iQig] HENDRICKS— TORSION 439 



part of the vine within the frame and above the upper clamp as 

 nearly vertical as possible in order to avoid changes in the moment 

 of inertia of the frame, and consequent error in the period of oscilla- 

 tion. Determinations of diameter are least accurate, only to 

 about 3 or 4 per cent, and as the fourth power is used in these 

 calculations, this introduces a further error of about 12-16 per cent 

 in the values for coefficient of rigidity. Moreover, the vines are 

 not perfectly circular in cross-section and are not of uniform 

 texture, as would be the case with a metal wire. For this reason 

 there are objections to the use of the terms rigidity and coefficient of 

 rigidity, in that they do not have as definite values as in physics. 

 When, however, these limitations are borne in mind, and it is 

 remembered that they express relative average values only, it is 

 believed that there should be no confusion. I believe that these 

 errors do not vitiate the conclusions which I have drawn from the 

 curves. It may be well to mention that similar rigidity curves 

 were obtained when the former method of balancing twist in the 

 vine against twist in a fine wire was used. 



It may be objected that the material was studied under 

 unnatural conditions, and that stretching introduces an unknown 

 quantity. This is to a great extent true, but it is also true that 

 many other investigations are made under unnatural conditions; 

 for example, histological studies are usually made upon material 

 that has been killed and stained. As a matter of fact, twining 

 plants, when they can get necessary support otherwise, are some- 

 times observed to twist freely without much coiling or actual twin- 

 ing. This was observed, for example, several times in the green- 

 house when a flowering bean grew near a tomato plant and found 

 support among its leaves and branches. It must also be remem- 

 bered that in ordinary twining the stem is probably subject to more 

 or less tension. What effect, if any, is produced by the amount 

 of tension used, and what effect, if any, the pressure of the clamps 

 has upon the vine can be determined only by further studies. 



The first part of the twist curve, which shows only a little slope, 

 may be taken to correspond with that phase of growth which has 

 to do largely with circumnutation, and in which there is but little 

 twist. Torsion of the stem (but not twining proper) is generally 



