CEREAL INVESTIGATIONS AT THE NEPHI SUBSTATION. 43 
A remarkable uniformity is observed in the length of heads from all 
sizes of seed sown at all the different rates. 
In yield the large seed gave the best results. This was true quite 
generally throughout the test. With all sizes of seed the 2-inch rate 
of seeding gave the highest yields. This evidently was due to the 
facts that (1) the percentage of survival from the 2-inch spacing was 
equal to that from the other rates, thus leaving more plants in the 
row, and (2) though the number of heads per plant was smaller there 
was no great difference in the average length of the heads. 
The general summary of Table XVI shows further that with one 
exception the Turkey variety surpassed the Koffoid variety on every | 
pomt. Only in the case of the length of the head the Koffoid variety 
had the advantage. But what the Turkey variety lacked in length of 
head it made up in percentage of survival, in number of heads per 
plant, and hence in yield. 
TREATMENT OF SEED WHEAT FOR BUNT. 
Tests comprising 76 nursery rows, each 2 rods long, for the purpose 
of studying the effects of various seed treatments on bunt infection, 
were started in the fall of 1909 and continued through 1912. The plan 
of the experiment was made by Mr. Edward C. Johnson, then patholo- 
gist in charge of cereal-disease work. 
The object of the experiment was to determine the effect on bunt 
infection of (1) the time of planting winter wheat, (2) treatment of 
the seed with different solutions of varying strength, and (8) the re- 
moval of smut balls during seed treatment. The last two points, being 
closely related, were studied simultaneously. 
Two varieties of winter wheat were used, viz, Turkey (C. I. No. 
2998) and Odessa (C. I. No. 3274). The Turkey is a bearded, white- 
chaffed, hard, red-grained variety, while Odessa is a beardless, brown- 
chaffed, soft, red-grained variety. Both varieties were heavily inocu- 
lated with smut spores taken from the Odessa wheat grown on the farm. 
TIME OF SEEDING AS AFFECTING BUNT INFECTION. 
Seed of each variety, which previously had been inoculated, was 
sown without treatment on six different dates in 1910 and 1912. In 
1911 this part of the test was omitted. One 2-rod row of each 
variety was sown on each of the following dates: August 15, Septem- 
ber 1, September 15, October 1, October 15, and November 1. The 
seed was sown by hand and at a much higher rate than that at which 
a machine would have sown it. 
When the grain was ripening the next fall, the percentage of bunt 
in each row was determined by the following method. At least 200 
