142 JOURNAL OF ECONOMIC ENTOMOLOGY [Vol. 3 



trol and have used many devices to secure this end. Among other 

 things ice cream freezers and domestic refrigerators have been drafted 

 into service, the cooling agent in all cases being ice with or without 

 salt. In this way they were able to secure temperatures down to the 

 melting point of ice, but anj^thing like constant temperatures below 

 about 42° or 43° Fahrenheit were out of the question. This method 

 is very unsatisfactory at best, as it requires a great deal of personal 

 attention and introduces the factor of human fallibility. As we 

 wished to go below 32°F. and to maintain the temperature constant 

 for long periods of time we soon saw the necessity of artificial refrig- 

 eration. 



The list of active agents used in artificial refrigeration is not an 

 extensive one, hence our choice of the gas we would use was not 

 difficult. Anhydrous ammonia, which is probably the most widely 

 used agent, was soon eliminated for the reason that our plant was to 

 be installed in the basement of the agricultural building, in Avliich, in 

 addition to the laboratories for the regular Station workers, are located 

 class rooms and laboratories for the accommodation of a large number 

 of students. The presence within this building of any considerable 

 quantity of so penetrating and irritating a gas as ammonia could 

 scarcely be attended with an\i:hing less than serious inconvenience. 

 Then, too, in our investigations we are dealing with living animals, 

 hence small leaks, which are almost sure to occur when working with 

 high pressures, might result disastrously. 



Sulfur dioxid was soon disposed of for the reason that the gas is 

 highly poisonous and also that the compressor using it operates with 

 a partial vacuum on the low pressure side. While this machine is 

 quite efficient when air can be kept out of the system every one 

 recognizes the fact that a vacuum is very difficult to maintain and a 

 little air leaking in seriously lowers the efficiency. 



This process of elimination left only carbon dioxid, which was 

 finally chosen. Safety and non-offensiveness of the gas were the prin- 

 ciple factors in this determination. With this gas leaks of consider- 

 able magnitude can exist without serious inconvenience to the operator 

 or other occupants of the building. There is also practically no 

 danger that the insects or other life under investigation will be killed 

 off in the middle of an experiment requiring several months to com- 

 plete, even if the whole charge of gas were to escape into the room. 



The plant (PI. 6) consists of a compressor of two tons refrigerat- 

 ing capacity, a Ti/o-horse-power motor, a brine tank of 200 gallons 

 capacity containing the expansion coils, a brine pump and an insu- 

 lated box containing four chambers, each supplied with cooling coils 



