RICHARDS AND MARK. — THERMAL EXPANSION OF HYDROGEN. 119 



provided above with a suitably pointed platinum wire to make the current 

 when the pressure of the warming gas reached a certain point.* An 

 electric fan suspended near the ceiling directed a current of air downward, 

 and thus kept all parts of the room at approximately the same tempera- 

 ture. This arrangement was found to give a temperature constant for 

 hours to the tenth of a degree centigrade and seldom to allow a change 

 of more than a few tenths. The great advantage of a constant-tempera- 

 ture room for this work can scarcely be over-emphasized. Not only are 

 the troublesome corrections for temperature in the pressure readings and 

 the errors due to the unequal expansion of the supports of the apparatus 

 eliminated, but all other errors due to thermal effects, both those foreseen 

 and those possibly overlooked, are also thus diminished, if not made quite 

 improbable. 



Here was set up, on very firm supports, the apparatus indicated in the 

 diagram, taken from the previous paper. In this apparatus a definite 

 mass of gas was made to expand between two temperatures under almost 

 constant pressure, the slight deviations from perfect constancy being 

 determined with great exactness by a Lord Rayleigh barometer. 



The receptacle for containing the gas consisted of two parts, — a large 

 bulb A for holding the total volume of the gas at 0°, and a supplemen- 

 tary bulb (between a and b) for accommodating the increase of volume at 

 32.383°. These volumes were made as nearly as possible 0.273 liter 

 and 0.0324 liter, proportional to the absolute temperatures used; and 

 each was fixed by a mercury meniscus set by making exact electrical con- 

 tact with a very finely pointed stout platinum wire firmly fused into the 

 glass. 



The bulbs were surrounded by a large vessel, in which were placed in 

 succession the two substances, ice and sodic sulphate, used to fix exactly 

 the lower and upper temperatures. 



The framework which supported this vessel or bath was made of 

 stout, well-seasoned pine scantling, mortised and diagonally braced so 

 as to offer great rigidity in all directions. The purpose of this bracing 

 was to prevent any change in the relative heights of the platinum points 

 in the bulb and the side tube. This support was held iu place on the 

 floor by four heavy bolts set in cement and imbedded in the ground 

 under the concrete. A bar of two-centimeter channel-iron, similarly 

 bolted down, was forged into the form of a large inverted U extended up 



* Regaud and Fouillard, Z. wiss. Mikroscop., 20, 138 (1903) ; also Journ. Anat. 

 etPhys., 36, 574 (1900). 



