444 Journal of Agricultural Research voi. xxu, no. s 



Shortly after fertilization the percentage of ash commences to drop, 

 falling very rapidly for a few days and then more gradually until com- 

 plete maturity. This is obviously not a clear statement of what occurs. 

 The percentage of ash on the dry-matter basis is a perfectly accurate 

 statement, but the plotted curve of such percentage does not give a 

 graphic idea of what is taking place in the kernel. There is a daily 

 increase in total ash. This increase is almost uniform. The total ash 

 content of the kernel when plotted is an ascending nearly straight line. 

 Whether more of this ash is contained in one part of the kernel than 

 another is not apparent. The ash at flowering time must be in solution 

 and in the protoplasm. There has not been time for any deposit in the 

 newly formed cell walls. Until several days after flowering the ash 

 content must be in the cell sap, the proteids, and such penetration of cell 

 walls as probably would occur if the tissue were not living. 



As about So per cent of the content of a newly formed kernel is water, 

 it was thought at first that calculating the percentage of ash on the basis 

 of water would be the best method of comparison. In the very early 

 stages, before any deposit could occur in the cell walls, this might be 

 true. However, as the development of the kernel proceeds, the water 

 occupies a smaller and smaller percentage of the kernel. Not only does 

 the proportion of cell walls increase, but the proportion of the proteid 

 matter in the active tissue probably is increased by the growth of starch 

 grains. These starch grains, being formed in the cells, must occupy 

 space previously largely occupied by cell sap. 



If the ash is to be accounted for entirely on the basis of cell sap, the 

 concentration of the cell sap must show a progressive increase to account 

 for the total ash. This is highly improbable. The curve of percentage 

 of ash based on water content is, however, more regular than the one 

 based on dry matter and is in the direction of the actual ash deposit. 



The ash was finally computed on the basis of the wet weight of the 

 growing kernel. By computing it on this basis, allowance was made for 

 both the ash in the cell sap and that in the organized components of the 

 cell. The use of such a method assumes that the ash in the dr}^ matter 

 would be a mechanical infiltration from the cell sap which would eventu- 

 ally show the same percentage throughout the cell. When computed in 

 this way a striking uniformity is revealed (fig. 3). Although the pro- 

 portion of water and dry matter varies over a range of 40 per cent during 

 the growing period, the percentage of ash on the basis of wet weight is 

 almost constant. In Table VI are given the analyses of kernels from 

 various plots. These plots differ in irrigation, in the years grown, and 

 in the variety used. The awns from the same samples from which the 

 kernels were taken show a variation of 15 per cent under the radical 

 changes of conditions of growth. The variation in the percentage of 

 ash on the basis of wet weight of kernel is a matter of tenths of a per cent. 

 Many of the apparent fluctuations have plausible explanations. At 



