EXPERIMENT STATION BULLETINS. 555 



Should the insecticide, then, cause an increase or a decrease in the car- 

 bon dioxide excretion, and should the oxygen absorption actually change 

 in like proportion, the respiratory quotient would be found to remain 

 constant — i. e., just as it had been in pure air. On the other hand, if 

 either rate increased or decreased in value faster than the other, it 

 would follow that the necessary change appearing in the respiratory 

 quotient would show that condition. 



The small size of insects and their peculiar mode of respiration 

 through many stigmata made it impossible to collect expired air alone 

 (unmixed with any other) for analysis. Besides, a method was re- 

 quired which would permit the oxygen used and the carbon dioxide 

 given off by the insect to be accurately estimated even when a volatile 

 insecticide was present. (For, it must be remembered that all these 

 contact insecticides given in the list, are more or less volatile). The 

 carbon dioxide might be estimated from absorptions made with a stand- 

 ard solution of barium hydrate (an extremely accurate means already 

 described) but the oxygen would have to be estimated volumetically. 

 Absorption of oxygen by phosphorus or by potassium pyrogallate seem- 

 ed to be the most promising, in this connection, of the different methods 

 recommended by gas analysts. 



The following method for obtaining the respiratory quotient was 

 tried first: I i i 



An insect respiration-chamber was connected up in a circuit between a series of 

 standard barium hydrate flasks and the air chambers of two vessels containing mer- 

 cury with an air-space above the mercury in each. Connections were made in 

 such a manner that by raising or lowering one of these mercury vessels every 

 forty to sixty minutes, the mercury would run from one vessel to the other and 

 pump the air very slowly through the respiration chamber, thence through the 

 barium hydrate flasks, back into the air space of the bottle which the mercury 

 was leaving. This circulation could be repeated ov^ and over. In that way, 

 the carbon dioxide was caught and could be determined by titration at the end 

 of the experiment. Before the standard barium hydrate was introduced, the 

 mercury pump was started and the air in the apparatus made to circulate through 

 a carbon dioxide filter until entirely free from that gas. Also, arrangements 

 were made so that samples of air could be taken from the apparatus for the 

 estimation of the oxygen percentage at the beginning and at the end of the 

 experiment. Knowing the volume of air at the beginning, it was therefore pos- 

 sible to determine how much oxygen had been taken up during the experiment. 

 Very good constant results in carbon dioxide estimation were obtained with this 

 apparatus, but the oxygen estimations varied. Rubber tube connections were 

 necessary in part of the apparatus and they seemed to permit some diffusion — 

 especially in case the vapors of an insecticide like gasoline were present. 



The above method had some good points and might have been im- 

 proved to reasonable efficiency for pure air experiments. Its practice 

 was discontinued, however, because it did not seem well adapted to 

 the use of insecticide vapors or to the exact estimation of the oxygen 

 used by the insect. 



Further preliminary experiments soon showed that in studying the 

 effects of such volatile insecticides as gasoline, kerosene and the like, 

 it was desirable (if not necessary) to be able to estimate the vapor 

 of the insecticide present. At least, it was necessary to be able to 

 remove the vapor-fumes entirely before estimating the oxygen in order 

 to get accurate results. As is well known, when phosphoinis is being 

 used, the presence of even a trace of the fumes of carbon disuilphide. 



