630 GENERAL COMPARISON OF THE FILICALES 



leaf and axis as branches of a dichotomy. The leaves always originate 

 monopodially. Secondly, other Pteridophytes, such as the Sphenophylls 

 and early Calamarians, exist with bifurcate leaves, but without any 

 suggestion of an origin of axis and leaf from a common dichotomous 

 system. These grounds, over and above the inherent improbability of 

 the comparisons with Fucoids introduced by Potonie, or with the game- 

 tophyte of Liverworts by other writers, should suffice to show that the 

 suggestion which I threw out in 1884 is untenable, as I very soon realised 

 it to be. All developmental evidence shows that the axis in Ferns, as 

 in other plants, was throughout descent a part of distinct origin from 

 the leaves which it bears. 



The dichotomous theory of origin of the whole shoot, including axis 

 and leaf, has been supported also by Tansley on the basis of anatomy ; and 

 it has been pointed out that analogies exist between the structure of axis 

 and of leaf in certain early fossils. 1 Especially it has been shown that there 

 is an approach to a radial type of construction of the lower region of the 

 leaf in certain cases. It need be no surprise that such similarities to 

 the structure of the axis should exist in an appendage which is a part 

 of the same shoot as the axis ; as the leaf became larger and more 

 important its requirements would become similar to those of an axis : to 

 meet these a structure analogous to that of the stem would then be 

 probable, such as is actually seen. In the facts adduced I see nothing 

 stronger than structural analogies : this class of evidence carries little 

 weight as against the objective fact that in living Ferns the leaf is always 

 seen to arise monopodially. Thus the dichotomous theory, which is based 

 on analogies, appears to break down in the absence of developmental 

 fact. 



It is possible now to institute a comparison of the shoot of Ferns with 

 that of other Pteridophytes, and to consider its relation to the theory of the 

 strobilus. In its original radial structure, with derivative dorsiventrality, 

 and in its occasional dichotomous branching it corresponds to other 

 strobiloid types. The genetic relation of leaf to axis as actually observed 

 is the same, and in point of fact it is in the proportion of leaf to axis 

 and in the architecture of the leaf that the chief difference lies. But among 

 strobiloid types, and especially among their fossil representatives, the leaf 

 is not always small or simple : the leaves of certain living Lycopods 

 (L. serratum and Isoetes) are relatively large, as were also those of some 

 of the fossils, notably Sigillaria. The branched leaves of the Spheno- 

 phylleae and Psilotaceae, and even of some of the Calamarians, such as 

 Archaeocalamites, and notably of Pseudobornia, are instances of branching 

 of leaves in strobiloid forms. Again, in our view a great leaf-enlargement 

 in a fundamentally strobiloid type has resulted in the Ophioglossaceae. Thus 

 variety in size and complexity of the leaves existed in other Pteridophytes 

 besides the Ferns. Even the dichotomy which is so frequent in the first 



1 Ni'w riiylologist, 1907. 



