37 ° 
The Ohio Naturalist. 
[Vol. Ill, No. 4 , 
ATAVISM IN THE WATERMELON. 
John H. Schaffner. 
In the summer of 1895 I noticed a peculiar variation in the 
leaves of a watermelon vine, growing in a patch in Clay county, 
Kansas. The plants were of the variety known as the “ Georgia 
Rattlesnake,” and, excepting the single plant mentioned, were of 
the usual type. 
□ The leaves of the watermelon seem to be quite constant in form. 
They are usually described as palmately five-lobed, the lobes 
being mostly sinuate-pinnatifid, with all the segments obtuse 
(Fig. ib). But in this plant the lobed condition of all the leaves 
was almost entirely absent, the border being only moderately 
undulate (Fig. ic). 
Some of the seed 
from this individual 
were planted in 1896, 
and the same leaf pe- 
culiarity was report- 
ed. The form has 
been successfully cul- 
t i v a t e d every year 
since that time, al- 
though it was usually 
planted in patches 
with the ordinary 
kind and much cross- 
pollination must have 
resulted. 
Whether this con- 
dition of entire leaves 
is common in the wa- 
termelon I do not 
know, but I regard it 
as a good example of 
atavism, or reversion 
to a more primitive 
type. Such reversions 
may perhaps be of frequent occurrence in the species. It is a 
well-known fact that the leaves of many fossil plants from the 
Cretaceous have entire borders, while the modern representatives 
of the same genera are often serrate, denticulate or lobed. Turn- 
ing now to the seedling of the ordinary watermelon (Fig. ia), 
we find that the leaves develop in succession from the entire to 
the mature, lobed form. The cotyledons are oval and entire, 
while the first leaf is almost an exact counterpart of the variety 
under consideration. I11 the second leaf the lobed condition is 
Fig. 1. a, A young seedling of the usual form. 
b , Leaf of usual form. 
c, Leaf of special form . 
