192 KLINOSTAT. [CH. VII 



is equally distributed. An experiment of Elfving's 1 shows 

 that, in some cases at least, the latter is the explanation. 

 If a grass-haulm which has finished growing is kept in a 

 vertical position, the pulvinus undergoes no change, but 

 growth does take place in the pulvinus of a grass-haulm 

 kept in slow rotation on the klinostat. This seems 

 to prove that just as the gravitation-stimulus acting on a 

 horizontal stationary pulvinus produces one-sided growth, 

 so an equally distributed stimulus produces a symmetrical 

 growth, i.e. a simple increase in length of the pulvinus. 

 The pulvini must be marked, and measured microscopically, 

 and they may then be fixed inside a large corked test-tube 

 and kept in rotation for 3 or 4 days. If the test-tube 

 contains a little water the haulms will be kept sufficiently 

 damp. As a control, similar haulms must be placed 

 vertically in a stationary test-tube. This precaution is 

 necessary because the pulvini may not have completed 

 their growth so that the control specimens will show 

 a certain amount of elongation. 



(220) Tlie peg or heel of Cucurbita. 



A similar conclusion mav be drawn from the 



i/ 



behaviour of germinating cucurbits on the klinostat. 

 When a seed of Cacurbita ovifera (vegetable marrow) is 

 allowed to germinate in normal conditions the well-known 

 peg or heel, shown in fig. 38, A' 2 , is developed on the 



1 Ofoersigi of Finska Vetenskaps Societets Forhandlingar, 1884. 



2 Fig. 36 A is from the Power of Movement in Plants, p. 102. 



