430 BOTANICAL GAZETTE [decembek 



of vine per average unit diameter. As in most of the experiments 

 no accurate measurements of the diameter were made, the expres- 

 sion more commonly used is L-^T\ I shall call this simply the 

 rigidity, or the relative resistance to twisting through a unit angle of 

 an average unit length of the vine. Certain objections to these terms 

 will be discussed later. As will be seen, they are not constant, 

 but vary. 



These experiments were performed in a greenhouse and have 

 extended over several seasons. Generally speaking, conditions 

 were found satisfactory for growth only during May, June, and the 

 first part of July, at least in this particular greenhouse. 



The apparatus is set up on a bench and leveled so that the 

 upright is plumb. A weight placed on the base and wires from near 

 the top of the standard to the roof of the greenhouse give the 

 apparatus stability. By means of slots and screws in the top 

 piece and in the arm of the clamp {K), adjustments are made so 

 that the vertical line to be taken by the silk and the part of the vine 

 used passes through the center of the semicircular scale. The 

 proper weight is then applied to the silk, and the upper part of the 

 growing internode near the tip is clamped in the frame. The lower 

 part of this internode, or the upper part of the next internode, is 

 then very carefully secured by the clamp {CI). 



Measurements of diameter were made by means of a gauge 

 similar to one form of wire gauge. This consists of a pair of straight- 

 edges, one of them graduated, held securely together so as to make 

 a V with a very small angle. Usually many readings were taken 

 (about 10-40), but irregularities in the stem prevented much 

 accuracy. It was not found possible to make many such measure- 

 ments in the case of the black bindweed. General observations 

 of weather and temperature were recorded, but were found to be 

 of only limited value. 



Data and curves 



Table I contains in an abridged form the data of a typical experi- 

 ment. The dates and times of the readings are given with the 

 corresponding lengths and the corresponding amounts of total 

 twist in angular degrees. The average periods of oscillation of the 



