AUTONOMIC MOVEMENTS IN SUCCULENTS 339 



The devise for fastening the knife edge to the pot, which is shown in the illustra- 

 tion, was found to be much more satisfactory than the cement sealing recommend- 

 ed by the authors. The glass pots used by these same authors were not suitable 

 for the present work. Consequently, the plants were allowed to become estab- 

 lished in galvanized iron cylinders in the bottoms of which holes were punched 

 for drainage. At the time of the experiment the bottoms of the cylinders were 

 sealed with galvanized iron caps held in place and made airtight by plastocene. 

 The sealing for the tops of the pots was made with heavy cardboard cut to fit 

 and made airtight by a thick layer of plastocene. A rubber stopper was fastened 

 with wire around the woody base of the plant so that the plastocene might not 

 come in contact with the plant surface. All plastocene was completely covered 

 with tin-foil, in order to avoid errors from dust, knocks etc. 



Results: The same facts appear as in former experiments (1) Under normal 

 conditions, the plant loses more water at night than it absorbs, and in the day 

 time it absorbs more than it loses. This means that its turgidity is less at night 

 than in the day time. (2) In all cases lessened turgidity is accompanied by a down- 

 ward movement of joints and increased turgidity by an upward movement. 



The agreement of the results from this experiment with those obtained in 

 experiment 9, where the actual amount of water absorbed by the roots was meas- 

 ured directly, shows the correctness of the assumption made above, namely, 

 the conducting tissue must be able to pass up water as fast as it is absorbed by 

 the roots. Consequently the amounts under the heading A in table 7 are a true 

 approximation of the water absorbed by the plant. 



Although it is aside from the subject considered in this paper' 

 it seems worth noting that this last experiment shows that when 

 wilting point determinations of cacti are made by the Briggs 

 and Shantz method, care must be taken that readings are made 

 at the corresponding hour each day, since any reading taken 

 for an interval in which the night occupied the greatest part of 

 the period might show an apparent wilting point when the 

 plant was still obtaining its normal amount of water from the 

 soil. 



Such changes in turgidity as the last experiments have shown 

 to exist, would probably cause both longitudinal and radial 

 expansion and contraction of joints. Some joints are for a 

 time in a vertical position and at that time show no upward and 

 downward movement. Joints in a vertical position were chosen 

 as subjects on which to detect contraction and expansion. Meas- 

 urements of length with a meter stick and of diameters with a 

 vernier caliper gave no evidence of such changes ; but the auto- 

 matic method of magnification used in the next experiment 

 showed them clearly. 



