RATIO (JF SIM'XUFIC IllOATH. 7 



range of ju'cuiacy, pressure excesses of 3.:} cm., G.() cm., etc., must be read off 

 within .01 cm., .02 cm., etc., respectively, along the mercury column ol)served. 



To reduce the registry of the suli)huric-iicid maiiometer to that of the mercury 

 manometer we brought to bear on both the same pre.ssure which iiad been ol)served 

 during the course of a (complete experiment ; and lor some time after its termination 

 this pressure was maintained at the given value. Not until all ])ressure variations had 

 fully subsided and permanent constancy of level was assured, did we proceed to the 

 measurement of the equivalent mei'cury column. Tlie telescope of the cathetometer 

 was then alternately directed to the two meniscuses of the mercury nnmometer. 

 In view of the large diameters of the tubes used, the iui(hne parts of the terminal 

 mercury surfaces were a[)])reciabiy plane. Sharj) lines of dcmarcalioii being desir- 

 able in the field of the catlietometer (preferably separating a dark shadow from a 

 bright background ), it was found expedient to illuminate the tulies of the manometer 

 from behind. An incandescent lamp shining through a film of translucent tissue 

 paper was therefoi'e [)laced at such a level, that no light was reflected from either 

 mercury meniscus into the cathetometer. Lamp and screen were movable up oj' 

 down, so that the best illumination corresponding to an}' height of meniscus was 

 attained. It is advisable to place the upper edge of the screen but very little 

 above the mercury meniscus. At the same time the unavoi(h-ibU! specks of dust 

 which float on the mercury surface are of much value in sighting. Our measure- 

 ment of differences of level was correct to .01 cm. For an initial pressure of 3.3 cm. 

 therefore, the value of >t ol)tained ^vould not from this cause be discrejiant by 

 moi'e than 1 %. A mercury thermometer suspended in the ojien tube of the 

 mercury manometer showed its temperature. The excellent cathetometer used in 

 these measurements was placed at our disposal by Mr. (i. Hansemaun, a courtesy for 

 which we take pleasure in expressing our indebtedness. 



d. MEASUREMENT OF THE FINAL TEMPEKATUKE, 71,. 



Two distinct measurements are necessary to complete the determination of 11. 

 The decrement of resistance experienced by the Ijolometer strip while the pressure 

 of the expanding gas falls from ^J>, to^A, is fiikst to be found. Thereafter this elec- 

 tric datum is to be expressed in its equivalent of teinperature. 



1. Measurement of the initial 'preanKre /h, corresponding to the resistance incre- 

 ment Wj, to.,. 



Suppose the initial conditions to be fully established, and let tlie fiolometer 

 strip be at the temperature '1\ oi the gas, conqiressed as far as i)ressurey>i. Let 



