peters: moisture requirements of seeds. 



35 



soil and with very short distances, comparatively, to pull the 

 water ; transpiration was going on ; and wilting gave a more 

 or less definite end-point ; while here, there were practically no 

 roots, just as many absorbing centers as there were seeds. 

 There was no transpiration to be a factor, and the end-point 

 was not even approximately fixed, making this problem really 

 in no way comparable to theirs. Yet, in a series from the corn 

 tests where the moisture supplied was above the wilting co- 

 efllicient, there remained at the close of the tests, 0.48, 0.51, 

 0.68, 0.67, 0.69, and 0.51 per cent, respectively, and with the 

 crude apparatus used, with the lack of soil temperature control, 

 and with the variations in the end-points reached, these do not 

 reallv diff"er a great deal. 



taf. 



6o 



40 



Flfi. 2. Curves showing increase in the surface forces of soils as drying proceeds; 

 to the left, for subsoil of the Oswego silt loam: to the right, for Xo. 2/0 sand. 



But, in contrast, in those tests which started with just about 

 this amount of water, the corn grains showed absorptive power 

 sufficient to pull the water down to 0.29, 0.38, and 0.41 per cent, 

 respectively. Dead plants, as shown by Briggs and Shantz (1) , 

 would have done this, or more, if extending through the seal, 

 but here it went into the seeds. This is especially interesting 

 in view of the fact shown by Shull (8) in his graph reproduced 

 here, that the soil forces tending to retain moisture increase 

 enormously as the soil becomes drier and drier, especially when 

 approaching air-dry conditions. In these three instances there 

 is shown a tremendous absorptive power which is evidently 

 not present in the six cases given above, or thej^ would have 

 pulled more moisture from the sand. 



