Floristic Interchanges Between Formosa and the Philippines

HUI-LIN LI I

INTRODUCTION However, though the basicconstituents of the two floras are quite different, indicating THE FLORAS of Formosa and the Philippines their separate origins, their close geographic are, in spite of their geographic proximity, proximity allows certain floristic interchanges, distinctive in their general nature. The twBashi Channel), probably in the early Luzon , is about 350 kilometers. Stretching in Tertiary, long' before their connections with the sea between the two is a chain of small other respective neighboring areas were sev­ islands, parts of the same volcanic line. The ered. Thus the bulk of the Formosan flora was first group consists of three islands off the derived from China or from other floras that southeastern coast of Formosa, Hoshaotao spread to Formosa through the Chinese main­ (Kotosho in' J apanese), now called Lutao, land, while the bulk of the Philippine flora Hungtauyu (Katosho inJapanese), now called was derived from Mal aysia and other lands Lanyu and commonly known as Botel Toba­ south of the Asiatic continent. go, and the tiny island, Little Botel Tobago. The distinctiveness of the two floras has These islands are separated from the Baran been amply demonstrated by Merrill (1923). (or Batanes or Bashi) Islands by the Bashi It will suffice to mention here the many tropi­ Channel at a distance of about 100 kilo­ cal and southern families that are present in meters . The Baran Islands are separated from the Philippines but absent from Formosa, the Babuyan Islands, which are close to the such as the Dipterocarpaceae, Centrolepida­ northern coast of Luzon, by the 'Balintang ceae, Monimiaceae, Nepenthaceae, Cunonia­ Channel, about 70 kilometers wide. The dis­ ceae, Erythroxylaceae, Dichapetalaceae, Da­ tance is nowhere great enough to make plant tiscaceae, Clethraceae, Epacridaceae, and Sal­ migration difficult. At South Cape, the south­ vadoraceae. On the other hand, many other ernmost tip of Formosa, the small island families, clearly of northern origin, occur in Y'Arni of the Baran group, the northernmost Formosa but are entirely absent from the ' of the Philippine Islands, is actually within Philippines, such as Cephalotaxaceae, Taxo­ sight. diaceae, Cupressaceae, Betulaceae, Lardiza­ These small islands show remarkable affini­ balaceae, Trochodendraceae, Papaveraceae, ties with the Philippine flora, even the north­ Geraniaceae , Callitrichaceae, Pyrolaceae, .Dia­ ernmost ones, Lutao and Lanyu, which are pensiaceae, Valerianaceae, and Dipsaceae. closest to Formosa . The conspicuous rela­ tionship of the floras and faunas of the two 1 Research Associate, M orris Arboretum, Unive rsity of Pennsylvania, Philadelphia. Manuscript re cei ve ~l)ul y latter islands to the Philippines has led some 1 1952. The author is indebted to Dr. A. C. Smith of to suggest extending the Neo-Wallace Line the Smithsonian Institution for his kindness in reading northward from the Philippines, passing it the manuscript and offering sug gestions for its im­ provement. between the main island of Formosa and these

179 180 PACIFIC SCIENCE, Vol. VII, April, 1953

two small islands (Kano, 1931). However, it largely confined to medium and high altitudes has been shown that the southern tip of For­ in northern Luzon. Such characteristically mosa, commonly known as the Hunchuen northern or Asiatic elements, mostly common (Koshun) Peninsula, has a flora quite dis­ genera in the mountains of Formosa, such as tinct from that ofthe Formosan mainland but Lilium, Liriope, Saururas, Tbesium, A rmaria, very close to that of these two small islands, Sedam, Duetzia, Rosa, Skimmia, BlIXlIS, Pistacia, showing similarly. distinctive relationships Androsace, Hoppea, Salvia, Ellisiophylillm, Hemi­ with the Philippine flora. There is a marked phragma, Peracarpa, Aster, Solidago, Anisopap­ infiltration of austral elements from the Phi­ pllS, and Artemisia, occur nowhere in the Ma­ lippines to Formosa, particularly in the ex­ layan region outside ofnorthern Luzon (Mer­ treme southern part of the island and in these rill, 1923-26). It is. apparent that Formosa is t~o small islands (Li and Keng, 1950). an important route of migration of these Asiatic types to the regions in the south (Van ' SOUTHWARD MOVEMENT OF Steenis, 1946). TEMPERATE ELEMENTS We may picture that during the Pleistocene Before the inception of this northward mi­ or earlier, when the general temperature was gration of relatively recent occurrence, as will much lower than now, these montane species be discussed later, there occurred another ex­ would have inhabited lower elevations and pansion of floristic elements in the opposite thus had greater ranges. Their ranges might direction, that is, from Formosa to the Phi­ have been more or less continuous, extending lippines . These elements, in contrast to the from the Himalayasand western China through southern tropical elements, mostly ofthe low­ the mountains of southern China to Formosa lands, are cold-temperate in nature and lim­ and Luzon . Subsequent rising of tempera­ ited to montane or alpine regions. In Formosa, tures forced these plants to higher elevations because of the massive and lofty mountain until finally they occupied the present iso­ ranges which occupy most of the central por­ lated mountain regions suitable for their exist­ tion of the island, there developed an exten­ ence. Thus, species like Hemiphragma betero­ sive alpine-montane flora rich in number of phyllllm Wall. and Ellisiophylillm pinnatum species. These plants show distinctively close (Wall.) Makino now have widely disjunct relationships with those ofwestern China and ranges on the high mountains of Luzon, For­ the eastern Himalayas and also to some ex­ mosa, western China, and the eastern Hima­ tent with the montane flora ofJapan in the layas, etc. north. These Philippine plants of northern origins There are over 40 peaks in the Formosan are sometimes specifically or subspecifically mountain .systems that exceed 1,000 meters distinct from their congeners in Formosa, in height. The highest peak is Yu Shan, or while others may be specifically identical. Mt. Morrison, towering to a height of 1,950 That some of these show morphological dif­ meters, the highest of all eastern Asia. Such ferentiation indicates that their separation an extensive, high mountain chain permits the must have been for a considerable length 'of existence of an alpine flora that otherwise time. As such plants are now confined to exists only in the Himalayas, western China, isolated and distant mountain regions and as and other mountain regions far north of For­ there were drastic .changes in climate during mosa. the geologically recent past, such a south­ From Formosa, a large number of tem­ ward migration of the montane flora may be perate elements, mostly of mountain regions , considered largely a matter of the past, there extend southward to the Philippines. In the being little possibility that the same process Philippine Islands, these Asiatic elements are is goin g on extensively at the present. Floristic Interchanges- LI 181

NORTHWARD MOVEMENT OF . the coastal lowlands. Their wide ranges are TROPICAL ELEMENTS due apparently to their more aggressive na­ In contrast to the southward migration of ture and especially to their wider tolerance of these ternperate elements , the northwardmove­ temperature extremes. ment oftropical elements from the Philippines In the second group are species like Acacia to Formosa is limited practically to the low­ con/usa Merr., Ipomoea polymorpha R. & S., and lands and particularly to the southern ex­ Oreocnide trineruis Wedd., which occur widely treme of the island of Formosa. Favored by in the Philippines and extend also to Formosa, the gradual rising of temperature since the especially along the coasts and lowlands, from Pleistocene, these tropical, or southern, plants south to north. The number of such species that find their way to Formosa are able to with wide ranges in Formosa is few, indicating establish themselves there. the insignificance of such elements and pos­ Apparently many strand plants migrate by sibly also their recent and scattered arrival. the help of ocean currents. In the case of There are many more species common to Formosa, the main current is supplied by the the Philippines and Formosa but confined in Japan Stream from the south which passes the latter island to the southern part. These along the island chain connecting the Philip­ species are sometimes limited only to the pines and Formosa. This chain ofsmall islands southernmost tip of the island-the Hun­ also acts as a series of stepping stones for the chuen Peninsula and the small islands of Lu­ migration of plants, especially from south to tao and Lanyu. Such species as Illigera luzon­ north. This northward migration of plants is ensis Merr ., Scbizostacbyum diffttsttm Merr., also aided by wind . The prevalent monsoon GoniothalamttS amttyon Merr. , Aglaia elliptici­ during the summer and fall and the general folia Merr. , Maba bttxifolia Pers., Guettarda track of typhoons rampant in the eastern Paci­ speciosa 1., Morinda citrifolia 1., Pempbis acidula fic during this same period of the year are also Forst., etc., are often important elements in in a northeastern direction from the tropics, the local vegetation. That their occurrence is often leading from the Philippines to For­ often limited to the seashore and coastal areas mosa. in Formosa distinctly indicates that they are That these islands serve as an efficient mi­ relatively recent arrivals. gration route of the northward movement of The flora of the two small islands, Lutao southern plants can be demonstrated by the and lanyu, off the southeast coast of Formosa presence of the following floral elements in is much more intimately related to the south­ Formosa: (1) species of wide tropical distri­ ernmost .part of Formosa than is generally bution, (2) species common to the Philip­ understood (l i and Keng, 1950). The south­ pines and Formosa, (3) species common to ernmost tip of Formosa manifests close floris­ the Philippines and the southernmost part of tic relationships with the small islands as well Formosa, and (4) species common to the as with the Philippines in general. Its flora is Philippines and the ' islands of Lutao and quite distinct from that of northern Formosa . Lanyu. The fact that some species are present on these In the first group are many paleotropic two small islands but absent in the southern­ species that extend to the whole island of most part of Formosa is due apparently to Formosa but more commonly only to the human factors, as the mainland of Formosa southern part of the island . Such common has been more severely exploited. As a result, species are Dodanea viscosa (1.) Jacq., Croton there are a number of Philippine species that cumingi« Muell.-Arg., Litsea cubeba Pers., and are present on these islands but do not occur jltfurrya paniculata (1.) Jack. Those that also on the mainland of Formosa, such as Myris­ occur in the northern part are mostly along tica cagayanensis Merr., Boerlagiodendron pecti- 182 PACIFIC SCIENCE, Vol. VII, April, 1953 natum Merr., Dysoxylam cumingian« DC., . The floras of Formosa and the Philippines Styrax kotoensis Hay ., Turpinia lacid« Nakai, have been studied, in most cases, independ­ Macaranga dipterocarpiftliaMerr., etc. ently of one another. Many of the species The importance in floristic migration played proposed from Formosa are later found to be by the chain of small islands between the conspecific with earlier-named Philippine spe­ Philippines and Formos a can be illustrated cies. Such scattered findin gs are increasing in more. precisely by a few cases. Among the number. More recently, with ample reference wide paleotropic species, chiefly along or near collections from the Philippine Islands at the coast and common throughout the Philip­ hand, the present writer restu died many For­ pine Islands, are M orinda citrifolia 1. and mosan plants in connection with related spe­ Maba btlxifolia (Rott.) Pers. ( = Diospyrosfem a cies from neighboring regions. It was dis­ Bakh.). These species extend through the covered that man y species, especially those Baran and Babuyan Islands to Formosa, lim­ considered endemic to southern Formosa or ited in the latter case only to lutao and lanyu lanyu (Botel Tobago), are synon ymous with and the southernmost Hunchuen Peninsula. certain generally widespread species of the Among the less wide species are Aglaia for­ Philippine Island s, especially of luzon and mosana Hay . and A . ellipticiftlia Merr ., both the northern small islands. Some of these ofthe northern part of luzon. These species notes have been publi shed in a few scattered also extend northward, following the same papers (l i, 1950, 1952). A number of recent route, through the Baran and Babuyan Is­ findings are enumerated below. Cited speci­ lands to Hunchuen Peninsula in Formosa and mens are selected from the herbarium of the to lutao and l anyu. National Taiwan University, Formosa (NTU) , That these southern tropical elements are and the U.S. National Herbarium , Smith­ relatively recent arrivals in Formosa can be sonian Institution (US). attested by their occurrence in coastal regions and in second ary forests as well as by their Urticaceae specific identity with Philippine plants. In 1. l aportea batanensis C. B. Robinson in most cases, the species in Formosa are exactly Philip. Jour. Sci. 5. Bot .: 481, 1910. the same as in the Philippin es. These species Laportea kotoensis Hay., Gen . Ind. Fl. For­ have not been isolated long enough to have mos. 70, 1916, nomen; Kanehira, For­ undergone any morph ological differentiation. mos. Trees 529, 1917; ibid., rev. ed. 170, It may be presumed that this northward mi­ 1936. Syn. nov. gration is not only very recent but is still Batan Islands , Botel Tobago. continuing at the present. Success of estab­ Neither Hayata nor Kaneh ira cited speci­ lishment of these southern species in Formosa mens, but Kanehira's very brief description is controlled mainly by climate factors. With shows that the plant is identical to the Philip ­ probable continued rise of temperature in the pine species, L. batanensis, known from the future, there may be mo re southern species Baran Islands only. entering into the flora of Formosa . At the same time, many of the species will be able M yristicaceae to extend their range to the northern part of the island. 1. Myri stica cagayanensis Me rr. in Philip. Jour. Sci. 17: 255, 1920; Kanehira, Formos. SOME FLORISTIC NOTES Trees rev. ed. 193, f. 141, 1926. In the past, this southern affi nity of the Myristica glomerata Kudo & Masamune in Formo san flora has not been properly empha­ Ann . Taihoku Bot . Gard. 2: 89, 1932. sized because of our inadequate information . Syn. nov. Floristic Interchanges - LI 183

Luzon ; Formosa, in southern part; and i ~ III. 4: 185, f. 106, 1895; K anehira, Formos. Botel Tobago. , Trees rev. ed. 314, f. 269, 1936. Formosa: Cuntin, Hunchuen Peninsula, H , Toluiftra cocbincbinensis Lour. Fl. Cochinch. Keng, Oct. 31, 1950 (US). . 262, 1790. Myristicaglomerata Kudo & M asamu ne is a Glycosmis citrifolia Lindley in Trans. H ort. name based on Sterculia glomerata Blanco, Fl. Soc. 6: 72, 1826. Filip. ed. 1. 764. 1837, which had earlier been Glycosmi: pentaphylla Correa in Ann. Mus. made into the combination Knema glomerata Nat. Hist. Paris 8: 386, 1806 ; M atsum. (Blanco) Merr. in J our. Str. Branch. Roy. As. & H ay. in Jour. Coll. Sci. Tokyo 22: 73, Soc. 76: 81. 1917, Sp. Blanco. 151. 1 918, a 1906; H ay. Icon. PI. Formos. 1: 121, species of Borneo and the Philippin es but not 1911. of Formosa. Kudo and M asamune were mis­ Citrus erytbrocarp« H ay. Icon. PI. Form os. led by the previous misidentification of the 6: 13, 1916. Syn. nov. Formosan plant in question as Myristica beter­ Glycosmis erythrocarpa H ay. Icon. PI. For­ ophylla sensu Ha y., Gen. Ind. PI. Formos. 61. mos. 8: 14, 1919. Syn. nov. 1917, etc., non F.-ViiI. Myristica heterophylla India to M alaysia, Philippine Island s, Hai­ F.-Vill. is a synon ym of Knema glomerata nan ; Formosa, common in forests at low (Blanco) Merr. The two plants in qu estion altitudes. differ greatly and are now placed in two dif­ Formosa: Taipei, H. Keng 1044 (US), K. ferent genera. Odashima 13607 (US), Kei-Dai Lin, N ov. 12, Rutaceae 1930 (NTU) ; Tamali, 'faitung, H. K eng 1376 (US); Bankin sing, A . Henry 1487 (US), 1587 1. Evodia confusa Merr. in Philip. J ou r. Sci. (US) ; Kuraru, K oshun, E. H. Wilson 11038 20: 391, 1922. (US). Evodia merrillii Kanehira & Sasaki in Ka ne­ Th is is a polymorph ous tree, very variable hira, Formos. Trees rev. ed. 313, f. 267, in its features. Ha yara's plant is described with 1936. smaller leaves which are usually tern ate, bu t Evodia roxburghiana sensu M atsurn & Hay. the characters are within the range of varia­ in J our. ColI. Sci. Tokyo 22: 70, 1906 tion of the species. (Enum .P I. Formos.); H ay. Icon. PI. For­ Anacardiaceae mos. 1: 118, 1911, non Benth. 1. Semecarpus gigantifolia Vidal, Sinopsis Celeb es and Philippines, widely distrib uted Adas 22, t. 36, f.A. 1883. and commo n in forests at low and medium Semecarpus verniciftra Hay. & K awakami in altitudes; Form osa, mostly in northern part. Hay. Icon. PI. Formos. 2:' 108, 1912; Formosa: Shirin, Taihoku, T. Tanaka & K aneh ira, Form os. Trees rev. ed. 367, Y . Shimada 11163 (US); K angu, H. Keng, Oct. f. 322, 1936. Syn. nov. 26. 1950 (US); Sinten, H. Keng 1811 (US); Wi dely distributed in the Philippines, in T yku ran, Taih oku , Sou-Gen Lin, Aug. 30, forests at low altitudes ; Form osa, along sea­ 1932 (NTU) . shore of the south and east coasts, and on K anehira and Sasaki cited no specimen for Botel Tob ago Island; cultivated in other their species. The descriptio n and illustration places. are distinctive. This charac teristic plant with Formosa: Taihoku , cult., E. H. Wilson9910 large leaves is in all respects referable to the (US), Kei-Dai Lin, N ov. 5. 1936 (NTU) . common Phili ppine species Evodia confusa Th e Formo san plant proves to be identical Merr., kn own earlier as "E. glabra." with this com mon Phil ippine species, many 2. Glycosmis cochinchinensis (Lo ur.) specimens of which are available for compar­ Pierre ex Eng l. & Prand, Nat . Pflanzenfam. ison . 184 PACIFIC SCIENCE, Vol. VII, April, 1953

Aquifoliaceae Luzon, Baran Islands; Formosa, in forests along the coast, rare in Hunchuen Peninsula 1. Ilex asprella (Hook. & Am. ) Champ. in only . Kew Jour. 4: 329, 1852; Kanehira; Formos. Formo sa: Koshun, Kuraru, Y. Kudo & S. Trees rev. ed. 370, f. 324, 1936; S. Y. H u Suzuki 15889 (NTU) ; Kuraru, B. Hayata & in Jour. Arnold Arb. 30: 269, 1950. Sasaki 15621 (isotype of G. -diospyrosifolium Prinos asprella Hook. & Am. Bot. Beech. Ha y., photo US). Voy. 176, pI. 36, f. 1, 2, 1833. Gonocaryum diospyrosifolium Hay . proves to Ilex arisanensis Yamamoto, Suppl. Icon . PI. be the same as the Philippine G. calleryanum Formos. 1: 30. f. 10, 1952; Kanehira, (Baill.) Bece. (G. tarlacense Vidal). The plant Formos. Trees rev. ed. 370, 1936. Syn. is a common tree in Luzon and is also found nov. . on the Baran Islands between Luzon and For­ Luzon to southeast China ; Formosa, in the mosa. In Formosa , it is a rare plant grown thickets from low to medium altitudes onl y in the shore forests of Kuraru, Hun­ throughout the island . chuen Peninsula, sometimes on coral rocks. Formosa: Sozan, Taih oku-shu, T. Tanaka & Y. Shimada 11009 (US), Y. Kudo ,'S. Suzuki Melastomataceae & K. M ori, April 21, 1929 (NTU) ; Mt. Kwan­ 1. Astronia ferruginea Elm. Leafl.Philip. non, Taihoku-shu, T. Tanaka & Y. Shimada Bot. 4: 1205, 1911; Merr. in Philip. Jour. 11094 (US); Taipei, H. Keng 1024 (US), T. Sci. 8: Bot. 342, 1913. Tanaka 76 (US) ; Shirin to Sozan, E. H. Wilson Astroniaformosana Kanehira, Formos. Trees 10293 (US) ; Musha, Nanto, E. H. Wilson 259, 1917. Syn. nov. 10065 (US); South Cape, A. Henry 1334 (US); Astronia pulchra sensu Hay. in Jour. ColI. Arisan, Faurie 186 (cotype of I . arisanensis Sci. Tokyo 30(1): 114, 1911; (Ma t. FI. Yamamoto, photo US). Formos.), Icon. PI. Formos. 2: 25, 1912; Photograph of Yamamoto's type together non Vidal. with his original description prove that 1. Astronia cumingiana sensu Kanehira, For­ arisanensis Yamamoto is clearly conspecific mos . Trees rev. ed. 507. f. 469, 1939; non with I. asprella. This species has characteristi­ Vidal. cally thin leaves with caudate tips and long Philippine Islands, from Mindanao to Lu­ slender pedicels. Faurie's plant , a mature fruit­ zon ; Formosa, in forests, Hunchuen Penin­ ing specimen, has relatively larger leaves than sula, Lutao and Lanyu (Botel Tobago). others and also was collected at higher alti­ Formosa: South Cape, A . Henry 658 (US) ; tudes than most others . Ilex arisanensisYama­ Kuskus, Y. Kudo & S. Suzuki 16066 (NTU). moto was not considered in S. Y. Hu's study The Formosan plant is neither a distinct of the genus , and it was maintained as a dis­ species nor referable to either A . pulchra or tinct species by Kanehira. A . cumingiana of the Philippines. The species Icacinaceae is actually identical with A . ferruginea Elm., a mo re or less comm on species of the Philip­ 1. Gonocaryum calleryanum (BailL) Bece. pines. The ferrugineous lepid ote inflorescence in Malesia 1: 123, 1877. and undersurface of the leaves are very dis­ Phlebocalymna calleryanum Baill. in Adan ­ tinctive and characteristic of this species. sonia 9: 147, 1896. Boraginaceae Gonocaryum diospyrosifolium Hay. Icon. PI. Formos. 2: 106, 1912; Kanehira, For­ 1. Cordia cumingiana Vidal, Phan. Cum­ mos. Trees rev. ed. 400 f. 359, 1936. ing. Philip . 187, 1885, Rev. PI. Vase. Filip. Syn. nov. 192, 1886. Floristic Interchanges -LI 185

Cordia kanehirai H ay. Icon . PI. Form os . Phil ippine Island s; Form osa, in southern 6: 31, 1916; H ou in Taiwania 1: 207, part, along the seasho re. 1950. Syn. nov. Formosa: Takao, A . Henry s. n, (US), K. Luzon ; Formosa, in the thickets in southern M oritani 2308 (NTU); Kuraru, E. H. Wilson part only,SCarce. 10967 (US). Form osa: K oshun, E. Matuda 1553 (NTU) ; The Form osan plant proves to be the same Kuraru, K oshun, R. Kanehira 7 (isotype of as th e Philipp ine E. navesii Vidal, a charac­ C. kanehirai, photo US). teristic plant readily distinguished by the H ayata originally placed th is plant as a long-linear and persistent calyx-lobes which close ally of C. cumingiana. In comparing spe­ enclose the globose fruit at maturity. cimens from the two regions , it is found that Verbenaceae the Formosan plant cannot be different iated 1. Premna nauseosa Blanco, FI. Filip. 489, from the Luzon plant and must be considered 1837. as conspecific. Premna integrifolia Blanco, Fl. Filip. ed. 2. 2. Cynoglossum lan ceo latum For sk , FI. 342, 1845; M atsum. & Hay. in J our. Aegypt. Arab. 41, 1775; Hou in Taiwania Coll . Sci. Tokyo 22: 299, 1906 (Enum . 1: 217, 1950. PI. Formos.) ; non Linn . Cynoglossumalpestre Ohwi in Acta Ph ytotax. Premna obtasifolia sensu Sasaki, List: PI. Geob ot. 2: 150, 1933; H ou in Taiwania Formos. 353, 1938; M ori in M asamune, 1: 218, 1950. Syn nov. Short FI. Formos. 181, 1936; non R . Br. India, Southern China to the Philippines ; Premna odorata sensu Yam amoto in J our. Formosa, mountain regions. Soc. Trop. Agr. 6: 554, 1934, p. p.; K a­ Formosa: No precise locality, A .Henry nehir a, Formos. Trees rev. ed. 654, 1936, 1015 (US); Arisan, T. Tanaka 520 (US); S. p. p.; non Blanco. Suzuki,July 17,1937 (N TU) . Philippine Islands; Form osa, in southern No isotypes of Ohwi's species are available , .,part only. but, among the cited specimens, Suzuki s. n. Formosa: N o precise locality, A . Henry 613, most closely approaches the original descrip­ 791 (US). tion of C. alpestre in having lon ger, dens er This is an exceedingly common species in hairs and smaller, thicker leaves than other the Philippine Islands, generally known as spe cimens of C. lanceolatum. These characters "P. integrifolia."This species was confused by are, howev er, evidently due to high er eleva­ many autho rs, including K anehira, with P. tions, whereas there are no structural differ­ odorata, ano ther Philippine species that occurs ences between plants of high er and lower also in southern Formosa. In P. odorata the altitudes. Cynoglossamalpestre Ohwi can at most leaves are densely tomentose beneath. In P. be considered an alpine form of the species nauseosa the leaves are more or less glabro us C. lanceolatum, but, in view of the wide dis­ beneath. . tribution and variable natu re of the species, 2. Prernna odorata Blanco, FI. Filip. 488. it seems not desirable to recogn ize it taxo­ 1837; Yamamoto in J our. Soc. Trop. Agr. nomically. 6: 554, 1934, p. p.; K anehira, Form os. 3. Ehretia navesii Vidal, Rev . PI. Vasc. Filip. Trees rev. ed. 654, f. 610, 1936, p. p. 194, 1886. Premna vestita Schauer in DC. Prodr. 11: Ehretia resinosa Hance in J our. Bot. 299, 631, 1847; Henry in Trans . As. Sci. J ap. 1880; H ou in Taiwania 1: 203, 1950. 24. Suppl.: 70, 1896. Syn. nov. Widely distributed in Luzon; Formosa, in Ehretia formosana Hernsl. in Jour. Linn. forests near seashore in northern and southern Soc. Bot. 26: 144, 1896. 5yn. nov. parts. 186 PACIFIC SCIENCE, Vol. VII, April, 1953

SUMMARY REFERENCES The floras of Formosa and the Philippines KANO, T. 1931. Notes on Pachyrrhynchides are distinctive in their general nature. The from Kotosyo and their zoogeographical former is related especially to the Chinese significance. Biogeogr. 5 0(;. J ap., Bul. 2: mainland and the latter to Malaysia. Their 193- 208. [In Japanese.l close geographic proximity, however, has al­ lowed certain floristic interchanges in relative­ LI, H . L. 1950. Studies in the Scrophulariaceae ly recent times. An earlier movement was the of Taiwan. Quart. J our. Taiwan M us. 3: southward migration of temperate elements 55- 80. from Formosa to the Philippines. These are --.- 1952. Notes on some families of For­ mostly montane and alpine elements and, in mosan phanerogams. Wash. Acad. Sci., Jour. the Philippines, largely confined to the moun­ 42 : 39-44. . tains of northern Luzon. These elements are --- and H. K ENG. 1950. Phytogeographi­ now of wide disjunct distribution and are cal affinities of southern Taiwan. Taiwania related to the alpine floras of western China 1: 103-128. and the Himalayas. A more recent movement . MERRILL, E. D . 1923. Die pflanzengeographi­ is the northward expansion of trop ical ele­ sche Scheidun g von Formosa und den Phi­ meI1;ts from the Philippines to Formosa. These lippinen. Bot. J ahrb. 58: 599-604. are usually lowland plants and are found in Formosa limited mostly to the southern ex­ --- 1923- 26. An enumeration of Philippine treme of the island and the adjacent small flowering plants. 4 vols. Bureau of Printing, islands Lanyu and Lutao. Floristic notes on Manila. certain species common to the Philippine and VAN STEENIS, C. G . G . J. 1946. Preliminary Formosan floras but considered by former revision of the genus Lonicera in Malaysia. authors as distinct are given. A rnold Arboretum, J our. 27: 442- 452.