252 SUMMARY OF THE WORKING HYPOTHESIS 
sufficiently trustworthy guides in the study of the origin of members, 
except perhaps between closely allied organisms. The present tendency 
is to study the embryo biologically, rather than as an embodiment of 
early historical fact: and to recognise that the various appendages of the 
embryo originate in such positions, and at such times as are most. suitable 
for the performance of their functions. The demonstration that “free- 
living” leaves or roots may occasionally exist, suggests that some such 
degree of freedom may rule also in the first stages of the embryo. 
There is, however, one relatively constant and fixed point in the 
embryology of Pteridophytes; it is the position of the apex of the axis 
in close proximity to the intersection of the octant walls in the epi-basal 
hemisphere. This at once defines the polarity of the embryo, whether 
or not the axis may assert itself early by active growth. But when once 
the more plastic stage of the embryo is past, and the characteristic form 
of the plant established, this would seem to be a more reliable basis 
for comparison than the first phases of the embryo (Chapter XIV.). 
A general comparison of the shoot in the sporophyte generation as 
regards symmetry leads to the conclusion that it was originally radial 
(Chapter XVI.). In the Bryophytes the sporogonium is a body which 
shows polarity, but retains with very few exceptions the radial symmetry. 
In the Pteridophytes many retain the radial symmetry also; but others 
depart broadly from it, some at an early period of the individual life, 
others at later periods. These changes may be referred to the unequal 
incidence of external conditions, and it has been shown experimentally 
that a radial structure may be influenced towards dorsiventrality by such 
external causes as unequal incidence of light, or of gravity. This has 
been the probable origin of the dorsiventrality as seen in the sporophyte. 
A comparison of the representatives of the same phylum among them- 
selves frequently indicates that those genera or species which are radial 
are less specialised in other respects than those which are dorsiventral : 
this is particularly clear in the Lycopodiales, as also in the large-leaved 
Ophioglossaceae and Marattiaceae. A careful review of the various phyla 
of Pteridophytes leads back constantly to the radial type as primitive. 
The fact that the radial construction is predominant in the Equisetales, 
Sphenophyllales, and Lycopodiales, while it is prevalent also in the 
Palaeozoic Filicales, shows how strong a hold it had among the earliest 
types of Vascular Plants. 
There is little evidence from plants of the present day of the existence 
of a primitive, permanently free-living, but rootless state of the sporophyte 
(Chapter XVII.). There is no certain knowledge how the root originated: 
it is clear, however, that in the Lycopodiales the structure of the root 
is more like that of their axis than in other plants, while the comparison 
may also be strengthened by the fact of its occasional exogenous origin 
in those plants, and its dichotomous branching. Further in the same 
phylum there exist in the Stigmarian trunks, and the rhizophores of 
