464 Benedict : The genus Ceratopteris 



after tentative groupings on the basis of the sporangial characters. 

 In the Engler & PrantI, Die Natiirlichen Pflanzenfamilien, the 

 number of spores in each capsule is given as sixteen, and Miss 

 Ford also records this number for the material she examined, but 

 in most of the numerous specimens I have seen, the number was 

 certainly more than sixteen, and for many plants was definitely 

 determined as thirty-two. Only in C. dtltoidea was sixteen 

 found to be the maximum number. Indeed in some of the ma- 

 terial provisionally identified as this species (Porto Rico, Sintenis)^ 

 sporangia with fewer than sixteen spores were apparently present, 

 although this count was not verified. The accurate counting of 

 the spores was rendered difficult by the fragile texture of the 

 sporangium walls. In many well-fruited leaves the capsules 

 seemed to be all ruptured, or would become so in transferring to 

 a slide, and in such cases the full number accredited to the species 

 was often not certainly determined, but an intermediate number, 

 I, ^., more than sixteen, was assumed to indicate the larger number. 

 The genus is peculiar in that it varies as previously noted in 

 characters usually very constant in fern groups of higher rank 

 than genera, viz.^ the number of spores per sporangium, and the 

 development of the annulus. The discovery of the form with the 

 vestigial annulus so impressed Hooker that he made it the basis 

 of a new genus and two new species, but he later withdrew both 

 genus and species. Now, although there is no question as to the 

 generic identity of all the plants, the specific identity of all seems 

 hardly probable, and one of Hooker's species, pteridoides^ has again 

 been accepted by Hieronymus and by Underwood for the American 

 plants, which with few exceptions differ markedly from the Old 

 World plants in leaf-form, habit, and habitat 



But the difficulty does not end here. C. pteyidoides as thus 

 recognized itself includes a complex of forms which differ in form 

 and cutting and in sporangial characters, and probably also in habit 

 and habitat as well, and it was with a view of finding out whether 

 there exists any correlation of these differences among the various 

 forms that the present study was undertaken. The results ob- 

 tained are so simple that it seems strange that* they have not been 

 obtained before. However, as stated in the title, the present re- 

 vision is merely preliminary, and the indications are that the genus 



