90 EXPERIMENTS WITH PLANTS 



netting) to one-half its depth with damp sawdust; on 

 this lay the seeds, well soaked, and add enough saw- 

 dust (equally damp with the first) to fill the box even 

 with the top. On top of this put several layers of 

 well-soaked cloth or blotting paper, and keep this wet 

 during the experiment; let the edges press closely 

 against the box. Hang the box up so that the under 

 side may be easily observed. Air, light and warmth 

 now come mostly from below instead of mostly from 

 above, while moisture, ordinarily more abundant be- 

 low the seed, is now about equally abundant above 

 and below, and the same is true of any food the 

 moisture contains. Since the sawdust is alike in depth 

 above and below, it cannot influence the roots to 

 rn-ow in one direction more than in another. 



If, now, the roots grow downward (and the stems 

 upward) it must be in obedience to some influence 

 other than these. 



Perhaps this is gravity : How can we test this sug- 

 gestion? What would happen if we should make the 

 force of gravity of no effect by placing the seeds on a 

 revolving w^heel? If the wheel turns always at the 

 same rate, each side of the root will in its turn be 

 underneath for a short time, but no side more than 

 another. So when all sides are equally affected by 

 the downward pull of gravity, will it influence the root 

 to grow in any one direction more than in another ? 

 We may use the apparatus shown in Fig. 75 to answer 



