I36 MAINE AGRICULTURAL EXPERIMENT STATION. I916. 



head of the Banner oat has a tendency to distribute its branches, 

 in such a manner as to leave a part of the circumference vacant 

 or filled only with one or two branches. Thus the Banner 

 head is not perfectly symmetrical. This lack of symmetry is 

 still more intensified by the different angles which the branches 

 in the different whorls form with the main axis. 



While the Banner oat has a typical stiff-branchea head, the 

 pure lines of the Irish Victor variety show characters of a 

 transitional type of head. A comparison of figure 19 with figure 

 20 clearly shows the difference between the two types. The 

 panicle of the Irish Victor lines suggests a greater symmetry 

 caused by the more proportionate dimensions of the branches 

 and the more uniform angles between the branches and the 

 main axis. It will be noted that the branches are more droop- 

 ing, thus causing the head to appear shorter and more extended 

 on its circumference. The branches of the Irish Victor head 

 spread more symmetrically about the whole circumference and 

 are somewhat finer than those of the Banner oat. The mam 

 axis ascends straight upwards and is rich in spikelets at the top. 



The type of head of No. 247 of the Imported Scotch variety 

 shows still more clearly the drooping habit. The general picture 

 of this type is given in figure 22. The head is markedly smaller 

 than that of the two former types, the branches are shorter 

 and decidedly more drooping. The main axis is straight but 

 with a slight tendency to bend over at the top. The branches 

 are proportionate and symmetrically spreading. 



While the figures 19 to 24 give the general appearance of the 

 three types of head, a closer analysis of the details of these 

 head characters is shown in Table 10. 



This table does not include all the 12 lines. However it con- 

 tains the best representatives of the three types. The given 

 data represent the means of measurements of a number of in- 

 dividuals of each line. From this table it may be noted that 

 there is a correlation between the length of the head and the 

 total number of spikelets. This correlation and several others 

 have been established by v. Tschermak 28 and by Love and 

 Leighty. It has been shown that with an increase in the length 

 of the head there is a corresponding increase in the number of 



3 Cf. Fruwirth, loc. cit., p. 332. 



