34 STRUCTURAL BOTANY 



is interesting, however, to know that in some Gymno- 

 sperms the membrane of the embryo-sac likewise be- 

 comes cuticularised. It has also been found that in 

 some plants of this class (Cycadeae) the development 

 of the endosperm is only completed while the seeds are 

 lying on the ground, and in a few cases the endosperm 

 has been observed to burst through the embryo-sac and 

 seed-coats and to become green, just like the prothallus 

 of Selaginella. 



A more serious difficulty is that there are four 

 megaspores in Selaginella, and only one embryo-sac in 

 the Fir, though there are Gymnosperms which have more 

 than one, as Gnetum. In most Gymnosperms the 

 sister-cells of the embryo-sac become abortive at an 

 early stage of their development, as is the case also 

 with the sister-cells of the fertile megaspore in some 

 fossil relations of Selaginella, and in the recent hetero- 

 sporous Water-Ferns. 



The organ in which the megaspores are produced is 

 a megasporangiurn ; that in which the embryo - sac 

 develops is the ovule. Both organs arise in the same 

 way from a group of cells near the growing-point. 

 The similarity of their development has already been 

 pointed out (see p. 17). We infer, then, that the 

 megasporangium corresponds to the ovule, or more 

 strictly to the nucellus of the ovule, for the megaspor- 

 angium has no integument. 



We have found, however, that the megasporangium 

 and microsporangium are just alike in the earlier stages 

 of their growth ; the former, as we have seen, corresponds 

 to the nucellus of an ovule, the latter to a pollen-sac. 

 Hence we must draw the conclusion that a pollen-sac 

 and the nucellus of an ovule are equivalent structures a 



