36 GYMNOSPERMS 



bors. Let us assume that all disorganize except one tetrad of spores 

 and that, even of this tetrad, three spores develop only slightly be- 

 fore they, too, break down (fig. 28). The growing megaspore then 

 absorbs the disintigrating megaspores, the tapetum, and finally all 

 of the wall cells of the sporangium, except the outer, epidermal layer. 

 Even the lower sporangia of the sorus disorganize, and the mega- 

 spore in the terminal sporangium germinates. In the early stages in 

 the phylogeny of heterospory, the megaspores must have been small 

 and there might not have been any free nuclear stage in the develop- 

 ment; but, as the megaspores became larger, a free nuclear stage 

 would appear, just as it is likely to appear anywhere, when the mass 

 of protoplasm is large in proportion to the nuclear figure; and then 

 cell walls would be formed (tig. 29). In the earlier stages, in phylog- 

 eny, the female gametophyte would probably break out, extend be- 

 yond the spore coats, and become green, just as it sometimes does in 

 living heterosporous Pteridophytes (fig. 30). 



Later, as the megaspore became larger, the gametophyte would be 

 retained within the spore, exposed for the sperms only at the triradi- 

 ate crack. Fertilization might occur at this stage and the embryo 

 might begin to develop; but if, even at such a late stage, the mega- 

 spore, with its embryo, should be shed, the plant would still be a 

 heterosporous Pteridophyte (fig. 31). 



As the megaspore became larger and larger, it would reach a more 

 advanced stage before shedding, until, finally, it would not be shed 

 at all, and the true seed stage would be achieved (fig. ^,2). 



With the permanent retention of the megaspore, the thick spore 

 coat, so necessary while the megaspore was in the discharged condi- 

 tion, would not only be unnecessary, but would be disadvantageous. 

 Consequently, they became thinner and thinner, while the sporangi- 

 um wall, especially at the top, became thicker and the protective 

 and nutritive functions were gradually transferred from the mega- 

 spore itself to the sporangium (figs. 32 and T,i,). 



The formation of a pollen chamber is practically universal in 

 gymnosperms. It is formed by the disorganization of cells in the re- 

 gion above the eggs, the product of the disorganization being the 

 sparkling pollination drop which catches the pollen. The drying of 



