1 86 HOMER E. CHENOWETH. 



mammals and that if these processes were sufficiently interfered 

 with death of the animal resulted. 



i . Method of Establishing Evaporation Gradients. 



The apparatus for this experiment was designed by Dr. V. E. 

 Shelford and E. O. Deere, and for descriptions and drawings and 

 method of establishing evaporation gradients see BIOL. BULL., 

 No. 25, '13. Descriptions will be made only where changes 

 were made in the apparatus. 



The experimental cages for the gradient experiment were de- 

 signed by Dr. V. E. Shelford. Since a much greater rate of air 

 flow was used in these experiments the cages were made about 

 three times as large as the designer used in his experiments. An 

 extra division of wire, mesh was inserted between the fish-tailed 

 burner-shaped introducers and cage apartment. This served to 

 make a more even distribution of the air over the cage. One 

 cage was used for both the experiment and the control. When 

 a control experiment was being made the air was shut off; no 

 moving untreated air was used. 



In the beginning, considerable difficulty was encountered in 

 obtaining a constant flow of air through the apparatus and into 

 the cages, due to the fact that the air supply was taken from the 

 university air mains. The pressure in the mains ran about 85 

 pounds per square inch and as the experiments were conducted 

 at a much lower pressure, from one to five pounds, a reducing 

 valve was placed in the branch line to the apparatus, to reduce 

 the pressure and keep it constant for any particular pressure that 

 was required for the experiment on hand. On account of the 

 well-known acute sensibility of man and mammals to slight dif- 

 ferences in air movement it was thought best to pay considerable 

 attention to air movement. 



At the outset it was decided to obtain the flow of air in terms of 

 a number of liters per minute as this can easily be determined. 

 This was done by filling a 1 3-liter bottle with water and finding 

 the time required for the air to displace the water. By pro- 

 portion the flow in liters per minute was readily obtained but as 

 the operation had to be repeated for each change in the air 

 pressure the work for finding the flow for each different pressure 



