28 THE UNIVERSITY SCIENCE BULLETIN. 



tides of dry sand were so light that unless extreme care was 

 used in covering and uncovering the bottles, some of these par- 

 ticles would be carried out on air currents and so give false 

 results upon subsequent weighings. From the two figures 

 obtained by these weighings, the per cent of moisture in the 

 corresponding sand was secured. 



At the end of the two-week period the seal was broken and 

 the contents of the tumbler were dumped upon a glass plate. 

 A sample was taken quickly for determining the moisture con- 

 tent. Germination was noted and the seeds were separated 

 from adhering sand grains by being gently brushed with a 

 camel's hair brush, were at once dropped into a weighing 

 bottle, and their loss of moisture then determined by weigh- 

 ing and drying to a constant weight. 



Seeds were considered to have germinated when .5 cm. of the 

 rootlet extended through the seedcoat, and to be "incipient" 

 when a shorter length was to be seen. This is another arbi- 

 trary standard, but some such point had to be chosen. 



It is realized that with no means available for controlling 

 the soil temperatures during the tests, considerable error may 

 have crept in, but with all allowance for such in the results 

 following, it is felt that it would not alter the conclusions 

 drawn. 



PRELIMINARY TESTS. 



An early step taken as a guide to the amount of absorption 

 to be expected was to determine the approximate curve of 

 water absorption of various seeds when conditions were favor- 

 able for germination. It was thought this might be used in 

 comparison with results obtained in the tests as an indicator, 

 suggesting nearness of approach to necessary amounts of 

 water to be furnished. Although of little assistance in the 

 way planned, the results later obtained tallied fairly closely. 

 To get these, ten weighed seeds were placed upon wet sand, or 

 on or between pads of wet cotton, in Petri dishes at room tem- 

 perature (averaging 19.5° C.) and weighed at intervals until 

 germination had taken place. 



The results are shown in the following tables : 



