284 
AMERICAN JOURNAL OF BOTANY 
[Vol. 10, 
Another difference in variability which is noteworthy is that between 
the length of the midrib and of the lateral ribs, especially in the horizontally 
oriented leaves. Here there is an average difference of 6.4 percent, which is 
more than four and a half times its probable error and thus evidently 
significant. 
Other Evidence 
Aside from these gross measurements, the authors examined the terminal 
buds of horizontal shoots during the winter and found (as has been noted 
by others) that the primordia of the four leaf groups were essentially equal 
in size, indicating that visible differentiation, at least, does not begin until 
the shoot starts to grow. The experiments of earlier workers on twig- 
reversal were also repeated and confirmed, it being found possible to reverse 
the anisophylly completely by twisting a horizontal stem through an angle 
of 180 0 , provided this was done before the buds opened. 
Discussion 
We should now consider the bearing of our results on the various theories 
which have been put forward to account for anisophylly. 
The evidence with regard to light is significant. It is a matter of 
common observation that the thickness of the blade is generally directly 
proportional to the amount of light it receives during development. The 
fact that the four leaf groups are identical in blade thickness therefore 
justifies us in concluding that they must receive essentially the same amount 
of sunlight and that light cannot be a factor of very great importance in 
determining the marked differences which exist between them. 
The case is different with regard to gravity, however. Not only are 
all the foliar structures on the lower half of the branch much larger than 
on the upper, but the shape of these organs is what we should expect to 
result if the downward pull of gravity is really effective in development; 
for the vertical upper leaves are “telescoped,” the blade being relatively 
short in proportion to its width and the petiole relatively short in proportion 
to the blade; and the vertical lower leaves are extended or drawn out, the 
blade being relatively long in proportion to its width and the petiole rela¬ 
tively long in proportion to the blade. 
It may be objected that these results are really due to some other factor 
and that neither these facts nor the experimental data gathered by ourselves 
and others at all prove that gravity is really the dominant factor concerned, 
in this connection the results as to relative variability are of interest. We 
have shown that the vertical upper leaves are the most variable of all in 
their size characters (the vertical lower ones probably ranking next), and 
that the lateral veins in the horizontal leaves are more variable in length 
than the midrib. This is difficult to explain until we study individual 
twigs and find that the intensity of the anisophylly which they display is 
