396 Journal of Agricultural Research voi. xix, no. 9 



While time of heading in barley is doubtless significant, it is very 

 difficult to determine. A barley spike may be visible two or three days 

 before it is fully exserted from the sheath. In some varieties the spikes 

 are never completely exserted. In a study of this difficulty it was noticed 

 that the emergence of the awns offered opportunity for a tangible obser- 

 vation. Upon trial it was found to be a very accurate index of the stage 

 of the development of the spike. With the observation as a basis, spikes 

 tagged as uniform before flowering were of so nearly the same stage of 

 development that, despite individual fluctuations, growth in as short 

 periods as 12 hours was evident in the data for many days; and almost 

 until maturity the individual variations in samples of only two spikes 

 did not obscure the growth in 24-hour periods. The accuracy of the 

 method and the spectacular uniformity of Idaho seasons is well shown in 

 figure 2, where the percentages of moisture in kernels in the seasons of 

 1 91 6 and 191 7 essentially coincide throughout the entire period of growth. 



Three or four days after the tips of the awns are visible on the earliest 

 culms a large number of culms are to be found with tips visible. At 

 this time the plots are carefully inspected and the requisite number of 

 culms is marked. The marking is done by tying a piece of wool yarn 

 about the culm. Culms are selected in which the awns are protruding 

 X to J^ inch above the sheath of the uppermost leaf. A sufficient num- 

 ber of culms is tagged to insure against accident. As soon as the spikes 

 are partially exserted a sample is taken. This sample and the one on 

 the following day usually have several florets which have not yet been 

 fertilized. The samples taken in the first few days consist of three 

 spikes in order to secure a greater quantity of material, but later the 

 number is reduced to two per sample. In most cases only one sample is 

 taken each day, but in the cases furnishing the data reported in Table I 

 two samples were taken, one in the morning and one in the evening. 

 The samples are taken in the field by cutting the culms near the ground. 

 These culms and spikes are wrapped in a moist towel and taken to the 

 laboratory. As a protection against evaporation in the laboratory the 

 spikelets are removed one at a time, the remainder of the spike being 

 left in the towel. To secure the data rapidly and satisfactorily two men 

 work on the same sample. The kernels are taken from the florets by 

 the operator of the calipers, who measures the length, lateral diameter, 

 and dorsoventral diameter in tenths of millimeters and records these 

 measurements. The kernels are then passed to the operator of the 

 balance and weighed to tenths of milligrams. Only the kernels on a 

 single side of each spike are measured and weighed individually. The 

 kernels of the other side of the spike are added to those measured indi- 

 vidually and weighed to obtain a larger sample. These are placed in 

 small vials and dried in a water-jacket oven. The vials are then corked, 

 and the material is preserved for later analysis. 



