140 THEORY OF THE STROBILUS 
It is a fact of importance that, in the individual life, the one or the other 
anatomical type is usually constant; but in certain Ferns the progression 
may be traced from the cladosiphonic in the young plant to the phyllosi- 
phonic in the mature, thus suggesting a similar progression in descent, 
viz. that the large-leaved phyllosiphonic Ferns were derived from a smaller- 
leaved cladosiphonic stock. Of the converse, viz. the progression from 
the phyllosiphonic to the cladosiphonic state in the individual life, I know 
of no example among the Pteridophytes, though it is true that there is 
some approach to it in the Marsiliaceae. Thus the anatomical evidence 
indicates a probability that, even in large-leaved Ferns, the cladosiphonic 
was the primitive type; but that the phyllosiphonic, once initiated, is as 
a rule maintained: this is shown by its persistence in the Seed-Plants, 
even where the leaf has been reduced in size. 
Having thus gained a valuable sidelight from anatomy, indicating 
that small-leaved types were probably primitive, we may now return to 
our central question of the initial relation of leaf to axis. Of the three 
theories already noted, the theory of overtopping as applied to the origin 
of the leaf may, in my opinion, be dismissed, as it is not based upon com- 
parison of nearly related forms, while the facts of embryogeny and of 
leaf-origin do not support it: and further, the sympodial development of 
a dichotomous system, on which it is founded, is a general phenomenon 
of branching, restricted neither to leaves nor to the sporophyte generation. 
As to the other two, the facts, whether of external form or of internal 
structure, seem to me to indicate this conclusion: that the strobiloid 
condition, was primitive for certain types, such as the Equisetales, Lyco- 
podiales, and Sphenophyllales; that in them the leaf was from the first 
a minor appendage upon the dominating axis, and anatomically they have 
never broken away from the cladosiphonic structure which is the internal 
expression of their microphyllous, strobiloid state. That the Filicales and 
also the Ophioglossales were probably derived from a microphyllous 
strobiloid ancestry, and achieved the phyllosiphonic structure as a conse- 
quence of leaf-enlargement, this being the derivative rather than the 
primitive condition; its derivation is even illustrated in the individual 
life of some Ferns. From the Filicales the phyllosiphonic structure was 
probably handed on to the Seed-Plants, and by them retained notwith- 
standing the subsequent leaf-reduction which followed on their adaptation 
to an exposed Jand-habitat. Thus a strobiloid origin may be attributed 
to all the main types of Vascular Plants. It seems to harmonise more 
readily with the facts than any phytonic theory does. 
A prototype, which was probably a prevalent, though perhaps not a 
general one for the Pteridophytes, may then be sketched as an upright, 
radial, strobiloid structure, consisting of a predominant axis, showing con- 
tinued apical growth, and bearing relatively small and simple appendages. 
On our theory the origin of these appendages in descent would be the 
same as it is to-day in the individual development, viz. by_the outgrowth z 
