gradually increasing suppression of that of the oophore. This never, however, as in the succeeding group, com
pletely loses its independence of the sporophore.The whole direction of the attempts at classifying the Pteridophyta, which correspond to the so-called vascular Cryptogams, has been completely changed by Fankhauser s discovery of the long-sought reproduction of Lycopodium. It proves to be totally unlike that of Selaginella, which, from the close agreement that exists between the two types in every other respect, it might have been expected to closely resemble. On the contrary, in the formation of a monoeci ous subterranean prothallus, it may be compared to Ophio- glossece. Yet, as Sachs has pointed out, it is impossible to use this striking divergence for the purpose of breaking up a group of plants which in all respects is perfectly con gruous. The anomaly of forms so closely allied, exhibiting two modes of gamogenesis, has now, however, found a parallel amongst the Filicales, in which group RhizocarpeoB seem to find their proper place.
The Pteridophyta are divisible into three probably natural classes, which may be briefly defined by the following characters, in themselves, no doubt, artificial:—
i. Filicales.—Leaves highly developed and bearing numerous sporangia.
ii. Equisetaceæ.—Leaves rudimentary, reduced to sheaths on the barren stems, the fertile ones bearing 5-10 spor angia.
iii. Dichotomæ.—Leaves small, simple; sporangia soli tary.
The Filicales include, according to the arrangement pro posed by Sachs, three orders. In Stipulates the sporangia are not trichomes, but are either wholly endogenous pro ducts as in Ophiocflossu?n, or are formed from cellular pro tuberances, and the tissue beneath the epidermis takes part in the formation of spores (Marattiacece). In Filices pro per the sporangia are, as already mentioned, trichomes, and these are developed upon the normal fronds. In Rhizocarpea} the sporangia are borne by metamorphosed leaves, which are united into a capsular body in a way com parable to that in which carpels bearing ovules are united into a syncarpous ovary. The germination of the spores also presents important differences in the different orders amongst the Stipulates. In Ophioglossacece it gives rise to a tuberous monoecious oophore (prothallium), destitute of chlorophyll. In Osmunda the prothallium exhibits some tendency to become dioecious ; it resembles the normal typo of prothallium in ferns, but has a kind of mid-rib, and produces adventitious shoots from marginal cells. These shoots become detached, and constitute a mode of agarnogenesis. Such a structure cannot fail to recall the thalloid Hepaticce, which also possess a mode of agamo- genesis by means of gemmules (e.g., Marckantia). In Rhizocarpeae the dioecious condition of the oophore has been carried so far that it is indicated even in the spores themselves ; the microspores are male, the macrospores female. In Salvinia there is a rudimentary prothallium bearing the anthericlia ; and a last trace of such a structure appears to exist in Marsilea and Pilularia. The macro- spore also develops a prothallium, in which numerous archegonia are produced, one only of which is fertilized. The development of the prothallium is so far external to the macrospore that it effects its rupture at the apical papilla. In the Rhizocarpece the development of the oophore has been almost entirely compressed within the limits of the spore.
Equisetaceæ call for little remark as regards the oophore. It is dioecious, and irregularly branched, but in other respects there is a close agreement with the oophore of ferns.
The Dichotomæ present in the morphology of the oophore a parallel series with the Filicales. The Lycopodiacece develop from the single form of spores which they possess a monoecious subterranean prothallium like that of Ophio- glossece. The Ligulatce (comprising Selaginella and Isoetes) produce, like Rhizocarpecc, microspores and macrospores. In this group the suppression of the oophore is carried still further ; the microspores possess, according to Millardet, the merest rudiment of a male prothallium. The macrospores produce a female prothallium, which is an endogenous structure in even a higher degree than in the Rhizocarpeoe. The first divisions of the oospore have not as yet been sufficiently studied. But while in Ferns proper, Equise- tacece and Ophioglossacece, the first septum is more or less inclined to the axis of the archegonium, it is parallel with it in Rhizocarpece (a position into which a very oblique inclination would readily pass), while in Selaginella it is at right angles with it.
3. Phanerogamæ. In at any rate two of the three groups of the Pteridophyta we see that the progressive tendency of the oophore (prothallium) is to lose its inde pendence. In Opliioglossece, Filices, and Equisetacece, its growth, independent of the spore, often continues fora con siderable period. In Rhizocarpeoz and Lycopodiacece, where male and female spores are produced, the oophore remains attached to the spore, although, protruding from it. In Isoetes it fills the cavity of the macrospore as a mass of tissue, and the bounding wall is only ruptured to allow the access of the antherozoids to the archegonia. In the Phanerogamce even so small an assertion of independence as this is suppressed. The macrospore (embryo-sac) is never detached from the sporophore previous to fertiliza tion, and the oophore, which is moderately developed in the Gyinnospermw, but in the Angiospermce is reduced to the merest rudiment, always remains entirely enclosed within the macrospore.
In the Gymnospermæ the endosperm is the homologue of the prothallium (oophore). The so-called "corpuscula" are the archegonia, or rather the central cells of reduced arche gonia (secondary embryo-sacs of Henfrey). The neck of the archegonium is represented by the "rosette" first described by Hofmeister. Strasburger has pointed out that a small portion of the contents of the central cells is divided off at the upper end, and this is the rudiment of the canal cell which, running the length of the neck of the archegonium in the Eryopliyta and Pteridophyta, leaves a pervious track by its deliquescence for the access of the antherozoids to the central cell. Strasburger considers all the rest of the contents of the embryo-sac to be the equivalent of the oosphere. The result of fertilization is to cause a trans verse partition of the lower part of the oo^iphere, as in Selaginella. From the lower cell thus constituted, by suc cessive longitudinal and transverse partitions, the suspensors are developed, the ends of each of which bear an embryo. It is noteworthy that the development of a suspensor as an intermediate structure between the fertilized oosphere and the embryo is also met with in Selaginella.
the comparable structures in Angiospermce. Here the pro- thallium (oophore) has completely disappeared, unless we regard, with Sachs, the " antipodal cells " as a last rudi mentary trace; their appearance is, however, inconstant. Within the embryo-sac (macrospore) "embryo-vesicles" arc formed, usually two or three in number. One of these is the oosphere, and may be compared to the central cell of an evanescent archegonium, to which another of the " embryo- vesicles " performs the function of a canal cell; for while it is that which is usually in nearest proximity to the pollen tube, it only apparently transmits the fertilizing influence. In a few cases this dormant embryo-vesicle is replaced
by a somewhat more highly developed arrangement the
