— 40 — 



opposite, as in the tipper pinnae, and the outlines of the pinnae 

 are wavy. It is these characters plus the scale characters, the sub- 

 falcate long segments as compared with the blunt, triangular 

 short segments of the cristata group, and the lack of triangularity 

 of the lower pinnae which show the close relationship of goldicana 

 and cclsa and their difference from the other group. 



The cause of the reduction of these basal pinnules on the 

 lower pinnae seems undoubtedly largely due to the effects of light 

 in flexing the pinnae, the species not being able, as in the cristata 

 forms to increase the distance between the pinnae, or rather has 

 adopted a different method in its evolution in adapting itself to 

 similar light conditions. In cclsa we have an extreme condition 

 of the goldicana type which has adapted itself to a greater amount 

 of light and also a greater poverty of root moisture, for as far as 

 known it does not grow in soil. Its pinnules are smaller and 

 wider spaced, the pinnae are smaller and wider spaced, the frond 

 is relatively taller and narrower and the reduction of the basal 

 pinnules is greater, as shown in my specimens and figures. With 

 these characters there is a greater flexing of the lower pinnae and 

 most of their pinnules are alternate. So in the cristata group, we 

 find clintoniana, Uoridana, and boottii, with values in the order 

 named as divergents in different directions from cristata and 

 ecologically they have a different habitat. 



The great difference between D. floridana and D. g. celsa 

 show that their evolution has been quite different, but their sim- 

 ilarities suggest the ecological values of their environments. The 

 questions involved are largely ones of plant mechanics and re- 

 quire for their solution not only field study, but a large series of 

 specimens of different ages and conditions. The relation between 

 average daily light and root moisture and the movements of the 

 frond parts, both during the life of the frond and the evolution of 

 the species, is similar to the distribution and plane direction of 

 the leaves of a tree where none touch but all occupy definite 

 space, determined as it were by a tacit agreement not to intrude 

 on another's territory. Hence leaves are ecologically comparable 

 with pinnules. 



In the following key I have left out the generic and minor 

 characters and have arranged the wording so that the characters 

 given are strictly comparable. 



