Box.— Vol. I . ] CA MPBELL—SPA R GA NItJM. 319 



by a membrane — whether of cellulose or not was not deter- 

 mined. Three antipodal cells having the appearance of the 

 normal ones in a recently fertilized sac could be distin- 

 guished. The structures at the micropylar end of the sac 

 were not clearly distinguishable. The extremely broad 

 form of the sac suggested the possibility of there having 

 been two embryo-sacs formed in the ovule, but if this was 

 the case, they had become entirely confluent. 



Another puzzling form in S . Greenii is shown in fig. 69. 

 Here the antipodal cells were apparently normal, but the 

 polar nuclei, which were in close contact, were separated 

 from the cavity of the embryo-sac by an evident membrane, 

 and the upper portion of the sac was similarly shut off. 

 Occupying the extreme upper part of this was a body look- 

 ing like a very large nucleus, but the nucleolus (?) did not 

 appear homogeneous, and it is possible that this apparent 

 nucleus was really a cell, as the nucleolus-like body was 

 quite similar in size and appearance to the nuclei of the 

 surrounding nucellar cells. No other nuclei could be cer- 

 tainly made out in this region, except a deeply staining 

 small body close to the wall of a vesicle which lay below 

 the large apical nucleus. If the latter is really a nucleus it 

 probably means that one nuclear division in the upper part 

 of the young embryo-sac had been suppressed, and no egg- 

 apparatus developed. 



In another specimen of S . Gi'ecnii the embryo-sac was 

 seen to be completely filled with granular cytoplasm, and a 

 similar condition has also been seen in S . eurycarpum. 



VII. Recapitulation. 



1. The stamen in Sparganiiim is of the usual type: the 

 ripe pollen-spore shows but one generative nucleus, but in 

 S. simplex there is regularly a structure present which prob- 

 ably represents a vegetative or prothallial cell which was 

 not observed, however, in S . Greenii. 



2. The early development of the embryo-sac follows the 

 normal course ; the egg-apparatus is small in S . simplex 

 and the polar nuclei fuse completely before the &^^ is 



