STUDIES ON OAT BREEDING. 7 



plants were three inches apart in the row and the rows ni 1911 

 and 1912 were alternately one foot and two feet apart. That 

 is, between the first two rows there was a one-foot space; 

 between rows 2 and 3 there was a two-foot space ; between rows 

 3 and 4 a one-foot space and then again a two-foot space. In 

 this way each plant was allowed the same amount of space, 

 viz., 54 square inches, or two and two-thirds plants to each 

 square foot. 



As soon as ripe the plants from each row were pulled and 

 tied in bundles bearing- a tag with the row number. These 

 bundles were hung in the curing shed until thoroughly dry. 

 Then each plant was threshed separately by hand and various 

 data recorded concerning it. (cf. p. 8). 



In most cases not more than 20 plants were threshed from a 

 single row. These were taken entirely at random just as they 

 came from the bundle. In no case were mutilated plants 

 included. 



For planting the next year individual plants with excep- 

 tionally high or exceptionally low characters were chosen. 

 These were planted in the same kind of garden rows and were 

 harvested and threshed as before. With the exception of one 

 particular the procedure has been the same every year. The 

 one exception is that in 1913 and 1914 the rows were all planted 

 one foot apart instead' of alternately one and two feet as used 

 in 191 1 and 1912. This change in the method of planting was 

 made through a mistake in 19 13. For certain reasons it seemed 

 advisable to continue this method in 1914. This change in the 

 amount of space allotted to each plant undoubtedly afifected the 

 absolute yield and other characters of the plants. However 

 it is also found that seasonal variation so affected many of the 

 characters of these plants- that it is not possible to deal with the 

 absolute figures from one generation to the -next. Instead re- 

 sort must be had to some measure which eliminates the abso- 

 lute measurement to a large extent. 



It is extremely difficult to get measurable characters of oat 

 plants which are not materially affected by environmental con- 

 ditions. In order to carry this work along with other problems 

 it was necessary to take characters which could be fairly easily 

 measured or counted. After a careful study of the available 

 characters it w^as decided to use the more obvious characters 



