Corn Investigations 81 
Under what conditions and to what extent the kernel weight 
is influenced by the degree of kinship between pollen and ovules 
is not so well understood. It has been suggested by other inves- 
tigators (1) that variety tests as commonly conducted, permit- 
ting promiscuous pollination between neighboring varieties, are 
unreliable because of the immediate effect of foreign pollen upon 
the yield and (2) that farmers may increase their yields by 
planting a mixture of varieties so that cross-pollination will 
occur. The theory upon which these statements are based is 
that the developing corn kernel receives an immediate stimulus 
thru being fertilized by foreign pollen. On this subject the 
following tests have been made at the Nebraska Experiment 
Station. 
METHODS OF DETERMINATION 
Since the effect of foreign pollen is confined to the indi- 
vidual kernels so fertilized, a direct comparison of hybrid and 
pure kernels on the same mature ear is possible. Pure, in this 
case, refers to kernels fertilized by pollen of the same variety, 
and hybrid- refers to those fertilized by another variety. 
For experimental purposes, varieties are used which differ in 
endosperm color, so that the two sorts of kernels' may be dis- 
tinguished by the phenomenon of xenia. 
This mixture of pure and hybrid kernels of known parent- 
age may be produced either by planting two varieties in close 
proximity, thereby permitting natural mixing of pollen, or by 
artificially collecting, mixing, and applying the pollen to recep- 
tive silks which have been previously covered to avoid chance 
fertilization. 
The method of comparing the weights of pure and hybrid 
kernels on the same ear has been commonly employed by other 
investigators and has been used in these investigations. The 
number and location of the hybrid kernels upon the ear is more 
or less a matter of chance. On some ears they are uniformly 
scattered thru the ear (fig. 21) ; on others they will tend to be 
localized in certain portions of the ear. Since the kernels 
usually become systematically smaller with an approach toward 
the tip of the ear, it is apparent that precautions must be taken 
to avoid experimental errors easily resulting from place effect on 
the ear. Pure and hybrid kernels should, for comparison, be re- 
moved in adjacent pairs. Thus the hybrid kernel should be 
taken for the test only when there is an adjacent pure kernel in 
the same row, which may also be removed. If a high de- 
gree of accuracy is desired, equal numbers of pure kernels should 
