Manchester Memoirs^ Vol. xliii. ( 1 899), No. 8. 3 



of those under test weighed, the glycerine films washed 

 off, the disc dried and again weighed. From the decrease 

 of weight an approximate value for the thickness of each 

 glycerine film was obtained. 



The temperature indicated by the thermometers not 

 being identical with those of the surfaces of the discs 

 tested, a small correction has to be applied to these 

 indications. Let 0i and d^ be the temperatures indicated 

 on the thermometers, the centres of the bulbs of which are 

 tx and h cms. from the surfaces in contact with the 

 glycerine film, and let t be the thickness of each of the 

 two glycerine layers between the surfaces of steel and 

 material. Then if r is the radius of the discs, and H the 

 amount of heat transmitted per second, the heat trans- 

 mitted per sq. cm. = HJTrr^, and if k\ is the thermal 

 conductivity of steel and k that of glycerine, this flow of 

 heat will cause falls of temperature Htijirr-ki and HUl-nr^k^ 

 in the steel discs, and Htfirr^k in each of the glycerine 

 layers. If fJ^ is greater than 0i the temperatures of the 

 surfaces of the disc tested are therefore 



^(1+0 ""'"■ 50+0 



respectively.* 



It was found extremely difficult to secure a constant 

 temperature of the outer steel discs owing to the slow 

 change of temperature of the water from the mains, and 

 this necessitated the treatment of the flow of heat and 

 distribution of temperature in the pile of discs as an 

 " unsteady " one. 



* In the experiments mentioned at the end of this communication, H never 

 exceeds "4, and / never "005, hence taking k for steel = "15 and for glycerine 

 •007, since !rr^ = 5o, and /^=^2 = '5) we see that the correcting terms never 

 exceed -027° for the steel and •006° for the glycerine, while ^^ — 9^ was 3* or 

 4". In the deduction of the results given, these corrections have therefore 

 not been applied. 



