THE CHEMICAL WARFARE SERVICE 167 



as boron trifluoride, were prepared, but no fluorine compound 

 found any use in chemical warfare. 



ANALYTICAL RESEARCH 



As a preliminary to the study of the properties of toxic gases 

 and of the means of defending against them, it is necessary to 

 be able to detect and determine these gases. An analytical 

 and testing section was, therefore, one of the first to be estab- 

 lished and it was always busy in spite of the fact that all the 

 other sections cooperated in developing methods of analysis 

 and testing. The details of analytical methods are not specially 

 thrilling to anybody except a technically trained man, so it will, 

 perhaps, be sufficient to say that satisfactory methods were 

 worked out for analyzing every toxic gas with which the 

 Chemical Warfare Service had to deal. Three typical cases 

 may be mentioned, however: the testing of canisters, the field 

 tests for mustard gas, and the special paint for shell. 



Canisters are tested on men and on machines. Multiple 

 machines have been developed which will test eight canisters 

 simultaneously at continuous flow of the gas-air mixture or at 

 intermittent flow. The continuous flow machines are the 

 easiest to construct and were made first. Since the man 

 breathes through the canister intermittently, the results with 

 the intermittent flow machines resemble more closely those 

 encountered when masks are actually worn in gas. The inter- 

 mittent flow machines are capable of wide variation both as to 

 volume of air passing through and as to number of oscillations 

 per minute. They can, therefore, be adjusted to simulate any 

 type or rate of breathing. Comparison tests on men have 

 shown that the intermittent machines give results in excellent 

 agreement with man tests, are easier to run, and are much more 

 accurate, because they do away with the personal idiosyncrasies 

 of the men. This does not mean that man tests should be 

 abolished. They must always be kept to provide for unex- 

 pected contingencies but they can be reduced to a minimum 

 with a great saving of time and friction. 



