RICHARDS AND MARK. — THERMAL EXPANSION OF HYDROGEN. 123 



at 0°; the volume from the lower stopcock to the upper contact point (a) 

 was found by weighing the mercury which tilled this at 0°, and the volume 

 between the lower stopcock and the lower point (b) was found by weigh- 

 ing the mercury which filled this at a higher temperature. The linear 

 coefficient of expansion of the glass was then determined with great care. 



From these data the volumes sought may be calculated with as great 

 accuracy as is possible under the circumstances. The details of the 

 measurements are given below. 



The bulb, filled with air dried by passage over potassic hydroxide, was 

 balanced against a sealed counterpoise of approximately the same volume. 

 Observations of temperature and atmospheric pressure necessary for the 

 calculation of the weight of air contained in the bulb were also made. 

 The weight of water of the corresponding volume at 0° C. was next 

 obtained as follows: A levelling bulb was connected with the temporary 

 stopcock at the lower end of the bulb by a rubber tube. Through this 

 the bulb was filled with pure water to the upper stopcock. The lower 

 stopcock was then closed and, in order to remove the dissolved air, the 

 water was boiled in vacuo in the bulb. After the amount of water had 

 again been adjusted by opening the lower stopcock so that it filled the 

 bulb to the upper cock, this latter cock was closed and all was packed in 

 ice for one and a half hours. In a special experiment this time had been 

 found amply lono; enough for the required degree of accuracy in temper- 

 ature. The lower stopcock was then closed (the upper one still remain- 

 ing open), the bulb and projecting tube were cleaned and dried, and the 

 levelling bulb was removed. When the same temperature as that within 

 the balance case had been obtained, the bulb and water were weighed 

 and the volume from stopcock to stopcock was calculated. 



These weighings of the bulb filled with water indicated 391.900, 

 391.904, and 391.903 grams ; average, 391.902 grams. The weight of 

 the bulb filled with pure dry air under the same conditions of tempera- 

 ture and pressure (20^ and 752.9 mm.) was 85.192, and the air which it 

 contained must have weighed 0.366 erram. Therefore the water which it 

 contained must have weighed 39 1.902 + 0.367 - 85.192 = 307.077 grams. 

 If the density of water at 0' is taken as 0.999868*, the total volume of 

 the bulbs is found to be 307.118 milliliters ± 0.001. This measure- 

 ment was made when the bulb was full of water, but in other respects 

 under exactly the same conditions of temperature and pressure as in an 

 actual gas measurement. The average pressure of the column of water 

 inside during the measurement of volume was less than 8 millimeters 



* Chappuis, Trav. et Mem. du Bureau des Poids et Mes., 13 (1904). 



