Ma?ichester Meincirs, Vol. xlii. (1898), No. 6. 53 



III. Air dissolved in the water used in the trials. 



This had an effect in raising the apparent specific heat 

 of the water, on account of the bubbles of air expelled 

 being saturated with vapour at 2I2^F. I made experi- 

 ments to determine the volume of the bubbles, and from the 

 data so obtained a correction was calculated at —000021. 



IV. Reduction of tJie weighings to vacuo. 



Bearing in mind that the density of the air affected 

 equally the cast iron load on the brake, and also the rider 

 weights of the weighing machine, this correction becomes 



— 0-00I20. 



V. Varyifig specific heat of the zvater. 



If Regnault's formula for the specific heat of water 

 were accepted, the value of K obtained would require a 

 correction of —000006, introduced by the fact that the 

 initial temperature of the water was never exactly 32^F., 

 as assumed in the specific heat used in the calculations. 



So little is known of the true value of the specific heat 

 at any temperature that this correction has not been 

 applied. 



VI. Correction due to pressures on thermometer bulbs. 

 This has been stated to be —0-00037. 



VII. Correction due to loss of head in stream of zuater 



between supply and discharge pipes. 

 A certain amount of heat would be generated by this 

 loss of potential energy. 



The correction required is -f O'oooo/. 



VIII. Engagement of revolution counter. 

 This has been already given = +0'0000i. 



Adding all these corrections together, we get as a 



final correction factor 



1—0-00125. 



