34 



THE UNIVERSITY SCIENCE BULLETIN. 



TABLE 6. Result of tests with wheat. 



Of these, No. 103 gives illuminating: results with Nos. 101 and 104 close seconds. 



DISCUSSION. 



Some interesting things are shown in these tables. Num- 

 bers 22-35 started with moisture contents above that of the 

 wilting coefficient of this sand, 1.31 per cent; the remaining 

 ones quoted were below it. Numbers 36, 38, 59, 60 and 103 

 showed satisfactory germination in a soil given less than the 

 wilting coefficient of moisture. Others are very close, not 

 listed simply because fewer of the seeds germinated. Some 

 are very suggestive: Numbers 28 and 29, for example, fully 

 germinated and with original moisture content but 0.08 and 

 0.05 per cent, respectively, above the limit. There seems abun- 

 dant evidence in the results shown here to indicate that seeds 

 can germinate at or below the wilting coefficient of the soil. 



Why germination did not take place in some instances is 

 still a problem. For example, in number 4, with 1.55 per cent 

 of moisture on the start, the seeds became slightly swollen with 

 one rotted, and 1.30 per cent of moisture remained in the sand 

 at the close of the test. In the light of the other tests, it hardly 

 seems that five infertile seeds were selected for this particular 

 one. 



Further, germinating seeds pull the moisture content down 

 to surprisingly low figures, the average, as already given, being 

 0.584 per cent for corn, 0.42 per cent for beans, and 0.83 per 

 cent for wheat. This evidently depends considerably upon the 

 rapidity with which water moves through the soil, as referred 

 to above. In this connection, while Briggs and Shantz found 

 the same amount of moisture remaining in the soil at perma- 

 nent wilting regardless of the kind of plants grown in it, re- 

 sults here show quite the contrary, as just pointed out. Of 

 course their plants had root systems distributed through the 



