TESTS OF SELECTIONS OF OATS. 
13 
The advantage from the small plats is not only in point of accuracy, but also in 
the area of land required. Seven one-tenth-acre plats covered an area of 30 492 
square feet, while 70 of the 17-foot rows (10 rows of each variety) required only 1^190 
square feet. The use of the row method in variety testing is commended by the 
results of this test. 
When the crop is ripe and ready for cutting, each row is cut and 
tagged separately, as shown in figure 4. The rows are then shocked 
together under canvas to dry for thrashing. The tin-ashing is done 
by means of a machine constructed by Mr. H. W. Teeter, superin- 
tendent of field work in the plant-breeding department. This con- 
sists of a cylinder built into a box lined with galvanized iron. The 
Fig. 4.— Harvesting the 15-foot rows of oats at the Cornell University experiment station. 
box is so constructed that the top and back may be lifted up, allow- 
ing the grain and straw to shake down on a coarse screen placed 
over a receptacle which receives the grain, as shown in figure 5. The 
advantage of lifting the top is that a full view of the interior is pos- 
sible, so that if any kernels remain behind they can be seen and 
brushed out. The cleaning is done by means of a small counter seed 
cleaner. 
Since 1908 every tenth row has been used as a check. All the 
check rows are sown with the same kind of seed, so that a check on 
soil differences may be obtained. The same variety, selection 
50al-18-4 from the hybrid Sixty-Day X Probsteier, was used as a 
check from 1908 to 1911, inclusive, but in 1912 a new check, Canada 
