222 
Hill and de Fraine. On the 
strands entirely separate one from the other. The bundles of one of these 
partly-divided cotyledons behaved in exactly the same manner as obtained 
in the similar case of P. contorta , var. Murrayana (Series K), mentioned 
above. The bundles of the other bifurcated cotyledon behaved differently ; 
one took no part in the transition and was therefore subsidiary, while the 
other underwent the equal division and rotation associated with the strands 
of a whole-cotyledon. Thus we have an example of the formation of two 
half-cotyledons from a whole-cotyledon, and also the formation of a 
subsidiary seed-leaf from a whole-cotyledon. 
Another plant (Series C) of this same species provides a similar 
instance of the formation of a subsidiary cotyledon from a pre-existing 
single structure; in this case, however, there is a further advance, for the 
cotyledon in question was split more deeply. Araucaria Ctmninghamii in 
which the two cotyledons are deeply bifid, may also be cited. 
These examples, taken in conjunction with the other facts mentioned 
above, furnish a proof as complete as may be expected, but our hypothesis, 
on first thoughts, apparently does not lead us any great distance ; for we 
can, so far, only derive a seedling with, say, ten half-cotyledons from a form 
which had five whole-seed-leaves, which is some way removed from the 
simpler condition of two cotyledons obtaining in Taxus , &c. 
The examination, however, of the foregoing plants has provided several 
instances which show that this difficulty is more apparent than real. 
Many seedlings exhibited cotyledons, the bundles of which behaved 
differently from what they should if they had been of an absolutely rigid 
nature. In other words, it does not follow that because a seed-leaf is 
a subsidiary cotyledon, or a half-cotyledon, that therefore it will always 
remain as such ; a subsidiary seed-leaf may, in the course of events, be pro- 
moted, as it were, to the rank of a half-cotyledon ; while a half-seed-leaf 
may be raised to the dignity of a whole-cotyledon. 
The following examples have a bearing on this particular point : — 
Cedrus atlantica (Series A). One bundle commenced to rotate in the 
manner usually associated with that of a half-cotyledon, but the movement 
came to nothing, and the strand in question finally behaved as that of 
a subsidiary seed-leaf. 
Pinus Coulteri (Series A and B). A bundle from a seed-leaf bifurcated 
and commenced to rotate exactly like the strand of a whole-cotyledon, but 
finally it acted like a trace of a half-cotyledon. 
Pinus australis , P. insignis , and Abies balsamea showed the same 
thing in a more marked degree. 
Pinus contorta , P. contorta , var. Murrayana , P. insignis , and P. syl- 
vestris , provided many examples of the bundles of cotyledons dividing and 
rotating within the seed-leaves in the manner they should if they belonged 
to whole-cotyledons. The bifurcation was sometimes very unequal (see 
