116 BRITISH EOCENE FLORA. 



their structure reveals, more clearly than do any existing seeds, the passage from the 

 cryptogamic stage whence these early " Progyumosperms " had, so to speak, then but 

 barely emerged. 



Even the pollen grains themselves are not without interest when viewed from an 

 evolutionary standpoint. Unicellular in Angiosperms, bi- or tri-cellular in Gymnosperms, 

 they are seen to be distinctly pluri-cellular in " Progymnosperms," for the subdivision 

 into many cells is said by Saporta and Marion to be discernible in all the silicified 

 pollen grains of Carboniferous age yet studied. 1 The included male prothallium is 

 supposed to be represented by the cells, and in that case has become merely rudi- 

 mentary. Their relative size, reaching to half a millimetre, is remarkable, for this 

 exceeds eight and a half times those of the Larch, the largest among living Coniferse 

 and twelve times those of Cycads and Ginkgo. But whatever the size, they are seen 

 to have been divided into eight, twelve, or eighteen cells, the dividing septa of which 

 are still perfectly visible. A considerable increase in bulk accrued to them, as we have 

 seen, during their sojourn in the pollen chamber of the seed they had entered ; while 

 the female organ progressed simultaneously with the development of its Corpuscula, 

 upon whose completion impregnation depended. That these cells are really the homo- 

 logues of a rudimentary male prothallium has frequently been suggested, and it is due 

 in a great measure to this structure and the position and development of the pollen- 

 sacs, that Gymnosperms are admitted to occupy an intermediate position between vascular 

 Cryptogams and Angiosperms. To this progressive and gradual obliteration of the 

 independent sexual existence, characteristic of vascular Cryptogams, the evolution of 

 Phanerogams is mainly to be traced. 



Next, however, in relative development to the Dadoxyloid groups stands the 

 "Progymnospermous " group or genus Dolerophi/ttum. This group deviates more or less 

 considerably from the types we have hitherto considered, and in it we may hope to find 

 the more immediate ancestors of true Gymnosperms. 



The Dolerophylla were arborescent plants of large size, provided with leathery or 

 thick, broadly ovate or orbicular leaves, with simple outer margins, but notched or auri- 

 culated inferiorly, sessile, and leaving a transverse scar of attachment on the stem which 

 they closely embraced. These leaves were of considerable size and shed either singly or 

 adhering to branches. They have been found principally at St.-Etienne, and also in the 

 Permian of Russia. The leaves and branches were produced in great conical buds with 

 convolute vernation. The veinlets are crowded, radiating, and dichotomous. 2 The lower 

 epidermis was dense and provided with stomata, while embedded in the parenchyma and 

 covering the veinlets are several rows of large elongated cells which contained gum or 

 resinous sap. The organs of fructification are still incompletely known, but the 



1 Prof. Williamson thinks that the supposed cells may be but modifications of the sculptured extine. 



2 The veinlets are said to possess a duplex structure characteristic of the " Progymnosperms " and now 

 only retained by the Cycadacece. 



