488 Holden and Daniels . — Observations on the 
for regarding it as of phyletic importance. That it is occasionally of very 
brief duration or that in a few species it may be altogether omitted is only 
to be expected, and in no wise affects the conclusions drawn from its general 
occurrence. 
A further illustration of the value of vascular evidence of a somewhat 
different type is furnished by the varying behaviour of the cotyledonary 
strands in teratological polycotyls. In some such the midrib retains its 
collateral structure and, with a similar strand from a neighbouring cotyledon, 
constitutes a single root-pole. Such cotyledons are held by us to have 
been primarily produced by the qualitative division of the growing apex of 
the seed-leaf. In other cases the cotyledonary strand develops ‘ double ’ 
structure, and may either unite with a similar strand at varying levels in the 
hypocotyl or root or remain entirely separate and constitute an additional 
root-pole. These types are interpreted by us as having arisen by the 
quantitative division of the cotyledonary apex, and have led to absolutely 
symmetrical triarchy and tetrarchy in normally diarch species. Such 
an hypothesis provides a perfectly logical and consistent explanation of the 
origin of teratological polycotyly, and appears to be equally applicable to 
the case of those Gymnosperms which possess polycotyledonous seedlings. 
If this hypothesis is rejected it remains to be explained why growing points 
which, according to Coulter and Land’s theory, are presumably equivalent 
should capriciously produce strands which behave in transition in such 
varying ways. 
Moreover, it is surely illogical to confine the argument to the special 
case of seed-bearing plants, and its absurdity is evident when we attempt 
to apply it to the vascular Cryptogams. As an example we may take the 
ontogenetic sequence in the Filicales, which is familiar enough to need no 
detailed description and the phyletic value of which is undoubted. Evidence 
of yet another kind is provided by the discovery of petrified plant remains 
from the Old Red Sandstone (of Middle or Lower Devonian age) of 
Aberdeenshire. The aerial parts of these plants, the anatomy and morpho- 
logy of which have recently been described by Kidston and Lang (20), are, 
with one exception, entirely devoid of leaves and consist of dichotomizing 
assimilatory branch systems with apical sporangia. The existence of such 
types, which represent the oldest known vegetable petrifactions, would 
seem to lend support to the contention that vascular differentiation phyleti- 
cally antedated differentiation into stem and leaf. 
Proceeding, however, from general criticism to an examination of the 
special evidence upon which Coulter and Land’s theory rests, it will be seen 
that much depends on the interpretation placed upon the dicotyledonous 
seedling of Agapanthtis. It seems to us that the weakness of Coulter and 
Land’s interpretation lies chiefly in the facts which it does not attempt to 
explain. It is not clear, for example, why one of the two growing points, 
