102 Nebraska Agricultural Exp . Station , Research Bui . 20 
If we assume that other cells differ as much in size on an 
average as the pith cells do, we may make such comparisons as 
follow. 
The approximate volumes of inbred and hybrid stalks, 
Table 39, were 130,848 cubic mm. and 439,036 cubic mm., giving 
a ratio of 100 :335. The volumes of average inbred and hybrid 
pith cells for the same strains and hybrids were .0016037 cubic 
mm. and .0020230 cubic mm., giving a ratio of 100:126. These 
data suggest that 10.6 per cent of the increased size due to 
crossing results from an increase in cell size and 89.4 per cent 
of it from increased numbers. The cell size increased 26 per 
cent and the cell number 209 per cent. Cell measurements in 
other tissues suggest that the average cell size of the plant was 
not increased to a greater percentage than were the pith cells. 
EAR-TO-ROW BREEDING 
TESTS WITH HOGUE’S YELLOW DENT CORN 
Ear-to-row breeding of corn consists in the separate com- 
parative yield determination of individual ears of a commercial 
variety with the object of isolating the more productive strains. 
An ear-to-row strain is the progeny in direct line of descent 
from an individual ear of corn. A number of plans for con- 
tinuing the apparently superior strains after once being deter- 
mined have been compared at the Nebraska Experiment Sta- 
tion. These plans are as follows : 
1. Continuous ear-to-row selection in which progressive im- 
provement is attempted annually by ear-to-row selection within 
the most productive ear-to-row strains. This work has been 
continued with three of the most productive “Class I” ear-to- 
row strains selected in 1903 from among 104 ears tested. 
In 1907 two separate selections were made from one of these 
and continued as separate strains. For this reason, strains 425 
and 459 as shown in chart 4 were identical during the first three 
years. The pedigree record of yields which indicates the choice 
of plats from which seed selections were made for continuing 
the strains is given in chart 4. In several instances selections 
were not made from the highest yielding plat, because quality 
as well as yield was taken into consideration. 
Seed for the comparative yield test reported in Table 40 
was prepared each year bv mixing the well-developed ears of 
the highest yielding rows of the four strains. 
(2) Continuing the strain by increasing the original ear 
