VEGETABLE KINGDOM 129 and in which the claters are rudimentary. Sphagnum and Archi- diurn among Mosses resemble the Liverworts in that the sporo- gonium is enclosed in the calyptra until near maturity. Although there are so many points of resemblance between them, there is no evidence to show that the Mosses and Liverworts form one continuous series. This might perhaps be assumed if the gametophyte alone were considered ; but it cannot when the sporo- phyte is taken into account as well. The more important features in the comparative morphology of the sporophyte are as follows. Taking first the Liverworts, in the Riccicee, proper (Riccia, Oxymitra) both the epibasal and hypobasal halves of the oospore (fertilized female cell) form sporogenous tissue. In the allied Corsinicse, the Marchantices, and the Antlwccrotese. the epibasal half alone gives rise to sporogenous tissue, the hypobasal half forming an organ of absorption and attachment which is the true foot. In the Junger mannicse, likewise, the sporogenous tissue is developed entirely from the epibasal half of the oospore, which also gives rise to the seta or stalk of the sporogonium ; but the hypobasal half is aborted and gives rise to only a rudimentary foot ; in some foliose forms a. false foot is formed by a dilatation of the base of the seta. In the Mosses there is not in any family such a simple sporophyte as that of the liiccicse, among the Liverworts ; but there are forms which present the same morphological difference as that which in the Liverworts distinguishes the Marckantieee and the Anthocerotcie, on the one hand, from these Jungermannicse, on the other. Thus, among Mosses the Sphac/naccee, Andreeeaccx, and Phascaccse have a true foot, whereas in the Bryinese there is a spurious foot at the base of the seta, as in some Jungermannicse. Taking all the foregoing facts into account, it is clear that the Ricciese are the most primi tive of existing Muscinese, and it is probably from them that the true Mosses, as well as the other Liverworts, have been derived. The following phylogeny may be suggested. As regards the Liverworts, there can be no doubt that the Marchantiese are to be traced back through the Corsinicse to the Riccicse,, and it is probable that, as Leitgeb points out, the frondose Jungermannicse were de rived from the Corsiniese through the Ricllcse,. With regard to the Antlwccrotese, Leitgeb seems to suggest that they may have been derived through Notothylas from the Jungcrmanniess ; but in opposi tion to this must be set the difference between the two groups as regards the morphology of the foot. Probably the Anthocerotcfe arose independently from the Corsiniese. The Mosses are to be traced back to the Anthoccrotese, on account of the general similarity in the structure of the sporogonium. The genus Anthoccros shows its special relation to the Sphagnacess in the development of the sporogenous cells from the amphithecium, whereas the genus Noto- Unjlas resembles the other Mosses in that the sporogenous cells are developed from the endothecium. The following scheme will serve to illustrate these remarks. Rlcciass
Corsiniefe I Marchantiex. Ridle, Anthoceroteas I I I (Anthnceros) (Notothylas) I I _, (Sphagnacem. Andreseacex % J Phascaceas ] C Jungcrmannieiv. ^ v Bryinete. The phylogeny of the Muscinese can be traced with exceptional clearness in both generations. With regard to the gametophyte, it must be remembered (see MtTSCINE.fi) that at its first develop ment from the spore it is a protonema, which is either filamentous, as usually in Mosses, or a flattened expansion, as generally in Liver worts, the shoot which bears the sexual organs being a secondary development on the protonema. Both these forms of protonema bear a striking resemblance to various Confervoid Chlorophyccse, and there can be little doubt that it is from this group that the Muscincse have arisen. This view is supported by the close resem blance of the simple sporophyte of Riccia among the Muscinese to that of Colcochfetc among the Confervoid Chloroj)hycese. There is, however, this difference, that in Riccia the cells resulting from the division of the oospore are differentiated into an external sterile layer, forming the wall of the capsule, and an internal mass of sporo genous cells, whereas in Coleochsete all the cells are sporogenous. SUB-KINGDOM ill. PTERIDOPHYTA (VASCULAR CRYPTOGAMS). In these plants the gametophyte is a filamentous or tuberous, or more commonly a membranous, flattened pro- t/iallium, exhibiting no differentiation into stem and leaf. The sexual organs are antheridia and archegonia, as in the Muscinese. The sporophyte becomes quite independent of the gametophyte, constituting the plant. The shoot is always differentiated into stem and leaf, and the tissues are highly differentiated. The sub-kingdom is divided into three classes, FILICIX.S:, EQUISETIN/E, LYCOPODIN^E. Each of these classes includes two series of forms, the homospor- ous, in which all the spores are alike, and the hctcrosporous, in which the spores are of two kinds, namely, large spores (macroyores) and small spores (microspores). On germination the spore of the homo- sporous forms gives rise to a well-developed prothallium, which be comes quite free from the spore, and usually bears both antheridia (male) and archegonia (female). The spores of the heterosporous forms give rise on germination to comparatively inconspicuous pro- thallia, which remain connected with the spores, and bear either exclusively male or exclusively female organs. Thus, in the hetero sporous forms the gametophyte is represented by a male and a female organism, the former being much the smaller. The male prothal lium is always developed from a microspore, the female prothallium from a macrospore (see REPRODUCTION). CLASS 1. FILICINJF;. The Filicinse are characterized by having relatively large and few leaves. The spore-bearing or fertile leaves (S2wrophylls) are generally similar to the foliage leaves, and are not aggregated on special shoots, so as to form flowers. The embryo sporophyte has no suspensor ; but it has generally (except iu Sal- vinia) a primary root. Series A. Homosporous Filicinse. This series consists of the sub-class Filiccs, which includes the Ferns. It seems to be certain that all existing Ferns are homosporous, and apparently all the known fossil forms are homosporous also, though it is not impossible that there may have been heterosporous Filiccs. The orders of existing Ferns are arranged in two categories, according to the mode of development of the sporangia. Lcptosporangiate Filiccs : the sporangium is developed from a single epidermal cell : Order 1. Hymenophyllaceai. Order 4. Gleichcniccse. ,, 2. Cyatheaccse. ,, 5. Schizseacese. ,, 3. Polypodiaccfe, ,, 6. Osmundacese. Eusporangiate Filiccs : the sporangium is developed from a group of epidermal cells : Order 1. Opliioglosscicew. Order 2. Marattiaccse-. Scries B. Heterosporous Filicinse. As far as is known at present, this series contains the sub-class Hydroptcrideie, or Rhizo- carpse. In addition to the various peculiarities connected with their heterospory, its members are characterized by the development of an investment round the clusters of sporangia (sori), the whole forming a sporocarp. They are all Leptosporaugiate. The HyclroptcridcK are arranged in two orders, according to the structure of the sporocarp. Order 1. Salviniacese : wall of sporocarp formed as a superficial outgrowth of the sporophyll ; each sporocarp contains a single sorus either of macrosporangia or of microsporangia (Salvinia, Azolla). Order 2. Marsileaccse : the wall of the sporocarp is formed by a portion of the sporophyll ; each sporocarp contains several sori ; and each sorus includes both macrosporangia and microsporangia (Marsilca, Pihdaria). It is possible that Isoctes represents the heterosporous euspor- angiate Filiccs (see infra). CLASS 2. EQUISETINJE. These plants are characterized by their rudimentary foliage leaves arranged in whorls at the nodes, and by their highly modified sporophylls, which are aggregated together at the ends of shoots, so as to form cone-like flowers. The embryo sporophyte has no suspensor, but a primary root. Series A. Homosporous Equisetinse. All the existing forms belong to this series and constitute the order Equisctaccse ; they all belong to the genus Equisctum. The gametophy te is a green flattened prothallium, like that most commonly found in the Filiccs; but it is much branched, and generally dioecious. Series B. Heterosporous Equisetinse. There are no existing heterosporous Equisctinx, ; it is probable that some fossil plants, such as Annular ia and Astcrophyllites, represent the extinct hetero sporous forms. CLASS 3. LYCOPODINJE. The Lycopodinse are generally charac terized by their small and numerous foliage leaves, which are not arranged in whorls, though they are almost entirely .wanting in Psilotum. The sporophylls are not highly modified, but they are frequently aggregated at the ends of shoots, so as to form cone-like flowers. So far as the embryology of this class is known at present, the embryo sporophyte has a suspensor, but no primary root. Series A. Homosporous Lycopodinse. Order 1. Lycopodiaccx. : the sporangia are borne on sporophylls (Lycopodium, Phylloglossum). Order 2. PsUotacex : the sporangia are borue on the stem (Psilo tum, Tmesipteris). Scries B. Heterosporous Lycopodinse. Order 1. Sclagincllaccse (Sclaginella). The existing heterosporous Lycopodinse, are usually considered to consist of a family, termed the Ligulatse, which includes the two
XXIV. 17