32 EXPERIMENTS WITH PLANTS 



germination) and placing them in an atmosphere satu- 

 rated with water (to prevent them from losing any 

 water), together with untreated seeds as control. The 

 openings may be closed and tested as described on 

 page 10. 



The seeds may be placed on wire netting, in a fruit- 

 jar containing water, as shown in Fig. 25 (the cover 

 must be tightly screw^ed on) . At least a dozen seeds 

 should be treated and the same number used as a 

 control. The precautions against mould mentioned 

 on page 28 should be observed. Set the jar in a 

 place warm enough to promote germination. If the 

 seed depends largely upon the opening for the admis- 

 sion of air, we shall expect to find the germination of 

 the treated seeds noticeably delayed. 



How much air is necessary for germination? We 

 may get an approximate idea by taking six bottles of 

 the same size and filling them to various heights with 

 moist sand, as shown in Fig. 29 (the first, one- sixth 

 full; the second, two-sixths, etc.). To exclude air 

 from the sand, the bottle may be first filled with water 

 and the dry sand slowly poured into it to the desired 

 height; the superfluous water may then be poured 

 off. The seeds (well soaked) may then be put in (the 

 same number in each bottle), the cork may then be 

 pushed down below the rim and sealed air-tight with 

 vaseline or sealing-wax. A closer approximation to 

 the correct amount could easily be made if the experi- 



