THE CLASSIFICATION OF GYM~ OSPERMO S PLANTS
DIVYA DARSHA'[ PANT Botany Department, The University, AHahabad
ABSTRACT But soon after the discovery of motile sperms in thi:i genus (HIRASE, 1896), Engler ( 1897) Various systems of classification of the gymno spermous plants are briefly reviewed. Arnold's created a special cla.· , the Ginkgoales for classification (Bot. Gaz.. 1948) is tentatively Ginkgo and its fossil representatives. By the modified as follows: beginning of the present century the knowl Division: Cycadophyta Class 1. Pterictosper edge of three more types of fossil gymno mopsida. Order:;: 1. Lyginopteridales. 2. Mec\ul losales. 3. Glossopteridales. 4. Peltaspermales. spermons plants had also become sufficiently S. Corystospermales. 6. Caytoniaks. Class 2. Cyca clear and had led to the establishment of dopsida. Order: 7. Cycaclales; Class 3. Pentoxylo three more class.s, the Cordaital s, the psida, Order: 8. Pentoxylalcs; Class 4. Cycadeol Pteridosperm:i and the Bennettitales. Engl r deopsida, Order 9. Cycackoicleales. Division: Chlamydospermophyta - Class Gneto ( 1897 ), Coulter & Chamberlain (1917) . psida. Orders: 1. Gnetales (restricted to Gnetum). Engler & !'rant! ( 1926), H.endle ( 1930) and 2. Welwitschiales. others all recognize the Gymnosperms a pri Division: Coniferophyta - Class 1. ConiIcro mary division of the Spermatophyta-Phane psicla, Orders: 1. Cordaitales. 2. Ginkgoales; 3. Coniferales; Class 2. Ephedropsida. 4. Epheclrales; rogamia or Embryophyta - Siphonogama. Class 3. Czekanowskiopsicla. Order: S. Czekanows They then directly divide the group into kiales; Class 4. Taxopsida. Order: 6. Taxales. classes or orders of co-ordinate rank, viz. The position of the Pentoxylales, Ephedrales. (1) Pteridospermae (or Cycadofilicales), Glossopteridales and Czekanowskiales is discussed in the light of recent research. (2) Cycada1cs, (3) Bennettitales, (4) Cordai tales, (5) Ginkgoales, (6) Coniferales, and (7) Gnetales INTRODUCTIO In contrast, Berry (1917) divides the The history of gymnosperm classification gymnospermous plants into the Pterido begins with the year 1827 when Robert spermophyta, Cycadophyta and Conifero Brown recognized the naked seed of the phyta without classing them as Gymnosperms. Cycads and the Conifers. and called them Gymnosperms. Later, Endlicher (1836 DEVELOPMENT OF PHYLOGE. ETIC 1840) gave them the same rank as the SYSTEMS three divisions of Dicotyledons under his Acramphibrya, Adolphe Brongniart ( 1843 ) In 1919, Sahni (1920) for the first actually included them in the Dicotyledons time clearly recognized two main phyletic and Bentham & Hooker ( 1862-1883) placed lines in the orders of the G mnosperms: them between the Dicotyledons and the (1) the Phyllo.-perms consisting of (a) the ~10nocotyledons. ltimately, Hoffmeister's Pteridospermae, (b) the Cycadalcs and prob epoch-making work ( 1851 ) on the develop ably (c) the Bennettitales, and (2) the ment and embryology of diverse plants led Stachyosperms consisting of (a) the Cordai Van Tieghem ( 1898) to remove them from tales, (b) the Ginkgoales, (c) the Coniferales, this intermediate position and he installed and (el) the Taxales (separated from tIle Coni them as one of the two primary divisions of f'rales and including Tax%s, Torreya and the Spermatophyta1 Cephalotaxu,s). As their names indicate, the Bentham & Hooker (1862-1883) recog ovules of the Phyllosperms are borne on nized only three orders of the living Gymno leaves or regarded to be so, while those of the spermae, viz. Gnetaceae, Coniferae and Cyca Stachyospc:rms are helieved to be stem 2 daceae. The genus Ginkgo was in those days borne . Besides the differences Jl1 the traditionally included in the Coniferae, e.g. nature of the ovule-bearing organs, Sahni it was included by Eichler ( 1889) in Taxeae along with Taxus, Torreya and Cephalotaxus. 2. This difference between the morphological nature of the ovule-bearing organ:; o( Cycads a.nd 1. The term Spermaphyta which is sometimes Conifers was first pointed out by Strasburger used is etymologically incorrect. (1879 ). 65 66 THE P,\LAEOBOTANIST
also pointed out the differences between the berlain's names3 . The acceptance of Cham generally large and much-divided leaves berlain's nomenclature for the two lasses of the former and the simple leaves of of Gymnosperms is, however, not against the latter. Subsequently, Margaret Henson Sahni's recognition of the two main phyletic ( 1921) suggested that Sahni's two main lines in this group, only the basis of their lines may have been totally distinct and distinction is different. she associated the Stachyosperms with the Another feature of Sahni's classification Sphenopsida and the Phyllosperms ,vith which seems to have come to stay is the sepa the ferns. ration of Taxus and its allied forms from Almost simultaneously with Sahni, Cham the Coniferales and their inclu ion in a new berlain (1920, 1935) also recognized two order, the Taxalcs. Florin ( 1948) has sub main groups among the Gymnosperms and sequently upheld Sahni's separation of the called them by the old names Cycadophyta Taxales as an order of co-ordinate rank with and Coniferophyta, Although the composi the Cordaitales, Ginkgoales and Coniferales tion of Chamberlain's Cycadophyta is the ( restricted), but he includes only Taxus, same as that of Sahni's Phyllosperms and Torreya and three other newly discovered that of his Coniferophyta essentially similar genera (Nothotaxus, A mentotaxus and A us to Sahni's Stachyosperms (except that trotaxus) in the group retaining Cephalotaxus Chamberlain also includes the Gnetales in wi thin the Coniferales, the Coniferophyta ), yet the emphasis on the The enigmatic Gnctales have been left out primary characteristics in the two classifica in a doubtful position by Salmi while Cham tions is different. berlain includes them in the Coniferophyta The main basis of Salmi's classification is but without connecting them to the other the morphological nature of the ovule-bear coniferophytes (see CHAMBERLAIN, 1935, ing organs. As Scott ( 1923) pointed out p, 4, FIG. 2). Pulle ( 1938) included them soon after Sahni proposed his scheme, it was in a separate class, the Chlamydospermae, mainly a theoretical distinction for some of which is made co-ordinate with his other the groups, e,g, the Cordaitales, Bennetti classes of seed plants, viz, Pteridospermae, tales, and Conifers, Schoute ( 1925), who first Gymnospermae and Angiospermae. The showed that the ovule in Cordaianthus was customary inclusion of the three genera attached on a lateral appendage, pointed out Ephedra, Welwitschia and Gnetu11'~ in the that this did not support Sahni's division of single order Gnetales has also become ques the Gymnosperms into the Stachyosperms tionable in view of the contributions of and the Phyllosperms. Since then Florin's Pearson ( 1929), Florin ( 1931, 1933, 1934), extensive work ( 1938-45, 1951) on the stro Eames ( 1952) and others. Arnold ( 1948 ) bili of both Cordaitales and Conifers has made proposed the separation of Ephedra in a it abundantlv clear that the ovule in both separate order, the Ephedrales, Eames these groups" is borne on the lateral append ( 1952) goes further and suggests that the age of a short axis. Florin does not pre Ephedrales are nearest to the Cordaitales and cisely state the nature of the ovulate organ, the Coniferales and rather widely separated but the fact that he often refers to it as a from Gnetttm and Welwitschia. He also sug megasporophyll suggests that he regards it gests the cff~ation of two separate orders, as a foliar structure, No doubt he compares Gnetales restricted to Gnetum and WeIwits it with a telome, but even foliar structures chiales for Welwitschia. are fundamentally telomes, Eames ( 1952 ) In a new classification of the Gymnosperms, has already pointed out that the stachyo Arnold (1948) suggested that the terms spermous character of the Cordaitales and the Spermatophyta and Cymnospermae could not Conifers cannot stand, indicate natural relationships mainly because Chamberlain ( 1920, 1935), on the contrary, they are based on single characters. He emphasizes mainly the factual details of the emphasized the differences between the differences in habit, stem anatomy and leaves Cycadophyta, Coniferophyta and Chlamydo of the two groups. The names chosen by spermophyta (Pulle's Chlamydospermae) Chamberlain are also non-committal. Ac and in his classification raised them to the cordingly, despite the slight priority of the names proposed by Sahni orne recent authors, 3, Takhtajan (1953), however, retains Sa-hni's e.g. Arnold (1948), Engler, Melchior & nomenclature and Lam (1950) extends the cI is Werdermann ( 1954), have preferred Chal1l tinction even into the Angiosperms. PAXT THE CL.\.SSIFIC:\TION OF GYMNOSPERMO S PLANT 67
status of classes equal in rank with his They have incorporated s me of the features Pteridophyta (ferns only) and Angiospermo of Arnold's classification, but have introduced phyta under the division Pteropsida. In the following changes: addition to the separation of the Ephedrales 1. The name Gymnospermae is reincor (see above) under his Chlamydospermo porated as a division of the plant kingdom phyta he accepted the separation of the co-ordinate with the Bryophyta, Pterido Taxales under the Coniferophyta. He reject phyta, Angiospennae and others. ed the names Spermatophyta and Gymno 2. Two class names have been changed spermae as terms which had" outlived their to Cycadopsida and Coniferopsida, and a usefulness". new class Taxopsida has been introduced to include the order Ta.·ales. 3. The class name Chlamydospermophyta EMB 'RGE 'S PREPHANEROGAMS AND PHANEROGAMS is changed to Chlamydospermae and instead of dividing it into two orders all the three In the meantime, the use of the term genera are included in the single order Phanerogams (and its counterpart Crypto Gnetales as in older classifications. gams) had automatically fallen in disuse, 4. The Caytoniales have been separated mainly because it was a misnomer. Coulter from the Pteridospermae and th' Nilssoniales & Chamberlain (1903) are against the use from the Cycadales and introduced as dis of the term because it artificially combines tinct orders in the class Cycadopsida. the Gymnosperms with the Angiosperms al 5. A new order Pentoxylales, including though actually they are somewhat more the Pentoxyleae (SRIVASTAVA, 1946; SARI r, closely comparable with the Pteridophytes 1948 ), is incorporated in the Cycadopsida. from which it separates them. However, 6. The Ginkgoales have been shifted from Emberger ( 1944) recently reintroduces the the Coniferopsida and included under Cyca term Phanerogamae along with a new name clopsida. Prephanerogamae to include the Pterido While some of the changes, introduced by spermae and Cordaitales which have no em these authors, may appear justifiable, there bryos in their seeds at the time of shedding. are others, e.g. the inclusion of the Gink According to Emberger, the Prephanero goal s under the Cycadopsida, the grouping gamae shed only megasporangia wrapped in together of Ephedra, Welwitschia and Gnetum integuments and not true seeds which, ac in the order Gnetales and a different ending cording to this author, should also have for the class name Chlamydospermae, embryos at that time. Favre-Duchartre which are difficult to und rstand. Clearly ( 1943 ) and haudefaud ( 1944) and others Arnold's classification, though earlier, is more have pointed out that Ginkgo and the Cyca systematic. dales should also be included in the Pre phanerogamae because in their seeds too A NEW PHYLOGENETIC SCHEME embryos are absent at the time of shedding ( this was known long before, e.g. SCHNARF, A modified form of Arnold's classification 1937). Emberger's classification, which is is, therefore, suggested below. The main based on a single character, is unnatural be changes proposed here are: cause it introduces such anomalies as the 1. The names Cycadophyta, Chlamydo separation of Pteridosperms and Cordaitales spermophyta and Coniferophyta are raised from the related Cycadophytes and Conifero to the divisional rank. phytes respectively. It also groups together 2. The endings of the class and order names such dissimilar plants as the Cordaitales and have been changed to conform with the re the Pteridosperms. The absence of embryos commendations of the latest International in the seeds of the two groups merely Code of Botanical omenclature, Utrecht indicates that their sccJs are in the same ( 1956). stage of evolution. 3. The order Ephedrales has been trans ferred to the Coniferophyta under Conifer opsida in view of the recent contributions of ENGLER, MELCHIOR AND WERDERMANN'S YSTEM Florin (1931,1934) and Eames (1952). 4. A new class Czekanowskiopsida with Engler, Melchior & Werdermann (1954) an order Czekanowskiales is formed for have suggested yet another classification. Czekanowshia and its nearest allies ( HARRIS, 68 THE PALAEOBOTAI TIST
1951 b). Leptostrob~ts is indeed very different Class 3 - Czekanowskiopsida from any other known Ginkgoalean fructi Order 5 - Czekanovvskiales fication and together with its deciduous short Class 4 - Taxop,;ida shoots and forked linear leaw,; it presents such Order 6 .-- Taxales a unique combination of reproductive and It will be noticerl that in the above scheme vegetative characters that it would in any case the Caytoniales are retained within the appear preferable to class it separately. Pteridospcrmopsida. Harris (1940, 1951a) 5. The class Chlamydospermophyta in has already pointed out that our knowl cluding the Gnetales and the Welwitschiales edge of the Caytoniales is far from being has been placed between the Cycadophyta complete and all we know about them makes and Coniferophyta to bring it nearest to the them quite respectable Pteridosperms. Sepa Cycadeoid opsida which the two genera of rating them from the other Pteridosperm this clas. resemble in their syndetocheilic as done by Engler, Melchior & Wer stomata (FLORIN, 1931, 1934) and, if Pearson dermann ( 1954) but retaining the Corysto (1909 & 1929, p. 185) is right, in the spermaceae ( TH REFERENCES ARNOLD, C. A. (1948). Classification of the CHAMBERLAIN, C. J. (1920). The living Cycads gymnosperms from the viewpoint of Palaeo and the phylogeJl y of seed plants. American botany. Bot. Gaz. 110(1).2·12. J. Bnl. 7: 146-153. BE:-isON, iVI. (1921). The grouping of vascular Idem (1935). Gymnosperms, structure and evo plants. New Phyt. 20: 82-89. lution. Chicago. BENTHAM, G. & HOOKER, J. D. (1862-1883). CHAUD!>;FA UD, YI. (1944). Les Prephanerogames. Genera Plantarum. ](ew. Rrll. Scient. Anee Paris. 82. BERRY, E. VV. (1917). The classification of CO LTER, J. M. & CH.-\wl!3ERU1:-i, C. J. (1903). vascular plants. Nat. Acad. Sci. 3: 330-333. Morphology of Angiosperms. Chicago. BRONGNJART, A. (1843). Enumeration des genres Idem ( 1910). Morphology of Gymnosperms. Chicago. de plantes & c. Paris. EAMES, A. J. (1952). Relationships of the BROWN, R. (1827). Character and description of Ephedr'l.k~. PhylolllOrpholngy. 2: 79-100. Kingia, with observations on the structure of EICHLER, A. VV. (1883). 5)!I'abllsrlcr VCW!riSul1gC1t. Ed. 3. its unimpregnated ovulum on the female flower :rOMBERGER, L. (1944 )...Lcs Plantes fo~siles dans of Cycadaceae and Coniferae. leurs rapports avec los veg0taux vivants. Paris. 70 TllE PALAEOBOT.-\~TST El'DLlCHER, S. (1836-1840). Genera. P1ant"rum. Idem (1956). Bisexu,,1 fructifications borne on ENGLER, A. ( 1892). Syllabus del' Vorlesungell iib"r Glussvpteris leaves from S. Africa. Pa/aeonta speciellc undmedicinish-pharmaeeutische Botanik. graphica. BIOO: l-25. ENGLER, A., l\lELCIlIOR, H. & 'vVEI