OP ARTS AND SCIENCES. 201 



in the largest bulb of the bulb-tube shown in the accompanying dia- 

 gram, and the tube was about half filled with cold water which had 

 been thoroughly boiled. The whole was then connected with the 

 water pump, and freed from adhering air bubbles by agitation in tin- 

 partial vacuum, — the water being allowed to boil gently under the 

 reduced pressure. Subsequently, the tube was filled with boiled water, 

 and returned to its normal horizontal position. Pure acid was now- 

 run in from a pipette, a short piece of glass rod within the large bulb 

 furnishing a means of agitating the oxide during its solution. The 

 gas which was set free was finally collected and measured in the sealed 

 end of the apparatus. In the concluding experiments, where greater 

 accuracy was desired, the gas was transferred, and remeasured in a tube 

 which had been carefully calibrated by means of mercury. 



The gas was analyzed with the help of a small Hempel's apparatus 

 made for the purpose. The burette with which the first ten analyses 

 were made was about seven millimeters in internal diameter and could 

 be read to a fiftieth of a cubic centimeter, while that used for the later 

 analyses was only four millimeters in diameter and could be read with 

 reasonable certainty to the hundredth. The portion of the gas absorbed 

 by caustic potash was assumed to be carbon dioxide, and the further 

 portion which was absorbed by alkaline pyrogallol was assumed in 

 like manner to be oxygen. The residue was perfectly inert, and was 

 undoubtedly nitrogen. One sample of this residue obtained from 

 zincic oxide, was mixed with oxygen and subjected to the spark of an 

 induction cod, without diminution in volume. The apparatus a d 

 chemicals were tested from time to time by analyses of air. 



In every case, the gas while saturated with aqueous vapor was 

 measured at the ordinary temperature and atmospheric pressure of the 

 room. The variations from the mean values of 22' and 76.0 centi- 

 meters were not sufficient to need attention in crude work of this kind, 

 when working upon such very small quantities of . 



