. C. amirantei (Coutiere) species out of their normal sequence. Wher­ 14. C. nanus Banner ever possible the substitution of other char­ 15. C. hawaiiensis Edmondson acteristics for these was made, especially for 16. C. paragracilis (Coutiere) sexually dimorphic characteristics. Finally, ac­ Crinita Group, Obesomanus Subgroup: cessory characters are given where it was Not represented. thought that they would be useful.

Crinita Group, Crinita Subgroup HAWAIIAN SPECIES OF CRANGON 17. C. paralcyone (Coutiere) 1. Anterior margins of orbital hoods 18. C. brevipes (Stimpson) with acute teeth 2 19. C. clypeata (Coutiere) Anterior margins of orbital hoods Crinita Group, Diadema Subgroup* rounded, flattened, or keeled but 20. C. paracrinita (Miers) never bearing acute teeth 13 21. C. paracrinita (Miers) var. bengalensis 2(1). Outer face of large chela with sev­ (Coutiere) eral longitudinal grooves and 22. C. gracilipes (Stimpson) crests; dactylar articulation of 23. C. pugnax (Dana) chela flanked with strong teeth. 3 24. C. diadema (Dana) Large chela rounded, no grooves, 25. C. pseudopugnax Banner crests, or teeth 10 26. C. percyi (Coutiere) 3(2). Inferior distal margin of merus of Brevirostris Group third leg rounded. (Some aber­ 27. C. rapax (Fabr.) rant specimens of some species 28. C. platyunguiculata Banner in this group lack acute teeth on the orbital hoods; they can be * This, name has been changed from Insignis to reflect the placement of C. insignis (Heller, 1861) as a recognized easily, however, by synonym to C. diadema (Dana, 1852); although the the twisted and crested nature of rules of zoological nomenclature do not apply to this the large chela.) 4 subgeneric category, it was thought that the change would reduce confusion. Inferior distal margin of the merus f Two species of this group, C. audouini (Coutiere) of the third leg produced into an and C. strenua (Dana), which have been reported from acute or obtuse tooth 8 the Hawaiian Islands, are not included in this list nor 4(3). Crest leading to the palmar adhe­ in the key because the records are doubtful (see pp. 141 and 142). sive plaque on large chela sharp; Hawaiian Crangonidae — BANNER 49

dactylus of large chela with small large and small chelae almost lobes on inner distal face but not glabrous; small chela slender bulbous at tip 5 with fingers bent at right angles Crest leading to palmar adhesive and crossing; dactylus of third plaque on large chela rounded; leg biunguiculate dactylus of large chela with tip C. paradentipes (p. 72) bulbous 6 10(2). Dactylus of third leg simple, 5(4). Second carpal article of second leg curved, conical, with acute tip. . 11 0.5 the length of the first Dactylus of third leg with tip either C. hailstonei vz.t. paucispinata (p. 51) obtuse or strongly hooked. ... 12 Second carpal article of second leg 11(10). Large chela about 2.5 times as long almost 1.5 times the length of as high, with depressions on su­ the first. . . .C. brachymerus (p. 58) perior and inferior margins. Me- 6(4). Dactylus of third leg with one or rus of third leg 3-4 times as long more superior serrations but with as broad; external spine of uro- no trace of an inferior unguis. . 7 pod usually black Dactylus of third leg with no supe­ C. gracilis var. simplex (p. 75) rior serrations but with a strong Large chela 3.0-3.5 times as long inferior unguis. C. oahuensis (p. 64) as high, without depressions on 7(6). Crest leading to the superior tooth margins. Merus of third leg 6-7 at the dactylar articulation of the times as long as broad; external large chela interrupted by a trans­ spine of uropod never black. . . verse notch. . C. albatrossae (p. 60) C. edmondsoni (p. 78) Crest leading to the superior tooth 12(10). Dactylus of third leg with abrupt at the dactylar articulation of taper and abrupt hook in distal the large chela continuous from 0.3; tip acute. Second article of the superior margin of the carpus of second leg broader palm C. tuthilli (p. 63) than long; mature specimens 8.5 8(3). Tips of the orbital teeth reaching mm. long C. latipes (p. 82) about half the length of ros­ Dactylus of third leg almost trum. Dactylus of third leg bi- straight and very blunt. Second unguiculate; inner face of large article of carpus of second leg chela with a moderate growth of about twice as long as broad; setae; lateral spine of basicerite mature specimens up to 42 mm. longer than stylocerite long C. ventrosa (p. 84) C. collumiana (p. 67) 13(1). Large chela with only 1 or 2 simple Tips of the orbital teeth reaching longitudinal grooves on outer at least near the end of ros­ face extending proximally from trum, or beyond 9 articulation of dactylus 14 9(8). Tips of orbital teeth not reaching Large chela with or without to tip of rostrum. Large and small grooves,-but never with only 1 chelae densely hirsute on inner or 2 longitudinal grooves 17 faces; small chela heavy and with 14(13). Second article of carpus of second straight fingers; dactylus of third leg over twice length of first. leg simple. . . . C. deuteropus (p. 70) Large chela subcylindrical Tips of orbital teeth extending be­ C. amirantei (p. 87) yond rostrum. Inner faces of Second article of carpus of second 50 PACIFIC SCIENCE, Vol. VII, January, 1953

leg shorter than first. Large chela of visible portion of first anten­ compressed 15 nular article, with interorbital ca­ 15(14). Scaphocerite reduced, not reaching rina extending to tip; orbital to end of antennular peduncle; hoods evenly rounded anterior­ merus of third leg unarmed. ly; chela compressed 22 Margins of large chela entire. . . 21(20). Second carpal article of second leg C. nanus (p. 90) about 1.5-2.0 times length of Scaphocerite longer than the an­ first; without tooth on basicer- tennular peduncle; merus of ite; inferior internal margin of third leg armed 16 merus of large chela unarmed; 16(15). Depression on lower margin of no sexual dimorphism in small large chela with both sides grad­ chela . C. brevipes (p. 103) ual; second carpal article of sec­ Second carpal article of second leg ond leg broader than long about equal to first, not over 1.3 C. hawaiiensis (p. 95) times length of first; basicerite Depression on lower margin of with tooth; tooth distally on in­ large chela proximally delimited ferior internal margin of merus by a lobe; second carpal article of large chela; dactylus of small of second leg longer than broad. . chela of female broadened, twice C. paragracilis (p. 96) as long as broad 17(13). Dactylus of third leg biunguicu- C. clypeata (p. 107) late 18 22(20). First carpal article of second leg Dactylus of third leg simple 19 1.75-1.85 times length of sec­ 18(17). First carpal article of second leg ond; merus of both large and about twice as long as second. . small chelipeds with strong C. hawaiiensis (p. 95) tooth C. paracrinita (p. 110) Second carpal article of second leg First carpal article of second leg about 3 times as long as first. about equal to length of second; Large chela subcylindrical, taper­ tooth on inferior internal margin ing; margins entire of merus of large chela reduced, C. paralcyone (p. 99) none present on small chela. . . 19(17). Large chela without notches, C. paracrinita var. bengalensis (p. 110) grooves, or crests, usually quite 23(19). Large chela with only transverse cylindrical . 20 groove proximal to dactylus, Large chela with either transverse without lobes on either margin; groove, proximal to the dactylus, chela subcylindrical to com­ or pronounced lobes and depres­ pressed 24 sions, especially along upper and Large chela with transverse groove lower margins; chela sometimes and also marked lobes on both cylindrical 23 superior and inferior margins and 20(19). Rostrum very small or absent; dor­ depressed areas on faces; chela sal carina of carapace not extend­ always compressed 30 ing to tip of rostrum when pres­ 24(23). Dactylus of third leg conical, large ent; orbital hoods with narrow chela subcylindrical 25 to broad vertical keel anteriorly; Dactylus of third leg triangular in chela subcylindrical 21 section, inferi6r surface broad Rostrum reaching about to middle and flat, somewhat twisted, large Hawaiian Crangonidae — BANNER 51

chela strongly compressed 29 Second carpal article of second leg 25(24). Merus of third leg without tooth over twice as long as first. Pro­ on inferior distal margin. All ap­ podus of third leg with numer­ pendages long and slender; ros­ ous long setae tral base narrow and flattened C platyunguiculata (p. 130) dorsally; merus of large cheliped 30(23). Length of dactylus of third legs with few movable spines about twice breadth of merus. C. gracilipes (p. 115) Large chela with transverse Merus of third leg with strong groove across top, connecting tooth on inferior distal margin. 26 triangular depressed area on each 26(25). Second carpal article of second leg face; lower margin with strong twice as long as first. Merus of lobe; small chela of male prob­ large cheliped with series of ably "balaeniceps-shaped" .... spines, subterminal tooth "on in­ C. leptochirus (p. 133) ferior internal margin; rostral Length of dactylus of third legs base very narrow; rostrum not about equal to breadth of merus. 31 depressed C. pugnax (p. 116) 31 (30). Posterior margin of groove behind Second carpal article of second leg dactylus of large chela sloping about as long as first 27 back gradually (best seen in pro­ 27(26). Rostral base broad, at least 0.7 as file), not projecting over floor of broad as long. Merus of large groove; with more or less con­ cheliped without spines; rostrum spicuous tooth distally on inferi­ somewhat depressed or internal margin of merus of C. diadema (p. 118) large cheliped; small chela "ba­ Rostral base narrow, less than 0.5 laeniceps-shaped" in male, slen­ as broad as long 28 der in female; usually with spine 28(27). Anterior portion of base of ros­ on ischium of third legs trum sharply carinate with slop­ C. crassimanus (p. 134) ing margins; margins overhang­ Posterior margin of groove behind ing orbital-rostral grooves only dactylus of large chela overhang­ in posterior 0.3. Large chela with ing floor of groove (best seen in depressed areas on superior por­ profile); no tooth on inferior in­ tion of inner and outer faces; ternal margin of merus of large merus of cheliped with few cheliped; small chelae of both spines but numerous serrations. sexes with slender conical fin­ C. pseudopugnax (p. 122) gers; usually without ischial Anterior portion of base of ros­ spine on third legs trum flattened, with only slight C. patifica (p. 138) carina; margins of rostral base overhanging orbital-rostral groove for entire length. Large Megacheles Group chela without depressed areas as Crangon hatlstonei (Coutiere) var. pauci- in C. pseudopugnax, merus with spinata var. nov. similar serrations. C. percyi (p. 124) Fig. 16 a-j 29(24). Second carpal article of second leg shorter than first. Propodus of Alpheus hatlstonei Coutiere, Fauna and Geog. third leg bearing only light setae Maid, and Laccad. 2(4): 879, pi. 74, fig. 18, C. rapax (p. 126) 1905. 52 PACIFIC SCIENCE, Vol. VII, January, 1953

TYPE SPECIMEN: A 16 mm. ovigerous fe­ Total length of small cheliped 0.7 length male, "Albatross" Station 4128, off Kauai, of large cheliped, chela about 0.6 length of 68-90 fathoms, U. S. N. M. 91602. larger. Chela almost 5 times as long as broad DESCRIPTION: Rostrum well developed, with fingers 0.4 length of entire chela. Palm reaching almost to end of first antennular ar­ ovoid, articulation of dactylus with strong ticle and almost twice as long as broad at teeth on each side. Dactylus narrow, crested. base; top rounded, not carinate. Orbital hoods Carpus rather small. Merus 0.7 as long as somewhat inflated, with acute orbital teeth chela, 1.3 times as long as corresponding ar­ reaching beyond middle of rostrum. Orbito- ticle on large chela; slender, about 5.5 times rostral furrows slight, rounded. as long as broad; inferior internal margin Visible portion of first antennular article armed with 4 small spines and small tooth more than 0.7 length of second article; second distally. Ischium with single inferior spine. article twice length of third article, and 3 Carpal articles of second leg with ratio 10 : times as long as broad at base. Stylocerite 5-: 3 : 5.5 : 4. reaching to end of first article. Lateral spine Ischium of third legs about 0.5 as long as of basicerite nearly reaching to level of tip of merus, with strong movable spine. Merus 6.5 rostrum. Squamous portion of scaphocerite times as long as broad, unarmed. Carpus 0.6 narrow and almost 0.9 as long as lateral spine; as long as merus, with acute tooth distally on tip of lateral spine reaching to end of anten­ superior margin. Propodus 0.9 as long as me­ nular peduncle. Carpocerite reaching slightly rus, with 10 strong spines along inferior and past middle of third antennular article. distal margin; last spine bent distally, with Large chela slightly over 3 times as long as minute serrations on convex portion. Dac­ broad, strongly compressed, with pronounced tylus 0.23 as long as merus, only slightly ta­ crests and grooves. [For terminology of crests pering in proximal 0.7; basal portion de- and grooves, see Fig. He.] Superior crest marked from distal 0.25 by slight ridge pro­ strong, terminating distally as strong acute longed inferiorly as small projecting tooth; tooth above articulation of dactylus. Trans­ tip tapering, acute. verse groove deep, proximally overhanging, distally rounded. Superior groove only mod­ DISCUSSION: This particular form is repre­ erately deep. Plaque crest strong, with sharp sented in the collections studied by only two ridge, proximally merging with superior mar­ semi-intact specimens collected by the "Al­ gin of palm. Palmar groove well defined. In­ batross" in Hawaiian waters. There is a series ferior crest with lower margin better defined of additional individuals, from the same and than upper; ending at dactylar articulation as other stations, which are so fragmentary that tooth smaller than that of superior crest. In­ positive identifications cannot be made. ferior depression shallow but distinct, shoul­ The other specimen sufficiently intact for der rounded. Dactylus with high crest, over positive identification, a male 19 mm. long, 0.2 length of entire chela, thin, flat on outer varies from the type in the following charac­ face, rounded inner face with several distinct teristics: rostrum reaches end of first anten­ rounded lobes distally, extending over fixed nular article; large chela 2.9 times as long as finger. Carpus of usual form. Merus 3 times broad; merus of large chela 1.4 times as long as long as breadth of outer surface and twice as maximum breadth, with 5 spines, plus 2 as long as maximum breadth, superior and in­ on ischium; small chela lacking; carpal arti­ ferior external margins rounded; inferior in­ cle with ratio 10 : 5 : 2.5 : 6 : 4; articles of ternal margin with strong, acute tooth and third leg with ratio 2.5; 10: 6:9: 1.9; pro­ 3 small spinules; superior margin rounded podus with 14 spines; dactylus with short distally. Ischium with single spine. bristles on superior margin. These differences Hawaiian Crangonidae — BANNER 53

FIG. 16. Crangon haihtonei (Coutiere) var. paucispinata var. nov. a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large cheliped, medial aspect; e, large chela, distal end, dorsal aspect; /, small cheliped; g, second leg; h, third leg; /', /', third leg, dactylus. {a, b, scale A; c-h, scale B; /', /', scale C.) 54 PACIFIC SCIENCE, Vol. VII, January, 1953 are minor and to be expected in the normal from C. hailstonei assimulans in the nature of range of variation. the dactylus of the third legs. Other differ­ The varietal name refers to the few spines ences noted in the table are of more variable found on the merus of the large cheliped. characteristics and therefore are of lesser RELATIONSHIP: This form and the next four worth. described are apparently closely related to DISTRIBUTION: One of the positively iden­ each other, to the species C. hailstonei (Cou­ tified specimens has its collection data given tiere), and to its two tentative varieties de­ above; the other was collected by the "Alba­ scribed by de Man, C. hailstonei assimulans and tross" at Station 3868 in Pailolo Channel be­ C. hailstonei laetabilis. De Man was not sure of tween Molokai and Maui at 294-684 fathoms the separation of his forms from the parent (U. S. N. M. 63535). Other specimens, with form described by Coutiere, as the differences only one or several appendages attached or between them, although distinct, were rather with the appendages loose in the bottle so minor. Similarly, these five forms differ from that identification was more uncertain, were the previously described forms and from each collected as follows: other on minor but distinct points. It is likely, therefore, that when sufficient "Albatross" Station 3847 (U. S. N. M. intact specimens of these and the previous 63528): South coast of Molokai, 23-24 varieties are available and when the variation fathoms; 17 specimens, 14 of which are in a large population is studied it will be nec­ completely lacking thoracic legs. essary to consolidate this and de Man's varie­ "Albatross" Station 3872 (U. S. N. M. ties, and perhaps the following species as well, 63532): Auau Channel between Maui into the species described by Coutiere. But, and Lanai, 32-43 fathoms; 1 specimen, on the other hand, it may be necessary to lacking large chelae, second legs, and raise this and the previous varieties to specific some of posterior legs. rank. However, with only a few specimens "Albatross" Station 3875 (U. S. N. M. available, it appears to me that the separation 63527): Auau Channel between Maui given in this paper is for the present the most and Lanai, 65-34 fathoms; 2 specimens. wise, as this form is separated from C. hail­ "Albatross" Station 4128 (U. S. N. M. stonei and the other varieties by only more 63531): Data as given for type; 3 other minor characteristics, and the other forms, de­ fragmentary specimens. scribed as new species, by greater differences in more important characteristics. Other specimens from the "Albatross" col­ lection which undoubtedly belong to this Table 2 facilitates the separation of these eight species and varieties. The characteristics complex and were identified by Coutiere as given are not all of equal worth for the sep­ C. hailstonei (MS.), but which are too badly aration of the species. For example, little if broken to permit me to even guess as to their any faith can be placed upon the minor vari­ identity, were collected as follows: ations found in the relative lengths of the "Albatross" Station 3809 (U. S. N. M. carpal articles, except in C. brachymerus, or in 63533): South coast of Oahu, 51-125 the length of the rostrum, except possibly in fathoms; 1 specimen. C. tuthilli. "Albatross" Station 3982 (U. S. N. M. As can be seen from the table, this variety 63534): Near Kauai, 40-233 fathoms; 1 differs from C. hailstonei hailstonei and from C. specimen. hailstonei laetabilis chiefly in the number of "Albatross" Station 4.055 (U. S. N. M. spines on the margins of the merus of the 63530): Northeast coast of Hawaii, 50- large chela, and from C. hailstonei hailstonei and 62 fathoms; 1 specimen. TABLE 2 X CHARACTERISTICS DIFFERENTIATING Crangon hailstonei AND SOME RELATED SPECIES AND VARIETIES

A B c D E F G H CHARACTERISTIC C. hailstonei C. hailstonei C. hailstonei C. hailstonei C. brachymerus C. albatrossae C. tuthilli C. oahuensis hailstonei assimulans laetabilis paucispinata n E3 0*5 O D Size Not given To 19 mm. To 15 mm. To 19 mm. To 18 mm. To 33 mm. 23 mm. To 27 mm.

Rostrum Length To end first Same as A "Slightly 0.7-1.0 of Unknown 0.7 of first 0.5 of first Similar to F antennular shorter" first anten­ antennular antennular > article nular article article article Length/breadth 2 2.5 2(?) 2 Unknown 2 1 1.5 w

Orbitorostral margin Straight, Concave Similar to A Similarto B Unknown Straight, Similar to F Similar to F transverse about 30° to transverse

Antennular peduncle Lengths of articles 1 : 1 : 0.5 0.7 : 1 : 0.4 Similar to A 1 : 1 : 0.5 0.9 : 1 : 0.6 0.7 : 1 : 0.5 0.7 : 1 : 0.6 0.7 : 1 : 0.7

Length/breadth, 2nd article 3 3.5 Similar to A 3 3 3 2 1.7-2.0

Length stylocerite 0.7 of first Equals an­ Similar to B Similar to B Almost to end 0.8 of first Similar to B Subequal to antennular tennular first anten­ antennular first anten­ article article nular article article nular article

Antennal peduncle Scaphocerite Equals an­ Similar to A Similar to A Similar to A Exceeding Similar to E Similar to A Similar to E tennular antennular (variable) peduncle peduncle

Carpocerite Shorter than Equals Similar to B Similar to A Similar to B Slightly Similar to B Similar to B scaphocerite scaphocerite longer than (variable) scaphocerite Large cheliped Merits Length/breadth 3.0 Not stated Not stated 3.0 3.0 2.0-2.2 2.7 2.0

Length/maximum breadth 1.6 Not stated Not stated 1.4-2.0 1.2 1.7 1.6 1.3

Length/palmar height 1.0 Not stated Not stated 1.0 0.6 0.7-0.8 1.0 0.7 TABLE 2—Continued CHARACTERISTICS DIFFERENTIATING Crangon hailstonei AND SOME RELATED SPECIES AND VARIETIES

A B c D E F G H CHARACTERISTIC C. hailstonei C. hailstonei C. hailstonei C. hailstonei C. brachymerus C. albatrossae C. tuthilli C. oahuensis hailstonei assimulans laetabilis paucispinata

Inferior tooth Strong Not stated Not stated Moderate to Very strong Weak to Strong Strong strong moderate

Superior—inferior spines 3—0 0-4-6 4-5—8-9 0—3-5 3-5 0—6-10 0—8 0-9 (variable)

Ischial spines Superior—inferior 0—0 Not stated 2-3 0—1-2 0—0 0—1-3 2—5 0—5

Chela Length/breadth 3.3 .2.6-2.9 "Little 2.4-3.0 2.7 2.4 2.8 2.5 broader" than A

Superior crest Interrupted Interrupted Interrupted Interrupted Interrupted Interrupted Continuous Interrupted

Plaque crest Sharp Sharp Sharp Sharp Sharp Rounded Low, slightly Rounded angular >

Tip of dactylus 2 lobes on Similar to A Similar to A Similar to A Similar to A Expanded Similar to F Similar to F n inferior in­ and bulbous ternal face n en Small cheliped n Total length/length of w large cheliped 0.8 Similar to A Similar to A 0.7 0.8 Almost equal Lost Equal n w Merus' - < Length/width Not given Not given Not given 5.5 4.0 2.2-3.0 Lost 2.0 < Meral length/chela length 0.6 Similar to A Similar to A 0.7 0.45 0.7 Lost 0.4

Spines, superior—inferior 0—6 Similar to A Similar to A 0—3 4-7 0—9-11 Lost 0—13 (?) (?) (variable) Ischial spines Inferior 1 Similar to A Similar to A 1 1 0 Lost 7 (?) (?) TABLE 2—Continued CHARACTERISTICS DIFFERENTIATING Crangon hailstonei AND SOME RELATED SPECIES AND VARIETIES

A B C D E F G H D CHARACTERISTIC C. hailstonei C. hailstonei C. hailstonei C. hailstonei C. brachymerus C. albatrossae C. tuthilli C. oahuensis hailstonei assimulans laetabilis paucispinata n 3 CfQ Chela O D Length/breadth 5.5 Similar to A Similar to A 5.0 4.0 •3-4 Lost 4.4

Finger length/chela length 0.44 "Fingers Similar to A 0.4 0.5 0.5-0.6 Lost 0.6 slightly longer > than palm" % w Dactylus Broad, not Similar to A Similar to A Broad, Broad, high Rounded, Lost Very slender conspicu­ crested crest heavy and elon­ ously crest­ gate ed Second leg Ratio of carpal articles Not given 10:4-5:3:5:4 10:4-5:3:4: 10:5:2.5-3: 10:15:6:6:6 10:5-6:3-4: 10:6:4:4:6 10:7-9:3-4: 4-5 5.5-6:4 5-6:5-6 4-5:5-6

Third leg Ratio of articles 4.6:10:6:9:3 4:10:7:9:2.5 ?:10:6:9:2.5 2.5-5:10:6: 5:10:6:7:2.5 3:10:6:8:2.5 4:10:6:8:2.5 3:10:5:7.5:2 9:1.9-2.3

Ischial spine Absent Present Not given Present Present Present Present Absent

Meral length/breadth 10.5 9.5 6-8 6.5-8 9 5-6 7 3.4-4.0

Distal superior carpal tooth Absent Acute Not given Acute Acute Acute Acute Rounded

Carpal spines Absent Absent Absent Absent Absent Absent Absent 2 (weak)

Propodal spines 6 9 8 10-14 13 11-12 8 6-10

Dactylus Slender, Single Slight Similar to C Similar to C Superior Similar to F Strongly simple superior superior and serrations biunguicu- tooth inferior late, teeth inermous

The data for C. hailstonei hailstonei taken from Coutiere (1905: 879—880, figs. 18-18c); for C. hailstonei assirmilans from de Man (1911: 331-333, pi. 14, figs. 64a-64e); for C. hailstonei laetabilis from de Man (1911: 333—334, pi. 14, figs. 64d—64e). Where necessary, the individual characteristics or proportions were derived from the figures. C. brachymerus, C. albatrossae, C. tuthilli, and C. oahuensis are all new species described on the following pages. 58 PACIFIC SCIENCE, Vol. VII, January, 1953

"Albatross" Station undesignated (U. S. N. with 2 large rounded lobes distally that meet M. 84243): Hawaiian Islands; 1 speci­ cutting surface of immovable finger; distal men. lobe with several groups of short strong bris­ tles. Carpus small. Merus very short, only 0.6 Crangon brachymerus sp. nov. as long as height of palm, broad, 3.0 times as Fig. 17 a-j long as breadth of outer face, 1.2 times as long as maximum breadth at distal end, with 3-4 TYPE SPECIMEN: The type and only known small spines along superior margin and 5 specimen is an 18 mm. male, from "Alba­ stronger spines on inferior internal margin. tross" Station 3874 (U. S. N. M. 63536) be­ Superior margin of merus distally produced tween Maui and Lanai Islands, 21-28 fathoms. but rounded; inferior internal margin greatly DESCRIPTION: Frontal region of carapace produced distally as broad, flat process termi­ abnormal, distorted and strongly asymmetri­ nating as strong tooth. Ischium of usual form, cal, with right orbital tooth rather poorly de­ without spines. veloped and left entirely lacking; rostrum dis­ Total length of small cheliped 0.8 length placed and bent sharply ventrad. Second an- of large cheliped, with chela 0.6 as long as tennular article about 0.1 longer than visible large chela and merus 1.3 times as long. Chela portion of first, 1.7 times length of third, al­ 0.25 as broad as long, somewhat compressed, most 3 times as long as broad. Stylocerite with teeth, crests, and grooves similar to large reaching to end of first antennular article. chela but much more poorly developed. Su­ Lateral spine of basicerite about 0.8 as long perior crest developed but rounded; trans­ as stylocerite. Lateral spine of scaphocerite verse groove very shallow; plaque crest con­ reaching slightly beyond end of antennular tinued as rounded ridge across palm; inferior peduncle; spine moderately strong, squamous crest also rounded; superior and inferior crests portion narrow. Carpocerite also reaching terminating distally as small but acute teeth slightly beyond end of antennular peduncle, at dactylar articulation. Fingers 0.8 length of but slightly shorter than scaphocerite. palm; immovable finger thin and slightly Large chela similar to C. hailstonei var. pau- twisted; dactylus with high, blade-like crest cispinata Banner (p. 51), compressed, 2.7 continuing almost to end, cutting surface times as long as high, with fingers about 0.25 knife-like, whole article twisted into shallow as long as entire chela. [For terminology of S-shape. Carpus slightly larger than that of crests and grooves see Fig. 17^.] Superior crest large chela. Merus 4 times as long as broad, acute distally, terminating in strong, acute with superior margin bearing 4 short spines tooth above dactylar articulation, merging and scarcely projecting, rounded, distal ter­ proximally with superior edge of chela. Trans­ mination; inferior internal margin with 7 verse groove deep with proximal margin more small spines and simple terminal tooth. In­ or less overhanging, distally rounded. Superior ferior surface of ischium with single spine. groove deep. Plaque crest pronounced, with Carpal articles of second legs with ratio sharp ridge, proximally merging with superior 10 : 14.5 : 5.5 : 6 : 6. Chela 0.8 length of first margin of palm. Palmar groove shallower and carpal article. less well defined than superior groove. Inferi­ Third legs in specimen broken. Fourth or crest arising from middle of palm, termi­ legs with ischium 0.5 length of merus with nating distally in tooth below dactylar articu­ strong movable spine. Merus 9 times as long lation, smaller than superior tooth. Inferior as broad, distally unarmed. Carpus almost 0.6 depression marked, shoulder strong but length of merus; superior, margin projecting rounded. Dactylus strongly twisted, with high as short, acute tooth. Propodus 0.7 length of crest, narrow blade; inner surface of dactylus merus, with 13 moderately strong spines on Hawaiian Crangonidae — BANNER 59

FIG. 17. Crangon brachymerus sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, e, large cheliped from "Albatross" Station 3847, medial and lateral aspects;/, small cheliped; g, small chela, distal end, superior surface; h, second leg; /', third leg;/', third leg, dactylus. Abbreviations on e: ic, inferior crest; ID, inferior depression; PC, plaque crest; PG, palmar groove; s, shoulder; sc, superior crest; SG, superior groove; TG, transverse groove, (a, b, scale A; c,f-i, scale B; d, e, scale C;/, scale D.) 60 PACIFIC SCIENCE, Vol. VII, January, 1953 inferior margin, distal 10 paired. Dactylus 0.25 0.5 the length of the chela; third and prob­ as long as merus, curved, tapering to acute ably most important, the second article of the tip, with very slight accessory tooth about carpus of the second leg is 1.5 times as long 0.7 distance to tip; small tuft of short thick as the first instead of only 0.5 as long. In my spines or bristles opposite biunguiculation. opinion, these three pronounced differences Parts of third legs present similar to fourth are great enough to separate this form as a but larger. distinct species. DISCUSSION: A single chela from the "Al­ batross" Station 3847 (U. S. N. M. 63571), Crangon albatrossae sp. nov. off Laysan Island, 97-199 fathoms, undoubt­ Fig. 18 a-m edly belongs to this species. The chela, shown in Figure He, is almost identical to the form TYPE SPECIMEN: An ovigerous female 33 described above. The palmar grooves of the mm. long, "Albatross" Station 4031 (U. S. N. outer face are deeper and the inner face has M. 63588), Penguin Bank, west of Molokai, irregular depressed areas, but a closer exam­ in 27 fathoms. Paratypes listed below. ination shows that without question these are DESCRIPTION: Rostrum 0.7 as long as visi­ the result of shrinking during a previous des­ ble portion of first article of antennular pe­ iccation. The only difference is the armature duncle, somewhat rounded dorsally, twice as of the superior margin of the merus where the long as broad at base; upper surface some­ small spines are so closely pressed against the what depressed toward tip. Orbitorostral rounded edge that only two can be discerned. grooves broad and shallow. Orbital hoods Careful examination did not disclose any slightly inflated, with teeth reaching 0.5 length more spines; but possibly in a better pre­ of rostrum. Margin of carapace between base served specimen more would be visible. of rostrum and orbital teeth straight but at The specific name refers to the short, broad angle of about 30° to transverse line of body. merus of the large cheliped. Second antennular article 3 times as long RELATIONSHIP: The separation of this spe­ as broad, 1.5 times length of visible portion cies from the related species and varieties in of first article, over twice length of third arti­ the C. hailstonei complex is given in Table 2. cle. Tip of stylocerite reaching to end of first It is unfortunate that the only specimen of article. this apparent species available has the frontal Lateral spine of basicerite reaching approxi­ region of the carapace distorted and abnormal mately to end of rostrum. Lateral spine of but other characteristics are sufficient to sep­ scaphocerite strong, lateral margin somewhat arate it from the related forms. It may be pre­ concave; tip reaching slightly beyond end of sumed, however, as none of the seven other antennular peduncle; squamous portion nar­ species and varieties differ markedly in the row, 0.9 length of lateral spine. Carpocerite nature of the rostrum and the orbital hoods, slightly longer than scaphocerite. that in this species too the rostrum is rather Large chela heavy, 2.4 times as long as elongate and the orbital teeth are well devel­ broad, compressed, 1.7 times as broad as oped, probably quite like the development of thick, tapering and distally sculptured. [For C. hailstonei v&r.paucispinata Banner (Fig. 16 a). terminology of crests and grooves, see Fig. The decisive differences between this spe­ \le\ Superior crest rounded, terminating dis­ cies and the variety C. hailstonei var. paucispi­ tally as subacute tooth above dactylar articu­ nata are three: First, the merus of the large lation; merging soon with superior margin of chela is very short and broad, markedly short­ chela proximal to transverse groove. Trans­ er than the palm of the chela is high; second, verse groove deep and abrupt, proximally the merus of the small cheliped is less than overhanging. Superior groove deep. Plaque Hawaiian Crangonidae — BANNER 61

FIG. 18. Crangon albatrossae sp. nov. a, b, Anterior region, dorsal and lateral aspects (one pair of antennular flagella lost); c, large cheliped, lateral aspect; d, large cheliped, medial aspect; e, large chela, superior face; /, g, small chelipeds; h, second leg; /', third leg;/', k, /, third leg, dactylus; m, uropods and telson. (a-d, f, /',/', m, from type specimen; e, g, h, k, /, from specimens collected off southwest Oahu.) {a, b, m, scale A; c-i, scale B; k, I, scale C;/', scale D.) 62 PACIFIC SCIENCE, Vol. VII, January, 1953

crest strong but rounded, extending to su­ of base. Lateral margins slightly convex; pos­ perior margin of chela. Palmar groove short, terior margin very slightly arcuate. Postero- well defined on lower side. Inferior crest laterally with only one pair of short spines; strong but rounded, terminating distally in larger dorsal spines about 0.55 and 0.8 dis­ small, subacute tooth. Inferior depression tance from base. deep, shoulder abrupt but rounded. Dactylus Maximum size: 33 mm. (type specimen). less than 0.3 length of chela, closing almost DISCUSSION: Only slight variation was noted across end; with high crest proximally, dis­ in seven specimens of this species available. tally expanded to appear bulbous. Chela with There was the usual and expected variation in sparse, scattered setae distally, especially near the proportions and lengths of appendages: fingers. Merus 2.2 times as long as broad, 1.7 in some the stylocerite was slightly shorter times as long as broad across maximum than the first antennular article, and in some breadth, 0.8 height of palm; superior distal it was as long; the length of the lateral spine margin produced but rounded, inferior inter­ of the basicerite also varied within narrow nal margin with 7 short movable spines and limits; the proportions of the large chela it­ poorly developed distal tooth. Inferior sur­ self were quite constant, but the merus of the face of ischium with 4 spines. large cheliped was somewhat more variable in Small cheliped almost equal in total length length, breadth, and armature, with the in­ to large cheliped, small chela about 0.75 as ferior internal tooth more highly developed long as large chela. Chela 4 times as long as in some than in the type; the small cheliped broad in female, 3-5 times as long as broad in appeared to show the slight sexual dimorph­ male, fingers occupying 0.6 of length. Palm ism described above; the limits of the ob­ subcylindrical without pronounced sculptur­ served variation in the carpal articles is given ing; articulation of dactylus flanked by acute above; the merus of the third legs ranged to teeth on each side; plaque crest slight, almost six times as long as broad in some rounded. Fingers slightly broadened with tips specimens; the number of spines on the pro­ crossing. Carpus with short, broad, rounded podus was as high as 11; the dactylus may tooth on upper distal margin. Merus 3 times have one to three superior serrations, and in as long as broad in female, 2.2 times as long some cases one of these serrations was devel­ as broad in male, with 9 irregularly placed oped into a lobe-like extension on the anterior movable spines and small distal tooth on in­ surface; the telson was considerably more nar­ ferior internal margin. row and elongate, but still preserving the basic Second legs lacking in type specimen. On form. paratypes, carpal articles with ratio 10 : 5-6 : The species is named in honor of the 3-4 : 5-6 : 5-6. Chela about 0.8 length of first steamer "Albatross" of the United States Fish carpal article. Commission which collected the type speci­ Ischium of third legs with strong movable mens in 1902. spine. Merus 5 times as long as broad, un­ RELATIONSHIP: This species plainly is close­ armed. Carpus 0.6 length of merus, with su­ ly related to C. hailstonei and its varieties; perior margin projecting distad as acute tooth. therefore it has been included in Table 2. Propodus almost 0.8 as long as merus, with 11 There are a series of minor characteristics that spines on inferior and distal margins. Dac­ would aid in the separation of this species tylus 0.2 as long as merus, about 3 times as from the varieties of C. hailstonei and C. brachy- long as broad at base, curved, tip simple, su­ merus, but the most important ones are found perior surface with 2 notches. in the large chela, in which the crest leading Telson short and broad, length only 1.2 to the plaque is rounded instead of sharp and times breadth at base, breadth at tip 0.6 that the dactylus has a bulbous end instead of a Hawaiian Crangonidae — BANNER 6; narrow lobe-bearing end, and in the dactylus "Albatross" Station 3873 (U.S.N.M. of the third legs which has marked superior 63589), in Auau Channel between Maui serrations and no trace of inferior teeth. and Lanai Islands, 32-37 fathoms; a male This species is also related to C. staphylinus 23 mm. long. (Coutiere). From the short description and Off southwest Oahu, 40-350 feet, collected plates of Coutiere (1921: 418, pi. 62, fig. 13; by the "Makua" (specimens in the Ber- note that this later description is exactly the nice P. Bishop Museum); five specimens, same as the original description [1908: 14] and male and female, from 20 to 29 mm. in addition has figures of the specimen) only long. small differences appear. The frontal region of the carapace is straighter and the orbital Crangon tuthilli sp. nov. teeth are shorter in C. staphylinus; the large Fig. 19 a-g chela may be different although with the in­ adequate description and the drawing of the TYPE SPECIMEN: A 23 mm. ovigerous fe­ poorly rotated chela, it is difficult to tell; the male, collected off southwest Oahu, 40-350 small chela is more slender; and, most im­ feet deep, by the "Makua." Sole specimen portant, the dactylus of the third leg is longer known. Specimen.in the Bernice P. Bishop and without the superior notches which are Museum. found in this species. DESCRIPTION: Rostrum small, reaching only DISTRIBUTION: In addition to the type loca­ to middle of visible portion of first antennular tion given above, the species has been col­ article, as broad as long, upper surface strong­ lected as follows: ly depressed toward tip. Anterior carapace

2. mm j -^mm 0.5 mm

FIG. 19. Crangon tuthilli sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, superior face; e, second leg, distal end; /, third leg; g, third leg, dactylus. (a, b, scale A; c-f, scale B; g, scale C.) 64 PACIFIC SCIENCE, Vol. VII, January, 1953 smoothly rounded, without trace of orbito- as merus, with acute distal projection of su­ rostral grooves. Orbital teeth short and nar­ perior margin. Propodus 0.8 as long as merus, row, reaching 0.7 length of rostrum. with 8 inferior and distal spines. Dactylus 0.25 Second antennular article 1.3 times as long as long as merus, curved, acute, with 2 su­ as visible portion of first, 1.5 times as long as perior serrations, no trace of inferior teeth. third, twice as long as broad. Stylocerite DISCUSSION: This single specimen came reaching end of first antennular article. from the same dredge haul that produced the Lateral spine of basicerite strong, longer paratypes of C. albatrossae Banner and super­ than rostrum, extending over 0.7 of length of ficially is very similar to them. The basis for first antennular article. Lateral margin of sca- separation of the species from the other spe­ phocerite almost straight, tip of lateral spine cies of the C. hailstonei complex is given in reaching end of antennular peduncle; squa­ Table 2. Probably the most valid and least mous portion only 0.05 shorter than spine. variable characteristics that will serve for the Carpocerite considerably longer than anten­ identification of this species are the nature of nular peduncle. the anterior region of the carapace, the ros­ Large chela only moderately heavy, 2.8 trum being much shorter and broader than times as long as broad, compressed and with that of the other species, and the lack of the distal sculpturing reduced. [For terminology transverse groove on the large chela. of crests and grooves see Fig. 17?.] Superior Unfortunately, with only one specimen crest low, rounded, terminating as relatively available, it cannot be determined whether small acute tooth above dactylar articulation. this specimen represents a heretofore uncol­ Transverse groove lacking. Superior groove lected species or merely an aberrant individ' shallow and broad. Plaque crest low, rounded, ual, probably of C. albatrossae. Minor genetic but merging with superior margin of palm. mutations could account for the differences Palmar groove rounded, narrower than su­ between this and the typical C. albatrossae. perior groove. Inferior ridge slight, rounded, The fact that the specimen was collected with distally ending as rounded protuberance be­ a series of specimens of C. albatrossae would low dactylar articulation. Inferior depression incline one toward that belief, but of course slight, shoulder low, gradual, and rounded. in a half-hour dredge haul there is no reason Dactylus similar to that of C. albatrossae Ban­ to suppose that all the specimens came from ner, with proximal portion narrow and crested, exactly the same habitat. However, the char­ distal portion expanded and bulbous. Carpus acteristics mentioned above and some of the of usual form. Merus equal in length to height more minor characteristics, such as the thick­ of palm, 2.7 times as long as breadth of outer ness of the second antennular article, appear face, 1.6 times as long as maximum breadth; to be rather constant within a species of this superior margin rounded, unarmed; superior group. For these reasons it was decided to distal angle slightly projecting but rounded; designate this specimen a species. inferior internal margin with 8 small spines The species is named in honor of Dr. Leon­ and large terminal tooth. Ischium with 2 ard D. Tuthill, Department of Zoology and spines on superior margin, 5 on inferior Entomology, University of Hawaii. margin. Small cheliped lost. Crangon oahuensis sp. nov. Carpal articles of second legs with ratio Fig. 20 a-m 10 : 6 : 4 : 4 : 6. Third legs with ischium 0.4 length of me­ TYPE SPECIMEN: A male 27 mm. long col­ rus, armed with strong spine. Merus 7 times lected from coral in water about 20 feet deep as long as broad, unarmed. Carpus 0.6 as long off Waikiki Reef, Oahu (U.S.N.M. 93455). Hawaiian Crangonidae — BANNER 65

FIG. 20. Crangon oahuensis sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, anterior carapace, dorsal aspect, showing asymmetrical development of orbital hoods; d, large cheliped, lateral aspect; e, large chela, superior face;/, large cheliped, merus, internal face; g, small cheliped, lateral aspect; h, small chela, superior face; /', small chela, section of dactylus showing ridge; j, second leg; k, third leg; /, third leg, dactylus; m, telson. (a-c, m, scale A; d-h, j, k, scale B; /', /, scale C.) 66 PACIFIC SCIENCE, Vol. VII, January, 1953

DESCRIPTION: Rostrum almost 1.5 times as distal 0.5 of inner face. Merus twice as long long as broad, with tip reaching 0.7 of length as broad, with superior margin terminated by of first antennular article; rostrum and rostral short obtuse tooth; inferior internal margin base slightly rounded dorsally, not carinate. with 9 small movable spines and strong sub- Orbital hoods not inflated. Orbital teeth terminal tooth. Ischium with rounded pro­ reaching about 0.7 of length of rostrum. Or- jection on superior margin, inferior margin bitorostral furrows slight, shallow, and broad; with 5 spinules similar to those on merus. anterior margin between base of rostrum and Small chela in both sexes of peculiar form, base of orbital teeth very slightly convex, but with palm slightly over twice as long as broad, at angle of about 30° to transverse line of but with dactylus long, slender, and curved, body. 1.4 times length of palm, when seen in lateral Second article of antennular peduncle twice view (as in Fig. 20g) over 11 times as long as as long as broad, 1.4 times length of first ar­ broad in middle; fingers curved in both planes, ticle, 1.5 times length of third article. Stlyo- tips cross slightly. Superior side of opposing cerite reaching slightly beyond end of first face of fingers with thin ridge except where antennular article. Lateral spine of basicerite tips curve and cross; ridge with close series of acute, tip reaching to end of first antennular fine rugae (Fig. 20i); on interior side of ridge article. [In the type specimen, the spines are opposing faces with orderly series of short developed asymmetrically, with that of one stiff setae. Articulation of dactylus flanked on side as described and that of the other each side by strong acute teeth. Merus slightly rounded anteriorly, shorter than the rostrum.] over twice as long as broad, slightly heavier Lateral margin of scaphocerite only slightly than that of large chela, with similar armature concave; spine 0.8 longer than squamous por­ except spinules more numerous and subter- tion; squamous portion reaching end of an­ minal tooth smaller. Small chela 0.8 as long tennular peduncle. Carpocerite equal in length as large chela. to spine of scaphocerite. Carpal articles of second legs with ratio Large chela heavy, 2.5 times as long as high, 10 : 9 : 4 : 4 : 6. compressed, 1.6 times as high as thick, dis- Third legs with ischium unarmed. Merus tally sculptured. [For terminology of crests 3.4 times as long as broad, unarmed. Carpus and grooves, see Fig. He.] Superior crest well 0.5 length of merus, superior distal angle pro­ developed, terminating distally as strong acute jecting but rounded; inferior margin with 2 tooth above dactylar articulation, proximally nearly spine-like bristles in type specimen, merging with superior margin. Transverse with distal angle rounded, not projecting. groove deep, abrupt. Superior groove deep Propodus 0.75 as long as merus, with 10 long and well defined. Plaque crest rounded but slender spines on inferior and inferodistal heavy, merging proximally with superior mar­ margins. Dactylus 0.2 as long as merus, curved, gin of palm. Palmar groove with upper mar­ with small secondary unguis on inferior gin poorly defined, lower margin abrupt. In­ margin. ferior crest rounded, strong, terminating as Telson 1.5 times as long as broad, base 1.7 strong, acute tooth below dactylar articula­ times as broad as tip; lateral margins some­ tion. Inferior depression shallow; shoulder what convex; distal margin arcuate. low, gradual and rounded. Dactylus, occupy­ DISCUSSION: The 12 specimens of this spe­ ing distal 0.25 of chela, distorted, hammer- cies at hand show some variation, especially shaped, end bulbous, middle thin, crested, those of smaller size when compared to the and twisted; outer surface of dactylus with 3 large specimen designated as the type. The grooves armed with short stiff setae. Chela general form and appearance was the same in with numerous short to medium bristles on all cases, but in some of the smaller specimens Hawaiian Crangonidae — BANNER 67 the articles of the antennular peduncle were DISTRIBUTION: This species is represented shorter and broader (second article ranges to by 13 specimens which were collected on 1.7 times as long as broad); the stylocerite Oahu at Hanauma Bay, Kahala, Waikiki, and was shorter than the first antennular article; Nanakuli (all except the Kahala specimen the relative lengths of the antennular pedun­ from water 10 to 20 feet deep) and on Hawaii cle, the scaphocerite, and the carpocerite were at Halape on the Kau Coast, one specimen not constant; the number of spines on the from 10 feet, one from 40 feet deep, the last merus of the large and small chelae varied, but two collected by R. W. Hiatt. in all cases were numerous; the second carpal article was at times 0.7 or 0.8 the length of the Crangon collumiana (Stimpson) first, with the other articles varying to a simi­ Fig. 21 a-h lar degree; the merus of the third legs was more slender, up to 4.0 times as long as broad; Alpheus collumianus Stimpson, Acad. Nat. Sci. and finally, the number of spines on the pro- Phila., Proc. 12: 30, I860. podus of the third legs usually was less than DESCRIPTION: Rostrum acute, short, reach­ the 10 spines of the type specimen, often ing to middle of visible portion of first anten­ around six or seven. nular articles, demarked laterally from orbital The most remarkable variation was ob­ hoods by abrupt depressions, with small ros­ served in the orbital teeth of three specimens. tral carina that extends posteriorly between Two specimens, one shown in Figure 20c, had orbital hoods. Orbital hoods rounded with the orbital teeth asymmetrically developed, anteriolateral margins produced into acute with a tooth of the usual type on one side and teeth as long as rostrum. Margin between the other side terminating in an obtuse non- base of rostrum and orbital teeth convex. projecting angle. The third specimen was Lateral spine of stylocerite acute, reaching lacking in the orbital teeth on both sides. beyond rostrum nearly to end of first anten­ This species has been named after the island nular article. Second article of antennular from which it was collected. peduncle usually 1.8 times length of visible RELATIONSHIP: This species apparently be­ portion of first article, 1.7 times length of longs to the group of species designated in third article. Lateral spine of basicerite long, this paper as the C. haihtonei complex, and for acute, reaching beyond end of first antennular that reason its characteristics have been set article. Scaphocerite with strong lateral spine forth in Table 2. It can easily be separated that exceeds squamous portion by about 0.25, from all other members of this complex and usually reaching end of antennular peduncle. from C. seurati (Coutiere) (which probably Tip of carpocerite extending beyond end of also belongs in the complex) by the extraor­ scaphocerite. dinary elongation and great thinness of the Large chela 2.3 times as long as broad, lat­ fingers of the small chela and by the strong erally compressed and distally sculptured. [For biunguiculation of the dactylus of the third terminology of crests and grooves, see Fig. legs. Other characteristics which will separate 17

but not merging with superior margin of 2. In some specimens the scaphocerite was palm. Palmar groove narrow and deep, reach­ equal in length or slightly shorter than the ing only about 0.3 length of palm. Inferior antennular peduncle; in others it was slightly crest heavy, rounded, terminating distally as longer than the carpocerite. The carpocerite, strong, acute tooth below dactylar articula­ usually markedly longer than the antennular tion. Inferior depression moderately devel­ peduncle, in a few cases was equal to it. oped; shoulder heavy but rounded. Dactylus 3. In most specimens the small chela was short, about 0.25 length of entire chela, ro­ markedly smaller than the large chela, as in tated to close almost distally across end of the one drawn; in a few of the larger ones it chela. Chela moderately setose, with numer­ was almost 0.8 the length of the large chela ous groups of setae on inner face, superior and similar in form to the small chela of C. and inferior margins, and about edge of fin­ deuteropus (Hilgendorf) (Fig. 22e). gers. Merus less than twice as long as broad, 4. The second carpal article of the second without spines, with superior and inferior in­ legs was usually about 0.9 the length of the ternal margins slightly produced and rounded. first but varied from 0.8 to 1.0 times as long. Small chela somewhat heavy. Palm 1.5 5. The merus of the third legs in a few times as long as wide, distally with superior specimens carried one or two weak spines, crest well developed and ending as projecting and the inferior tooth, usually acute with an tooth which flanks articulation of dactylus. angle of 60 or 70 degrees, was in some cases Fingers about equal in length to palm, dis­ obtuse with an angle of more than 100 de­ tally strongly hooked. Carpus with flat tooth grees. on superior distal margin. Merus over twice The Hawaiian form of this species differs as long as wide, without spines. from Stimpson's original description chiefly Carpal articles of second legs with ratio in the nature of the third and fourth legs, 10 : 9 : 4 : 4 : 7. which he describes as, "Pedes tertii quartique Ischium of third legs with movable spine. . . . mero lato, inferne spinuloso et apicem Merus 3-3 times as long as wide, armed with unidentato. . . ." In the Hawaiian forms, the strong but rounded tooth on inferior distal few specimens with feeble spinules on the margin. Carpus with superior margin ending merus would hardly be called spinose. The distally in acute dentate projection. Propodus specimen described by de Man (1911: 334) with 5-6 sets of strong movable spines along had one or two small movable spines on the inferior margin. Dactylus biunguiculate, with carpus of the third legs which are lacking in inferior tooth small and located in distal 0.25 the Hawaiian specimens. Finally, in de Man's of article. plates (pi. 14, figs. 65, 65a) the superior ridge Largest specimens in collection 20 mm. of the large chela merges with the surface of long. In life body mostly transparent with red the palm without any abrupt groove demark- and yellow chromatophores, legs with yellow ing it as in the Hawaiian form. These two and green chromatophores; superior surface characteristics of de Man's specimen were not of chelae olive drab. remarked upon in the original description. DISCUSSION: This species does not seem to These differences are minor and probably in­ exhibit the wide range of variation found in significant. some. However, the following variations were The three specimens listed below, which noted: were collected by the "Albatross," had been 1. The length of the second antennular ar­ examined but not published upon by Cou- ticle varied from 1.5 to 2.0 times the length of tiere. The Laysan specimen ("Albatross" Sta­ the visible portion of the first article; in most tion 3962) was given a manuscript name for specimens about 1.7 or 1.8 times as long. a new species. The specimen had recently Hawaiian Crangonidae — BANNER 69

jZmm j0.5mm

FIG. 21. Crangon collumiana (Stimpson). a, b, Anterior region, dorsal and lateral aspects; c, large chela, lateral aspect; d, large chela, medial aspect; e, small chela; /, second leg; g, third leg; h, third leg, dactylus. (a-g, scale A; h, scale B.) 70 PACIFIC SCIENCE, Vol. VII, January, 1953 molted, and its chitin was soft and distorted; tral carina; base separated from orbital hoods however, close examination showed only one by depression. Orbital hoods rounded except difference from the specimens described for acute orbital teeth which equal rostrum in above: there were two spines on the carpus of length. Margin between orbital teeth and ros­ the third legs (therefore more like the speci­ trum straight to slightly arcuate. men described by de Man). The other two Antennular peduncle long and slender, specimens, which Coutiere had identified as with second article slightly less than 1.5 times C. dentipes (Guerin)—a species never reported as long as visible portion of first, about twice in the Indo-Pacific—showed no characteris­ as long as third. Lateral spine of stylocerite tics that would distinguish them from C. col- from 0.7 as long to as long as first antennular lumiana as identified in this paper. article. Entire antennular peduncle densely DISTRIBUTION: This rather common species hirsute. Lateral spine of scaphocerite strong, has been collected from the reef flats and from 1.25 times as long as narrow squamous por­ waters ranging to 20 feet deep at the following tion; outer margin strongly concave. Lateral localities around Oahu: Waimanalo, Hana- spine of basicerite strong, extending beyond uma Bay, Black Point, Waikiki, Mokuleia, end of rostrum. Carpocerite hirsute. Scapho­ Haleiwa, Kawela Bay. It was dredged from cerite usually not reaching end of last anten­ 100-200 feet off Diamond Head and from 180 nular article; carpocerite almost always longer feet off Kaena Point, where it was found liv­ than both antennular peduncle and scaphocer­ ing in sponges. Off Maui it was collected at ite. [The specimen drawn in Figure 22^? is ab­ Makena, Kalama Park, Lahaina, and two lo­ normal in these relationships.] calities south of Lahaina. The "Albatross" Chelae very densely hirsute on inner face specimens came from off Laysan, 16 fathoms and superior portion of outer face; structure (Station 3962, U.S.N.M. 63596), off French obscured by setae. Hirsute surfaces papillose Frigate Shoal in \AV2~l6V2 fathoms (Station where bundles of setae are attached. Large 3968, U.S.N.M. 63509), and off Nihoa, 24-40 chela strongly compressed, ovate in shape, fathoms (Station 4l63, U.S.N.M. 63510). Ed- about twice as long as high. Fingers approxi­ mondson (1925) has reported the species from mately 0.2 as long as entire chela. [For termi­ Pearl and Hermes Reef, Laysan, Lisianski, and nology of crests and grooves, see Fig. 17

FIG. 22. Crangon deuteropus (Hilgendorf). a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect, setae not shown; d, large cheliped, medial aspect, only bases of setae shown, length indicated by dotted line; e, small cheliped, medial aspect, setae not shown; /, second leg; g, third leg. 72 PACIFIC SCIENCE, Vol. VII, January, 1953 slightly over 0.5 length of palm. Distal por­ bear a more fixed relationship to the length tion of superior crest well developed, ending of the second article than to the length of the distally in very large, acute, projecting tooth, first article. ending proximally in transverse groove. Other 5. The number of spines of the propodus grooves and crests not developed. Fingers of the third legs varied from five to seven laterally compressed, thin at cutting edge. pairs. Carpus heavy, with rounded to acute flat tooth The smallest specimens, 8 millimeters long, superior to point of articulation of propodus. as well as the largest, 35 millimeters long, Merus heavy, only twice as long as wide, were within this range of variation, and no without teeth. particular variation was correlated with the Second legs only lightly hirsute. Carpal ar­ growth differences. ticles with ratio 10 : 5 : 2 : 3 : 4 (in specimen If one considers that Hilgendorf described drawn). the small chela as if it were the large chela— Third legs compressed and wide. Ischium if he did not, his was a most aberrant cran- with long spine. Merus 3.5 times as long as gonid—the Hawaiian form of this species wide, strong tooth on inferior distal edge, does not differ from the original description lined by long setae along superior margin. in any way that is not covered by the observed Carpus also lined by setae on both edges; each variation. distal margin projecting as acute tooth. Pro­ DISTRIBUTION: The species has been col­ podus with superior edge lined by setae, in­ lected, mostly in water from 6 to 20 feet deep, ferior edge with 5 pairs of movable spines and at the following localities about Oahu: Ka- unpaired proximal spine. Dactylus short, sim­ neohe Bay, Waimanalo, Hanauma Bay, Wai- ple, acute, strongly curved. kiki, Nanakuli, Kawela Bay. One specimen Specimens range to 35 mm. long. In life was dredged from water 100 feet deep off mostly colorless with scattered small red Waimanalo. Off Maui it was collected at five chromatophores on body and greenish-brown localities north and south of Lahaina. Ed- tracing of margins of anterior body regions mondson (1925) reported the species from and appendages; chelae with dark greenish French Frigate Shoal; the Bernice P. Bishop cast. Museum has two unreported specimens from DISCUSSION: The 50 specimens of this spe­ Pearl and Hermes Reef. cies from Oahu were examined in detail and Elsewhere the species has been reported in the following variations were noted: the Pacific only from the Netherlands Indies, 1. The rostrum varied from slightly shorter but within the Indian Ocean it has been col­ to slightly longer than the orbital teeth. lected at various points, including the Mal- 2. The frontal margin of the carapace be­ dive Archipelago and the east coast of Africa, tween the rostrum and the orbital teeth varied from almost straight to definitely arcuate. Crangon paradentipes (Coutiere) 3. The length of the stylocerite and the rel­ Figs. 23 a-i, 24 a-b ative lengths of the scaphocerite, carpocerite, Alpheus paradentipes Coutiere, Fauna and and antennular peduncle varied as noted in Geog. Maid, and Laccad. 2(4): 880, pi. 74, the description. Usually the scaphocerite fig. 17, 1905. reached to somewhat past the middle of the third antennular article. DESCRIPTION: Rostrum short, reaching only 4. The second carpal article of the second to near middle of first antennular article, nar­ legs varied from 0.55 to 0.85 times the length row, rather flattened on top, with carina ex­ of the first article, usually being about 0.6 to tending only short distance on anterior cara­ 0.7 its length. The distal articles appeared to pace. Orbital hoods not inflated, anterior Hawaiian Crangonidae — BANNER 73

j I m m 2mm 0.5mm

FIG. 23. Crangon paradentipes (Coutiere). a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, ventral aspect; e, f, small chela, female, lateral and ventral aspects; g, second leg; h, third leg; /', third leg, propodus and dactylus. {a, b, scale A; c-h, scale B; /, scale C.) 74 PACIFIC SCIENCE, Vol. VII, January, 1953

margin not continuous curve, with acute or­ tooth on inferior internal margin; ischium also bital teeth that extend beyond tip of rostrum. with spine. [No male in the collection with Anterodorsal surface of carapace somewhat the small chela intact, but according to Cou- depressed; orbitorostral furrows very broad tiere the male chela has the fingers broadened and shallow, not reaching as far posteriorly but not hooked.] as posterior edge of eyes; anterior margin be­ Carpal articles of second legs with ratio tween rostral base and base of orbital spines 10 : 7 : 3 : 4 : 5. almost straight, forming angle of about 30° Merus of third legs almost 6 times as long to transverse line of body, base of orbital as broad, armed with strong subterminal spines advanced. tooth. Carpus 0.6 as long as merus, with small Second article of antennular peduncle twice acute projection on superior distal margin; in­ as long as broad, equal in length to visible por­ ferior margin with 4 movable spines, most tion of basal article, 1.3 times as long as distal distal one on inferior distal angle. Propodus article. Spine of stylocerite reaching to end of 0.7 as long as merus, with 7 spines on inferior first article. Lateral spine of basicerite poorly margin but none distally. Dactylus 0.3 as long developed but reaching to end of rostrum. as propodus, biunguiculate with inferior un­ Scaphocerite equal in length to antennular guis shorter than superior. peduncle, slightly exceeded by carpocerite. Telson twice as wide at base as at tip, 1.8 Large chela heavy, 2.5 times as long as times as long as broad at tip; lateral margins broad, compressed, 1.7 times as high as thick, slightly convex; posterior margin arcuate. with pronounced sculpturing distally. [For Largest specimen, a female, 10 mm. long; terminology of crests and grooves, see Fig. color in life not noted. 17?.] Superior crest high and thin, continuing DISCUSSION: The few specimens in the pres­ distally as strong acute tooth lying above dac- ent collection show but little variation; in tylar articulation; soon merging with superior some the orbital teeth are slightly longer than margin of palm proximal to transverse groove. in the specimen described; in some the lateral Transverse groove broad and deep. Superior spine of the basicerite is better developed; the groove rather deep. Plaque crest heavy, number of movable spines on the merus of rounded. Palmar groove deep. Inferior crest the large chela is as great as five. The expected flattened with margins rounded, distally pro­ variation in the number of spines on the car­ jecting as strong, acute tooth below dactylar pus and propodus of the third legs was ob­ articulation. Inferior depression extensive and served. One interesting specimen from the deep; shoulder slightly overhanging but with­ "Albatross" collection had its sole remaining out sharp margins. Dactylus strongly com­ second leg with the carpus divided into six pressed, curved, distorted; only 0.2 length of rather than five articles; it appeared as if an entire chela because of distortion; chela ro­ extra article was inserted between the normal tated so finger closes horizontally almost first and second articles. across end of palm. Merus 3 times as long as Allowing for the variation observed, these broad, with superior margin terminated by specimens agree well with the descriptions acute tooth and inferior internal margin with and figures of Coutiere. I believe that the dif­ 2 spines and sub terminal tooth. Ischium also ference in the merus of the large chela, which with small spine. was described as bearing on "son bord supero- Small chela of female 3.5 times as long as interne [sic] . . . 7-8 dente ..." and which was broad, with fingers 0.5 of length. Articulation shown as being serrate on the inferior internal of dactylus flanked on each side by strong margin, probably normally carried movable teeth. Fingers with tips bent at 90° and cross­ spinules as do the specimens described above. ing. Merus with 3 spines and small distal Otherwise there are no marked differences. Hawaiian Crangonidae — BANNER 75

localities in the Indian Ocean, and from the Netherlands Indies.

Macrochirus Group Crangon gracilis (Heller) var. simplex var. nov. Fig. 25 a-k Alpheus gracilis Heller, K. Akad. Wiss., Wien, Sitzungber. 44: 271, Taf. 3, figs. 19-20, 1861. TYPE SPECIMEN: A male 12 mm. long, col­ lected on Waikiki Reef, Oahu (U.S.N.M. 93456). DESCRIPTION: Rostrum acute, rounded dor- sally, without carina; tip reaching from slight­ ly past middle of visible portion to end of first antennular article. Rostrum separated from orbital hoods by rounded shallow de­ pression. Orbital hoods rounded, with small acute orbital teeth, well demarked from curva­ ture of hoods; tips reaching almost to end of rostrum. Antennular peduncles with second article FIG. 24. Crangon paradentipes (Coutiere). a, Normal second leg from specimen from "Albatross" Station somewhat longer than broad and shorter than 4063; b, second leg with a carpus of six articles from visible portion of first or third. Stylocerite specimen from an undesignated "Albatross" Station large, lateral spine reaching past middle of (U.S.N.M. 63600). second antennular article. Lateral spine of DISTRIBUTION: This species is one of the basicerite from shorter to longer than ros­ few that have been found in sponges in trum. Scaphocerite with strong lateral spine Hawaii, a common habitat for some species slightly longer than antennular peduncle and in other parts of the world. However, of the shorter than carpocerite. four separate collections of the individuals Large chela strongly compressed, slightly noted below, only one was noted as coming over 0.5 as thick as high, 2.5 times as long as from a sponge; whether the others did or not high, without lobes or grooves. Shallow is not known. rounded depression on superior internal face All records of the species probably are from proximal to articulation of dactylus, similar deep water beyond the reefs; two were from concavity on inferior internal face inferior to Oahu, off Kaena Point, 180 feet (in sponge), point of articulation. Fingers 0.3 length of and off Diamond Head, 100-200 feet; one chela. Chela almost glabrous except for usual taken by the "Albatross" was from north­ bristles near tip of fingers and few scattered east of Hawaii in 61-63 fathoms (Station setae along superior and inferior margins. 4063, U.S.N.M. 63601); the fourth, also taken Merus without teeth, with 4-6 short spines by the "Albatross," merely states "Hawaiian along inferior internal margin. Small chela Islands" (U.S.N.M. 63600). with fingers longer than palm, with tooth at The species is also known from the Lacca- point of articulation of dactylus; merus with dive Archipelago, its type locality, from other armature similar to that of large chela. 76 PACIFIC SCIENCE, Vol. VII, January, 1953

Carpal articles of second legs with ratio tions of the telson where a series of six speci­ 10 : 6 : 3 : 3 : 6. mens gave differences in the ratio of the width Ischium and merus unarmed; merus 4-5 of the posterior border to the greatest width times as long as broad; propodus tapering of 1.6-2.0, and in the ratio of the width of the distally, with proximal end about 1.5 times posterior border to the length of 2.4-3.0. A width of distal; inferior margin with 6-8 long most peculiar variation was observed in the slender spines; dactylus slender, 0.25 length external spine of the uropods: in 40 specimens of merus, simple, with slight convexity on all but 5 have brown to black spines, 2 have inferior margin at point where lower unguis spines with faint color, and 3 have colorless would be expected. spines. This change of color is not the result Telson with sides rather straight and with of the solvent action of the preserving fluid pronounced taper, 1.8 times as wide as long (formaldehyde or alcohol), as some specimens at posterior margin. External spine of uro­ preserved over 14 years still have the spines as pods almost always brown to black. dark as the newly preserved ones. Perhaps the Specimens in collection range to 18 mm. in variation may be the result of moulting, for length. In life, body and chelae transversely several of the specimens with slightly colored banded with broad gray bands of red and blue or colorless spines have what appears to be chromatophores; eggs brilliant yellow. Trans­ softer chitin. verse bands red in preserved specimens. The varietal name refers to the lack of a sec­ DISCUSSION: This species varies in a num­ ondary unguis on the dactylus of the third leg. ber of characteristics. The rostrum, which in RELATIONSHIP: There have been four de­ most extends only to the middle of the visible scriptions published for this species and its portion of the first antennular article, in some varieties. The type specimen first described by specimens reaches to the end of that article Heller was subsequently redescribed by de and in one specimen extends to one third of Man (1911), who corrected some errors found the second article. In some specimens there is in Heller's description and also extended the a faint trace of an obtusely angular dorsal ca­ original description. Coupled with this rede- rina on the rostral base. The orbital teeth vary scription, de Man also described a tentative in breadth and length, sometimes being al­ variety, C. gracilis var. luciparensis which he most as long as the rostrum and sometimes distinguished from Heller's type specimen only half as long (contrast Fig. 25a and c)\ in chiefly by minor differences in the proportions all cases, however, the teeth arise abruptly of the various appendages and by the external from a uniformly curved anterior margin of spine of the uropods being black instead of the orbital hood. There are variations, as brown. It is interesting that the variations would be expected, in the lengths of the sty- noted above are sufficient to bridge these dif­ locerite and of the lateral spine of the basicer- ferences. Coutiere (1905) described C. gracilis ite, the latter in a few cases exceeds the length var. alluaudi from Mahe which was distin­ of the rostrum. The small and large chelae guished from the main species by the lack of vary somewhat in their proportions but their a secondary unguis, and on all other points general form and armature remain constant as "ces specimens me paraissent impossibles a shown. The second carpal article of the sec­ distinguer de YA. gracilis." One is at a loss to ond legs varies from 0.5 to 0.7 the length of know whether the specimens were exactly the first, with a similar variation in the other like Heller's erroneous description, therefore articles as well. On the third legs the variation different from the type specimen, or exactly of the breadth of the merus and the number like the type specimen. of spines on the propodus is noted above. The Hawaiian form here described is dis­ Variations were also noticed in the propor­ tinguished from the original species and from Hawaiian Crangonidae — BANNER 77

FIG. 25. Crangon gracilis (Heller) var. simplex var. nov. a, b, Anterior region, dorsal and lateral aspects; c, anterior carapace, dorsal aspect, showing a more typical development of the orbital teeth; d, e, large chela, lateral and medial aspects;/, g, large cheliped, merus and carpus, lateral and medial aspects; h, small cheliped; i, second leg; /, third leg; k, third leg, dactylus. (a-c, scale A; d-g, scale B; h-j, scale C; k, scale D.) 78 PACIFIC SCIENCE, Vol. VII, January, 1953

the two varieties by the following character­ Edmondson (1925) also recorded the spe­ istics: First, the other forms have a distinct cies from Johnston Island in the central Pa­ but subacute dorsal carina, while the Hawai­ cific. It and its two varieties have been re­ ian variety is usually evenly rounded dorsally. ported from the Netherlands Indies, the Indian Second, the base of the rostrum is separated Ocean, and the Red Sea. from the orbital hoods by rather shallow rounded depressions instead of deep narrow Crangon edmondsoni sp. nov. grooves as de Man describes them. Third, the Fig. 26a-p inferior internal margin of the merus of the large cheliped is without a small obtuse tooth TYPE SPECIMEN: A female 24 mm. long col­ at its distal extremity in the local specimens. lected at Kawela Bay, Oahu, by C. H. Ed­ From the original type for the species and mondson (Bernice P. Bishop Mus. 4059). from de Man's variety, this variety is distin­ DESCRIPTION: Rostrum narrow, longer than guished, as is C. gracilis var. alluaudi, by the breadth at base; concave sides with 3 or 4 lack of the secondary unguis on the dactylus. pairs of short bristles; produced as acute tip, Future workers may decide that the lack of reaching to or slightly beyond end of first an- the secondary unguis is sufficient to separate tennular article, distally depressed; slightly this form, together with Coutiere's, as a spe­ rounded dorsally, without carina. Orbital cies distinct from that of Heller. Except for hoods continuous with curvature of carapace the fact that this is not supported by other posteriorly, not sharply delimited from sur­ differences, I would be inclined to make this rounding carapace and separated from base of separation myself. However, as Coutiere did rostrum by only slight shallow depressions. not think of this as a specific difference, as Anterior margin of orbital hoods produced there are no supporting differences, and as into acute, dorsally flattened teeth, reaching there is the swelling on the dactyl which may almost to mid-point of rostrum. Margin of be the remnant of the secondary unguis, I orbital hoods between rostral base and apex have decided to leave the form within this of hoods straight to concave. species. The new variety for the Hawaiian Visible portion of first article of antennule specimens was created to emphasize this dif­ slightly larger than second and 0.25 longer ference and the fact that these specimens do than third. Stylocerite strong, reaching to not have the deep grooves between the orbital middle of second antennular article. Entire hoods and the rostrum. dorsal surface of antennular peduncle with oc­ C. gracilis var. simplex is most closely related casional short setae, with denser groupings of to C. edmondsoni Banner and C. latipes Banner longer setae at points of articulation. Inner of the other local species. The differences be­ flagellum about 7 times as long as peduncle; tween them are set forth in Table 3. outer flagellum 4 times as long as peduncle, DISTRIBUTION: This species has been col­ Lateral spine of basicerite of antenna acute, lected both from the shallow reefs and in tip not reaching to end of first antennular ar­ waters up to 20 feet deep at the following lo­ ticle. Scaphocerite as long as antennular pe­ cations on Oahu: Waimanalo, Hanauma Bay, duncle with lateral spines only slightly longer Black Point, Waikiki, Nanakuli, Mokuleia, than squamous portion, squamous portion ta­ Kawela Bay, and Punaluu. On Maui it was pering toward tip. End of carpocerite reach­ collected at Makena, Kalama Park, Lahaina, ing slightly beyond scaphocerite. Antennal and four localities north and south of Lahaina. flagellum longer than body, about twice as Edmondson (1925) reported it (as C. gracilis) long as inner antennular flagellum. from Laysan and Lisianski Islands. No speci­ Large chela elongate, tapering, margins mens were collected by the "Albatross." rounded, laterally compressed, 3.4 times as Hawaiian Crangonidae — BANNER 79

FIG. 26. Crangon edmondsoni sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, d, e, large cheliped, lateral, ventral, and medial aspects;/, g, large cheliped, merus, lateral and medial aspects; h, small cheliped, lateral aspect; i,j, small chela, ventral and ventrolateral aspects; k, small cheliped, merus, medial aspect; /, second leg; tn, third leg; », third leg, dactylus; o, fifth leg, propodus and dactylus;/), uropod and telson. (a, b,/ g, o, p, scale A; c—e, h—k, scale B; /, m, scale C; n, scale D.) 80 PACIFIC SCIENCE, Vol. VII, January, 1953

long as high, 5.0 times as long as thick. Mar­ as long as broad, with rather straight margins, gins entire, with slight trace of longitudinal without distal teeth. Carpus 0.6 as long as groove proximal to dactylus. Fingers com­ merus, with inferior and superior margins pro­ pressed, slightly less than 0.5 as long as palm, jecting distally as small rounded teeth. Pro­ slightly depressed from axis of palm, rotated podus 0.7 as long as merus, with 6 movable so dactylus opens and closes on lateral plane. spines along inferior margin. Dactylus almost Dactylus 1.8 times as long as broad, 2.5 times 0.3 as long as propodus, curved, simple; su­ as long as thick near base, with the upper (or perior margin with indentation or tooth 0.7 lateral) margin slightly rounded, distal mar­ of distance from base, curve then continuing gin strongly rounded. Chela with occasional to tip, with tuft of short setae on each side of setae on superior and inner face, with usual dactylus at this break. Entire legs with occa­ grouping of short bristles on opposing faces sional setae. Fourth legs similar. Propodus of of fingers. Carpus of usual form. Merus with fifth legs with 7 movable spines, and with inferior margins acute, superior margin "brush" feebly developed, of but 6 rows of rounded. Superior margin with 6 feeble spi­ setae. nules on basal 0.7; distal angle with obtuse, Telson 3.0 times as long as width of pos­ projecting tooth. Inferior lateral edge of me­ terior margin. Sides tapering, with angle of rus with 13 feeble spinules, slightly larger than taper greater in distal 0.2. Anterior pair of those on superior edge, margin produced as dorsal spines located in middle of dorsal sur­ acute tooth. Inferior internal margin of merus face, posterior pair 0.7 of distance toward with 7 movable spinules, with bristles be­ posterior edge. Posterior margin somewhat tween them. Merus 2.7 times as long as wide convex. distally. Ischium of usual form, with 3 strong Specimens range to 27 mm. long. Color in movable spines on inferior internal face, and life not noted; some of preserved specimens on adjacent margin of merus. with 7 distinct longitudinal bands reaching Small chela almost as long as large chela from antennular peduncles to end of abdo­ and with same rotation. Chela almost as high men; in preservation all bands red. Middle as thick, 5.0 times as long as broad, with band very narrow; dorsolateral bands quite slight taper toward fingers. Inner margin of broad; lateral and ventrolateral bands, latter palm continued as obtuse projection at point along margins of abdominal pleura, each of articulation of dactylus. Fingers almost as about 0.5 width of dorsolateral band. Distal long as palm, tapering, definite hook on tip. 0.25 of large and small chelae violet in pre­ Merus somewhat similar to that of large chela, servative. No specimens with any dark pig­ but without spines or bristles on superior or mentation on external spine of uropod. inferior external margins; 5 movable spines DISCUSSION: Even with the small number of and several bristles on inferior internal mar­ specimens of this species at hand, 19 in all, gin. Ischium with 5 smaller movable spines. some interesting variation was noted. As Lateral face of merus slightly more than twice usual, the relative length of the rostrum varied as long as wide. in some specimens. This may be due more to Second thoracic legs about 0.7 as long as the degree of flexure of the antennular pe­ entire animal. Ischium 1.2 times as long as duncle than to the variation in length of the merus, 0.7 as long as entire carpus. Carpal ar­ rostrum. In all cases the inner anterior margin ticles with ratio 10 : 7 : 4 : 4 : 6; fifth article of the orbital hoods were concave. Slight dif­ twice as broad as proximal end of first. Chela ferences were noted in the relationships of the 1.5 times as long as first carpal article. parts of the antennular and antennal pedun­ Ischium of third legs 0.4 as long as merus, cles. The most conspicuous differences were with proximal movable spine. Merus 6 times in the large chela, where the length of the fin- Hawaiian Crangonidae — BANNER 81 gers varied from 0.3 to 0.5 the length of the RELATIONSHIP: This species appears to be palm, and the whole chela from less than 3-0 very closely related to C. gracilis and, there­ to more than 3-5 times as long as high. Both fore, should belong to the Macrochirus large and small chelae may be glabrous. In all Group. According to de Man's key (1911), specimens the chela was without the depres­ the characteristics of this species would leave sions of the upper and lower margins that it close to C. shearmei (Alcock and Anderson), were found in C. gracilis var. simplex Banner. but obviously they are not closely related, be­ The teeth on the superior margin of the me- cause C. shearmei, a deep-sea species, has a rus, not conspicuous even in large specimens, minute rostrum, and the large chela lacks en­ were scarcely visible in small specimens. The tirely the piston on the dactylus. ratio of the first to the second article of the This species shows many characters in com­ carpus of the second legs varied from 10 : 7 mon with C. gracilis var. simplex Banner and C. to 10 : 8. The merus of the third leg varied latipes Banner; the three species are distin­ from six to seven times as long as broad. guished in Table 3. The species is named in honor of Dr. C. H. DISTRIBUTION: Besides the type specimen Edmondson, Professor Emeritus of Zoology from Kawela Bay, Oahu, the species has been at the University of Hawaii and Curator of In­ collected on Oahu from Hanauma Bay at vertebrates at the Bernice P. Bishop Museum. 15 feet, off Kahala at 150 feet, and off Nana-

TABLE 3 CHARACTERISTICS DIFFERENTIATING Crangon latipes, C. gracilis VAR. simplex, AND C. edmondsoni

C. latipes C. gracilis var. simplex C. edmondsoni

Maximum observed size 8.8 mm. Maximum observed size 18 mm. Maximum observed size 27 mm. Orbital teeth 0.3-0.5 length of ros­ Orbital teeth usually 0.7 length of Orbital teeth 0.5 length of rostrum; trum; abruptly set off from orbital rostrum; abruptly set off from or­ merging with orbital hoods on hoods; rostrum reaching end of bital hoods; rostrum not reaching inner margin; rostrum reaching to first antennular article end of first antennular article or beyond end of first antennular article Large chela 2.2 times as long as Large chela 2.5 times as long as Large chela 3.5 times as long as broad; 2.2 times as broad as thick; broad; 2.0 times as broad as thick; broad; almost as broad as thick; without depressions on margins with depressions on upper and without depressions on margins lower margins Merus of large cheliped with 2 or Merus of large cheliped with 4-6 Merus of large cheliped with 7 spines less spines on inferior internal spines on inferior internal mar­ on inferior internal margin; su­ margin; none elsewhere gin; none elsewhere perior and inferior external mar­ gins also spinose Carpal articles of second legs with Carpal articles of second legs with Carpal articles of second legs with ratio 10 : 8 : 3 : 3 : 14; second ar- ratio 10 : 6 : 3 : 3 : 6; second ar­ ratio 10 : 7 : 4 : 4 : 6; second arti­ article as broad as long ticle twice as long as broad cle over twice as long as broad Third legs Third legs Third legs Ischium with spine Ischium unarmed Ischium with spine Merus 3-3.5 times as long as Merus 4-5 times as long as broad Merus 6-7 times as long as broad broad Propodus with 4-5 spines Propodus with 6-8 spines Propodus with 6 spines Dactylus 2.3 times as long as Dactylus almost 4 times as long Dactylus 3 times as long as broad, broad, distally strongly curved as broad, slightly curved and slightly curved and gradually and abruptly tapering gradually tapering throughout tapering throughout length, length with notch on superior margin External spine of uropod never dark External spine of uropod usually External spine of uropod never dark brown to black Coloration not noted Body with longitudinal stripes Body with transverse stripes 82 PACIFIC SCIENCE, Vol. VII, January, 1953 kuli at 15-20 feet; two other specimens came tions 2 very shallow grooves reaching proxi- from unspecified localities on Oahu, presum­ mally from articulation of dactylus; grooves ably from shallow water. One specimen in very faint, visible only by rotating cheliped. Bernice P. Bishop Museum came from Ma- Merus 1.3 mm. long, about twice as long as kena, Maui, and one was also collected in 3 breadth of inner face; inferior internal margin feet of water at Kalama Park on that island. with 2 movable spines, terminating distally R. W. Hiatt collected two specimens at Ha- as small, acute tooth; termination of superior lape, Kau Coast, Hawaii, from coral heads in margin rounded. the intertidal zone. Small chela 1.0 mm. broad, 2.0 mm. long; fingers 0.5 length of chela, hooked distally. Crangon latipes sp. nov. Merus similar in size and proportions to that Fig. 27 a~j of large chela but with 3 short spines on dis­ Alpheus sp. Coutiere, Fauna and Geog. Maid, tal portion of inferior internal margin. and Laccad. 2: 882, pi. 74, fig. 19, 1905. Second legs with ischium 0.7 mm. long; TYPE SPECIMEN: A male 8.5 mm. long col­ merus 0.9 mm. long, 0.24 mm. broad; articles lected off Waikiki Reef, Oahu, in 20 feet of of carpus with total length of 1.0 mm., with water (U.S.N.M. 93457). ratio 10 : 8 : 3 : 3 : 14; second secondary ar­ DESCRIPTION: Anterior carapace produced ticle as long as broad; chela 0.6 mm. long, as well-developed rostrum, tip reaching slight­ fingers 0.5 of length. ly beyond end of first antennular article; ros­ Ischium of third legs 0.4 mm. long, with trum about 0.3 longer than width at base. movable spine; merus 1.36 mm. long, 0.44 Rostral base laterally demarked from orbital mm. broad, inermous; carpus 0.6 mm. long; hoods by shallow depressions; depressions propodus 1.1 mm. long, with 5 movable deeper and more abrupt posteriorly; rostrum spines; dactylus 0.35 mm. long, 0.13 mm. and base without definite carina, rounded dor- broad in basal half, abruptly tapering to sharp­ sally. Orbital hoods inflated, anterior margin ly curved acute tip. rounded in both dorsal and lateral aspects. Fourth legs similar in structure. Fifth legs Orbital teeth small, acute, arising abruptly generally similar but more slender, propodus from curvature of orbital hoods; tips reach with only several distal spines and without about 0.3 length of rostrum. well-defined "brush." Antennular peduncle in dorsal view with Telson 1.4 mm. long, 0.75 mm. broad at second article about 0.2 longer than visible base, and 0.36 mm. broad at tip, sides with portion of first article and 0.3 longer than uniform taper. Outer distal spines 0.07 mm. third; second article 1.3 times as long as broad. long, inner spines 0.18 mm. long; anterodor- Acute tip of stylocerite reaching to middle of sal spines about 0.5 distance from tip, pos­ second antennular article. terior spines 0.32 mm. from tip. Lateral spine of basicerite reaching nearly Three additional specimens: male 6.5 mm. to end of first antennular article. Scaphocerite long; 2 females 8.8 and 8.0 mm. long. Color with lateral spine strong, lateral margin al­ of specimens in life not noted. most straight, squamous portion well devel­ DISCUSSION: Two of the paratypes agree oped; tip reaches well beyond end of anten­ quite well with the type specimen. The female nular peduncle. Carpocerite reaching almost from Nanakuli differs chiefly in that the ros­ to end of scaphocerite. trum is asymmetrically bifurcate at the tip Large chela smooth in contour, strongly and one of the orbital teeth points toward the compressed. Chela 1.8 mm. high, 0.8 mm. rostrum. This, obviously, is the result of an thick, 4.0 mm. long to end of dactylus; dac­ accident or of anomalous growth. The female tylus occupying distal 1.1 mm. Only indenta- from Lahaina also agrees with the type quite Hawaiian Crangonidae — BANNER 83

mm .0.4mm

FIG. 27. Crangon latipes sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, d, large cheliped, lateral and medial aspects; e, small cheliped, lateral aspect;/, second leg; g, third leg; h, third leg, dactylus; /", fifth leg;/, tel- son. {a, b, f, g, i, j, scale A; c-e, scale B; h, scale C.) 84 PACIFIC SCIENCE, Vol. VII, January, 1953

well, except that the large chela is somewhat According to de Man's key (1911) in the more slender and its merus does not bear the Macrochirus Group, this species keys out to two movable spines found in the type. the two species C. faceta (de Man) and C. However, the second male, from the island splendida (Coutiere). It can be separated easily of Hawaii, shows a few notable differences. from these two species on the basis of the The rostrum, the stylocerite, and the lateral breadth of the articles of the second and third spine of the basicerite are all shorter, with the legs; by the greater breadth of the large chela; rostrum and the stylocerite reaching only to and especially by the lack of the longitudinal near the end of the first third of the second groove on the palm of the chela (characteris­ antennular article, and the lateral spine of the tics taken mostly from de Man's key; Cou- basicerite reaching only to the middle of the tiere's original description, without figures, of visible portion of the first article. The orbital C. splendida was too brief to be of any value). teeth are half the length of the rostrum, and This species is apparently identical to Al- the orbital hoods are slightly less inflated. pheus sp. of Coutiere {loc. cit~.) which was left The large chela is 2.7 times as long as broad unnamed because his three specimens each instead of 2.2, and the fingers occupy slightly lacked the large chela. The form and propor­ over 0.5 of the length of the chela instead of tions of the anterior body region and the pro­ slightly more than 0.25. The ratio of the car­ portions of the second legs, both shown in pal articles of the second legs is slightly differ­ the plates, are identical to this species; the ent—10 : 10 : 8 : 6 : 16, with the second arti­ unique dactylus of the third legs was neither cle slightly broader than long. The merus of mentioned nor drawn. If the dactyli are the the third legs is 3-5 times as long as broad in­ same, the specimens undoubtedly belong to stead of 3.0 times, and the propodus of the this species. Incidentally, Coutiere believed same legs bears four instead of five spines. that the species would belong to the Mega- Other parts of the two specimens showed cheles Group; the large chela shows its affini­ close similarity. ties to be with the Macrochirus Group. This male appears to be either within the normal range of variation or to be a closely DISTRIBUTION: This rare species is evidently related variety. It is likely that all the differ­ widespread among the larger Hawaiian Is­ ences may be due to simple growth differences lands at least, for the four specimens are from combined with the inherent variation as shown three islands. The type came from water 20 in so many other species; in any case, the dif­ feet deep off Waikiki Reef, Oahu; an 8.8 mm. ferences are not great enough, especially with female came from 15-20 feet of water off only one specimen showing them, to warrant Nanakuli, Oahu; an 8.0 mm. female came the erection of a new species or variety. from 4 feet of water at Lahaina, Maui; and the aberrant 6.5 mm. male was collected by R. W. RELATIONSHIP: This species is closely re­ Hiatt from a head of Pocillopora meandrina lated to two other Hawaiian species, C. gra­ Verrill in 5 feet of water at Halape, Kau Coast, cilis var. simplex Banner and C. edmondsoni Ban­ Hawaii. Coutiere's specimen came from Hu- ner. The more conspicuous differences are lule Male Atoll, Indian Ocean. given in Table 3- Of these characteristics, the most readily Crangon ventrosa (Milne-Edwards) observed and decisive for rapid separation of this species are the carpus of the second legs Fig. 28 a-i and the dactylus of the third to fifth legs. The Alpheus ventrosus Milne-Edwards, Histoire Na- anterior region of the carapace also affords turelle des Crustaces, 2: 352, 1837. good characteristics, but it is more variable Alpheus laevis Randall, Acad. Nat. Sci. Phila., and therefore less reliable. Jour. 8(1): 141, 1839- Hawaiian Crangonidae — BANNER 85

FIG. 28. Crangon ventrosa (Milne-Edwards), a, b, Anterior region, dorsal and lateral aspects; c, large chela of female, lateral aspect; d, large cheliped of female, medial aspect; e, small cheliped, lateral aspect; /, second leg; g, third leg; h, i, third leg, dactylus, inferior and posteroinferior surfaces. (In drawings h and /', the portions shaded with dashed lines are of thin chitin. Appendages drawn from several specimens, so relative size is not indicated.) (a-g, scale A; h, i, scale B.) 86 PACIFIC SCIENCE, Vol. VII, January, 1953

Alpheus obesomanus Boone, Vanderbilt Mar. long as broad; articles with ratio 10 : 8 : 6 : Mus., Bui. 6: 135, 1935. [Part of distribution 6 : 12. records, only.] Ischium of third legs inermous. Merus in- ermous, over 3 times as long as wide, widest DESCRIPTION: Rostrum slender, acute, at 0.7 of length then tapering sharply to tip. reaching almost to end of first article of an- Carpus with inferior margin continued as tennular peduncle, straight to depressed at small tooth. Propodus with 5 to 7 movable tip; small groups of short bristles along lower spines, each with broadened and flattened tip. margin. Rostral base broad, flattened, not Dactylus heavy, blunt, laterally compressed, carinate dorsally, separated from carapace by with thick longitudinal ridge on anterior face deep and narrow sulci on each side. Orbital continuing around tip as transparent flange, hoods rounded laterally, extending anteriorly blunt tip carrying "fingernail" of hard chitin, as acute teeth from inner portion of hood, tip reinforced with another ridge of chitin in 0.3-0.5 length of rostrum. form of horse's hoof. Inferior and lower pos­ Antennular peduncle slender, with second terior portion of face of dactylus proximal to article variable but usually only slightly longer "nail" of soft and flexible chitin (see Fig. than visible portion of first or third. Stylocer- 28£, /). ite well developed, reaching past middle of Body in general large and strongly com­ second antennular article. Lateral spine on pressed. Large females in collection range to basicerite acute, equal to or exceeding length 42 mm. in length. Usually orange-red on up­ of stylocerite. Scaphocerite long, distally nar­ per surface, sometimes longitudinally striped row, with lateral spine only slightly longer with deeper red along dorsal surface of cara­ than squamous portion and reaching slightly pace and abdomen; mottled with spots of beyond antennular peduncle and tip of car- deeper red along superior and superolateral pocerite. portions of both large and small chela. Large chela usually less than 2.5 times as DISCUSSION: The usual variation was noted long as broad, compressed, with margins in this species; much of it has been remarked rounded, without grooves or crests. Dactylus upon in the description. Additional variation elongate, not strongly arcuate, 0.5 length of was found in the proportions of the small palm, tip usually acute in males, rounded in chela; in the ratio of the lengths of the carpal females. Merus with both superior and in­ articles of the second legs, especially in the ferior internal angles obtuse, although pro­ fifth article which varied from noticeably jecting; inferior internal margin with about shorter than the first to somewhat longer; and 5 spines. in the number of spines on the propodus of Small chela almost as long as larger but the third legs. The fingers of the large chela more slender, 2.5-3 times as long as broad. show sexual dimorphism, but a few specimens Dactylus strongly curved at tip, crossing fixed seemed to intergrade. One specimen in the finger when closed. Inner margin of cutting collection had an orbital tooth on the right face of fixed finger with distinct lamellar ridge but not on the left side. along entire length, series of short strong bris­ These specimens agree well with the brief tles along outer edge. Fingers about equal in description given by Milne-Edwards. De Man length to palm. Merus with superior distal (1911) remarked on the variation manifested angle subacute, often with several movable by the few specimens of this species in his col­ spines on inferior internal edge. lection and noted that the second article of Carpal articles of second legs totaling 1.7 the antennular peduncle varied from slightly times length of merus and broad in propor­ shorter than the first article to over 1.5 times tion to width; second article only 1.5 times as as long, the rostrum from shorter than to Hawaiian Crangonidae — BANNER 87 equal to the first peduncular article, and the from the Gulf of Akabah at the western tip of scaphocerite from shorter than to as long as the Red Sea to the Gulf of California in Mex­ the antennular peduncle. De Man also sug­ ico. Edmondson has reported the species from gests that the two forms of the large chela Johnston, Wake, and Palmyra Islands, and (one with a rather obtuse dactylus, the other others have reported it from Tahiti. with an acute dactylus) might be a sexually As C. obesomanus (Dana), Boone (1935: 135) dimorphic trait; in general this was true in the reported the following previous distribution­ Hawaiian specimens. al records (in part): "Gulf of California (Cou­ DISTRIBUTION: C. ventrosa, in Hawaii at tiere) ; Sandwich Islands (Randall); Honolulu least, has always been found in living heads (Bate); . . . Sandwich Islands and Fiji Islands of corals of the genus Pocillopora, usually (Dana); . . . Gulf of Aykab [sic] and through either P. meandrina var. nobilis Verrill or P. li- the Red Sea (deMan)...." She mentions none gulata Dana, and occasionally in P. damicornis of the records noted above for C. ventrosa (Linnaeus); in these corals it associates with (1935: 126). Inasmuch as the records here Synalpheus charon (Heller) and crabs of the quoted from Boone were made by the authors genus Trapezia. to which she attributes them not for C. obeso­ This, although not the most common Ha­ manus but for C. ventrosa {A. laevis), it would waiian species, is certainly the most spectacu­ appear that these records have been inter­ lar, both in color and in size, and consequent­ changed. However, other records listed by her ly there have been many records of it through­ under C. obesomanus are correct. out the Islands. The first record was under the name of A. laevis made by Randall in 1839; it Crangon amirantei (Coutiere) was also reported by Dana (1852), Stimpson Fig. 29 a-j (1860), and Bate (1888). In the Islands it Alpheus amirantei Coutiere, Soc. Philomath. probably occurs wherever Pocillopora is found; Paris, Bui. IX, 11(5): 15-16, 1908. specimens studied have come from the fol­ Alpheus amirantei Coutiere, Linn. Soc. London, lowing localities: on Oahu, both in shallow Trans. II (Zoology), 14(4): 421, pi. 63, fig. water and in water up to 20 feet or more deep, 16, 1931. [Text exactly the same as 1908 at Waimanalo, Hanauma Bay, Waikiki, Nana- kuli, and Kawela Bay; on Maui, at Makena reference; plates not previously published.] and Lahaina. Previously unreported specimens DESCRIPTION: Carapace with swollen, al­ collected by the "Albatross" came from the most hemispherical orbital hoods, anterior following locations: Honolulu (further data margin of hoods produced as rounded, sub- unrecorded); south of Molokai, 23-24 fath­ vertical keel slanting mesad ventrally, area be­ oms; two dredge hauls on Penguin Bank tween keel and rostrum flat. Frontal margin (west of Molokai), 27-29 fathoms; and off anterior to orbital hoods almost straight; small Kauai, 18-41 fathoms. "Albatross" Stations: notch near base of rostrum. Rostrum small, (not recorded), 3847, 4031, 4032, and 4023, extending less than 0.5 length of visible por­ respectively; U.S.N.M.: 63633, 63634, 63636 tion of first antennular article, with high ros­ (63637, 63638—lot divided into two vials), tral carina only slightly lower than orbital and 63635, respectively. Edmondson (1925) hoods. Deep depression between rostral base has reported the species from French Frigate and orbital hoods. Side of carapace between Shoal and Laysan Island; there are 15 speci­ base of antenna and orbital hood depressed. mens in Bernice P. Bishop Museum from Visible portion of first antennular article as Pearl and Hermes Reef. long as second; third article about 0.75 length Crangon ventrosa has been reported through­ of second; second article about 1.5 times as out the entire tropical Indo-Pacific region, long as broad. Stylocerite very short, reaching 88 PACIFIC SCIENCE, Vol. VII, January, 1953

only slightly beyond 0.7 of length of visible movable spines. Dactylus simple, strong, and portion of first article, and only slightly ex­ slightly curved, about 0.25 as long as merus. ceeding broadened portion. Outer flagellum Telson 1.5 times as long as broad at base, heavy, short, not over 1.5 times length of anterior breadth 1.7 times breadth at posterior peduncle. margin. Lateral margins slightly convex an­ Basicerite with strong, acute spine. Scapho- teriorly and slightly concave posteriorly; pos­ cerite strong, outer margin curved, spine dom­ terior margin arcuate. Length of larger pair of inant over reduced squamous portion; tip posterior spines approximately equal to reaching beyond end of antennular peduncle breadth of tip; dorsal spines as long as larger and carpocerite. Carpocerite as long as anten­ posterior spines, located approximately 0.3 nular peduncle. and 0.7 of distance from anterior end of tel­ Proximal portion of palm of large chela sub- son; distal border without spinules. Inner uro- cylindrical in section, as thick as high, entire pod with numerous strong spines on distal chela less than 3 times as long as thick; distal- and distolateral margins. ly tapering, more flattened. Chela proximal to Largest female 10.0 mm. long; color in life articulation of dactylus slightly inflated; lat­ not noted. eral face proximad of articulation with poorly DISCUSSION: In the four specimens avail­ denned but noticeable longitudinal groove. able, no marked variation was noticed. Unfor­ Dactylus of usual form, about 0.3 length of tunately, a second female did not have a small entire chela. Merus twice as long as broad, chela, so it was impossible to determine superior distal margin rounded, inferior inter­ whether the irregularity of the superior mar­ nal margin with strong sub terminal tooth, gin of the merus noted above was actually a inermous. Large chela sexually dimorphic in sexually dimorphic characteristic or merely an size, that of female 0.8 and that of male 1.4 individual variation. times as long as carapace. This form differs in several respects from Small chela sexually dimorphic in size and the type as described by Coutiere. In the first proportions. Appendage of female with fin­ place, in Coutiere's specimens the scaphocer- gers 0.4 as long as chela; chela 3 times as long ite was not longer than the carpocerite, and as broad. Carpus elongate, equal in length to both were longer than the antennular pedun­ fingers. Merus twice as long as broad, 0.8 as cle instead of as described here. The basicerite long as chela; superior margin irregular; in­ is described as, "a peu pres inerme," whereas ferior internal margin without tooth. Append­ in this form the basicerite has a relatively age of male about 0.3 larger proportionately strong spine. The greatest differences are in than that of female, with fingers 0.6 length of the small chelae; the differences in propor­ chela; chela 3 times as long as broad. Carpus tions of the finger and the palm between the 0.5 as long as fingers. Merus similar to that of specimen drawn by Coutiere and the one female, but 0.7 as long as chela and with su­ drawn here are merely sexual dimorphism; perior margin regular. however, Coutiere shows his specimens as Carpal articles of second legs with ratio carrying a continuous band of setae which are 10 : 23 : 5 : 5 : 10. lacking in all Hawaiian specimens, and he Third legs with strong movable spine on states that the merus bears a strong tooth ischium. Merus 2.6 times as long as broad, ("epine") which is lacking in the Hawaiian with very strong subterminal tooth. Carpus specimens. The second article of the carpus 0.5 as long as merus, with inferior distal mar­ of the second legs is over three times as long gin projecting as acute tooth; article without as the first, and in the local specimens it is movable spines. Propodus only slightly long­ only 2.3 times as long. Finally, the carpus and er than carpus, heavy, with about 10 strong propodus in Coutiere's specimens were slight- Hawaiian Crangonidae — BANNER 89

FIG. 29. Crangon amirantei (Coutiere). a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, ventrolateral aspect; e, large cheliped, merus, medial aspect;/, small cheliped, lateral aspect; g, second leg; h, third leg; ;', third leg, dactylus;/', telson and uropod. (Drawn from a female.) {a, b, i,j, scale A; c-h, scale B.) 90 PACIFIC SCIENCE, Vol. VII, January, 1953 ly longer in relation to the merus than they reaching beyond middle of second article. were in these specimens. Outer flagellum not bifurcate. These differences cannot be considered as Basicerite of antenna very heavy, reaching of major magnitude because in related species beyond end of first article of antennular pe­ (see C. paragracilis, C. brevipes, and C. clypeata) duncle, 0.3 thicker than long. Lateral spine of all of these characteristics were found to be basicerite acute, extending to tip of stylocer­ quite variable. Possibly the most important ite. Scaphocerite with squamous portion difference is the lack of the tooth on the merus broad and rounded, 0.8 as long as lateral of the small cheliped, but, even if this were spine; tip of lateral spine not reaching end of constant, it would not be sufficient for the third article of antennular peduncle. Carpocer- erection of a new species. ite heavy, subcorneal, almost 0.3 longer than DISTRIBUTION: All five specimens in the antennular peduncle. collection came from Oahu: two from Hana- Labrum massive. Incisor process of mandi­ uma Bay, 20 feet deep, one off Waikiki, 20 ble expanded, covering inferior portion of feet deep, and two from Nanakuli, 15 feet labrum and bearing numerous strong teeth; deep. mandibular palp also expanded, with strong Coutiere's two specimens, both males, came bristles in addition to teeth similar to those on from Amirante, 25-80 fathoms, in the Indian incisor process. Maxillule and maxilla normal. Ocean. First maxilliped with endite of basipodite ex­ tended and contours rounded; no visible ar­ Crangon nanus sp. nov. ticulations in endopod. Epipodite of second Figs. 30 a-n, 31 a-h maxilliped enlarged. Third maxilliped as shown in Figure 30c. Alpheus sp.?, Metalpheus, n. gen., Coutiere, Soc. Philomath. Paris, Bui. IX, 9(5): 23-26, Large chela strongly compressed laterally, 1908. twice as long as broad. Margins entire except Alpheus sp.?, Metalpheus, n. gen., Coutiere, for several small rounded eminences on in­ Linn. Soc. London, Trans. II (Zoology) ferior edge where bristles arise. Outer face 17(4): 419-421, pi. 62, fig. 15, 1921. [The with 2 shallow grooves extending 0.25 of same description as above, but with fig­ length proximally, superior starting at point ures.] of articulation of dactylus, inferior near socket for piston of dactylus. Palm almost 4 times as TYPE SPECIMEN: A female 9-0 mm. long, long as fingers. Dactylus laterally disposed, collected from coral in water about 10 feet upper margin strongly arcuate and closing deep at Hanauma Bay, Oahu (U.S.N.M. across end of chela. Fixed finger very short, 93458). with terminal indentation to accommodate DESCRIPTION: Rostrum short, acute, barely dactylus. Chela with scattered setae and bris­ reaching end of first antennular article, about tles distally on inner face and on proximal twice as long as wide at base, without distinct portion of dactylus. Merus over 0.3 as long as carina, separated from orbital border by shal­ chela, 1.5 times as long as broad, inermous, low rounded concavities. Orbital border pro­ triangular in section, with inferior margins jecting forward, rounded tips 0.7 as long as rounded. rostrum. Orbital hoods inflated. Small chela more than 0.7 length of larger, Antennular peduncle short and thick, with fingers as long as palm. Palm 1.5 times as long second article about 0.7 length of visible por­ as high. Fingers curved, hooked tips crossing, tion of first and slightly more than 0.7 length immovable finger with thin knife-like ridge of third; second article about 1.3 times as on inner edge. Distal portions of palmar edges broad as long. Stylocerite with tip of spine and fingers with scattered groups of short Hawaiian Crangonidae — BANNER 91

FIG. 30. Crangon nanus sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, third maxilliped; d, e, large cheliped, lateral and medial aspects; /, g, small cheliped, lateral and medial aspects; h, second leg; i, third leg; j, third leg, dactylus; k, fifth leg; /, fifth leg, dactylus; m, telson and uropods (c-g from a paratype). (a, b, j, /, m, scale A; c, h, i, k, scale B; d-g, scale C.) 92 PACIFIC SCIENCE, Vol. VII, January, 1953 setae and bristles. Carpus slightly longer than tles; inferior tooth of dactylus feebly devel­ that of large cheliped and with rounded pro­ oped. jection on inner distal edge. Merus similar in First pleopod of female reduced and with­ form and proportions to that of large out articulations in both rami; endopod lack­ cheliped. ing appendix interna; second and following Ischium of second legs almost as long as pleopods with endopods bearing separate merus. Carpal articles together 1.2 times broad lobes for appendix interna. Second pleo­ length of merus, with ratio 10:6:4:3:8. pods of male of peculiar form (see Fig. 31^). Fingers of chela as long as palm. Entire ap­ Telson 1.7 times as long as broad, tip about pendage shorter than third legs. 0.5 as broad as base; margins slightly convex; Third legs without spines on ischium. Me­ dorsal spinules at 0.4 and 0.7 of distance from rus tapering, inermous, 3 times as long as anterior articulation. wide. Carpus 0.4 as long as merus, without Branchial formula without mastigobranch spines but with both margins continued dis- on base of seventh thoracic appendage and tally to form rounded tooth. Propodus arcu­ setobranch on eighth; pleurobranch of third ate, 0.7 as long as merus, with 6 movable also lacking. spines. Dactylus biunguiculate with inferior Mature individuals of both sexes ranged hook larger in diameter but slightly shorter in from 6 to 10 mm. in length. Color in life not length than distal hook. Fourth legs similar recorded. but larger. Fifth legs with proximal 3 articles DISCUSSION: The 20 specimens of this spe­ shorter and of slightly different proportions; cies show little variation. One peculiar tend­ propodus and dactylus equal to those of third ency, similar to that observed in C. paragra- legs; brush of propodus of but 2 rows of bris­ cilis, is for the anterior portion of the body to

FIG. 31. Crangon nanus sp. nov. a, Second pleopod of male; b, first pleopod of female; c, d, e, endopods of second to fourth pleopods of female;/, anterior region of specimen showing slight protrusion of eyes; g, h, anterior region of specimen showing extreme protrusion of eyes, (a-e, scale A; f-h, scale B.) Hawaiian Crangonidae — BANNER 93 be displaced, exposing the eyes as is shown in lationship of the lengths of the carpal articles Figure 3lf-h. Five specimens of this species of the second legs are roughly the same; even show this condition to some degree. The pos­ the curvature of the propodus is similar in sible causes of the condition are discussed both species. The branchial formula is similar under C. paragracilis. and evidently different from all other mem­ Normally occurring variations were slight. bers of this genus. In one specimen the orbital hoods are equal In fact, there is sufficient similarity between to the rostrum in length. Slight differences in the two species so that they could easily be the proportions of the carpal articles of the confused unless they were closely examined. second legs and in the proportions of the ar­ The chief differences lie in the following char­ ticles of the third legs were observed but all acteristics: the large cheliped, in which the articles exhibited a generally consistent if not chela has the inferior margin entire and the exact relationship one to another. Finally, the merus unarmed in C. nanus, but with the chela articles of the antennal peduncle between the bearing a shoulder on its inferior margin and basicerite and the carpocerite are or are not the merus with spines on the inferior internal visible in lateral view, the differences evident­ margin and a spine on the superior distal ly caused by the degree of rotation of the car­ angle in C. paragracilis; the pleopods, which pocerite. are of the usual form in C. paragracilis and as RELATIONSHIP: Because of the close rela­ shown in Figure 31 in C. nanus; the merus of tionship this species bears to C. paragracilis the third leg which is unarmed in C. nanus (Coutiere) it has been placed in the Macro- and bears an acute tooth in C. paragracilis. chirus Group. According to de Man's key Furthermore, in C. nanus the scaphocerite is (1911) this species is a relative of C. macro- shorter than the antennular peduncle, the me­ chirus (Richters). However, it does not appear rus of the second leg is equal in length to the to be closely related when the characteristics first four carpal articles, and the inferior un­ of the latter species are considered, for macro- guis of the dactylus of the third leg is normal­ chirus has deep orbitorostral grooves, the sec­ ly about 0.75 as long as the distal unguis, ond antennular article is three times the length whereas in C. paragracilis the scaphocerite is of the third antennular article, and the fingers longer than the antennular peduncle, the me­ of the large chelae are proportionally larger. rus of the second leg is equal in length to the Certainly, of all Hawaiian species this spe­ first three carpal articles, and the inferior cies is most closely related to C. paragracilis unguis of the third legs is 0.25-0.3 the length (Coutiere) and to the close relative of C. para­ of the distal claw. The differences between gracilis, C. hawaiiensis Edmondson. As C. ha- the two species appear to be constant. waiiensis is distinguished from C. paragracilis There is one other description in the litera­ principally in the form of the large chela, as ture that appears to be of the same or a closely C. paragracilis is better known, and as C. para­ related species, that of "Alpheus sp.? Metal- gracilis is probably the parent species to both pheus n. gen.?" of Coutiere {loc. cit.). There this species and C. hawaiiensis, C. nanus is com­seems to be some confusion about this spe­ pared chiefly to C. paragracilis. In the two spe­ cies, for while Coutiere states it shows rela­ cies the form of the anterior carapace is almost tionship to C. paragracilis and considers it as exactly the same; the compact antennular pe­ possibly identical to the unfigured C. rostra- duncle and the stylocerite are similar; the long tipes (Pocock, 1890), he also considers it as a and heavy carpocerite is similar; the peculiar representative of a new genus. mouthparts are similar; the general shape of C. nanus is similar to "Alpheus sp." in many both the large and small chela are similar important respects: the form of the orbital (specific differences are noted below); the re­ hoods, anterior carapace, antennules and an- 94 PACIFIC SCIENCE, Vol. VII, January, 1953

tennae, mouthparts (with the exceptions described as having the superior margin pro­ noted below); small chela (the large chelae duced into a "conspicuous tooth," a descrip­ were lacking in Coutiere's specimens), second tion that would not suit the subacute projec­ and following legs, including the important tion of this species; finally, the articles of the dactyli; and the peculiar pleopods of the carpus of the second legs in C. rostratipes have male and female. Even the tendency toward the ratio 10 : 5 : 2 : 3 : 4, instead of 10:6: protuberant eyes, noted above, was found in 4:3:8. Coutiere's specimens, as in the plate the me­ Returning to the contention of Coutiere dial portion of the eyestalks appears beyond that his specimen may represent a distinct new the anterior margin of the carapace. genus, he states: C. nanus differs from Coutiere's specimen in Volume des fouets antennaires et de l'antenne, incom­ the rostrum, which is poorly demarked lat­ plete protection des yeux, volume du labre, forme tres e e erally even at the edge of the carapace and speciale des mandibules, du 2 maxillipede, du 3 max­ illipede, des pinces de la le paire (?), brievete de la which has its tip depressed in lateral view; 2e paire, reduction du nombre des epipodites, forme in the endopod of the third maxilliped, in tres speciale des pleopodes, tels sont les caracteres que which the basal article is narrower, the second Ton pourrait invoquer pour la separation de YA. rostra­ article broader, and the third article bearing a tipes et des formes amnes. Le nouveau genre pourrait distal tuft of setae; and in the small chela, recevoir le nom de Metalpheus s'il etait conserve. which lacks spines on the lower margin of the The prime basis for separation, not empha­ sized in this paragraph, was the supposition palm. In Alpheus sp. the rostral ridge is still of Coutiere that the large chela (lost in his discernible between the orbital hoods, the specimens) was similar in form to the small rostral tip is not depressed, and the palm of chela. This was an erroneous assumption. the small chela bears five (?) spines on its in­ Of these characteristics, all save the short­ ferior margin. When the wide range of varia­ ness of the second thoracic legs and the modi­ tion in the closely related C. paragracilis is fied pleopods are found in C. paragracilis as considered, these minor differences certainly well as C. nanus. The similarity of the two do not appear to be of specific worth. species, as pointed out above, is especially The possibility suggested by Coutiere that marked in the peculiar mouthparts and bran­ C. nanus is a synonym of C. rostratipes (Po- chial formula. This relationship is so close cock) can be ruled out. In the first place, Po- that C. paragracilis and C. nanus must be in­ cock's specimen came from Fernando Noron- cluded in the same genus. However, as C. ha in the Atlantic Ocean off Brazil; while this paragracilis shows definite affinities to other does not exclude the possibility that the two members of the genus Crangon, I do not be­ species are identical, it certainly reduces the lieve that the erection of the proposed genus probability, for almost no species of this would be justified. genus found off the Atlantic coast of the DISTRIBUTION: Most of the specimens of Americas is also found in the Indo-Pacific this species came from coral heads in water area. Anatomically, several characters separate 10-20 feet deep immediately outside the shal­ the species in spite of the fact that Pocock low portions of the reef. The few specimens published only a short description without from near the intertidal zone were collected plates: C. rostratipes is described as having the at the following places: from algal holdfasts rostrum "springing from center of a depres­ at Kawela Bay and on the open coast east of sion," which certainly could not be applied to Koko Head, on Oahu; from coral at Kalama the Hawaiian form; the articles of the anten- Park and at Lahaina, on Maui. The specimens nular peduncle are described as being equal from deeper water came from the following in length, whereas in C. nanus the third article localities: off Mokulua Islands, Hanauma is the longest; the merus of the small chela is Bay, Waikiki Reef, and Nanakuli, on Oahu; Hawaiian Crangonidae — BANNER 95

in the deeper water offKalama Park, on Maui. Ischium of third legs 0.5 length of merus, Coutiere did not record the locality in the unarmed; merus 3.3 times as long as broad, Indian Ocean where his specimens were col­ with small acute tooth or smaller obtuse pro­ lected. jection; propodus 0.3 as long as merus, almost twice as wide distally as proximally, with Crangon hawaiiensis Edmondson neither inferior nor superior margins project­ Fig. 32 a-h ing but terminating at approximately right angles without spines or teeth; propodus 0.9 Crangon hawaiiensis Edmondson, Bernice P. length of merus, curved, with 6 small spines; Bishop Mus., Bui. 27: 14, figs. 2a-i, 1925. dactylus biunguiculate, curved and heavy, less DESCRIPTION: Rostrum triangular with nar­ than 0.4 length of merus, 2.2 times as long as row acute tip, about 1.5 times as long as broad broad at base, with distal unguis longer but at base, tip probably reaching to near end of less broad at base than inferior unguis. first antennular article. Orbital hoods prob­ Sole specimen known female 12 mm. long. ably subacute, pointed, and somewhat shorter DISCUSSION: The type specimen was re-ex­ than rostrum. Orbitorostral grooves shallow. amined and redrawn for this study. Unfor- First antennular article heavy, probably with visible portion slightly longer than sec­ ond article. Second article about 1.3 times as broad as long, somewhat shorter than third article. Stylocerite reaching to or slightly be­ yond end of first antennular article. Lateral spine of basicerite longer than stylocerite. Lateral margin of scaphocerite slightly con­ cave; lateral spine 0.2 longer than squamous portion; tip reaching beyond end of anten­ nular peduncle. Carpocerite heavy, reaching about to end of scaphocerite. Mouthparts similar to C. paragracilis and C. nanus (see pp. 96 and 90). Large chela almost 2.5 times as long as broad, strongly compressed but not twisted, fingers occupying distal 0.3. Palmar surfaces smooth, rounded; slight concavity in inferior margin below and proximal to articulation of dactylus; faint groove on outer face of palm, extending short distance proximad from upper margin of dactylar articulation. Merus 1.8 times as long as maximum breadth; in­ ferior external margin rounded and not def­ inite; superior margin terminating in acute tooth. Scattered setae on inner face of chela FiG. 32. Crangon hawaiiensis Edmondson. a, Anterior and on inferior internal margin of merus. carapace, dorsal aspect (the right orbital hood is folded; the left is badly crumpled and its restored outline is Small chela lost. indicated by the dashed line); b, c, antennule and an­ Carpal articles of second legs with ratio tenna, dorsal and lateral aspects; d, large cheliped; e, second leg; /, third leg; g, third leg, dactylus; h, fifth 10 : 5 : 5 : 5 : 7; second article 1.1 times as leg, propodus and dactylus. (a-c, g, scale A; d-f, h, broad as long. scale B.) 96 PACIFIC SCIENCE, Vol. VII, January, 1953 tunately the type not only lacks the small inferior internal margin of the merus, by the chela, but the chitin of the anterior thoracic shorter spine of the stylocerite, and by the regions is so soft and distorted that the true ratio of the articles of the carpus of the second nature of the orbital hoods could not be as­ legs. It may .also be distinguished from C. certained. nanus by the shorter stylocerite and the longer Careful comparison of the specimen to the scaphocerite. original description and plate of Edmondson Schmitt (1939) has described a new subspe­ showed only two minor differences: first, cies, C. hawaiiensis clippertoni from Clipperton there was a faint longitudinal groove on the Island in the eastern Pacific. It was distin­ palm of the large chela that was not previous­ guished from the type by two characteristics: ly noticed; second, the inconsistency of the the supposed lack on the type of longitudinal specimen as to the armature of the merus of grooves on the palm and by the rostrum the third legs was not remarked upon. On this which in Schmitt's specimen did not reach last point the omission is probably of no great beyond the orbital hoods. The first of these moment, for in other species it has been ob­ differences is already removed, and it is likely served that occasional specimens may lack a that, if this species shows variation similar to tooth on one or more legs, as this one does, that of C. paragracilis, the other difference will yet the majority of the specimens will have it. be within the range of variation for the spe­ Of the Hawaiian species this stands mid­ cies. However, until such a time as more speci­ way between C. paragracilis (Coutiere) and C mens are examined, it would be best to leave nanus Banner. It is like both species in the the subspecies as it now stands. form of the rostrum and orbital hoods (prob­ DISTRIBUTION: The type locality for this ably), in the mouthparts, in the general form species is Lisianski Island, in shallow water. of the antennular and antennal bases, in the general form (but not the specific sculptur­ Crangon paragracilis (Coutiere) ing) of the large chela, in the general form of Fig. 33 a-h the third legs, especially in the curved pro- podus and the shape of the biunguiculate dac- Alpheus paragracilis Coutiere, Soc. Ent. de tylus. It is like C. paragracilis in the length of France, Bui. (7): 149-152, 1898a. the scaphocerite and in the tooth on the merus Alpheus paragracilis Coutiere, Fauna and Geog. of the third legs, although in this species it is Maid, and Laccad. 2(4): 883, pi. 76, fig. 22, more poorly developed. It is like C. nanus in 1905. [Description same as original; figures the heavy form of the basicerite although not added.] as heavy, in the breadth of the articles of the DESCRIPTION: Carapace produced into short carpus of the second legs, and in the lack of acute rostrum almost reaching end of first ar­ the deep sculpturing of the large cheliped. ticle of antennular peduncle. Rostrum and However, it can be easily distinguished carapace rounded dorsally, without carina. from these two species, as the lower concavity Shallow, rounded depression between base of of the chela of C. hawaiiensis is slight and rostrum and orbital hoods. Orbital hoods gradual, whereas in C. paragracilis it is delim­ rounded, without processes. ited posteriorly by a lobe, and in C. nanus the Stylocerite acute, extending to middle of entire depression is lacking; the second article second article of antennular peduncle. Pedun­ of the carpus of the second legs is broader cle proper short and heavy, with second arti­ than long instead of slightly or markedly cle shorter than visible portion of first or third. longer than broad, as it is in the other two Lateral spine of scaphocerite not over 0.2 species. From C. paragracilis it may be dis­ longer than squamous portion, as long as car- tinguished also by the lack of spines on the pocerite, definitely longer than antennular pe- Hawaiian Crangonidae — BANNER 97

FIG. 33. Crangon paragracilis (Coutiere). a, b, Anterior region, dorsal and lateral aspects; c, d, large cheliped, lateral and medial aspects; e, small cheliped;/, second leg; g, third leg; h, third leg, dactylus. {a, b, scale A; c-e, scale B; /, g, scale C; h, scale D.) 98 PACIFIC SCIENCE, Vol. VII, January, 1953 duncle. Outer margin of scaphocerite some­ in some other members of the genus. Like what concave. Lateral spine of basicerite ex­ most species studied, there is some variation tending beyond rostral tip. in the proportions of the appendages, for ex­ Large chela compressed, slightly over twice ample, the first 2 carpal articles of the second as long as high. Palm with 2 indentations on legs which vary in ratio from 10 : 5 to 10 : 7. outer face, longitudinal groove behind articu­ The tooth on the merus of the third legs varies lation of dactylus and vertical shoulder on in­ from a strong and acute tooth to a less well- ferior side slightly proximal to articulation. developed tooth with the point approaching Distal end of chela rotated almost 45° to an obtuse angle. The squamous portion of the proximal end, dactylus opening somewhat scaphocerite varies from broadly rounded as laterally. Fingers slightly less than 0.3 length shown in Figure 33^ to narrow as shown by of palm. In addition to usual bristles on im­ Coutiere for the type (1905, fig. 22). movable finger and dactylus, row of several When this variation is considered, there is groups of bristles on inner face near superior only one minor difference apparent between margin, occasional bristles on inner face of these specimens and those described by Cou­ chela. Merus with superior edge produced as tiere, the difference being that the propodus acute tooth, with series of 8-10 very short, of the third legs in the type bears only four conical spines along inferior internal margin. weak spines instead of the 6-10 in the Ha­ Small chela about 0.75 as long as large waiian form. chela, slender; fingers curved, about as long A peculiarity of this and the related species, as palm. Propodus with acute tooth at point C. nanus Banner, is that these species, seem­ of articulation of dactylus. Merus similar in ingly alone among the local crangonids, can form to that of large chela. have their eyes forced out from under their Carpal articles of second legs with ratio carapace. In the collection of both species 10 : 7 : 4 : 5 : 7. available about 5 per cent of the specimens Third legs with merus over 3 times, as long showed this condition, which is evidently as broad, with small acute tooth distally on caused by rough handling in collecting, a inferior margin; carpus with distal projections thing very easily done with specimens so of superior and inferior margins poorly devel­ small. Either too much pressure on the thorax oped, rounded; propodus shorter than merus, or too much pull on the antennules and an­ strongly arched, with 6-10 strong movable tennae will displace the anterior portion of the spines along inner edge; dactylus biunguicu- body so the eyes will be exposed partially or late, with both claws strong and acute, distal completely (see Fig. 31/-^ for C. nanus). As claw twice as long as and broader at base than there is usually no distortion of the body or inferior. the appendages accompanying this protru­ Branchial formula without mastigobranch sion, the condition could be confusing to an on seventh thoracic appendage, without seto- inexperienced worker. branch on eighth appendage, without auxil­ DISTRIBUTION: This is one of the most com­ iary pleurobranch at base of third maxilliped. mon species about the Islands and has been Largest specimens reaching 17 mm. in collected from most of the seaward reefs of length; body with bright blue and smaller red Oahu. It has been found in heads of coral, in chromatophores, blue predominating on coralline algae, and in the holdfasts of algae. body, especially on antennular and antennal Specific localities on Oahu where it has been bases, chelae, and caudal fan; tips of chelae collected, from water up to 20 feet deep, in­ yellow-green; eggs green. clude: Kaneohe Bay, Waimanalo, Hanauma DISCUSSION: This species shows some vari­ Bay, Black Point, Waikiki, Nanakuli, Moku- ations but they are not as marked as they are leia, Kawela Bay. On Maui, at a similar depth Hawaiian Crangonidae — BANNER 99 range, it was collected at Makena, Kalama of first, about twice as long as third (note that Park, Lahaina, and four localities ranging the elongation of the second article shown in from 12 miles south of Lahaina to 10 miles Figure 34^ is atypical). Stylocerite short and north of Lahaina. The only record of the spe­ variable with lateral spine not reaching end of cies from deeper water was a collection from first antennular article. Basicerite with acute 180 feet off Kaena Point, Oahu, where it was lateral spine of varying size, usually quite living in sponges. Edmondson (1925) has re­ small. Scaphocerite relatively narrow, with ported it from French Frigate Shoal, Pearl and strong lateral spine 1.3 times as long as squa­ Hermes Reef, Laysan, Lisianski and Ocean Is­ mous portion; lateral margins concave. Lat­ lands. The "Albatross" took only one group eral spine of scaphocerite usually longer than of specimens, those from Laysan Island, evi­ antennular peduncle, usually greatly exceeded dently from shallow water. by carpocerite. The species has been reported from Johns­ Large chela subcylindrical, tapering, with­ ton and Wake Islands (Edmondson, 1925) out grooves or lobes; distal portion slightly and from Tahiti in the central Pacific, and bent, thus lower surface concave in profile. from the Indian Ocean. It was not reported Fingers about 0.3 length of entire chela. Dac- by de Man from the Siboga Expedition ma­ tylus laterally disposed, with strongly arcuate terial, although the range of the species ex­ margin. Distal portion of inner face of chela tends to either side of the area investigated by sometimes sparsely covered with setae. Carpus that expedition. of usual form. Merus short, triangular, with superior margin ending distally as rounded to Crinita Group, Crinita Subgroup subacute projection, inferior internal margin Crangon paralcyone (Coutiere) with acute tooth distally. Fig. 34 a-i Small chela short, approximately 3 times as Alpheusparalcyone Coutiere, Fauna and Geog. long as broad. Fingers 0.6 as long as palm, Maid, and Laccad. 2: 895, pi. 80-81, figs. broad, flattened, lined with dense setae in 34a-h, 1905. male, narrower and conical in female. Palm Crangon laysani Edmondson, Bernice P. Bish­ somewhat triangular in section. Merus similar op Mus., Bui. 27: 17, figs. 3a-j, 1925. to that of large chela. Crangon bucephalus (Coutiere), Edmondson, Carpus of second legs variable, but usually Bernice P. Bishop Mus., Bui. 27: 14, 1925 with ratio 10 : 30 : 9 : 9 : 11. [at least in part]. Ischium of third legs with small blunt spine DESCRIPTION: Rostrum not reaching be­ (not shown in Fig. 34/6). Merus about 3-5 yond middle of visible portion of first anten- times as long as broad, with strong acute nular article, broad yet acute, with dorsal ca­ tooth on distal portion of inferior border. rina continuous from tip to behind orbital Carpus with 1-5 movable spines on inferior hoods. Orbital hoods with no trace of teeth, border, continued distally as acute tooth; su­ depressed anteriorly as flattened shelf expand­ perior margin with numerous long setae. Pro- ing as broad area between base of rostrum and podus with varying number, usually about 12, anterior portion of orbital hoods; frontal mar­ movable spines along inferior margin, most of gin of this area varying from arcuate to S- them in pairs, with many long setae on su­ shaped. Margin of carapace above base of an­ perior margin. Dactylus short, 0.2 as long as tenna higher than at base of the orbital hoods, propodus, with small secondary unguis ro­ forming slight but distinct depression. tated so as to be almost invisible, except when Antennular peduncle short, with second legs are turned slightly. Fifth legs with article somewhat longer than visible portion "brush" well developed. 100 PACIFIC SCIENCE, Vol. VII, January, 1953

Pleura of first abdominal segment showing the inferior internal margin is reduced and sexual dimorphism, in male ending posteri­ rounded (Fig. 34J). orly as dentate process that points slightly As noted by Coutiere, the small chela is posteriorly, in female rounded. Following sexually dimorphic, but the breadth of the pleura of usual shape. dactylus of the male is greater in proportion Telson 1.7 times as long as broad, tip 0.7 as in larger specimens, and in the largest speci­ wide as base; sides straight and posterior mar­ mens it is broader than that described by Cou­ gin somewhat arcuate; dorsal surface slightly tiere. depressed medially. Dorsal spinules large, pos­ In the second legs the second carpal article terior pair located near middle; posterior mar­ is usually nearly three times the length of the gin usually with series of about 10 very feeble first, but it varies more or less continuously spinules. Inner branch of uropod with strong from 2.2 to 4.0 times as long. The proportions spines on posterolateral margin. of the other articles also vary but not as much. Most specimens were 10 mm. long or less; The spines on the carpus of the third legs however, those specimens from Kaneohe Bay, usually are two in number but vary from one Oahu, listed below reach 17.6 mm. in length. to five. The spines along the inferior margin Color in life not observed. of the propodus vary from 8 to 13. Studies were not made on the size of the secondary DISCUSSION: This species, as well as the two unguis of these legs because it was so difficult other members of this subgroup from Hawaii, to rotate the leg to where the full unguis could has been found to vary considerably in char­ be seen; however, they, too, appear to vary, acteristics commonly accepted as of taxonom- being quite reduced in some specimens. ic importance. In none of the specimens examined does The ratio of the lengths of the articles of the the sixth abdominal somite bear spines. The antennular peduncle varies considerably, with telson was not measured over a series of in­ the second article varying from 1.4 to 2.0 dividuals, but examination did not show that times the length of the visible portion of the it varies greatly in general shape. How vari­ first article. The stylocerite varies from mark­ able the spinules on the tip of the telson are edly shorter than the first antennular article to could not be determined, as at best they are almost equal in length. exceedingly difficult to discern. In some speci­ The relative lengths of the antennular pe­ mens they may have been missing. duncle, the scaphocerite, and the carpocerite This form evidently has been described also vary. In some specimens the scaphocerite under two names: as C. paralcyone (Coutiere) is scarcely longer than the antennular pedun­ from the Maldives and Laccadives in the In­ cle, and in turn is greatly exceeded by the car­ dian Ocean and as C. laysani Edmondson from pocerite; in others the scaphocerite is almost the Hawaiian Archipelago. The two species equal to the carpocerite, and both extend far appeared to be separate and valid on the basis beyond the end of the antennular peduncle; of the original descriptions, but when the although the carpocerite is always longer than variation within the form is considered it ap­ the antennular peduncle, the amount by which pears that there are no important differences. it exceeds is variable. The lateral spine of the The original descriptions of the two species basicerite also varies in size and in some speci­ showed the following characteristics in com­ mens is so reduced that it is difficult to dis­ mon: the length of the stylocerite; the shape cern. of the front of the carapace except for the ros­ The large cheliped is constant in its general trum; the shape of the large and small cheli- form in all specimens but varies somewhat in peds; the shape and armature of the third legs; its proportions; in one specimen the tooth of and the shape of the telson. On the basis of

J Hawaiian Crangonidae •— BANNER 101

FIG. 34. Crangon paralcyone (Coutiere). a, b, Anterior region, dorsal and lateral aspects; c, basicerite, showing tooth (same specimen); d, large cheliped, lateral aspect; e, large chela, ventral aspect; /, small chela; g, second leg; h, third leg; ;', third leg, dactylus. (a-c, scale A; d, e, scale B; f-h, scale C; /, scale D.) 1

102 PACIFIC SCIENCE, Vol. VII, January, 1953

other characteristics, it appeared that the two differences can be anticipated to be bridged forms might be separated. For example, the by the study of additional specimens, it is ad­ relative lengths of the scaphocerite to the an- visable to place the species C. laysani as a syn­ tennular peduncle and carpocerite; the relative onym of C. paralcyone. length of the second article of the second The specimen from Laysan identified by thoracic legs, described by Coutiere as 2.43 Edmondson (1925) as C. bucephalus (Coutiere) times the length of the first and by Edmond- was re-examined and found to be a member son as 3.0 times the length of the first; the of this species. armature of the carpus of the third legs; and DISTRIBUTION: This species appears to be the presence of spinules on the posterior bor­ more abundant in deeper water than in shal­ der of the telson. However, all of these char­ low water. It has been collected at the follow­ acters are found to be variable and the differ­ ing locations in water up to 20 feet deep: at ences between the Indian and the Pacific forms Kaneohe Bay, Hanauma Bay, Kahala Beach, removed by the variation within the Hawaiian Waikiki Beach, off Nanakuli, on Oahu; at population. Makena, Lahaina, and 10 miles south of La- There are several differences that seem to haina, on Maui; at Halape, Kau Coast, Ha­ remain between this form and the form de­ waii, from coral at 35-40 feet deep. It was re- scribed by Coutiere: first, in the Indian Ocean TABLE 4 form, the rostrum is but slightly produced, DEEP-WATER COLLECTIONS OF Crangon paralcyone whereas in this form it usually reaches the end of the first article of the antennular peduncle; U.S.N.M. however, Coutiere has shown in his plates that LOCATION DEPTH STATION CATALOGUE NUMBER * NUMBER this varies in his specimens, and similar varia­ tion has been seen in the local specimens. Sec­ fathoms ond, the lateral spine of the basicerite is longer NE of Hawaii 29-26 4053 63623 26-50 4054 63619 in Coutiere's specimens than in the local speci­ 24-83 4061 63608 mens; however, in the latter it is quite vari­ Auau Channel 32-43 3872 63607 able. Third, the merus of the large chela was 34-65 3875 63617 28-43 3876 63609, 63612, shown by Coutiere without spines or teeth 63625 (although he described it as bearing "une forte Off Puaaloa, Maui 26-27 — — epine a son apex infero-interne"); in the Ha­ N of Maui 45-52 4070 63626 49-57 4075 63611 waiian form there is a pronounced and almost S of Molokai 23-24 3847 63618 always acute tooth. The external spine of the 44-73 3848 63613 uropod is shown to be bent in Coutiere's 43-73 3849 63621, 63624 Penguin Bank 27-29 4032 63616 plates, whereas in this form it is straight. Fi­ Oahu nally, de Man in his key (1911: 314) puts em­ Off Mokulua Island 20 — — phasis on the presence of two teeth on the 1 mile S of Kahala 25 — — Off Diamond Head 35 — — posterior dorsal margin of the sixth abdom­ Off Kaena Point 25 — — inal segment, teeth that are not found in the Off SW Oahu 6-60 — — Off Kauai 24-43 4024 63610 Hawaiian specimens; however, Coutiere re­ Off Kaula Rock, ported that, "... 2 prolongments epineux . . . SW of Niihau 35 — — pouvant parfois manquer ou etre tres peu Off Nihoa 26 4147 63622 27-29 4158 63614 marques." Off French Frigate Shoal 14-17 3968 63615 Therefore, as all the differences between the 17 3971 63620 Hawaiian forms and the type described by * Station numbers are those of the "Albatross"; Coutiere are either bridged by the variations collections without station numbers were made by the noted or are both minor and variable so the "Makua" and "Salpa." Hawaiian Crangonidae — BANNER 103

ported from shallow water at Laysan as C. lay- lected from a head of coral in 10 feet of water sani and C. bucephalus by Edmondson (1925). off Nanakuli, Oahu. A male paraneotype 9.6 It has been dredged at the locations given in mm. long has also been designated from the Table 4. same locality (U.S.N.M. 93459). The species has been reported from various DESCRIPTION: Anterior portion of carapace locations in the Indian Ocean and from the with dorsal carina high and thin, ending Netherlands Indies. abruptly at base of rostrum. Rostrum short, (?) Crangon paralcyone (Coutiere) triangular, seldom reaching past anterior mar­ A single 9 millimeter specimen, collected at gin of orbital hoods. Orbital hoods hemi­ Hanauma Bay, Oahu, in 20 feet of water, has spherical, continued forward as short, sub­ been referred to this species with doubts. It acute, subvertical keel that slants mesally agrees with C. paralcyone in the form of the toward ventral side; anterior margin of hoods rostrum and the orbital hoods, the stylocer- almost vertical when seen in profile, reaching ites, the spine on the basicerite (on one side higher than rostral carina. Area between or­ bital hoods and base of carina flattened, with only; the spine is lacking on the other side), anterior margin varying from concave to the relative lengths of the scaphocerite, car- straight. Lateral margins of carapace with in­ pocerite, and antennular peduncle; somewhat dentation between lateral margin of orbital in the form of the large and small chelipeds; hoods and covering of base of antennal pe­ in the relative lengths of the articles of the duncles. carpus of the second legs; and somewhat in the form and armature of the third legs. It dif­ Second article of antennular peduncle from fers in the following points: the articles of the 1.5 to 2.0 times as long as visible portion of antennular peduncle are broader; the dactylus first article; third article about as long as first. of the large chela is relatively shorter; the me- Stylocerite short, broad, lateral spine reaching ral tooth of the third legs is reduced to a very only 0.8 length of visible portion of first an­ short rounded protuberance, the spines of the tennular article. Basicerite unarmed. Scapho­ propodus are much longer, almost comb-like cerite narrow, tapering, squamous portion in form. about 1.1-1.2 times as long as lateral spine; lateral margin slightly concave. Scaphocerite Most of these differences could be individ­ usually slightly longer than antennular pedun­ ual differences in a variable species, the most cles and slightly shorter than carpocerite. notable being the dactylus of the large chela and the armature of the third legs which has Large chela without notches or grooves, the reduced tooth on the merus and the longer subcylindrical, tapering. Dactylus heavy, be­ propodal spines. It is possible that these too tween 0.25 and 0.3 length of entire chela, may be individual differences, or they may be strongly arched dorsally. Merus about 1.5 constant and adequate criteria to separate this times as long as broad, unarmed, distal end of specimen as a species. However, with only superior margin rounded. Chela sexually di­ one such specimen, and with the parent spe­ morphic in both size and proportions, in fe­ male noticeably shorter than carapace, in male cies showing such marked variation, it is con­ longer than carapace, with relatively greater sidered advisable to place the specimen in C. thickness and shorter fingers than in female. paralcyone, with reservations. Small chela about 3 times as long as fingers, Crangon brevipes (Stimpson) latter narrow and conical in both sexes; in Figs. 35 a-j, 36 a-e, 37 male fingers about twice as long as broad, in Alpheus brevipes Stimpson, Acad. Nat. Sci. female about 3 times as long as broad. Carpus Phila., Proc. 12: 30, I860. 0.3 length of chela. Merus 3 times as long as NEOTYPE: A female 13.4 mm. long, col­ broad but otherwise similar to large chela. 104 PACIFIC SCIENCE, Vol. VII, January, 1953

Carpal articles of second legs variable, but that of the anterior margin of the carapace be­ ratio usually within following range: 10 : 15- tween the ridges of the orbital hoods and the 21 : 5-7 : 5-7 : 9-12. rostrum; this varied from a concave margin in Third legs with spine on ischium. Merus which the tip of the short rostrum did not 2.7-3.1 times as long as wide, with strong sub- reach the anterior margins of the orbital apical tooth on inferior margin. Carpus with hoods, to a straight margin with the rostrum acute tooth on distal end of inferior margin, reaching beyond the hoods; in some speci­ without movable spines. Propodus with 10 mens the frontal region was asymmetrical. movable spines along inner margin. Dactylus Antennules and antennae: The observed range simple, curved, acute. in proportions of the articles of antennular Pleura of abdomen rounded in both sexes. peduncle is reported in the description. The Telson broad, flat, slightly tapering distally, relative length of the scaphocerite also varied, point of greatest width 1.3-1.8 times as wide but without exception it was no shorter than as posterior margin, length 2.7-2.9 times the antennular peduncle and no longer than breadth of posterior margin; both pairs of the carpocerite. In all specimens the basicerite dorsal spines heavy, 0.3 and 0.6 of distance lacked a lateral spine. from point of articulation. Chelipeds: Some of the variation of this sex­ In addition to usual setiferous bristles pos­ ually dimorphic appendage is indicated in the terior margin of telson with about 10 small description. It was also found to differ in the spinules, these difficult to discern. Inner degree of compression, with the breadth al­ branch of uropod also with series of strong most equaling the height in some specimens movable spinules, about 20, along postero­ while in others the breadth was only about 0.7 lateral margins. of the height. The palm was found to vary Specimens reaching 16.0 mm. in length; from 1.5 to almost 3.0 times the length of the color in life not observed. merus. In none of the specimens did the merus DISCUSSION: Inquiry at the major American have an acute tooth on the end of the inferior museums revealed no records of Stimpson's internal margin, but although the margin was type specimen, so it is presumed to have been straight in most cases, a few specimens were lost. Unfortunately from Stimpson's descrip­ found to bear a small obtuse and rounded pro­ tion it is not possible to determine the exact jection. type locality, for all that he specified was, "in- The small cheliped was quite uniform ex­ sulas Hawaienses; inter ramos madreporar- cept for the slight sexual dimorphism de­ um." The neotype meets those specifications. scribed above. All the specimens available were examined Second legs: The range of variation of the carefully, but a group of specimens from Wai- carpal articles has been given; a further dis­ kiki Reef which exhibited the usual range of cussion of this variation is given below where variation were studied in greater detail. In this C. brevipes is contrasted to C. clypeata (Cou- group, the variable characteristics were meas­ tiere). ured as accurately as possible and the results Third legs: These appendages were quite are here summarized. constant in armature and proportions, with Anterior region of the carapace: The dorsal the ischium always bearing a spine; the merus carina varied from broad and dorsally rounded varying from 2.7 to 3.0 times as long as broad, to narrow and acute; in no specimen did it that of the smaller specimens being more nar­ reach the anterior tip of the rostrum. The vari­ row than the larger; the merus always with-a ation in thickness of the anterior ridge of the strong distal tooth; the propodus always bear­ orbital hoods may be seen in the drawings. ing strong spines; the dactylus always simple. One of the most variable characteristics was However, in the smallest specimens (those Hawaiian Crangonidae — BANNER 105

FIG. 35. Crangon brevipes (Stimpson). a, Anterior region, dorsal aspect (asymmetry of rostrum and scaphocerite found in specimen); b, anterior region, lateral aspect; c, large cheliped, lateral aspect; d, large chela, ventral aspect; e, large cheliped, merus and carpus, medial aspect; /, small cheliped, lateral aspect; g, small cheliped, merus and carpus, medial aspect; h, second leg; /', third leg; j, third leg, dactylus. (a, b, scale A; c-i, scale B;/, scale C.) 106 PACIFIC SCIENCE, Vol. VII, January, 1953

C. clypeata (Coutiere), a species found in the same habitat; if that had been so, C. clypeata would have been relegated to synonymy. The two species agree in most characteristics: in the general shape and configuration of the anterior portion of the carapace; the propor­ tions of the antennular peduncle and antennal peduncle; in the large chela; in the third legs; in the telson. When a large number of specimens was studied, however, the following characteris­ tics were found to be reliable for their separa­ tion: 1. Sexual dimorphism of the small chela of males. In C. clypeata the dactylus of the small chela of males, although somewhat variable, is always noticeably broadened or subspatu- late and surrounded by dense setae; in C. brev­ ipes it is never subspatulate. 2. Tooth on basicerite. This tooth, or spine, is almost always present in C. clypeata and is always lacking in C. brevipes; however, in a few FlG. 36. Crangon brevipes (Stimpson). Variation in of the specimens of C. clypeata it is very re­ the frontal region of the carapace, antennules, and an­ duced, and sometimes it is entirely lacking. tennae; specimens drawn in a, b, and c came from Wai- 3. Tooth on merus of large cheliped. This kiki Beach, Oahu; the specimen in d and e came from Mokuleia, Oahu. {a, b, scale A; c, scale B; d, e, scale C.) tooth is usually large and acute in C. clypeata and entirely lacking in C. brevipes; however, about 5 millimeters long) the propodus bears this characteristic alone is not adequate to only 4-5 well-developed spines instead of the separate the species, as sometimes in C. usual 8-11 spines. clypeata it may be reduced and obtuse or Telson: The measured range of variation of rounded, or even entirely lacking, whereas the telson has been given. No attempt was in C. brevipes the margin may bear an obtuse made to determine whether the small spinules projection. on the posterior border of the telson or on the 4. The relative length of the second article of posterolateral margin of the inner uropod carpus of the second legs. In an attempt to find varied in number, as it was too difficult to the extent of variation in the relative lengths count them accurately. In all specimens that of the first and second carpal articles, they were studied carefully the spinules were pres­ were measured in 103 specimens of both spe­ ent. cies; the result of this study is shown in Figure These specimens agree perfectly with my 37. This shows that most of the specimens of interpretation of the short description given, C. clypeata have the ratio of the second to the without figures, by Stimpson; the description first articles of 1.0 and that most of the speci­ is adequate to separate this species from all mens of C. brevipes have the ratio of the same other species known to occur in the Hawaiian articles between 1.7 and 2.0. However, speci­ Archipelago. mens that on the basis of the other character­ Until this species was thoroughly studied, istics were definitely C. clypeata did reach the it was felt that probably it was identical with maximal ratio of 1.3, and similarly specimens Hawaiian Crangonidae — BANNER 107 that were definitely C. brevipes did reach the tical with this species. It was plainly a differ­ minimal ratio of 1.4. Therefore, it is likely ent species as C. parabrevipes had grooves and that, if enough specimens were measured, the ridges on its chela which are entirely lacking edges of the normal distributional curve would in C. brevipes. C. parabrevipes (Coutiere) was overlap. Another graph, not reproduced here, considered a synonym of C. acuto-femorata was made to ascertain whether the ratio be­ (Dana) by de Man (1911). tween the articles varied with the size of the DISTRIBUTION: The species appears to be specimens; in this the ratio of the lengths of moderately plentiful on Oahu, where it was the first two carpal articles was plotted against collected both in shallow water of the reef flat the length of the carapace. This study showed and outside the reefs in water to about 20 feet no correlation whatsoever, with both species deep at the following locations: off Mokulua having great variation in ratios for all size Islands, Waimanalo, Hanauma Bay, Waikiki, ranges. Nanakuli, Mokuleia, Kawela Bay. It was col­ On the basis of these four variable charac­ lected at Halape, Kau Coast, Hawaii, by R. teristics, satisfactory separations of the two W. Hiatt, in coral at a depth of 30-40 feet. species may be made. If, for example, the ratio The only record of the species from deep water of the first two articles of the carpus of the is two specimens collected in 30 fathoms off second legs was between 1.3 and 1.5, in cases Waikiki Beach. where it was undecisive, then the specimen could be assigned to its correct species on the Crangon clypeata (Coutiere) basis of the teeth of the merus of the large Figs. 37, 38 a-k, 39 a-u chela or of the basicerite. In none of the well Alpheus clypeatus Coutiere, Fauna and Geog. over 200 specimens examined was any doubt Maid, and Laccad. 2: 897-898, pis. 81-82, encountered as to the identity of an individual figs. 36-36g, 1905. specimen. DESCRIPTION: Frontal border, seen from The belief that these are indeed separate above, truncate and projecting, with margin species is confirmed by the study of paired from slightly concave to slightly convex. Ros­ specimens. Both of these species live in algae- trum either small or lacking. Dorsal carina covered tubes on old coral; in each tube there high and definite, sharp or rounded, anteriorly is a single pair of specimens, a male and a terminated abruptly before reaching edge of somewhat larger female which are evidently mates. During routine collecting, whenever it frontal area. Area between orbital hoods and was possible to get both members of this pair, they were removed and preserved for special study. In the small number of these pairs available, none consisted of mixed species— either both were plainly C. brevipes or both were plainly C. clypeata. This endemic species apparently has been derived from the Indo-Pacific C. clypeata. As the differences between the species are neither constant nor of great magnitude, as the hab­ itat remains the same, the separation of the two species can be presumed to have been in •Ratio of second to first article relatively recent time. of carpus of second legs Coutiere described a species, C. parabrevipes FIG. 37. Comparison of carpal ratio of Crangon cly­ (189&z: 151), which he thought might be iden­ peata (Coutiere) and Crangon brevipes (Stimpson). 108 PACIFIC SCIENCE, Vol. VII, January, 1953 rostrum convex dorsally. Orbital hoods hemi­ Dactylus simple, curved. Fourth legs similar spherical. Depression between lateral margins to third legs except meral tooth less pro­ of orbital hoods and portion of carapace nounced. above base of antennae. Telson with sides of uniform taper 2.2-2.9 Antennular peduncle with second article 1- times as long as width of posterior margin, 2 times as long as visible portion of first arti­ greatest breadth 1.3-2.2 times as wide as pos­ cle; third article equal to or somewhat longer terior margin. Pairs of dorsal spines about 0.4 than visible portion of first; second article 1.5- and 0.7 distance from articulation. Posterior 2.0 times as long as broad. Short lateral spine margin between posterolateral spines arcuate, of stylocerite reaching nearly 0.7 length of vis­ with about 10 small spinules; inner branch of ible portion of first article. Basicerite almost uropod similarly with spinules along distolat- always bearing thin, acute lateral spine. Sca- eral margin. phocerite, antennular peduncles, and carpo- Mature specimens usually reaching maxi­ cerite subequal in length, all variable, but car- mum length of 18 mm.; one female, however, pocerite almost always slightly longer. 23 mm. long. Color of live specimens variable, Large chela of male subcylindrical, usually body transparent, covered dorsally with red, slightly compressed, entire and smooth, ta­ yellow, and blue chromatophores in reticu" pering toward fingers. Entire chela about 4 lated pattern; abdomen with reticulations in times length of fingers and about 2.5 times transverse bands; legs usually transparent; its greatest height. Chela rotated, with dac- large chela with more red and blue chromato­ tylus lateral in position. Merus about 1.5 phores than body; eggs bright green. times as long as wide, with superior distal DISCUSSION: Like C. paralcyone and C. brevi- angle not greatly produced; inferior internal pes, this species shows great variation in many margin usually with strong acute tooth dis­ characteristics. Some 10 specimens were tally. Chela relatively shorter and thinner in studied intensively to determine the range of female than in male. variation; unfortunately, time did not permit Small chela of male with dactylus expanded, the making of the complete series of measure­ about twice as long as broad, with dense ments for the almost 100 specimens available. fringe of short setae on margin; fingers slight­ The specimens not measured, however, were ly shorter than palm; height of palm slightly all examined to determine any marked varia­ over 0.3 length of chela. Merus twice as tion. long as wide, about 0.7 as long as chela, with The ranges of variation of the frontal region all distal angles rounded. In female dactylus of the carapace, of the antennular and antennal narrow, 3 times as long as broad, uniformly peduncles, of both chelipeds, of the third legs, tapering; palm about 1.5 times as long as fin­ and of the telson either have been given in the gers; chela slightly over 3 times as long as description or are shown in Figures 38 and 39. high; merus thinner than in male. The range of variation of the second legs is Ratio of carpal articles of second legs vary­ given in the description and shown graphical­ ing thus: 10 : 9-13 : 2.9-4.7 : 2.9-4.7 : 6.3- ly in Figure 37. A few additional variations are 10.0 (see Fig. 37). noteworthy: The lateral spine of the basicer­ Ischium of third legs armed with short ite, usually well developed, is sometimes re­ strong movable spine. Merus 3.1-3.5 times as duced or absent. The large chela, already long as wide, with strong tooth distally on noted as sexually dimorphic, also varies from inferior margin. Carpus prolonged distally on almost circular in cross section to distinctly inferior border as rounded tooth, otherwise compressed. The merus of.the large cheliped unarmed. Propodus broad and flattened, with almost always carries a well-developed tooth 9-11 movable spines along inferior margin. on the inferior internal margin, but in some Hawaiian Crangonidae — BANNER 109

FIG. 38. Crangon clypeata (Coutiere). a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, male, lateral aspect; d, large cheliped, merus and carpus, medial aspect; e, small cheliped, male;/, dactylus of e showing maximum breadth; g, small cheliped, female; h, second leg; z, third leg;/, third leg, dactylus; k, telson and uro- pods. (a, b, e-i, k, scale A; c, d, scale B;y, scale C.) 110 PACIFIC SCIENCE, Vol. VII, January, 1953 specimens it is reduced and rounded and in a As I have seen no specimen of this species, very few completely absent. On the third legs the original description is given. of very small specimens (less than 5.0 mm. Rostrum triangular, acute, arising from the frontal long) there are only 4-5 spines on the pro- margin of the carapace (which is slightly concave on podus instead of the usual 9~11- each side of its base), but not prolonged backward as a dorsal carina. Orbital arches entire, arcuated, without The observed range of variation of this spe­ spinules; anterior margin of the carapace sinuated on cies is great, and if a few specimens on the the sides, without spines. Postabdominal segments fringes of the normal curve of variation were smooth, entire, with the lateral margins broadly rounded; terminal segment not three times as long as examined alone they probably would be inter­ broad at the base, with its distal end subtruncated. preted as separate but related species. How­ Eyes completely concealed beneath the carapace. An- ever, in the large number of specimens exam­ tennules with three joints of the peduncle exposed, of which the middle one is slightly the longest, with a ined, the gaps between the extremes were small spine-like scale at base, reaching nearly to the end closed so the collection presented an almost of the basal joint. Basal scale of antennae about reach­ continuous spectrum of variation from one ing to the end of the antennal peduncle, with the outer margin convergent towards it and clothed with long extreme to the other. hairs. Anterior legs or chelipeds having the merus and The description of the type specimen of carpus slender; merus with a small tooth or spine at the Coutiere easily falls within the range of varia­ distal end of its under margin; palm of larger cheliped rather more than twice as long as broad, smooth, with­ tion noted above; there should be no doubts out notches or ridges, largest at its rounded basal end, as to the specific identification of this species. with an impressed curved line on its upper and proxi­ Locally this species can easily be confused mal end; fingers nearly half as long as the palm; the upper with its superior margin arcuated. Smaller cheli­ with C. brevipes; the separation of the two spe­ ped with the carpus rather longer, and chela very slen­ cies is discussed in detail on page 106. der, fingers hairy. Second legs with the first joint of the carpus longer than the second, the last three joints of DISTRIBUTION: This species is usually found nearly equal length, the last a little the longest, the in the same habitat as C. brevipes; like C. brevi- joint preceding these somewhat longer. Ambulatory pes, it is moderately common around Oahu. legs somewhat hairy. Distal ends of the rami of the In the shallow water of the reef surface and on uropods clothed with long hairs. Color light yellowish (in spirit). Fingers of larger cheliped pinkish. Length the outer sides of reefs to the depth of about 7 lines (nearly 15 millim.). 20 feet it has been collected at the following Only one specimen has been reported from localities: off Mokulua Islands, Waimanalo, the Hawaiian Archipelago, that by Edmond­ Hanauma Bay, Waikiki, Nanakuli, Mokuleia, son (1925) from Laysan Island. on Oahu; at Lahaina and at 10 miles north of Lahaina, on Maui. Edmondson (1925) has re­ Elsewhere in the Pacific the species has been ported this species from French Frigate Shoal, recorded from Johnston and Palmyra Islands Pearl and Hermes Reef, and Ocean Island. by Edmondson. It has also been reported from One chela, collected by the "Albatross" at Djibouti in the Indian Ocean. The type local­ Honolulu—without further designation— ity is Senegambia, West Africa. probably is this species (U.S.N.M. 63633). Crangon paracrinita (Miers) var. Edmondson has also reported the species bengalensis (Coutiere) from Johnston and Wake Islands. Its original Fig. 40 a-k description was based on specimens from the Indian Ocean, but it was not collected by the Alpheus paracrinitus var. bengalensis Coutiere, Siboga Expedition. Fauna and Geog. Maid and Laccad. 2: 901, pi. 32, figs. 37-37e, 1905. Crinita Group, Diadema Subgroup DESCRIPTION: Rostrum triangular, acute, Crangon paracrinita (Miers) longer than width at base, tip reaching ap­ Alpheus paracrinitus Miers, Ann. and Mag. proximately to middle of visible portion of Nat. Hist. V, 8: 365, pi. 16, fig. 6, 1881. first antennular article, rounded dorsally, not Hawaiian Crangonidae — BANNER 111

FIG. 39- Crangon clypeata (Coutiere). Showing variation in front of carapace, antennules, antennae, and cheli- peds. (a,.b, v, are of one specimen; c, d, j-l, of another; e, m-o, of a third; /, g, p-r, of a fourth; h, s-u, of a fifth.) {a, b, e, i-l, v, scale A; c, d, f, g, scale B; h, scale D; m-u, scale C.) 112 PACIFIC SCIENCE, Vol. VII, January, 1953 carinate, separated from anterior orbital hoods process distally). Carpus slightly over 0.5 as by short shallow round-edged depressions. long as merus with neither distal angle greatly Orbital hoods not conspicuously inflated, projecting or acute. Propodus about 0.8 as rounded anteriorly. Frontal margin between long as merus, with 6-8 spines. Dactylus orbital hoods and base of rostrum concave. simple, curved and elongate, 0.3 as long as Visible portion of first antennular article propodus. subequal to second; third article shortest. Tip Abdominal pleura of both sexes rounded. of stylocerite reaching beyond tip of rostrum, Telson 1.7 times as long as broad at base, 1.7 almost to end of first antennular article. Lat­ times as broad near base as at tip. Proximal eral spine of scaphocerite about 1.1 times as 0.6 of telson broad, only slightly tapered; dis­ long as squamous portion. Basicerite with tal 0.4 abruptly tapered anteriorly, gradually acute spine. End of carpocerite greatly exceed­ tapered posteriorly. Posterior margin almost ing both scaphocerite and antennular pedun­ straight. Paired dorsal spines 0.3 and 0.6 cle; scaphocerite equal to or slightly longer of length from base. than antennular peduncle. Length up to 14 mm. Body usually trans­ Large chela compressed, with smooth regu­ lucent white with definite narrow transverse lar margins and faces except for very slight bands of red on abdomen and somewhat simi­ concavity in upper margin proximal to articu­ lar bands, poorly defined, on thorax; bands lation of dactylus and longer, more gradual across bases of antenna and antennules more concavity in lower margin below articulation definite; chelae with red mottling; eggs dark of dactylus. Chela 3 times as long as broad, yellow in early stages, more greenish in ad­ fingers about 0.3 length of chela. Finger 1.5 vanced stages. times as long as greatest height. Merus of DISCUSSION: No marked variation in the large cheliped 2.0 times as long as broad, su­ specimens in the collection was noted, al­ perior edge distally rounded, inferior external though a detailed study of the ratios of the edge rounded in basal portion, inferior inter­ parts of the appendages was not undertaken. nal edge with small acute tooth distally. Slight differences were noted, as would be ex­ Small chela subcylindrical, almost 5 times pected, in the proportions of the antennules, as long as broad, fingers 0.5 length; shoulders antennae, large and small cheliped (especially at articulation of dactylus rounded. Carpus in the length-width relations of the merus, elongate, slightly over 0.3 length of chela. which varied from 1.8 to 2.7 times as long as Merus 3.4 times as long as broad, similar in broad), in the merus of the third legs, and in form to that of large cheliped but with inferior the telson. It should be noted that in the 26 internal spine either poorly developed or specimens of this variety in the collections the lacking. second article of the carpus of the second legs Both large and small chelae sexually dimor­ varied from 0.85 to 1.3 times as long as the phic in size, large chela of male 1.3 times first. The merus of the large cheliped was length of carapace, equal in female. Small armed with teeth of varying size on the in­ chela with similar range; in addition, dactylus ferior internal margin, and the corresponding of male broader than that of female and bear­ margin of the small cheliped often had a ing heavy fringe of bristles. feebly developed tooth. However, in general the specimens were quite similar. Carpal articles of second legs with ratio 10 : 10 : 5 : 5 : 8. This variety was separated from the parent Ischium of third legs with movable spine. species by Coutiere on the basis of three char­ Merus 5 times as long as wide, rounded dis­ acteristics : first, the parent species has a small tally (in specimen drawn only left leg rounded tooth at the articulation of the dactylus of the distally, that of right leg with small acute small cheliped, whereas the variety has none. Hawaiian Crangonidae •— BANNER 113

FIG. 40. Crangonparacrinha (Miers) var. bengalensis (Coutiere). a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, medial aspect; e, large cheliped, merus and carpus, medial aspect; /, small cheliped, lateral aspect; g, small chela, ventral aspect; h, second leg; /', third leg (meral tooth abnormal); j, third leg, meral-carpal articulation normal, from same specimen as i; k, telson. {a, b, scale A; c-e, h~j, scale B; /, g, scale C; k, scale D.) 114 PACIFIC SCIENCE, Vol. VII, January, 1953

Second, the tooth on the inferior internal mar­ All specimens collected in the windward gin of the merus of the large cheliped is re­ Hawaiian Islands have been assigned to this duced, and the one at the corresponding loca­ variety. It is true that the armature of the me­ tion on the small chela is lacking in the va­ rus of the chelipeds has been overlooked in riety, whereas both are present and well devel­ making this separation, but first, the variation oped in the type described by Miers. Finally, of the armature indicates that too much faith the first carpal article of the second legs is al­ should not be placed upon it; second, the Ha­ most twice as long as the second in C. para­ waiian forms agree with the variety in the ratio crinita (Coutiere: ". . . rapport de 1.75 a of the articles of the carpus of the second legs 1.85 . . .," loc. cit), whereas in the variety the —at least, they lie within the expected range two articles are of almost equal length. of variation—and none had the tooth flanking The differences between the Hawaiian the articulation of the dactylus of the small specimens and those described by Coutiere chela. If the variation in these two forms is are slight except for the variations noted. similar to that found in C. brevipes (Stimpson)

TABLE 5 CHARACTERISTICS DIFFERENTIATING THE SPECIES OF THE DIADEMA SUBGROUP

CHARAC­ C. paracrinita TERISTIC C. pugnax C. diadema C. pseudopugnax C. percyi C. gracilipes var. bengaknsis

Rostral base Very narrow Broadly trian­ Simple ridge Slightly broad­ Similar to Rounded, triangle gular with ca­ anteriorly, er than in pugnax not well without ca­ rina; some­ triangular pugnax; with­ marked; rina; not what de­ posteriorly; out carina slightly depressed pressed strongly de­ anteriorly; depressed pressed strongly de­ pressed

Orbitoros- Narrow, deep Broad and Broad and Narrow, deep Narrow, deep Short, shal­ tral groove shallow deep low, round­ anteriorly ed

Large chela Transverse Transverse Transverse Transverse Transverse Without groove* groove groove and groove groove groove depressed areas

Large cheli­ Strong tooth Similar teeth, Similar teeth, Same as C. Similar teeth, Without su­ ped, merus on superior lacking meral with serra­ pseudopugnax 2-4 spines perior tooth, margin; in­ spines tions on all on inferior inferior ferior mar­ margins margin tooth weak; gin with no spines tooth and series of spines (?)

Carpal arti­ 10 : 25 10 : 10 10 : 11 10 : 10 10 : 10 10 : 10 cles of sec­ ond legs, ratio of first to second

Third legs, Strong tooth Strong tooth Strong tooth Strong tooth Without Without merus tooth tooth *De Man states in his key (19H) "Chela without a transverse groove near the articulation of the dactylus." However, Dana (loc. cit.) states "narrow emarginate above, near articulation," and shows in his figure what might be the groove. Hawaiian Crangonidae — BANNER 115

and C. clypeata (Coutiere), the diagnostic char­ ing to or slightly beyond end of first antennu­ acteristic afforded by the ratio of the lengths lar article. Scaphocerite long, with strong lat­ of carpal articles may prove to be of some­ eral spine about 1.1 times as long as squamous what questionable worth. portion; outer margin somewhat concave; To separate this from the related Hawaiian squamous portion somewhat reduced. Strong species, see Table 5. lateral spine of basicerite reaching tip of ros­ DISTRIBUTION: On Oahu four specimens trum. Scaphocerite slightly longer than anten­ were from unrecorded localities, presumably nular peduncle; tip of carpocerite reaching from the shallow water at the reef surfaces; slightly beyond end of second antennular ar­ specific localities were Kaneohe Bay, under ticle. Antennal flagellum about 1.5 times body coral heads on sand flats about a foot deep at length. low tide, where the species was common; Large chela subcylindrical, 1.3 times as high Waimanalo, 6 feet deep; Hanauma Bay, 20 as broad, 3.7 times as long as high, tapering feet deep; Black Point, 2 feet deep; and Wai- toward distal end. Palm with deep transverse kiki, 2 feet deep. On Maui, specimens were groove proximal to articulation of dactylus. collected at Kalama Park, 3 and 8 feet deep; Palm, at articulation of dactylus, with narrow 10 and 12 miles south of Lahaina, both at 6 depression to accommodate high crest of dac­ feet; and at Lahaina, 4 and 6 feet deep. There tylus when flexed. Dactylus 0.3 length of are no other records of the variety from the chela, compressed, strongly arcuate, longer Hawaiian Islands. than fixed finger. Merus about 2.5 times as long as broad; superior margin somewhat Crangon gracilipes (Stimpson) rounded but usually projecting as acute tooth Fig. 41 a-i (specimen drawn also with secondary tooth); inferior internal margin acute, with 2-4 spines Alpheus gracilipes Stimpson, Acad. Nat. Sci. and acute tooth distally. Phila., Proc. 12: 31, I860. Small chela not sexually dimorphic, long DESCRIPTION: Rostrum narrow, triangular, and narrow, over 5 times as long as broad, acute, flattened dorsally, without carina, tip fingers somewhat broadened, 0.7 length of slightly depressed, reaching 0.7 of length of palm. Dactylus with fringe of long setae run­ visible portion of first antennular article. Ros­ ning longitudinally from near point of articu­ trum separated from posterior portion of or­ lation over crown of article before apex. Small bital hoods by deep narrow depressions, more subacute tooth on propodus above articula­ anteriorly by wide flattened areas with abrupt tion of dactylus. Merus usually with acute sides against both orbital hoods and rostrum, teeth distally on inferior internal and superior latter overhanging groove. Orbital hoods in­ margins; inferior internal margin usually with flated, large, with abrupt medial margin, pos­ one to several spinules. teriorly merging with carapace, anteriorly with Carpal articles of second legs with ratio dorsoventral keel. Anterior margin of orbito- 10 : 10 : 4 : 5 : 6. rostral area concave.When seen in lateral view, Ischium of third legs with small movable anterior carapace without definite angle along spine; merus 8 times as long as broad, iner- dorsal surface at base of rostrum [as there is in mous; carpus 0.5 as long as merus, with tooth C. percyi (Coutiere)], and without tubercle at at inferior distal margin poorly developed; this point. propodus almost as long as merus, with 9-11 Antennular peduncle long and slender with slender spines; dactylus simple, curved, 0.2 second article over twice as long as broad, length of merus. longer than visible portion of first article, over Telson 2.1 times as long as broad at base, twice as long as third article. Stylocerite reach­ tip 0.6 width of base. Lateral margins almost 116 PACIFIC SCIENCE, Vol. VII, January, 1953

straight, posterior margin arcuate. Dorsal and Waikiki, Oahu, and one from Maalaea, spines moderately developed, 0.4 and 0.7 of Maui; these were all without records as to distance from articulation to tip. Length of depth, but presumably they were from shal­ larger pair of terminal spines less than 0.5 low water. Edmondson (1925) reported two breadth of tip. Tip of telson and distolateral specimens from Lisianski Island. The species margin of inner uropod with series of small has also been reported from the Hawaiian Is­ spinules. lands by Coutiere. Length up to 22 mm. One specimen ob­ C. gracilipes occurs from the Red Sea through served alive; dark, almost black, to unaided the Indian Ocean, the East Indies, and Japan, vision; color from greatly expanded blue and to Samoa, Tahiti, and Hawaii. The type local­ red chromatophores. Under microscope body ity is Tahiti. mottled, with clear areas over eyes and on branchiostegites over third maxillipeds and Crangon pugnax (Dana) elsewhere, and with darker areas like two "eye Fig. 42 a-h spots" on sides of abdomen. Antennules and antennae, second to fourth legs, and caudal Alpheus pugnax Dana, U. S. Explor. Exped. fan bluish; fifth legs reddish; chelae slightly 13: 554, pi. 32, figs. 6a-6h, 1852. darker than body with pink cast at dactylar Alpheus pugnax Dana, Acad. Nat. Sci. Phila., articulations and on tips of fingers. Proc. 6: 27, 1854. [Description without fig­ ures.] DISCUSSION: In a few specimens available for comparison no great variation was noted. As no specimens are available, the original The rostral length varied, with the tip reach­ description is given: ing from 0.7 of the length of the first article Beak [rostrum] acute, narrow triangular, flat above, of the antennular peduncle to beyond the end arising from between the bases of the eyes. Basal spine of the same article. In some cases the second of outer antennae [of basicerite] small, basal scale [scaphocerite] longer than base [carpocerite]. Base [pe­ antennular article was over twice the length of duncle] of inner antennae shorter than scale of outer; the first. The merus of the large chela varied basal spine [stylocerite] of inner antennae not shorter as noted above; almost always the distal teeth than the first joint, second joint short. Anterior feet unequal; larger hand long, smooth with rounded mar­ were well developed; the same is true of the gins, narrow emarginate above near articulation, fingers small chela. In some of the specimens the sec­ short (about one-third the length of hand), arm having ond carpal article of the second legs was a spine at both inner and outer apex. Feet of second pair long, first joint short, hardly half the second in slightly shorter than the first. The number of length. Third and fourth pairs slender, third joint with spines on the propodus of the third legs varied an acute tooth at lower apex. slightly. At Lahaina, Island of Maui, Hawaiian group. Length, twelve lines. The beak from its base is very The only possible difference between this narrow, triangular, with straight sides, and there is no and the original description of Stimpson lies carina prolonged down the back. Second joint of inner in the lateral spine of the basicerite which he antennae hardly longer than first. Larger hand partly pubescent, somewhat fusiform, but little compressed, specifies to be minute, whereas on all of these the finger turned out of plane of the hand; movable specimens it is large and well formed; how­ finger short, thin above, with arcuate dorsal margin. ever, exactly what Stimpson meant by "mi­ First joint of carpus of second pair of feet less than half the second; third shorter than fourth; fifth longer than nute" is not known. The specimens also agree fourth; hand as long as fourth and fifth. The fifth joint well with the specimens described by de Man of third pair of legs long, very much longer than fourth, (1911) from the southwest Pacific. about eight set of spinules on its lower side, rather long hairy above. To separate this from the related Hawaiian species, see Table 5. Dana further states in his key: "Base of DISTRIBUTION: The few specimens available rostrum rising between the eyes, deep sulci in were collected from Kaneohe Bay, Kaaawa, carapace either side of rostrum. Orbital mar- 1

FIG. 41. Crangon gracilipes (Stimpson). a, b, Anterior region, dorsal and lateral aspects; c, d, large cheliped, lateral and medial aspects; e, large chela, distal end, ventral aspect;/, small cheliped, male, lateral aspect; g, small chela, male (different specimen), ventral aspect; h, second leg; i, third leg. {a, b, g, h, scale A; c-f, i, scale B.) 118 PACIFIC SCIENCE, Vol. VII, January, 1953

ther attempt to obtain specimens of the spe­ cies, the reefs 5, 10, and 30 miles away from Lahaina were visited; none of these reefs du­ plicated the original conditions of the La­ haina reef, for they were either deeper, or more sandy, or less well developed, etc., and no specimens of C. pugnax were obtained. However, on the basis of the original de­ scription and figures alone, it is possible to separate clearly and decisively C. pugnax from all other species of Hawaii. Table 5 gives the criteria for its separation from the obvi­ ously closely related Hawaiian species of the subgroup Diadema. If this species is not extinct, and there is no good reason to presume it is until far more extensive collections are made about the Ha­ waiian Islands, it is valid and recognizable.

Crangon diadema (Dana) Frontispiece, Fig. 43 a-k FIG. 42. Crangonpugnax (Dana), "a, Front, side view, enlarged; b, upper view; c, part of outer maxilliped; d, Alpheus diadema Dana, U. S. Explor. Exped. e, larger hand, in different positions; /, smaller hand; g, part of leg of second pair; h, part of leg of third pair." 13: 555, pi. 25, figs. 7a-7e, 1852. (Figures and legend from Dana; figures redrawn from Alpheus diadema Dana, Acad. Nat. Sci. Phila., U. S. Exploring Expedition, Folio Atlas to Vol. 13, Proc. 6: 23, 1854. [Description without fig­ 1855, plate 35, figures 6a-h.) ures.] Alpheus insignis Heller, K. Akad. Wiss. Wien, gin inermous." [Translated from Latin; Dana, Sitzungber. 44:269, Taf. 2, figs. 17-18,1861. 1854.] NEOTYPE: A male, 22 mm. long, collected DISCUSSION: As far as I have been able to determine, this species has not been reported at Lahaina, Maui, from a head of coral {Po­ since its original description. Moreover, the ntes) in about 4 feet of water (U. S. N. M. type specimen has been lost. As no specimens 93460). in the collections available agreed with the de­ DESCRIPTION: Rostrum short, reaching to scription on all characteristics, a special field middle of visible portion of first antennular trip was made to the type locality, Lahaina, article; rostral base broad, flattened, extending Maui, to collect neotypes of this species and posteriorly between eyes; margins of rostrum of C. diadema (Dana). Unfortunately, the con­ and base anteriorly slightly concave and taper­ ditions at Lahaina are not the same as they ing, almost parallel in middle, posteriorly were 100 years ago when the collections were slightly concave and spreading; margins lat­ made by Dana; now the near-by sugar mill is erally overhanging orbitorostral groove, with discharging fresh water charged with large opposing face of groove almost touching ros­ quantities of red mud onto the reef area, kill­ tral portion; carina prominent only in middle ing most of the coral and the other life on the section; posteriorly marked by slight protu­ reef. It was possible to collect quite a few berance. Orbital hoods high, rounded, ex­ specimens in spite of the pollution, but none ceedingly clear, conspicuously demarked on had the characteristics of C. pugnax. In a fur­ all sides, anteriorly projecting as slight Hawaiian Crangonidae — BANNER 119 rounded vertical keel. Area between orbital about 0.8 as long as merus, tapering, with hoods and narrow orbitorostral furrow flat­ about 20 movable spines of varying sizes. tened; anterior margin arcuate. Dactylus simple, slightly curved, acute, as Antennular peduncle with first article long as merus. slightly shorter than second, correspondingly Telson almost twice as long as broad, pos­ longer than third; second article almost twice terior margin 0.75 as broad as anterior section; as long as broad. Stylocerite short and broad, lateral margins slightly concave. Posterior lateral spine reaching end of first antennular margin arcuate. Dorsal spines large, located article. Lateral spine of basicerite well devel­ about 0.3 and 0.7 of distance from articulation oped, shorter than stylocerite. Carpocerite to tip. Length of larger pair of posterolateral reaching end of antennular peduncle, ex­ spines 0.3 width of tip; smaller pair less than ceeded by scaphocerite. Scaphocerite with lat­ 0.5 as long as longer pair. Margin between eral spine dominant, squamous portion re­ posterolateral spines with 9 somewhat irreg­ duced; lateral margins concave. ularly placed spinules. Distolateral margin of Large chela subcylindrical, almost as broad inner uropod with 4 large spines and 9 spi­ as high, about 3 times as long as broad, taper­ nules. ing distally; surface without sculpture except Large females up to 26 mm. long; no males for transverse groove proximad of articulation exceeding 22 mm. of neotype. Color in life of dactylus; chela sparsely hirsute on upper variable, but usually dark, olive green, reddish and inner surfaces. Dactylus about 0.4 length brown, etc., with pronounced irregular mot­ of chela, strong, with margin arcuate. Merus tling of lighter color; specimens at times al­ about twice as'long as broad, about 0.25 as most transparent. long as chela; superior distal margin project­ DISCUSSION: All the specimens at hand were ing as acute tooth, inferior internal margin examined, but detailed studies on the varia­ with tooth distally. Large chela of female sim­ tion in proportions of the appendages were ilar in form but relatively much smaller. not made. However, variation was noted in Small chela of male 0.8 as long as carapace, three parts. First was in the anterior region of subcylindrical, tapering, 3 times as long as the carapace. Small specimens, 8 millimeters broad, without sculpturing. Fingers slightly long or less, had a markedly different rostrum over 0.3 length of whole chela, expanded, than the adults. In these specimens the ros­ about twice as long as broad, with dense fringe trum was narrow and of almost uniform taper, of setiferous bristles along both sides and over with its tip reaching almost to the end of the upper distal surface of finger ("balaeniceps- first antennular article; the rostral carina was shaped," see Fig. 43f-g). Carpus slightly well developed along the complete length; elongate, 0.25 length of chela. Merus similar moreover, in these specimens the area between to that of large chela but without inferior in­ the rostrum and the orbital hoods extended ternal tooth. Small chela of female quite simi­ relatively further forward, markedly surpass­ lar in form but much smaller, without balae- ing the anterior margin of the orbital hoods. niceps-shaped dactylus and with carpus rela­ In specimens of increasing size, the condition tively more elongate. of the rostrum approached that of adults. In Carpal articles of second legs with ratio mature specimens, variation was noticed es­ 10 : 10 : 3 : 4 : 5. pecially in the shape of the base of the ros­ Third legs with strong spine on ischium. trum between the orbital hoods; in some it Merus 3.5 times as long as broad, strong, was as described above with the margins of curved, acute tooth subterminally on inferior the middle portion straight and almost paral­ margin. Carpus 0.5 as long as merus, inferior lel, whereas in others the curve from the edge margin projecting as strong tooth. Propodus of the carapace to the posterior limits was uni- 120 PACIFIC SCIENCE, Vol. VII, January, 1953 form and gradual. Furthermore, although in hoods, instead of straight and parallel as they most large specimens the tip of the rostrum are in this specimen; however, as noted above, was depressed as shown in Figure 43^, in a this is a variable character. Second, his profile few of intermediate size it was found to con­ drawing shows the rostrum as turning up near tinue almost "level," and in one it was slight­ its apex, instead of being depressed as in this ly elevated. specimen; this, too, is a variable characteristic. A second field of variation was in the carpal Third, the margin between the base of the ros­ articles of the second legs; some had the ratio trum and the anterior orbital hoods is shown given above, but in some the ratio of the first as concave instead of convex; this may be due two articles approached 10 : 13. None were to faulty delineation in the drawing. Fourth, noted where the second article was shorter the second carpal article of the second legs than the first. was described as being much shorter than the The third variation was in the propodus of first and was shown to have the ratio (approxi­ the third legs. Small specimens had only four, mately) of 10 : 7 instead of 10 : 10 or 10 : 12, five, or six spines well developed; among the as found in all the specimens in the collection. large specimens some had a few more, some Finally, the propodus was described as bear­ had a few less, than the 20 described. ing "six sets of spinules on inner side," in­ Since the original description of this spe­ stead of approximately 20 found in the neo­ cies from the Hawaiian Islands by Dana, no type. This last difference could be attributed specimens have been reported. In 1861 Heller either to the immaturity of Dana's specimens described from the Red Sea what was appar­ or to inadequate observation. ently a closely related form, Alpheus insignis, Thus, the only significant difference be­ and all later workers, evidently because they tween Dana's description and this neotype is were not certain of the characteristics of in the ratio of the carpal articles of the second Dana's species, used Heller's name. As C. in­ legs. This may be due to a chance variation in signis the species was later reported from the the specimens or to faulty observation; but, Hawaiian Islands. Coutiere in 1899 (p. 500) even if an actual difference, it would not be a listed C. insignis as a synonym of C. diadema sufficient difference to warrant calling the lo­ but in 1909 (p. 899) reviewed the situation cal specimens other than diadema. and decided, like de Man, to reserve judg­ I can find no significant differences between ment until Dana's type was redescribed. Un­ this species and the form described by Heller. fortunately, the type of Dana's species was Heller shows the margin between the rostrum lost, and it was impossible to determine the and the anterior margin of the orbital hoods exact characteristics of the species. to be concave instead of convex (as did Dana) To resolve this uncertainty, a special trip and somewhat more narrow; he describes the was made to Lahaina, Maui, the type locality, propodus of the third legs as bearing seven or to collect replacement type specimens for eight spines. As these differences, if correct, Dana's description. This effort was successful would be at most of only subspecific value, as the reef at Lahaina had a larger proportion Heller's species is placed in synonymy. of this species than any other locality visited, Table 5 contrasts this with related Hawaiian and the specimens also were larger and more species. robust. One of them is described above. DISTRIBUTION: This species is common in This neotype differs in five minor ways the shallow waters of the Hawaiian Islands. from Dana's drawings and description. First, On the reef flats or in water to 20 feet deep it the margins of the base of the rostrum were has been collected from the. following locali­ shown as definitely and uniformly concave ties: on Oahu, Kaneohe Bay, Mokulua Is­ throughout their length between the orbital land, Waimanalo, the shore near Koko Head, Hawaiian Crangonidae — BANNER 121

FIG. 43. Crangon diadema (Dana), a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, ventral aspect; e, large cheliped, merus, medial aspect;/, small cheliped, male, lateral aspect; g, small chela, distal end, male, ventral aspect; h, small cheliped, female, lateral aspect; /, second leg;/', third leg; k, telson and uropods. (Drawings of neotype.) {a, b, k, scale A; c-j, scale B.) 122 PACIFIC SCIENCE, Vol. VII, January, 1953

Hanauma Bay, Black Point, Waikiki, Nana- Lateral spine of basicerite well developed kuli, Mokuleia, Kaaawa, and Kawela Bay; on but not reaching end of first antennular arti­ Maui, Makena, Kalama Park, Lahaina, and cle. Scaphocerite with lateral spine dominant 12 miles south of Lahaina. The "Albatross" and squamous portion reduced; lateral mar­ collected five specimens from "Honolulu gin strongly concave. Carpocerite as long as Reef" (U.S.N.M. 63552, 63553). There is only antennular peduncle; both exceeded by sca­ one dredged specimen, taken between 40 and phocerite. 350 feet deep off the southwest coast of Oahu. Large chela subcylindrical, 1.3- times as Edmondson (1925) has reported the species broad as high, 2.7 times as long as broad (as C. insignis) from Laysan and Lisianski Is­ [note: Figs. 44c and 44^ show neither maxi­ lands and Pearl and Hermes Reef. mal nor minimal diameters]; margins rounded; As C. insignis the species has been reported tapering toward fingers. Palm marked by deep throughout the tropical Indo-Pacific region transverse groove proximal to articulation of from the Red Sea to Samoa; Edmondson finger, which spreads and soon disappears on (1925) has reported it from Johnston Island superior face, but on inferior face continues in the central Pacific. into shallow, poorly defined depressed area extending distally to articulation of dactylus Crangon pseudopugnax sp. nov. and proximally an equal distance. Adhesive Fig. 44 a-i plaque on independent lobe, demarked by TYPE SPECIMEN: A female 15.6 mm. long, palmar depression on inferior side and by collected at Kalama Park, southeast Maalaea deep, short depression on superior side that Bay, Maui; it was collected from old coral at accommodates ridge of dactylus upon flexure. a depth of about 8 feet. Five paratypes, one Dactylus about 0.3 length of entire chela; from shallow water at Makena, Maui, one high, narrow, arcuate, with pronounced ridge from shallow water at Waikiki, Oahu, and along free margin; tip rounded. Fixed finger three from about 18 feet of water off Waikiki blunt at tip. Chela with occasional setae and Reef, Oahu (U.S.N.M. 93513). usual tufts on fingers. Merus triangular, sharp­ DESCRIPTION: Rostrum acute, twice as long ly angled; superior margin terminating as as broad at anterior margin of base, tip reach­ acute dentate projection, with 5 small inden­ ing to end of first antennular article. Rostral tations proximally from which very small carina well-developed crest continuing to mid­ bristles arise; inferior external margin with 7 dle of orbital hoods, then broadening into similar indentations; inferior internal margin dorsally flattened triangular area. Orbital with 9 indentations and strong subacute tooth hoods inflated, rounded anteriorly, well de- subterminally. Ischium with 3 indentations marked from surrounding carapace, except on for bristles. posterior margins merging with carapace; or­ Small chela cylindrical, 5 times as long as bital hoods very clear. Area between orbital broad, fingers about 0.7 as long as palm. Fin­ hoods and rostrum flattened; anterior margin gers narrow. Chela with usual setae. Carpus of arcuate. Anterior portion of carapace de­ usual form. Merus about 0.7 as long as chela, pressed, seen in profile, with small medial triangular, with indentations similar to merus protuberance or lobe. of large chela but fewer; inferior internal mar­ Second and third articles of antennular pe­ gin unarmed. duncle equal in length but both slightly Carpal articles of second legs with ratio shorter than visible portion of first article; sec­ 10 : 11 : 4 : 5 : 7. ond article about 1.3 times as long as broad. Ischium of third legs with movable spine Stylocerite well developed, tip reaching be­ on inferior margin (difficult to see). Merus 5 yond end of first antennular article. times as long as broad, with strong, acute Hawaiian Crangonidae — BANNER 123

FIG. 44. Crangon pseudopugnax sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, ventral aspect; e, large cheliped, merus and carpus, medial aspect; /, small cheliped, lateral aspect; g, second leg; h, third leg; /', telson and uropod. (Setae are not shown on the chelipeds or the second legs.) {a, b, i, scale A; c-h, scale B.) 124 PACIFIC SCIENCE, Vol. VII, January, 1953 tooth on distal inferior margin. Carpus 0.5 close to C. pugnax (Dana) that I tentatively as long as merus, unarmed except for tooth on identified it as that species in the field. inferior distal angle. Propodus 0.8 length of C.pseudopugnax can be separated easily from merus, with 10 movable spines in addition to these related species by a combination of usual setae. Dactylus simple, acute, somewhat characteristics. From C. diadema the species curved. may be separated most easily by the nature of Telson almost twice as long as broad, tip the base of the rostrum; from C. gracilipes by slightly more than 0.5 as broad as base; lateral the presence of a groove on the large chela margins slightly convex anteriorly, slightly and by the tooth on the merus of the third concave posteriorly; posterior tip strongly legs; from C. pugnax by the lengths of the first convex. Dorsal spines moderately developed, two articles of the second legs. (See Table 5.) anterior approximately 0.4 and posterior 0.7 This species shows a close affinity to an­ of length of telson from base; medial pair of other related Hawaiian species, C. percyi (Cou­ terminal spines narrow, acute, over 0.5 as long tiere). They can be distinguished by the na­ as breadth of tip of telson. Medial margin of ture of the rostrum, which in C. pseudopugnax tip with 3 irregularly placed, feeble spinules. bears an acute carina on the anterior portion Inner uropod with 5 spines of graduated size of the rostral base and does not overhang the along distolateral margin; spine of outer uro­ orbitorostral groove except at the posterior pod long, rather slender. end, while in C. percyi the anterior surface is In life, type specimen with broad, red, trans­ broad and flat, and the margins overhang the verse bands. orbitorostral grooves for their entire length; DISCUSSION: Only one of the paratypes had by the large chela, which bears a depression all its appendages; all were smaller than the on the lower face in C. pseudopugnax that is type specimen. In these few specimens not lacking in C. percyi; and by the small cheliped, much variation was noted. The most conspic­ which lacks teeth at the dactylar articulation uous difference was in the rostral base between and a tooth on the inferior internal margin of the eyes. Most of the specimens showed a the merus in C. pseudopugnax, whereas the dac­ narrow but carinate base that was set off lat­ tylar articulation is flanked by two teeth and erally by shallow grooves, while the type the inferior internal margin of the merus bears lacked these grooves. One specimen had a a distal tooth in C. percyi. slight spine on the merus of the small chela, and a specimen half the size of the type did This species is also related to C. philoctetes not have the depressed areas on either side of (de Man), but the latter has a series of mov­ the groove of the large chela. Aside from these able spinules on the merus of the small chela differences and slight differences in the pro­ and the merus of the third legs, which are ab­ portions of the articles, all the specimens were sent in C. pseudopugnax. Unfortunately the similar. Unfortunately there were no intact large chela was absent in de Man's type speci­ male specimens in the collection, but it is pre­ men. sumed that the small chela of the male is ba- DISTRIBUTION: All specimens known are re­ laeniceps-shaped as it is in the related species. corded above.

RELATIONSHIP: This species plainly belongs Crangon percyi (Coutiere) to the Diadema Subgroup, and shows rela­ tionship to C. diadema (Dana) in the nature of Fig. 45 a-g the orbital hoods and the transverse groove on Alpheuspercyi Coutiere, Soc. Philomath. Paris, the palm of the large chela; the rostrum is Bui. IX, 9(5): 21, 1908. , somewhat like that of C. gracilipes (Stimpson) Alpheus percyi Coutiere, Linn. Soc. London, and the general configuration is sufficiently Trans. II (Zool.) 17: 426, pi. 64, figs. 22- Hawaiian Crangonidae — BANNER 125

22f, 1921. [Same description as original, but Ischium of third legs with moderately with figures.] strong movable spine. Merus 5 times as long DESCRIPTION: Rostrum acute, reaching to as broad, with strong tooth distally on inferior end of first article of antennular peduncle; margin. Carpus 0.5 as long as merus, with posteriorly continued as broad base between usual inferior distal tooth. Propodus almost eyes; lateral margins almost straight, over­ 0.8 as long as merus, with 11 relatively long hanging orbitorostral furrow; base with low, spines on inferior and distal margins. Dactylus indistinct dorsal carina. Orbitorostral furrows 0.2 as long as merus, somewhat curved, acute, deep and narrow, flattening anteriorly, with simple. anterior margin somewhat sinuate. Orbital Telson of usual form, similar to that of C. hoods hemispherical anteriorly, high and of pseudopugnax, with series of small spinules on clear chitin; merging with carapace posterior­ posterior margin. Distolateral margin of inner ly. Anterior region of carapace, seen in pro­ uropod with about 12 small spinules. file, depressed; beginning of depressed region Specimens in collection 3 females, one 25 with small medial protuberance. mm., others about 20 mm. long. First and second articles of antennular pe­ DISCUSSION: The other specimens agreed duncle subequal; third article somewhat with the one described and illustrated on most shorter. Stylocerite strong, slightly exceeding characteristics. The points of difference be­ first article. Lateral spine of basicerite some­ tween one or the other of the two undescribed what shorter than stylocerite. Scaphocerite specimens and the one described above were: with strong lateral spine, squamous area re­ First, the "break," or angle, the anterior cara­ duced; lateral margins strongly concave. Ba­ pace and rostrum made to the dorsal surface sicerite as long as antennular peduncle, both of the carapace was less pronounced, and the exceeded slightly by scaphocerite. tubercle was less noticeable. Second, the ros­ Large chela rounded, subcylindrical with trum was shorter, reaching only 0.7 of the maximum diameter 1.2 times minimum diam­ length of the visible portion of the first anten­ eter, chela 2.7 times as long as maximum di­ nular article. Third, the second antennular ar­ ameter. Palm with deep transverse groove ticle was relatively longer, being 1.3 times the proximal to dactylus but without other sculp­ length of the visible portion of the first article. turing. Dactylus about 0.3 length of entire Finally, on the large chela there was a poorly chela, arcuate, high and narrow. Chela with marked shallow depression that ran from the only scattered setae except for tufts about fin­ transverse groove toward the articulation of gers. Merus over twice as long as broad, su­ the dactylus. These differences probably are perior and inferior internal margins with individual variations. strong acute teeth; all margins with series of The specimens of this species, which has small indentations from which fine setae arise. not previously been reported from the Ha­ Small chela subcylindrical with tapering fin­ waiian Islands, differ in only a few details gers, 2 poorly developed teeth flanking articu­ from the original description and figures by lation of dactylus; fingers in female conical, Coutiere. It is uncertain from Coutiere's draw­ in male balaeniceps-shaped, or bearing a dense ings whether the configuration of the base of fringe of setae over broadened dactylus (ac­ the rostrum and the orbitorostral furrows are cording to Coutiere; no males collected from exactly the same in the two forms, but in any Hawaii). Merus similar to that of large chela case they are very similar. Coutiere does not except more slender and superior and inferior- show as much depression of the anterior mar­ internal teeth more poorly developed. gin of the carapace as is shown in these speci­ Carpal articles of second legs with ratio mens, nor does he show the median tubercle. 10 : 10 : 4 : 5 : 6. The carpocerite may be somewhat shorter in 126 PACIFIC SCIENCE, Vol. VII, January, 1953

Coutiere's specimens. It should be noted that scribed and illustrated; see these and the on the left side of the described specimen following pages in this reference for a com­ (Fig. 45 £) the spine of the basicerite was [like plete synonymy.] that shown by Coutiere, but the one on the right was entirely absent. The large and small DESCRIPTION: Rostrum small, acute, equi­ chelae are similar, but the merus of these ap­ lateral, tip reaching to end of first 0.3 of pendages in the type may or may not have visible portion of first antennular article; ros­ the series of fine indentations found in the lo­ tral carina anteriorly sharp, posteriorly some­ cal specimens, for Coutiere's plates are not what rounded, extending to end of orbital clear at the point. The third legs are similar hoods. Orbital hoods large, inflated, rounded except that Coutiere does not indicate the anteriorly, posteriorly merging with the cara­ ischial spine, although he does state that the pace; higher in middle than rostral carina; appendage "est tres semblable [to C. dasy- transparent, hence from side rostral carina cheles (Cout.)] comme forme ..." and that C. visible through hood. Anterior margin of dasycheles has the spine. Finally, on the tip of carapace from orbital hoods to rostrum al­ the telson and on the inner uropod he indi­ most straight. cated no spinules; these, however, are very difficult to discern on the local specimens be­ Antennular peduncles elongate and thin, cause they are confused with the bases of the second article 3-5-4.0 times as long as broad, setiferous bristles, and it is likely that Cou­ 2.5 times length of first article or of third tiere merely overlooked them. article. Stylocerite rounded, leaf-like, anterior spine represented by very small tooth reaching In my opinion none of these differences are about 0.75 of length of visible portion of first of sufficient magnitude, constancy, or general antennular article. Antennular articles with reliability to warrant the erection of a new spe­ few short bristles near points of articulation; cies for the local form. stylocerite with fringe of short bristles on The characteristics which will distinguish margins. Basicerite with distinct but short this from closely related local species are tooth about same length as rostrum. Scapho- given in Table 5. cerite with strong lateral spine, squamous DISTRIBUTION: The three specimens col­ portion narrow; lateral margins concave. Sca- lected locally came from Oahu in 20 feet of phocerite slightly longer than carpocerite water at Hanauma Bay, in 15 feet of water off which reaches to end of antennular peduncle. Nanakuli, and at Halape, Kau Coast, Hawaii, in water 35-40 feet deep (last collected by R. Distal articles of third maxillipeds large, of W. Hiatt). usual form, densely bristled on inner face. Large chela high, compressed, margins Brevirostris Group rounded, 2.7 times as long as high, about twice as high as wide. Upper surface with Crangon rapax (Fabricius) rounded transverse groove extending short Fig. 46 a-i distance down each face. Dactylus strong, Alpheus rapax Fabricius, Sup. Ent. Syst. p. high, compressed, 0.35 as long as entire chela. 405, 1798. Carpus of usual form. Merus 2.6 times as long Alpheus malabarkus Hilgendorf, K. Akad. as broad, with superior distal angle rounded, Wiss., Berlin, Monats. p. 832, 1878. inferior internal margin with 5 movable spines Alpheus brevirostris de Man, Linn. Soc. Lon­ and acute terminal tooth of moderate size. don, Jour. 22: 261, 1888. Dactylus, margins of palm, superior and in­ Alpheus rapax de Man, Soc. Zool. France, ferior internal margins of merus with scattered Mem. 22: 147-155, 1909- [Species rede- long setae. Hawaiian Crangonidae — BANNER 127

FIG. 45. Crangon percyi (Coutiere). a, b, Anterior region, dorsal and lateral aspects; c, d, large cheliped, lateral and ventral aspects; e, small cheliped; /, second leg; g, third leg. {a, b, scale A; c-g, scale B.) 128 PACIFIC SCIENCE, Vol. VII, January, 1953

Small chela of male with palm almost 2.5 description in a few characteristics. The speci­ times as long as wide, fingers 1.5 times length mens are similar in general configuration, and of palm; margins without grooves. Dactylus the general relationship of the size, shape, and broad with opposing faces flattened and armature of the appendages. Some minor fringed with dense setiferous bristles, or ba- differences were noted in a comparison of laeniceps-shaped; both dactylus and fixed fin­ this specimen with de Man's description, for ger strongly hooked at tip. Carpus somewhat example: the rostrum reaches to the end of longer than broad. Merus narrower proxi- the first third of the visible portion of the mally than distally, maximum breadth 1.5 first antennular article in this specimen, and times minimum, article twice as long as broad from one half to two thirds of the length in at maximum breadth; armature like that of de Man's; the visible portion of the first large cheliped except tooth on inferior inter­ antennular article is slightly longer and the nal margin lacking. Small chela of female third article is slightly shorter, when compared reported (de Man, 1909) as similar in form to to the second article, than those described by that of male except lacking rows of setae on de Man; the merus of the large chela is 2.6 dactylus (therefore not of balaeniceps-shape) times as long as broad in this specimen and and slightly smaller in proportions. bears five spinules, while it is 3.0-3.5 times Carpal articles of second legs with ratio as long as broad and bears six to seven spinules 10 : 7 : 3 : 3 : 5. in the described specimens; the ratio of the Ischium of third legs with strong spine. carpal articles is 10 : 7 : 3 : 3 : 5 instead of Merus 4.4 times as long as broad, with few 10 : 8-9 : 3 : 3 : 4; and so on. Two other setae along margins. Propodus slightly less differences may be of greater importance: the than 0.5 length of merus, superior distal mar­ merus of the small cheliped of the male is gin with sparse long setae, superior distal broader and lacking the small terminal tooth angle projecting as tooth. Propodus 0.7 on the inferior internal margin found in the length of merus with 7 spines on inferior and specimens described by de Man; the carpus terminal margins, scattered long setae on in­ and propodus of the third leg in this specimen ferior and superior margins. Dactylus 0.5 bears only scattered long setae, while de Man length of propodus (or 0.3 length of merus), describes and depicts his specimens as being narrow, almost straight, tip acute, inferior densely hirsute on these articles. surface flattened, superior surface rounded; When considering these differences it should rounded surface with 2 short bristles. be remembered that de Man's description is Telson similar in form to that described for based upon two specimens, a male 51 mm. C. platyunguiculata (p. 130), 1.8 times as long long and a female 46 mm. long, while the as wide at point of maximum breadth; tip Hawaiian specimen is only 28 mm. long. 0.7 as wide as maximum breadth; postero­ Therefore, it appears that all these dif­ lateral spines slightly shorter and weaker than ferences are of minor importance and may be those of C. platyunguiculata. Tip of telson and growth differences, or differences normally distal lateral margins of inner uropod with occurring in a population, or, possibly, they weak spines similar to those of platyunguicu­ may be differences indicative of a geograph­ lata. ical subspecies of the parent C. rapax. How­ Sole specimen in collection, male 28 mm. ever, in my opinion, the description of this long; specimens reported by de Man {he. Hawaiian form as a new species or a new cit.) to reach 51 mm. Color in life not noted. subspecies is not indicated on the basis of the DISCUSSION: Crangon rapax (Fabr.) was re- present specimen. described and illustrated by de Man in 1909. This species can be separated from C. platy­ The specimen from Hawaii differs from this unguiculata, the only other Hawaiian species Hawaiian Crangonidae — BANNER 129

FIG. 46. Crangon rapax (Fabricius). a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, dorsal aspect; e, small cheliped, lateral aspect; /, small cheliped, merus and carpus, medial aspect; g, second leg; h, third leg; /', third leg, dactylus. {a, b, g, h, scale A; c-f, scale B; /, scale C.) 130 PACIFIC SCIENCE, Vol. VII, January, 1953 of the Brevirostris Group and the only other base visible through hoods in lateral view; Hawaiian species with which it could be con­ much larger than eyes. fused, by the relative lengths of the second Antennal peduncle long, rather slender. carpal articles of the second legs, almost 3.0 Second article 3 times as long as broad, 1.3 times the length of the first article in C. times as long as first article, twice as long as platyunguiculata and only 0.7 its length in this third article. Stylocerite large, with lateral species; by the excessive setae on the third spine poorly developed, reaching only slight­ legs in C. platyunguiculata and by the thicker ly beyond curvature of scale; tip reaching chela in that species. Other characteristics for almost to end of first antennular article. Later­ the separation of the two species are found in al spine of basicerite well developed, slightly the frontal region of the carapace, the relative shorter than stylocerite. Scaphocerite large, lengths of the scaphocerite and the anten- with heavy outer spine; lateral margins slight­ nular and antennal peduncles, and the small ly concave, tip relatively heavy and somewhat chelae of the males and females. incurved; squamous portion narrow. Carpo- DISTRIBUTION: This specimen was collected cerite reaching beyond end of scaphocerite, on the sand flats of Kaneohe Bay, Oahu, which slightly exceeds end of antennular where the water is only a few inches deep at peduncle. low tide. For the possible habitat of this Third maxillipeds of usual form, but inner species, see the discussion under C. platyun­ faces of terminal and subterminal articles very guiculata. densely clothed with setae and short bristles. It has been reported from Zanzibar, Dji­ Large chela compressed, 2.3 times as high bouti, and the Mergui Archipelago (Andaman as thick, 2.3 times as long as high, margins Sea) in the Indian Ocean, from the Nether­ rounded, broadest near middle, slightly taper­ lands Indies, Japan, and near Queensland in ing proximally, somewhat more distally. Up­ the Pacific, but not previously from the cen­ per margin proximal to dactylus with deep tral Pacific region. transverse furrow expanding into shallow de­ pressed area on each face. Margin above Crangon platyunguiculata sp. nov. articulation of dactylus with shallow excava­ tion adjacent to palmar adhesive plaque for Fig. 47 a-k accommodation of crest of dactylus when TYPE SPECIMEN: A female 25 mm. long, flexed. Dactylus 0.3 length of chela, crested, collected at Waialua Bay, Oahu, from old strongly arcuate, heavy, 2.4 times as long as coral in about 6 feet of water. Paratypes, two wide, compressed like chela. Chela with nu­ males, one 20 mm., the other 10 mm. long, merous long setae on margins. Carpus of from the same location (U. S. N. M. 93461). usual form. Merus 2.5 times as long as broad, superior distal margin rounded, inferior in­ DESCRIPTION: Rostrum small, triangular, as ternal margin with 2 strong movable spines, long as broad at base, tip reaching less than one about middle, other more distal, weak 0.5 length of visible portion of first anten- but acute subterminal tooth; long setae found nular article. Rostral carina distinct and con­ only on distal portions of article. tinuous from tip of rostrum to between orbital hoods, thereafter more rounded and less well Small chela of female somewhat com­ defined but continuing to posterior margin pressed, 3.5 times as long as broad, fingers of hoods. Orbital hoods large, rounded an­ between 0.5 and 0.6 of length; surfaces with­ teriorly, frontal margin reaching to middle of out grooves or depressions; articulation of rostrum; posteriorly merging with convexity dactylus flanked by small subacute projection; of carapace; in lateral view higher than rostral fingers quite rounded. Entire chela heavily base; transparent, hence outlines of rostral covered with long setae, especially along up- Hawaiian Crangonidae •— BANNER 131

FIG. 47. Crangon platyunguiculatus sp. nov. a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect (setae not shown); d, large chela, dorsal aspect; e, large cheliped, merus, medial aspect (setae not shown);/, small cheliped, medial aspect; g, small chela, male, lateral aspect; h, second leg; /', third leg;/, third leg, dactylus; k, telson and uropods. {a, b,j, k, scale A; c-i, scale B.) 132 PACIFIC SCIENCE, Vol. VII, January, 1953 per and lower margins and margins of op­ men and in narrow band across eyes; similar posing faces of fingers. Carpus 0.25 as long poorly defined bands on chelae. In one live as chela, 1.2 times as long as broad. Merus specimen blue chromatophores expanded, similar in form to that of large chela but giving over-all blue-gray appearance; in other, without terminal tooth on inferior internal red chromatophores expanded, giving reddish margin, with 4 small movable spines instead. cast. Small chela of male 3.7 times as long as DISCUSSION: No significant variation was broad, fingers occupying slightly more than noted in the three specimens. 0.6 length of chela. Dactylus not rounded, RELATIONSHIP: This species plainly belongs but with opposing face flattened, with short to the Brevirostris Group and can be separated dense fringe of stiff setae along margins of easily from all other Hawaiian species save face in middle 0.7; fixed finger also flattened, C. rapax (Fabr.) by the appearance of the with corresponding fringe of setae in opposi­ anterior margin of the carapace, the com­ tion; other setae similar to those of female. pressed chela with a transverse groove, and Carpus and merus in male similar to those of the dactylus of the third legs. female. Within the Brevirostris Group this species Carpal articles of second legs with ratio and C. sauvensis (de Man) can be distinguished 10 : 23 : 7 : 10 : 11. by the characteristics given in de Man's key Ischium of third legs with strong movable (1911: 322-325). From most of the other spine. Merus 4.0 times as long as broad, species in the group these two can be dis­ inermous with occasional short setae. Carpus tinguished by the following characters: the 0.5 as long as merus, distal angles not pro­ presence of a transverse groove on the palm jecting but rounded; superior margin with of the large chela; the balaeniceps-shaped series of long setae. Propodus 0.6 as long as dactylus of the small chela of the males; the merus, tapering toward tip, straight, with 7 lack of a tooth on the merus of the third leg. movable spines of moderate size, with nu­ Three more closely related species, C. djedden- merous long setae on both margins, some of sis (Coutiere), C. djiboutensis (de Man), and setae 0.5 as long as article. Dactylus long, C. pubescens (de Man), all have a longer and simple, almost 0.3 as long as merus, superior thinner merus on the third leg; in addition margin rounded, inferior margin broad and C. djeddensis has a transverse groove on the flattened, somewhat curved; tip acute; lateral third chela; in C. djiboutensis the terminal spine margins with short movable spine and short of the scaphocerite is much longer relative to curved bristle. the squamous portion; and in C. pubescens the Telson with anterolateral margin convex, rostral carina is longer. broadest at about 0.3 of length, posterolateral C. platyunguiculata can be separated from margin concave and narrowing; 1.8 times as C. sauvensis by the characteristics given in the long as broad, tip 0.7 as broad as maximum following tabulation: breadth. Posterior margin strongly arcuate, C. platyunguiculatus C. sauvensis reaching beyond tip of posterolateral spines. Rostrum reaching beyond Rostrum reaching beyond orbital hoods by 0.5 its orbital hoods by 0.8 its Posterolateral spines about 0.3 as long as length length breadth of tip; dorsal spines poorly developed, Stylocerite with tip Stylocerite with tip in- located 0.4 and 0.6 of distance from articula­ straight curved Scaphocerite shorter than Scaphocerite as long as tion to tip. Tip of telson and distal lateral carpocerite carpocerite margin of inner uropod with weak spinules. Large chela with palm 2.0 Large chela with palm 1.5 Body translucent white to gray base color, times as long as fingers times as long as fingers Carpal articles of second Carpal articles of second with red and blue chromatophores arranged legs with ratio 10 : 23 : legs with ratio 10 : 15 : in poorly defined transverse bands on abdo­ 7 : 10 : 11 6:6:6 Hawaiian Crangonidae — BANNER 133

Propodus of third legs 1.2 Propodus of third legs 1.5 rounded, extending slightly posterior of eyes. times length of carpus times length of carpus Dactylus of third legs not Dactylus of third legs spi- Orbital hoods somewhat inflated, rounded spiniform, flattened on niform, not flattened anteriorly, demarked from base of rostrum by inferior face shallow rounded grooves. Anterior margin of On almost all other points the two species are carapace somewhat emarginate between orbi­ similar , so similar in fact that, with only one tal hoods and rostral base. or two of the differences listed above, the Visible portion of first antennular article Hawaiian form could be considered at most as long as second article; second article about a variety. However, the great difference in the 1.8 times as long as broad; third article about dactylus of the third leg reinforced with the 0.7 length of second. Tip of spine of stylo- other more minor differences leaves little cerite reaching to end of first article. doubt but that this species is valid. Lateral spine of basicerite reaching to tip The differences between this species and of rostrum. Scaphocerite strong, lateral mar­ C. rapax, its closest relative locally, are dis­ gin slightly concave, tip reaching consider­ cussed under that species. ably beyond end of antennular peduncle. DISTRIBUTION: In addition to the type Carpocerite longer than antennular peduncle, series listed above, two specimens were col­ reaching to end of squamous portion of lected from Kaneohe Bay, Oahu, where they scaphocerite. were found underneath coral heads on shallow Large chela rather slender, 3 times as long sand flats, less than a foot deep at low tide. as broad, compressed, with grooves and de­ In many of the areas of silty sand in Kaneohe pressions. Transverse groove on superior Bay there are numerous burrows evidently margin, with distal edge rounded and prox­ made by this species or by C. rapax (Fabr.). imal edge overhanging floor of groove; Some of these burrows are also inhabited by groove expanding on external face to elon­ specimens of a small fish. Unfortunately, I gate, narrow, triangular depression with poor­ was unable to collect either the crangonids or ly demarked margins; on internal face con­ the fish, as the burrows penetrated through tinuous with similar but less extensive the veneer of sand into the underlying con­ depressed area. On inferior margin, opposite solidated coral of the reef. It was also im­ groove on superior margin, notch with abrupt practical to seize the animals in their burrows, proximal edge and gradual distal edge; on to entice them out, or to drive them out. The external face notch continues, well-defined, two that were collected evidently had burrows for 0.3 height of face; on internal face notch that led under large loose coral heads lying poorly defined and continuing for only 0.25 on the surface of the sand; when the heads of height. Finger of usual form, less than 0.3 were moved the shrimp were caught in a dip length of entire chela. Merus 3 times as long net. as broad, with superior margin ending in obtuse angle; inferior internal margin with 2 Edwardsi Group movable spines and distal tooth. Crangon leptochirus (Coutiere) Small chela of female almost 5 times as long Fig. 48 a-h as broad, without sculpturing, fingers occupy­ ing almost 0.5 length of chela. Articulation Alpheus leptochirus Coutiere, Fauna and Geog. of dactylus flanked by small obtuse projection Maid, and Laccad. 2(4): 914-916, pi. 87, on inner side. Carpus with small tooth pro­ figs. 54-54e, 1906. jecting over propodus. Merus 4.5 times as DESCRIPTION: Rostrum triangular, about long as broad, almost 0.3 longer than merus 1.5 times as long as broad, tip reaching almost of large chela, with 3 movable spines and sub- to end of first antennular article. Rostral carina terminal tooth on inferior internal margin. 134 PACIFIC SCIENCE, Vol. VII, January, 1953

Carpal articles of second legs with ratio Oahu, in 100-200 feet of water. Coutiere's 10 : 11 : 5 : 6 : 7. specimens came from the Indian Ocean. DISCUSSION: Unfortunately all of the three specimens in the collection were dredged from Crangon crassimanus (Heller) deeper water and only one is reasonably in­ Fig. 49 a-k tact, the others having lost most of their ap­ Alpheus crassimanus Heller, Reise der . . . No- pendages in the rough handling in the dredge vara. . . . Zool. Theil 2(8): 107, pi. 10, fig. net. In the Edwardsi Group, the largest of the genus, positive identification cannot be made 2, 1865. without reference to the small chela of the DESCRIPTION: Rostrum acute, reaching al­ male; and, as the small chela of the sole male most to end of first antennular segment, with was lacking in these specimens, they are re­ margins gradually curving out to merge with ferred to this species with some doubts. frontal portion of orbital hoods and carapace, The specimens compare quite well with the demarked from orbital hoods by broad, shal­ description and plates of Coutiere except on low depression. Subacute rostral carina arising the following points: between anterior portion of orbital hoods. C. kptochirus Hawaiian Form Orbital hood hemispherical, without keels or Second antennular article Second antennular article ridges. 1.4 times length of visi- as long as visible portion Lateral spine of stylocerite slightly longer ble portion of first of first Scaphocerite exceeding Scaphocerite exceeding than first antennular article. Antennular pe­ antennular peduncle by antennular peduncle by duncle slim, with second article equal to or 0.1 its length 0.2 its length Large chela with parallel Large chela with slight ta- longer than visible portion of first, from 1.5 sides; proximal edge of per; proximal edge of to 3.0 times as long as third. Lateral spine of superior groove round- superior groove abrupt, basicerite not as long as stylocerite. Scapho­ ed; depressed areas on overhanging; depressed upper portions of faces areas on faces broader cerite with heavy lateral spine; lateral margin very narrow concave; squamous portion narrow, with an­ Meri of large and small Meri of large and small terior margin not rounded but joining spine chelae each with 4 spines chelae with 2-3 spines Second carpal article of Second carpal article of at acute angle. Scaphocerite and carpocerite second legs 0.7 length of second legs 1.1 length of almost equal, slightly longer than antennular first first peduncle. Third legs without ischial Third legs with ischial spine, merus 7.5 times as spine, merus 5.5 times as Large chela heavy, 2.2 times as long as long as broad long as broad broad, laterally compressed, 1.7 times as high These differences appear to be decisive as thick, with grooves and depressions. when listed out in this fashion; however, Rounded transverse groove on superior mar­ studies on other species in this group and in gin proximal to articulation of dactylus, con­ other groups of the genus have shown that tinuous with irregular depression with each of these characteristics are variable at rounded edges on both inner and outer face. least to the extent here observed. Therefore no Proximal margin of transverse groove, in pro­ dependence can be placed upon any one of file, rounded and sloping back to superior the differences listed; possibly in aggregate margin of chela. Another groove on inferior they are significant, but without the important margin of chela, directly below superior small chela of the male, and with but a single groove, rounded only distally and demarked intact specimen to study, I do not believe by abrupt shoulder proximally, ending that the designation of this form as a distinct abruptly on outer face about 0.2 distance from species is warranted. lower edge, and on inner, face gradually di­ DISTRIBUTION: The three specimens • were minishing but still visible beyond mid-line. taken in a dredge haul off Diamond Head, Dactylus 0.3-0.4 length of chela, heavy and Hawaiian Crangonidae •— BANNER 135

FIG. 48. Crangon leptochirus (Coutiere). a, b, Anterior region, dorsal and lateral aspects; c, large cheliped, lateral aspect; d, large chela, medial aspect; e, small chela, female; /, second leg; g, third leg; h, third leg, dactylus. (a, b, scale A; c-g, scale B; h, scale C.) 136 PACIFIC SCIENCE, Vol. VII, January, 1953

curved. Merus twice as long as broad; superi­ in Figure 4$c, d, e, f. In some specimens, es­ or margin unarmed; inferior internal margin pecially the smaller ones, the sculpturing is with tooth distally. very much reduced and noticeable only in the Small chela of usual form in female, sub- area adjacent to the margins; in larger speci­ spatulate in male. In female 3.4 times as long mens the shallow depression may be quite ex­ as broad, without sculpture; fingers slender, tensive over both faces, usually with rounded tapering, about as long as palm; short acute and ill-defined margins but at times with def­ tooth at point of articulation of dactylus inite limits. The tooth on the inferior internal usually present, as illustrated, but sometimes margin of the merus of the cheliped is usually reduced or completely absent. In male small present although it is not as large as in many chela relatively larger and heavier. 2.9 times other species; however, in some specimens it as long as broad, without sculpturing except may be reduced or entirely absent. In no speci­ for slight depressed area on upper surface men seen was the upper groove other than proximal to articulation of dactylus. Dactylus gradually rounded on both margins. 0.5 total length of chela, broad, flat, 2.7 times Small cheliped: In some of the larger females as long as broad, with dense fringe of setifer- the dactylus, although it never approaches the ous bristles along margins, crossing top of extreme condition found in males, is often article proximal to tip; tip compressed and proximally broadened, not slender and taper­ hooked (dactylus balaeniceps-shaped). In ing as in younger females. The tooth on the both sexes carpus projecting over propodus inner side of the articulation of the dactylus is as broad flat tooth, merus without tooth on almost always present, although sometimes re­ inferior internal margin, otherwise similar to duced; occasionally it is reduced to a rounded that of large chela. shoulder. No specimens were observed with Carpal articles of second legs with ratio a tooth on the merus like that of the large 10 : 10 : 5 : 5 : 8. chela. No males were found without the ba­ Ischium of third and fourth legs usually laeniceps-shaped dactylus. with movable spine; merus inermous; distal Second legs: The second article of the carpus margin of carpus not produced into acute varies from 0.75 to 1.3 times the length of the processes; propodus usually with about 10 first, and the fifth article from 0.5 to 1.0 times movable spines; dactylus simple, slightly the length of the first. curved. It has been observed that those members of Largest specimen, female from Kaneohe this species collected from Kaneohe Bay and Bay 26 mm. long; color usually transparent especially those collected on the inner arms of with reddish to greenish transverse bands. Pearl Harbor, where the water is very warm, DISCUSSION: Marked variation was noted in quiet, and filled with silt, are larger, averaging C. crassimanus, as in several other Hawaiian about twice the length of, and have a redder species of which numerous specimens were color than, specimens collected from coral collected. In this species the careful append­ boulders of the open reefs. No other charac­ age-by-appendage study of size ratios was not ters distinctive of the two groups of specimens made, but the following variations were noted. could be found. Antennular peduncles: The second article The local specimens agree well with the varies in relative length from as long as the original description of Heller, differing only visible portion of the first and 1.5 times as in one possible point. Heller states, "Anten- long as the third to twice as long as the first narum spina externa basalis minima," and, in and three times as long as the third. his plate, does not show a spine at all, while Large cheliped: The amount of sculpturing in the Hawaiian specimens it is well formed. on this appendage varies markedly as is shown The ratio of the articles of the carpus of the Hawaiian Crangonidae — BANNER 137

FIG. 49. Crangon crassimanus (Heller), a, b, Anterior region, dorsal and lateral aspects; c, d, large cheliped, lateral and medial aspects; e, f, large chela from another specimen, setae not shown, lateral and medial aspects; g, small cheliped, female; h, small chela, male, long setae not shown, lateral aspect; /, second leg; /', third leg; k, third leg, propodus and dactylus. {a, b, k, scale A; c-j, scale B.) 138 PACIFIC SCIENCE, Vol. VII, January, 1953 second leg (10 : 6 : 2 : 2 : 4, as taken from the distant from Hawaii as Djibouti, at the mouth drawings) are almost within the range of vari­ of the Red Sea, and Cape York, Australia. ation of the Hawaiian form. De Man (1911) suggests that possibly C. Crangon pacifica (Dana) crassimanus is a synonym for C. lobides (de Fig. 50 a-i Haan), but a study of the literature following Alpheus pacificus Dana, U. S. Explor. Exped. the publication of the note reveals that the 13 (1): 544, pi. 34, fig. 5, 1852. question has not yet been settled and that C. crassimanus is still accepted as a valid species. Alpheus pacificus Dana, Acad. Nat. Sci. Phila., Proc. 6: 21, 1854. [Diagnosis without fig­ One specimen of this species in my collec­ ures.] tion and two in the collections of the Bernice P. Bishop Museum (No. 5057) bear an un­ NEOTYPE: A male 32 mm. long, collected usual parasite attached to the posterior ven­ at Kaiama Park, about 18 miles southeast of tral portion of the thorax. It is in the form of Lahaina, Maui, from a coral head in about 3 a more or less bean-shaped body, whitish in feet of water (U.S.N.M. 93462). color and about 5 mm. long, which is attached DESCRIPTION: Rostrum acute, triangular, al­ by a short stalk. The specimens from the Ber­ most twice as long as broad at base, reaching nice P. Bishop Museum were described as nearly to 0.7 of visible portion of antennular new by Nierstrasz and Brandis (1930) as Faba peduncle; margins straight, with few short glabra, while my specimen was identified as setae. Rostral carina rounded, arising slightly this species by Dr. Edward C. Reinhard. The posterior to eyes. Orbital hoods not greatly parasite is of unknown systematic position, inflated, rounded anteriorly. Orbitorostral but the original authors suggest it may be an groove moderately shallow with gradually epicaridan crustacean. It is interesting to note sloping margins; anterior portion of carapace that, of all the crangonids from Hawaii, the between rostral base and front of orbital parasite has been found only on this one spe­ hoods concave. cies and that on that species it is extremely Second article of antennular peduncle about rare. 1.5 times as long as broad, 1.2 times as long DISTRIBUTION: This species is one of the as visible portion of first or third articles. most common in the shallow inshore waters Spine of stylocerite reaching slightly past end of the reefs. There is, in Hawaii, no record of of first antennular article. its collection from water over 5 or 6 feet deep. Lateral spine of basicerite reaching almost On Oahu it has been collected from the fol­ as far as spine of stylocerite. Scaphocerite with lowing localities: Kaneohe Bay, Koko Head, squamous portion well developed, more than Black Point, Waikiki, Pearl Harbor, Moku- 0.9 length of slender lateral spine; lateral mar­ leia, Haleiwa, Kaaawa, Punaluu; on Maui it gins only slightly curved. Carpocerite slightly was collected at Kaiama Park and 12 miles longer than scaphocerite; scaphocerite 0.1 south of Lahaina; on Molokai at Kainalu. longer than antennular peduncle. Antennal Edmondson (1925) reported the species from flagellum longer than body. Laysan and Lisianski Islands. The "Albatross" Large chela massive, 2.2 times as long as collected specimens in Pearl Harbor (U.S.N. broad, compressed, twice as broad as thick, M. 63559) and Honolulu Reef (U.S.N.M. with lobes and crests. Superior margin cut by 63560, 63562). deep transverse groove proximal to dactylus, In the central Pacific the species was re­ proximal edge of groove overhanging, distal ported by Edmondson (1925) from Johnston edge rounded; groove continued on outer Island. It ranges through the Indo-Pacific re­ face as shallow, poorly defined, depressed gion and has been reported from places as far area extending downward about 0.3 breadth Hawaiian Crangonidae — BANNER 139

FIG. 50. Crangon pacifica (Dana), a, b, Anterior region, dorsal and lateral aspects; c, d, large cheliped, lateral and medial aspects; e, small cheliped, male, lateral aspect;/, small chela, female, lateral aspect; g, second leg; h, third leg; /', telson. (a, b, i, scale A; c-f, scale B; g, h, scale C.) 140 PACIFIC SCIENCE, Vol. VII, January, 1953 of face and posteriorly beyond middle of slightly shorter than width of merus. Merus palm, continued on inner face as much smaller with several short setae; superior margin of triangular depressed area immediately below carpus and propodus with scattered, long groove. Inferior margin with much deeper setae. Brush on fifth legs well developed. notch below upper groove, the proximal edge Telson tapering only slightly in posterior overhanging, the distal edge abrupt but third, anterior margins convex; 2.2 times as rounded; notch continues more than 0.3 dis­ long as broad, maximum breadth 1.4 times tance up outer face with proximal edge well breadth of tip. Margin of tip only slightly ar­ defined throughout but distal edge fading; cuate. Posterolateral spines feeble, scarcely depression on inner face similar to that on reaching beyond rounded portion of tip; dor­ outer face but with both edges gradual. Dac­ sal spines heavy, 0.4 and 0.6 of distance from tylus 0.4 length of chela, heavy, with distal articulation to tip. Middle of posterior border margin strongly curved. Scattered setae on and distolateral margin of inner uropod with superior margin and inner distal face of chela small, irregularly placed spinules. and on fingers. Merus about 0.25 as long as Neotype about maximal size observed. chela, twice as long as broad, superior distal Color in life, banded red and transparent, with margin obtuse and lacking inner distal tooth; blue-violet markings on appendages, especial­ with occasional setae along margins. Large ly large chela. chela of female similar. DISCUSSION: Although not as much varia­ Small chela of male 3.4 times as long as tion was noted in this species as in the closely broad, fingers 0.6 of total length. Palm com­ related C. crassimanus (Heller) the following pressed, without grooves, ridges, or depres­ points are noteworthy: The length of the sec­ sions except for slight indentation at base of ond antennular article was as much as twice fixed finger. Both fingers curved distally so the length of the visible portion of the first in that points cross; both fingers with dense some smaller specimens; the sculpturing on rows of marginal setae proximally, becoming the upper parts of both faces of the large chela more scattered distally; oppositional faces of varied in extent; the number of setae on the both fingers somewhat flattened. Articulation opposing faces of the fingers of the small of dactylus not flanked by tooth. Carpus with chela also varied from the neotype, some hav­ tooth extending over base of propodus. Me­ ing more, some less; the second article of the rus similar in form to that of large chela. Small carpus of the second legs was almost as long chela of female of same proportions and arma­ as the first article in some specimens; and the ture as that of male except perhaps with few number of spines on the propodus of the third less setae proximally on fingers; size relatively legs varied slightly from the 11 found in the less than that of male. type specimen. In none of the specimens, Carpal articles of second legs with ratio however, was there a tooth on the merus of 10 : 8 : 3 : 3 : 5.5; second article of carpus 4.4 the large chela, nor at the articulation of the times as long as broad; chela as long as first dactylus of the small chela; in no case was the carpal article. proximal margin of the superior groove of the Ischium of third legs with strong movable large chela other than overhanging the floor spine; merus 4 times as long as broad, taper­ of the groove nor was the small chela of the ing slightly distally, unarmed; carpus about males approaching the condition found in C. 0.5 as long as merus, neither superior nor in­ crassimanus (balaeniceps-shaped). ferior distal margin projecting distally as acute Dana's original specimen, now lost, was tooth; propodus 0.75 as long as merus, with very large, V>A inches (or about 45 mm.), and 11 moderately large spines on inferior and dis­ was collected at some undesignated spot in tal margins; dactylus acute, curved, simple, the Hawaiian Islands. The neotype, the larg- Hawaiian Crangonidae — BANNER 141

est in the present collection, unfortunately is does penetrate as far into Kaneohe Bay); it also somewhat smaller (32 mm.), and was collected reaches a bit further beyond the tidal zone on Maui near where Dana had collected other having been collected from water 15 feet deep Hawaiian types (see under C. diadema and C. at Nanakuli, Oahu. It has been collected from pugnax). It agrees well with Dana's description the following localities on Oahu: Kaneohe and plates except in the ratio of the articles of Bay, Waimanalo, Hanauma Bay, Black Point, the carpus of the second leg which according Waikiki, Nanakuli, Mokuleia, Kawela Bay, to his plate had a ratio of 10 : 5 : 3 : 2.5 : 4, and Kahana Bay; on Maui it has been col­ which is 10 : 8 : 3 : 3 : 5.5 in this specimen. lected at Makena, Kalama Park, Lahaina, and While this character was quite variable in the 12 miles south and 10 miles north of Lahaina. specimens in the present collection, in no case The "Albatross" collected specimens on Ho­ did it approach the extreme given by Dana; nolulu Reef and at Waialua, Oahu (U.S.N.M. however, it is likely that this difference still 63590 and 63593) and Napili Harbor Reef, can be individual variation or possibly a Maui (Station No. 3881, U.S.N.M. 63592). growth difference. A second more slight dif­ Edmondson (1925) has reported the species ference is that Dana's plates do not show the from Laysan and Ocean Islands. proximal margin of the superior groove of the large chela overhanging the groove; this char­ It is widely spread through the Indo-Pacific acter, too, is somewhat variable in these speci­ region, from the Red Sea and Madagascar to mens, and may well be either an individual or New South Wales and Wake Island. a growth difference.

This species resembles C. crassimanus (Hel­ Species of Doubtful Record ler) very much, and there is no positive way to distinguish between them except by the char­ acter of the first thoracic legs. The small chela Crangon strenua (Dana) in the male is subspatulate in C. crassimanus and of the usual form in both the male and Alpheus strenuus Dana, U. S. Explor. Exped. female in C. pacifica. On the large chela the 13(1): 543, pi. 34, fig. 4, 1852. most useful character is the presence of a tooth on the internal distal angle of the merus Originally described from Tongatabu, in C. crassimanus, with none in C. pacifica; Tonga, this species is of wide distribution in however, this tooth is variable in size and the south Indo-Pacific, from the Galapagos often very much reduced. The proximal edge on the east to Djibouti at the mouth of the of the upper depression is rounded in C. cras­ Red Sea on the west. It was reported from an simanus and usually overhangs in C. pacifica, unspecified locality in Hawaii by Stimpson but some specimens of both species approach (1861) as A. avarus Fabr. (he regarded A. the perpendicular. The superior depression on strenuus as a synonym for A. avarus). How­ the internal face usually extends posteriorly ever,- as the species has not been reported past the mid-line of the palm in C. crassimanus from this archipelago since that time, and as and scarcely back of the upper edge of the de­ it can be confused most easily with the com­ pression in C. pacifica. mon C. crassimanus (Heller) (the two species can be separated with certainty by the pres­ DISTRIBUTION: This species, like C. crassi­ ence of a balaeniceps-shaped small chela in manus, is common in shallow waters. It does both the male and female in C. strenua in not seem to penetrate the inshore portions of contrast to having the condition only in the the reefs nor the quieter waters of enclosed male in C. crassimanus), it is likely that Stimp­ bays as much as C. crassimanus (although it son made a mistake in identification. 142 PACIFIC SCIENCE, Vol. VII, January, 1953

C. audouini (Coutiere) seum. Soc. Ent. de France, Bui. 1898 (6): 149- 152; (7): 166. Alpheus audouini Coutiere, Fauna and Geog. 1898^. Note sur Synalpheus biunguicu- Maid, and Laccad. 2: 911-913, pi. 87, fig. latus Stimpson? deMan [Crust.]. Soc. Ent. de 52, 1905. France, Bui. 1898 (11): 232-233. This species was reported from Hawaii by 1899- Les "Alpheidae," morphologie Coutiere in his original description: "L'espece externe et interne, formes larvaires, biono- se trouve dans toute la Mer Rouge, l'Ocean mie. Theses presentees a la Faculte des Sci­ Indien, la Malaisie, et de la Nlle Zelande aux ences de Paris. . . . Ser. A, No. 321 [No. lies Sandwich." Unless this species is con­ d'Ordre 980]. 599 pp., 409 text figs., 6 pis. fused with the closely related C. pacifica (in Masson et Cie., Paris. [Also in: Ann. Sci. which case the name C. pacifica would take Nat., VIII, Zool. 9: 1-560.] priority), it has not been reported from the 1903. Note sur quelques Alpheidae Islands since, nor was it found in the present des Maldives et Laquedives. Soc. Philomath. collections. Therefore it seems likely that Paris IK, 5(2): 72-91. Coutiere's sweeping distributional list was a 1906. Marine Crustacea, XV. Les Al­ little too inclusive. pheidae. In: Fauna and Geography of the Mal- dive and Laccadive Archipelagoes. [J. Stanley REFERENCES Gardiner, ed.] Vol. 2, Rpt. 21, pp. 852-921, ANONYMOUS. 1904. Records of the dredging pis. 70-87, text figs. 127-139. University and other collecting stations of the U. S. Press, Cambridge. Fish Commission Steamer Albatross in 1901 1908. Sur quelques nouvelles especes and 1902. Report of the Commissioner for the d'Alpheidae. Soc. Philomath. Paris, Bui. IX, Year Ending June 30, 1902, U. S. Commission 11(5): 1-26. of Fish and Fisheries 1902: 399-432. 1909. The American species of snap­ BALSS, HEINRICK. 1927. Decapoda. In: Hand- ping shrimps of the genus Synalpheus. U. S. buch derZoologie [Willy Kiikenthal, ed.] Bd. Natl. Mus., Proc. 36: 1-93, text figs. 1-54. Ill, Pt. 1: 841-1082. W. de Gruyter u. Co., 1921. Les especes d'Alpheidae rap- Berlin u. Leipzig. portees de l'Ocean Indien par M. J. Stan­ BATE, C. SPENCE. 1888. Report on the Crus­ ley Gardiner. Linn. Soc. London, Trans. II tacea Macrura dredged by H. M. S. Chal­ (Zool.) 17(4): 413-428, pis. 60-64. [Also lenger during the years 1873-76. In: The titled: Reports of the Percy Sladen Trust Expe­ Voyage ofH. M. S. Challenger, Zoology 24. xc dition to the Indian Ocean in 1905. . . . 6(10): + 942 [157 pis. in separate vol.]. Eyre & 413-428, etc.] Spottiswoode, London. DANA, JAMES D. 1852. Crustacea. In: United BOONE, LEE, 1935. Scientific results of the • States Exploring Expedition, During the Years world cruise of the yacht "Alva" 1931, Wil­ 1838, 1839, 1840, 1841, 1842, under the liam K. Vanderbilt, Commanding. Crusta­ Command of Charles Wilkes, U. S. N. Vol. 13, cea. . . . Vanderbilt Mar. Mus., Bui. 6: 1- viii + 685. [Folio Atlas of 96 plates pub­ 264, 96 pis. lished in 1855.] COUTIERE, HENRI. 1896. Note sur quelques • 1854. Conspectus Crustaceorum, etc. genres nouveaux ou peu connus d'Alphe­ Conspectus of the Crustacea of the Explor­ ides, formant la sous-famille des Alpheops- ing Expedition under Capt. Wilkes, U. S. N. ides. Paris Mus. d'Hist. Nat., Bui. 2(8): 380- Acad. Nat. Sci. Phila., Proc. 6: 6-28 [Proc. 386. for 1852]. 1898-2. Note sur quelques Alpheides DE MAN, JOHANNES GOVERTUS. 1909. Note nouveaux de la collection du British Mu­ sur quelques especes du genre Alpheus Hawaiian Crangonidae — BANNER 143

Fabr. appartenant au Group Brevirostris. fehlen des Commodores B. vonWullerstorf-Ur- Soc. Zool. de France, Mem. 22: 146-164, pis. bair. Zool. Theil. Bd. II. 280 pp., 25 Taf. 7-8. HILGENDORF, FRANZ MARTIN. 1878. Die von 1910. Diagnoses of new species of Herrn W. Peters in Mozambique gesam- macrurous decapod Crustacea from the "Si- melten Crustaceen. K. Akad. Wiss. Berlin, boga Expedition," V. Ned. Dierk. Ver., Monatsber. 1878: 782-851, 4 Taf. Tijdschr. II, 11(4): 287-319. HULT, JORAN. 1938. Crustacea Decapoda 1911. The Decapoda of the Siboga Ex­ from the Galapagos Islands collected by pedition. Part II. Family Alpheidae. In: 1 Mr. Rolf Blomberg. Arch, for Zool. 30(5): Siboga-Expeditie 39a (2): 133-327 [Livre 1-18. 60]; 1915, Supplement [Plates and explana­ tions] 39a1 (2): 23 pis. [Livre 74]. INTERNATIONAL COMMISSION ON ZOOLOGI­ CAL NOMENCLATURE. 1910. Opinions [1- 1922. The Decapoda of the Siboga Ex­ 25] rendered by the International Commis­ pedition. On a collection of macrurous dec­ sion on Zoological Nomenclature. Smithsn. apod Crustacea of the Siboga Expedition, Inst., Pub. 1938: 1-61. [Opinion 17, pp. chiefly Penaeidae and Alpheidae. In: Siboga- 40-42.] Expeditie 39a4 (5): 1-51, pis. 1-4 [Livre 93]. 1922. Opinions rendered by the In­ EDMONDSON, CHARLES HOWARD. 1923. Crus­ ternational Commission on Zoological No­ tacea from Palmyra and Fanning Islands. menclature, Opinions 68 to 77. Smithsn. Bernice P. Bishop Mus., Bui. 5: 3-43. Inst., Misc. Pub. 73(1): 1-73. [Opinion 74, • 1925. Crustacea. Marine zoology of pp. 32-34.] the tropical Central Pacific. Bernice P. Bishop Mus., Bui. 27: 3-22. LENZ, HEINRICH. 1901. Ergebnisse einer Reise 1930. New Hawaiian Crustacea. Ber­ nach dem Pacific. Crustaceen. Zool. Jahrb., nice P. Bishop Mus., Occas. Papers 9(10): Abt. f. System., Geog. u. Biol. Tiere 14(5): 3-18. 429-482. 1933. Reef and Shore Fauna of Hawaii. MIERS, EDWARD JOHN. 1875. On some new 295 pp. Bernice P. Bishop Museum, Special or undescribed species of Crustacea from Publication 22, Honolulu. [Revised 1946.] the Samoan Islands. Ann. and Mag. Nat: FABRICIUS, JOHANN C. 1898. Supplementum En- Hist. IV, 9: 341-344. tomologia systematica emendata et aucta .... 1881. On a collection of Crustacea 572 pp. Hafniae [Copenhagen]. made by Baron Hermann Maltzan at Goree Island, Senegambia. Ann. and Mag. Nat. GURNEY, ROBERT. 1938. The larvae of the decapod Crustacea. In: Great Barrier Reef Hist. V, 8: 204-377, 4 pis. Expedition 1928-1929, Scientific Reports. 6(1): 1884. Crustacea. In: Report on the Zoo­ 1-60. British Museum (Natural History), logical Collections made in the Indo-Pacific London. Ocean during the Voyage ofH. M. S. "Alert," 1881-2, xxv + 684 pp., 54 pis. British Mu­ HASKELL, DANIEL CARL. 1942. The United seum (Natural History), London. States Exploring Expedition 1838-1842 .... xii + 188 pp. The New York Public Li­ MILNE-EDWARDS, HENRI. 1837. Histoire natu- brary, New York. relle des crustaces, comprenant I'anatomie, la HELLER, CAMIL. 1861. Beitrage zur Crustace- physiologie et la classification de ces animaux. en-Fauna des Rothen Meeres. K. Acad. Tome II. 532 pp. Roret, Paris. Wiss. Wien, Sitzungber. 44: 241-295, 3 Taf. NIERSTRASZ, H. F., and G. A. BRENDER A 1865. Crustaceen. [Abth. 8] In: Reise BRANDIS. 1930. Three new genera and five der Osterreichischen Fregatte Novara um die new species of parasitic Crustacea. U. S. Erde in den Jahren 1857-58-59 unter den Be- Natl. Mus., Proc. 77(9): 1-9- 144 PACIFIC SCIENCE, Vol. VII, January, 1953

POCOCK, R. I. 1890. Crustacea of Fernando York under the earlier date as Vol. X of Noronha. Linn. Soc. London, Jour., Zool. Alaska.] 1890: 506-529, pi. 20. RICHTERS, F. 1880. Crustacea Decapoda of RAMADAN, M. M. 1936. Report on a collec­ Mauritius and the Seychelles. In: Beitrage tion of Stomatopoda and Decapoda from zur Meeresfauna der Insel Mauritius und der Ghardaqa, Red Sea. Egypt. Univ., Faculty Seychellen. . . . [K. Mobius, ed.] vi + 352 Sri., Bui. (6): 1-43, pis. 1-2. pp., 25 pis. Gurmann, Berlin. SCHMITT, WALDO L. 1939. Decapoda and RANDALL, JOHN W. 1839- Catalog of the other Crustacea collected on the presiden­ Crustacea brought by Thomas Nuttall and tial cruise of 1938. Smithsn. Inst., Misc. Col­ J. K. Townsend, from the west coast of lect. 98(6): 1-29. North America and the Sandwich Islands STIMPSON, W. 1861. Prodromus descriptions .... Acad. Nat. Sri. Phila., Jour. 8(1): 106- animalium evertebratorum, quae in expedi- 147. tione ad Oceanum Paciflcum Septentrion- RATHBUN, MARY J. 1904.2. Some changes in alem. . . . Pars VIII. Crustacea Macrura. crustacean nomenclature. Wash. Biol. Soc, Acad. Nat. Sri. Phila., Proc. 12: 22-47. Proc. 17: 169-172. [Proc. for I860.] 1904£ and 1910. Decapod crustaceans WEBER, FRIEDRICH. 1795. Nomenclator ento- of the northwest coast of North America. mologicus secundum entomologiam systematicam Smithsn. Inst., Harriman Alaska Series 10: ill. Fabricii. . . . v'm + 171 pp. Chilonii et 1-210. [First privately published in New Hamburgi. News Notes

The Congress of the Republic of the Philip­ scientist with many years' experience in agri­ pines, in a special session, has made funds avail­ cultural development in Indonesia. able for expenses in connection with the Eighth In association with Dr. E. M. Ojala, Deputy Pacific Science Congress. The Eighth Congress Chairman of the Research Council, Dr. Kroon is to be held in Quezon City, The Philippines, will be responsible for carrying out the Com­ from November 16 to 28, 1953. Dr. Vidal A. mission's projects for fostering the economic Tan is President of the Congress, and Dr. Patro- development of Pacific territories. He will work cinio Valenzuela is Secretary-General. (From in collaboration with the Executive Officers in the Pacific Science Association Information Bul­ charge of the two other main fields of Com­ letin.) mission activity — health (Dr. Emile Massal) and social development (H. E. Maude). According to Dr. Gilbert Archey, Secretary- General of the Seventh Pacific Science Congress, From the summer number of the American the final volume of the Proceedings, of the Sev­ Council of Learned Societies Newsletter we re­ enth Congress is now in the hands of the print­ print, by permission, this delightful poetic ers. All volumes should be issued by February comment by Kit Berry: or March of 1953. The Meteorology and Oceanography volume is being distributed at O SYLLABUS, O SYLLABI the present time. Enquiries and orders for the Servitudo Scholastica Grandiflora set of the Proceedings or for the separate vol­ umes are being handled by the Secretary, Royal "The Oxford don, with brow so high, Society of New Zealand, Victoria University Pronounces it 'gladioli'", College Buildings, Wellington. (From the Pa­ And Cambridge, Mass., aspiring higher, cific Science Association Information Bulletin.) Builds Babel's tower the heavens nigher. O syllabus, O syllabi. O octopus, O octopi. The antique hybrids still appear, Dr. Alexander Spoehr has been appointed With Grecian front and Latin rear, Director of the Bernice P. Bishop Museum, Embodied in our English tongue Honolulu, succeeding the late Sir Peter Buck. By no true graft, but stiffly hung, Dr. Spoehr, who is Curator of Oceanic Eth­ Mere fossils from an ancient lore, nology at the Chicago Natural History Museum, Which learning with less effort wore, will assume the directorship in January, 1953. And, when it grafted, grafted true, He is an authority on native kinship and social And made a hybrid live and new. organization, and is recognized as an expert on O syllabus, O syllabi. field and museum technique. (From the Pacific O omnibus, O omnibi? ? Science Association Information Bulletin.) Yet it has been said by a liberal scholar that The South Pacific Commission has appointed "Culture is forgotten learning," and by a as its Executive Officer for Economic Develop­ prophet that "The letter killeth, but the spirit ment, Dr. A. H. J. Kroon, a Dutch agricultural giveth life."

[145]

Index to Crangonids Names marked with an asterisk (*) are either new or redefined, those in italics are synonyms. Page numbers in bold type indicate descriptions.

Alpheus (see Crangon or Synalpheus for species). *pacifica, 48, 51, 138 9, 46 parabrevipes, 107 Alpheopsis, 10, 14, 21, 30 paracrinita, 48, 50, 110 key, 15 bengalensis, 48, 50, 110, 114 aequalis, 15 paradentipes, 48, 49, 72 * biunguiculatus, 15, 18 paragracilis, 47, 48, 50, 92, 96 chilensis, 15 paralcyone, 48, 50, 99 equalis, 15 (?) paralcyone, 103 fissipes, 20 percyi, 48, 51, 114, 124 idiocarpus, 14 philoctetes, 124 Arete, 20, 30 *platyunguiculata, 48, 51, 128, 130 Areteopsis, 20 *pseudopugnax, 48, 51, 114, 122 Athanas, 20 pubescens, 132 Athanopsis, 20 pugnax, 48, 51, 114, 116 Automate, 20 rapax, 48, 51, 126, 133 rostratipes, 93, 94 Betaeus, 20 sauvensis, 132 harrimani, 5 seurati, 67 truncatus, 5 shearmei, 81 splendida, 85 Crangon, 9, 10, 20, 46 staphylinus, 63 key, 48 strenua, 48, 141 acuto-femorata, 107 *tuthilli, 48, 49, 55, 63 *albatrossae, 48, 49, 55, 60, 64 ventrosa, 48, 49, 84 amirantei, 48, 49, 87 Crangonidae, 8 audouini, 48, 142 key to Hawaiian genera, 10 avarus, 141 *brachymerus, 48, 49, 55, 58 Jousseaumea, 10, 21 *brevipes, 48, 50, 103 key, 10 brevirostris, 126 brevirostris, 10, 12 bucephalus, 99, 102 mauiensis, 10, 12 clypeata, 48, 50, 106, 107 collumiana, 48, 49, 67 Metalpbeus, 90, 93 crassimanus, 48, 51, 134, 141 dasycheles, 126 "Neoalpheopsis, 10, 20 dentipes, 70 key, 21 deuteropus, 48, 49, 70 euryone, 21, 25 *diadema, 48, 51, 114, 118 *hiatti, 21 djeddensis, 132 djiboutensis, 132 Ogyris, 20, 30 *edmondsoni, 48, 49, 78 Parabetaeus, 20 faceta, 84 Pterocaris, 20, 30 gracilipes, 48, 51, 115 gracilis, 75, 76 Synalpheus, 10, 20, 26 alluaudi, 76 key, 29 luciparensis, 76 albatrossi, 28, 29, 30, 36 *simplex, 48, 49, 75, 77, 81 *biunguiculatus, 29, 32, 36 hailstonei, 51, 54 charon, 28, 29, 37, 40 assimulans, 54, 55 *coutierei, 37 hailstonei, 54, 55 gravieri, 40 laetabilis, 54, 55 helleri, 38 paucispinata, 48, 49, 51, 60 lockingtoni, 42 hawaiiensis, 48, 50, 92, 95 macromanus, 29, 45 clippertoni, 96 paraneomeris, 28, 29, 36, 40, 46 insignis, 118 halmaherensis, 42 laevis, 84 oxyceros, 42 *latipes, 48, 49, 81, 82 praedabundus, 42 laysani, 99 prasalini, 42 leptochirus, 48, 51, 133 prolatus, 42 lobides, 138 seychellensis, 42 malabaricus, 126 prolificus, 28, 29, 38 * nanus, 47, 48, 50, 90 * redactocarpus, 26, 29 obesomanus, 86 townsendi, 28, 40, 44 *oahuensis, 48, 49, 55, 64 brevispinis, 28

1147]