12 BULLETIN 137, U. S. DEPARTMENT OF AGRICULTURE. 



stance, of 242 selections of 6-rowed barley, the culms of 153 meas- 

 ured i 5 mm. in thickness and only 33 deviated more than 0.1 mm. 

 from this figure. 



THE EXSERTION OF THE SPIKE. 



The exsertion of the spike is closely related to the length of culm 

 because it depends upon the elongation of the peduncle. Some bar- 

 leys clear the boot much more completely than others. That this is 

 a true varietal character is shown by the number of varieties in 

 which it has been described. The Princess in Sweden is often in- 

 cluded at the base. The same is true of this variety in Minnesota 

 and California. The Smyrna seldom clears the boot completely in 

 more than one or two culms on each plant. An interesting fact was 

 noted in this variety with reference to location. In Minnesota, half 

 the head often remains in the boot, and the same condition prevails 

 over the whole of the Plains area. In California, however, the heads 

 are completely exserted. The exsertion is still short as compared with 

 most varieties, but it is perfect. Like other physiological charac- 

 ters, the exsertion of the spike is variable, but its range of variation is 

 sufficiently limited to occasionally determine a variety. That it is 

 not more often useful is due to the fact that almost all barleys are 

 of the type in which the spike is completely exserted. 



NUMBER OF NODES PER CULM. 



The number of nodes to the culm is naturally identical with the 

 number of leaves to the culm and is discussed under that heading. 



NUMBER OF CULMS PER PLANT. 



The number of culms per plant seems to be a varietal character, 

 but one which is so dominated by environment as to make it impos- 

 sible to determine when it is given true expression. It is probable 

 that all students of the cereals have gone through the same process of 

 diminishing confidence to final doubt as to the utility of this factor. 

 In this investigation the number of tillers was recorded on over 

 20,000 plants without being able to discover a method of using such 

 information for minor distinctions, as was possible, for instance, 

 with the time and method of tillering. The broad groups vary as 

 groups in this character and occasionally a variety deviates suffi- 

 ciently from its group to become distinct, but the mass is, for the 

 most part, inseparable. 



Two causes of variation were studied in detail, viz, spacing and 

 geographical location. In Minnesota a selection of Smyrna, a heavy- 

 tillering variety of a 2-rowed group, and a light-tillering selection 

 of Manchuria of the 6-rowed group were planted at three different 

 spacings. The results obtained are shown in Table II. As will be 



