778 Journal of Agricultural Research voi.vi.No.20 



The seeds in the drinking glasses were so placed that they could be 

 examined through the glass. The drinking glasses were kept on a shaded 

 window ledge outside the laboratory during two periods of 10 days and 

 14 days, respectively. Each of these periods included several cold days 

 during which the soil became frozen clear to the bottoms of the glasses. 

 During an intervening period of 9 days and again after the second period 

 on the window ledge they were kept in the laboratory at ordinary room 

 temperature. 



The greenhouse flats were kept outdoors during the entire experiment. 

 The soil in them was alternately frozen and thawed at intervals during 

 the first two or three months. 



Check tests were made in a germinating chamber at about 20 C. 

 without previous treatment of the seeds. 



All of the tests were continued until March 24, 1910, and those in the 

 greenhouse flats until May 7, 1910. 



Table VIII shows the percentages of the seeds which germinated or 

 produced seedlings in the different tests, the percentages which remained 

 impermeable, and the percentages which softened but did not germinate 

 in all of the tests except those which were conducted in greenhouse flats. 



1. Subjection to a freezing temperature previous to the germination 

 test slightly increased the percentages of the impermeable red-clover, 

 white-clover, alfalfa, and black-locust seeds which germinated. In 

 some cases, especially when the seeds had been frozen in water, this 

 treatment increased also the percentages which softened but did not 

 germinate. This latter effect is partly the result of the fact that a part 

 of the seeds which softened after the first period of freezing were killed 

 by the second freezing. 



The impermeable sweet-clover seeds were wholly unaffected by this 

 treatment. 



2. The effect of freezing the impermeable seeds in soil in drinking 

 glasses was similar to but greater than the effect of subjecting them to a 

 freezing temperature previous to a germination test in a germinating 

 chamber. Nearly all of the seeds which softened following each period 

 upon the window ledge softened during the first few succeeding days in 

 the laboratory. 



3. The percentages of the impermeable seeds of the clovers and alfalfa 

 which produced seedlings in the greenhouse flats with frequent freezing 

 and thawing were much greater than the percentages which germinated 

 in the germinating chamber or in the drinking glasses. This was par- 

 ticularly noticeable with the sweet-clover and alfalfa seeds. Nearly all 

 of the seedlings appeared during warm days immediately following 

 freezing weather. 



Only 8 per cent of the impermeable black-locust seeds produced seed- 

 lings, and these few seedlings appeared after settled warm weather had 

 begun late in March. 



