IO2 A. O. WEESE. 



reactions of the animals to air of different evaporating powers. 

 The evaporation gradients were established by varying the 

 humidity, the temperature and the rate of flow of the air to which 

 the animals were subjected. The apparatus used in the gradient 

 experiments was essentially the apparatus described by Shelford 

 and Deere (1913) and Shelford ('14.0) with the following modi- 

 fications. The source of the air was the compressed air system of 

 the University of Illinois. The measurement of the flow of the 

 air was made more accurate by the insertion in the supply line 

 of delicate draft gages described by Hamilton ('17). The 

 gradient cages used in this series of experiments were also larger 

 (45 X 8 X 5 cm.) to accommodate the larger animals, and the 

 floors were covered with a layer of sand I cm. in thickness. 



The experiments ordinarily covered each a period of twenty 

 minutes, each space on the recording blank representing ten 

 seconds, as indicated in the accompanying Chart I. The figures 

 at the top of each column represent, respectively, the relative 

 humidity, computed from readings of standard wet and dry bulb 

 thermometers, and the evaporation per twenty-minute period, 

 measured with porous cup atmometers, of the air supplied the 

 corresponding third of the gradient cage. In the graphically 

 recorded experiments an animal was placed in the cage, generally 

 in the center section, and its movements recorded by means of 

 tracings. In the statistically recorded experiments ten indivi- 

 duals were distributed through the cage and records of their 

 positions made at the indicated intervals. Experiments des- 

 ignated by the same, number were performed consecutively and 

 with the same animal or group of animals. Otherwise the num- 

 bering is entirely arbitrary. For a more complete account of 

 the method of recording, etc., see the explanation of Chart I. 



In Experiments 12, 13 (Table I.) ten individuals were placed in 

 the gradient cage and readings as to position were taken every 

 thirty seconds, the final results being expressed in percentage 

 figures. Experiment 12 was performed with individuals from 

 Cage No. I, in which the temperature varied between 24 and 

 35. while those used in experiment 13 were from Cage No. 2, 

 where the temperature rose to 40 during the warmer parts of the 

 day. There was no marked difference in the behavior of the two 



