124 PROCEEDINGS OF THE AMERICAN ACADEMY. 



of mercury, a pressure which would produce an increase of volume in 

 the globe of less than 0.001 milliliter,* hence this distention during 

 measurement may be neglected. If mercury had been used instead of 

 water it would have been a very serious possible cause of error. 



In order to obtain actually the volume occupied by the gas at 0° the 

 volume below the upper platinum point must be determined and sub- 

 tracted from 307.118. Therefore mercury was next run into the bulb 

 from below to the upper platinum point, the air bubbles were removed 

 by exhaustion, and after equilibrium of temperature had been reached in 

 the ice bath, the exact adjustment of the mercury to the level of the 

 platinum point was accomplished by a fine screw and indicated by elec- 

 trical contact. The stopcock was then closed, the rubber tube removed 

 as before, and the mercury inclosed in the bulb run out and weighed. 

 Three successive experiments gave 472.21, 472.14, and 472.19 grams; 

 in average, 472.18 grams. If mercury has a density of 13.5956 at 0°, 

 this volume is then 34.730 milliliters ± 0.001. The mercury was so 

 shallow and this bulb so stout, that it could not have been appreciably 

 distended during measurement. The actual volume occupied by the 

 gas at 0° must therefore have been 307.118 — 34.730 = 272.388 milli- 

 liters, ± 0.002. 



To ascertain the total volume of the supplementary bulb and the large 

 bulb taken together, —namely, the volume from the upper stopcock to 

 the lower platinum point, — the small volume between the lower stop- 

 cock and the loiver platinum point was found and subtracted from the 

 total volume 307.118. Instead of measuring this small volume at 0° C, 

 it was done at a higher temperature, in order that the shape of the menis- 

 cus at this point might be more nearly that which it assumed during the 

 performance of the main experiment. f After dividing by the density of 

 mercury and making correction for the contraction of the glass, the vol- 

 ume at 0° was found to be in five determinations, — 1.944, 1.941, 1.946, 

 1.944, 1.943; average, 1.944 ± 0.0008 milliliters. 



Thus the volume of the large bulb and the supplementary bulb to- 



* Lord Rayleigh found in a 2 liter globe a distention of 0.6 milliliters, and 

 Cooke and Richards found in a 5 liter globe a distention of 1.7 milliliters for 

 each atmosphere of internal pressure. Our bulb was at least as stout in pro- 

 portion as these, hence the above conclusion. 



t This determination should have been made at 32.4°. It was, as a matter of 

 fact, through some oversight, made at 18.3°, where the average amount of mercury 

 found was 20.350 grams. The breaking of the tube made a redetermination at 

 32.4° impossible, but the above value can hardly be essentially wrong. 



