434 . BOTANICAL GAZETTE [December 



of a right-handed screw in the latter, while it is in the same sense 

 as that of a left-handed screw in the former. The rigidity curves 

 for the bindweed are much smaller in dimensions than for the bean. 

 It will also be noticed that differences in diameters account to a 

 great extent for differences in rigidity curves in general. 



Curve E-1 is interesting in that it shows that when the healthy 

 vine is limited in its growth by unfavorable conditions the same 

 rules as to twisting apply. Even in this case there was, as a matter 

 of fact, a slight reverse twist at the end, although the figure does 

 not show it. On the other hand, under distinctly adverse condi- 

 tions (as when the greenhouse became too hot and too dry), not 

 only was the growth itself stunted, but the torsion was quite 

 irregular. The resulting plant in such a case could hardly be 

 regarded as healthy. 



Curves £-11 are also interesting. Here at the beginning of 

 the experiment the growing tip was purposely removed, but not the 

 leaf at the upper node, which was just above the clamp in the 

 frame. This produced no apparent effect upon the two curves. 

 On the other hand, in some experiments the vine just below the 

 upper node was pinched accidentally, after which that internode 

 died. In this experiment, however, it was observed that the 

 stump above the upper node grew 2 or 3 cm. during the experiment 

 and exhibited apparently a normal amount of twist. In the 

 meantime a new bud appeared at this node. 



One more interesting point is shown by curves F. Two adjacent 

 internodes were fastened in the apparatus and were measured for 

 length and twist in the usual way, while the behavior of the node 

 between them was studied by means of the mirror clamp (fig. 4). 

 From the resulting data three curves were obtained: one for the 

 part below the node, one for the part above, and one for the total. 

 These curves are plotted together and then a straight line is drawn 

 from the "zero" point to the top of the total curve. The tops of 

 the other two curves fall approximately on this line also. This 

 shows that in a limited portion of the vine, where torsion is taking 

 place freely, the final amount of torsion per unit length is uni- 

 formly distributed. If a single internode had been studied in this 

 way, presumably a similar result would have been obtained. This 



