72 PLANT MORPHOLOGY 



Much of the difference of view as to foliar origin centres round 

 the question whether originally the leaf was relatively large, or small. 

 Those who hold that the larger-leaved forms were the more primitive 

 will be naturally disposed towards the view of the original preponder- 

 ance of the leaf over the axis, and will favor some phy tonic theory; 

 those who hold the smaller-leaved forms to be the more primitive 

 will probably adopt a strobiloid theory of origin of the leafy sporo- 

 phyte. I propose to offer some remarks on the relative probability 

 of these alternative views. 



If large-leaved prototypes be assumed generally for vascular plants, 

 this naturally involves a widespread reduction, since small-leaved 

 forms are numerous now, and have been from the earliest times of 

 which we have any record. Reduction is a ready weapon in the 

 hands of the speculative morphologist, and it has often been used 

 with more freedom than discretion. But reduction should never be 

 assumed in order to meet the demands of convenience of comparison, 

 nor as a cover for doubt. The justification of a view involving re- 

 duction must be found in its physiological probability in the case 

 in question, and this should be backed by comparisons of form, and 

 of anatomical structure: the conclusion should also be in accordance 

 with the paleontological record. All suggested cases of reduction 

 where such justification is absent should be looked upon as doubtful. 



Convincing evidence of reduction of leaf-complexity in an evolution- 

 ary sequence, supported on all these grounds, has been adduced in 

 the progression from ferns, through cycado-filicinean forms, to the 

 cycads, and it applies with special force in the case of their sporo- 

 phylls. Ferns, which are essentially shade-loving and typically zoidi- 

 ogamic, or amphibious, may be understood to have given rise to 

 the cycado-filices and cycads, which are more xerophytic, and show 

 that essential character of land-plants, the seed-habit. Not only 

 is such a progression physiologically probable, but it is supported 

 by paleontological evidence, as well as by detailed facts of anatomy, 

 and of reproductive morphology. The case for reduction of leaf- 

 complexity seems to be here fully made out: and somewhat similar 

 arguments will also apply for other types of gymnosperms. 



7, Caen. 1903) is analogous to that of Potonie", though differing from it in detail. 

 It involves the ranking of the lycopod leaf as a " phylloid," the leaf of the fern as a 

 true leaf, or "phyllome," differentiated from an indifferent system of "cauloids," 

 on which the "phylloid" appendages had become abortive. It regards the leaves 

 of equiseta and sphenophylls as phyllomes, reduced from the larger-leaved fern- 

 type. The argument is chiefly based on comparisons as to branching and anatom- 

 ical structure. I do not think that these grounds suffice to override the probability 

 that the leaves of lycopods are essentially of the same nature as those of the sphe- 

 nophylls, or equiseta. (Compare my Studies, no. v.) Professor Lignier's view 

 further involves the acceptance of homologous alternation, while he makes no 

 mention of the chromosome-differences of the two generations. Such difficulties 

 do not arise if the leaves of the sphenophylls and equiseta are regarded as being 

 in the upward rather than the downward scale of development, a view of them 

 which would equally harmonize with the anatomical comparisons of Prof. Lignier. 



