_ the species, where the season is short, is, as we have seen, 1 
514 
kind of. triangular cob bearing two rows usually, when many ker- 
nels are found, although single kernels may occur in a scattered ~ 
way. If four of these dichotomous kernelled branches be col- 
lected and tied together to form an ear we would. have an ! 
eight-rowed ear with distinct openness between the pairs of rows. ‘ 
If search be made in a crop of flint corn we shall very likely find 
ears that are but partially kernelled, the kernels being scattered 
here and there only over the cob. These kernels are of the type 
of their species, but rounded through the removal of pressure 
against adjoining kernels, and often as thick as broad. The re- i 
lations of these kernels to the cob are however of the greater im- 
terest. Where no kernels are found the cob is rather undevel- 
oped; when the kernels are absent over a considerable space, even — 
flimsy, but thickening again as the kernels appear. The cob — 
hence varies in diameter according to the presence of kernels. 
If the observer is very fortunate he may find a cob with kernels — 
_ in groups of different size. In this case he will observe that there 
is a tendency in the cob to develop according to the size of the 
kernels, as well as according to their number. We here have a 
plausible explanation for the taper in ears where the kernels di- ; 
minish in size towards the tip; and this decrease in size of kernel, : a 
together with slight dichotomous openness in the upper portion, 
may give an ear of a most decided taper. I have frequently MO” 
ticed in varieties of flint and sweet corn that the unripe ©af> 
always are tapering. This seems to be accounted for by the a 
velopment of the kernel from the base upward; the butt kernels 
fertilizing first, the tip kernels last, and often quite an interval of 
time between. If these unripe ears be plucked at the period or 
cident with the commencement of the hardening process, OF eM 
in some cases slightly before, and allowed to continue the process 
of ripening from the stored material it will be noticed that the eat : 
ripens but the taper remains, as the kernels do not ince 
equally in size. This explains, perhaps, the number of tapering 
ears in a belated crop of a variety whose normal ear would be 
scribed as cylindrical, but tapering towards the tip. This qua 
| flint grown at a point somewhat near the extreme northern limit 
ing, while the apparently same variety as grown in the Nor 
