W. E. Brencitlf.v and A. D. TTall 



201 



steadily until about six days before cutting, after which it falls off: the 

 dry weight rises steadily, though there is little increase in the last six 

 days. The riper Square Head's Master even shows a slight but per- 

 ceptible decrease in the weight of 1000 grains in the last three days. 

 This is pi'obably real and due to the continuance of respiration after 

 migration had ceased, though the loss is so small that it falls within the 

 limits of experimental error. It is impossible to obtain quite con- 

 cordant results in drying material like grain, which will continue to lose 

 water in the drying oven at 100" C. for an indefinite period. 



Fig. 4 shows the relationship of green to dry weight — all three 

 samples in 1907 follow a very parallel course, the notable features of 



— . —^y^ ' 



_^ 



Plot 10 

 R.F. 

 Plot 3 



3 6 9 12 15 18 21 24 27 30 33 36 39 days 



Fig. 4. °\^ dry weight to green weight, 1907 only. 



which are a change of curvature after the third period, and another 

 change about six or nine days before cutting. Both these breaks are 

 symptomatic ; as will be seen later the first marks the final contraction 

 and drying up of the pericarp, the second indicates the beginning of 

 desiccation and the conclusion of the migration. 



Fig. 5 shows the actual water contained in 1000 grains and is highly 

 instructive : the water rises until the third or fourth period, then 

 it remains approximately constant in amount until six days from 

 cutting, after which it falls rapidly. Again the two critical dates 

 are about twelve days after the first sampling and six days before 

 cutting. 



Nitrogen. The percentage of nitrogen in the dry matter of the 

 grain (Fig. 6) falls rapidly at first but after the first six periods becomes 



