Box.— Vol. I.] CAMPBELL— NAl AS AND ZANNICHELLIA. 57 



differences, does not admit of any very close connection 

 between such forms, and it would be useless to attempt a 

 detailed comparison of the Monocotyledons and Pterido- 

 phytes until a great many more types of the former have 

 been critically examined, especially those whose embryos 

 are more like those of the Pteridophytes than is that of 

 Naias. 



The occurrence of deviations from the ordinary structure 

 in the embryo-sac in both Naias and Zannichellia suggests 

 the importance of further investigations on this point, as it 

 is here that the surest clues to relationship with the Pteri- 

 dophytes, if such exist, are most likely to be found. Recent 

 investigations in this direction have shown that the type of 

 structure is not so uniform as has generally been supposed, 

 this being especially true with respect to the number of the 

 antipodal cells. In both Naias and Zannichellia we have 

 also seen that the number of nuclei, apart from the antipodal 

 cells, may be considerably augmented. A few observations 

 were made by the writer upon Spazganiutn eurycarpuni^ 

 which showed that in this plant also the structure of the 

 embryo-sac is not always constant. In fig. 129 is shown a 

 case where there were about ten free nuclei in the upper 

 part of the sac, and four very prominent antipodal cells. 

 The latter may reach an extraordinary development in this 

 species, as is shown in fig. 130. Other variations from the 

 type were seen in the same species, although as yet these 

 have not been critically studied. These instances are suffi- 

 cient to show that there is much more variability than is 

 usually supposed, and a further examination of these same 

 parts in other low monocotyledonous types may be expected 

 to yield further instances of the same kind. 



The question of the position of the Monocotyledons in 

 the system, and the relative positions of their different orders, 

 is by no means settled. There is, however, a growing ten- 

 dency to regard them as the most primitive of the Angio- 

 sperms, and not as Strasburger has suggested, a group 

 derived originally from the Dicotyledons. So, again, as to 

 the relative positions of Gymnosperms and Angiosperms, the 



