THE A WAKE XI NO OF THE SEED 31 



to harden, then test the joint by placing the sealed end 

 of the tube under water and blowing into it forcibly. 



The weight of the water- column tends to draw air 

 through the cover. If any air enters it will appear as 

 a bubble at the top of the tube; should this happen, 

 empty the tube and test the joint as before by blow- 

 ing forcibly. If no leak is detected, refill the tube 

 and invert it as before. Test different sorts of seeds, 

 especially those whose covers seem most porous and 

 permeable to air. It should be remembered that in 

 this case we are testing a seed -cover which is in con- 

 stant contact with water; peihaps in a dry condition 

 it would admit more air. This may be tested by fill- 

 ing the tube only partly full, so as to leave a large 

 air- bubble at the top of the tube; mark its limit 

 exactly on a piece of paper pasted on the outside of 

 the tube; the entrance of air can then be detected by 

 the increase in the size of the bubble. We may, if we 

 wish, avoid any contact of water with the cover, by 

 sealing on a cover, warming the tube well and placing 

 the unsealed end in water ; as the air in the tube 

 cools, water will be drawn in. Which of these tln^ee 

 methods approximates closest to the condition of the 

 cover during germination ? 



Does air enter the seed through the openings? You 

 may test this matter by closing the openings of seeds 

 which have been previously thoroughly soaked (in 

 order to ensure a sufficient supply of water for 



