442 STATE BOARD OP AGRICULTURE. 



of treated (with insecticide) and untreated extract samples. (See Fig. 

 1, Plate I.) So many factors were involved that it was necessary to 

 use extreme precaution in order to be at all sure that a difference in 

 volume of the oxygen absorbed by hydroquinone-extract-mixture in air 

 and hydroquinone-extract in gasoline-vapor-air was due alone to the 

 influence of gasoline vapor. It was necessary always to thoroughly mix 

 a fairly concentrated extract and then to remove two equally measured 

 samples for the experiment at as nearly as possible the same time. 

 Furthermore, exactly equal measures of a standard hydroquinone solu- 

 tion must be added to the two samples at the same time, and then the 

 tests with the two equally prepared mixtures must necessarily be run 

 (the one in air and the other in gasoline-air) under like conditions of 

 temperature for the same period of time. It would not do to run the 

 tests for equal periods, merely. They must cover the same period as 

 nearly as that might be made possible. The reason for these precautions, 

 if not apparent here, will be noticed later in the discussion. Another 

 difficulty presented itself. Oxygen was absorbed rather slowly in the 

 tests, and so it seemed advisable to find some method of measuring 

 directly the entire amount of carbon dioxide that might be given off, 

 as well as the entire amount of oxygen taken up, rather than to depend 

 upon percentage estimations from samples. No method was available for 

 removing and accurately estimating carbon disulphide vapor from air 

 used in a test, and so it seemed necessary to be able to measure the oxy- 

 gen taken up and the carbon dioxide given off in the presence of the in- 

 secticide employed during the test. Finally, an apparatus was devised 

 which complied very well with the requirements just outlined. The es- 

 sential features of the apparatus are represented in Figure 1, and a brief 

 description follows here. 



The apparatus (Fig. 1), consisted of two systems which were almost 

 exact duplicates, each made up of a gas-burette "a" connected with a 

 mercury-mug "b"; a gas-container "c"; a potash absorption flask "d"; 

 a mercury manometer "e"; a burette "f" (graduated to O. 1 c. c.) for the 

 hydroquinone solution; and a circulating pump "g". The gas container 

 was of about 300 c. c. capacity, with two two-w^ay cocks above and a 

 wide mouth (2i^ inches) beneath. The wide rubber-stopper for this con- 

 tainer was connected with a mercury-mug "h", and a hooked U-tube also 

 passed from the stopper to connect with the hydroquinone burette "f". 

 Mercury could thus be used not only to seal the mouth of the gas- 

 container, but to float the stender dish up under the hooked end of the 

 U-tube. The gas-burette of each system was connected by tubes (with 

 airtight sealed joints) through the corresponding gas-container, absorp- 

 tion-flask, mercury manometer and circulating pump of its system, as 

 represented in the figure. Between the "pump-gas-container connecting 

 tube" and the ''absorption-flask-pump connecting tube" was a cross tube 

 with a stopcock ''St." Fig. 1. This cock was kept closed when the pumps 

 were running. It was opened as soon as the pumps were stopped, and 

 always, before the manometers were adjusted, the circulating wheel of 

 each pump was given a quarter turn backward. Thus, any air that might 

 be compressed in a pocket was released into the upper air-chamber, leav- 

 ing the pockets filled with mercury. The circulating pump of each 

 system was actuated by belts from the same double pulley of a reducing 



