510 Mr. A. W. Warrington on 



point. So fine was this point that a minute globule of mer- 

 cury could be made to escape from the bottle only by sharply 

 inverting the latter so that its contents rushed with some 

 considerable force to the narrow orifice. By this means the 

 weight of mercury added could be controlled within 5 milli- 

 grammes. 



Hydrometer A was counterpoised on the balance by the 

 similar one B, not in use. A was removed and the mercury 

 transferred to it. A was then replaced on the balance-case, 

 and the amount of mercury added from the weighing-bottle 

 determined with every care. 



The weight of the quartz is 32*45768 ±0-000015 grammes 

 (see Table VIII.). 



Tables IV. and V. give the data for the calculation of 

 P — P l5 and Table VI* the data from which the specific gravity 

 of the piece of quartz actually used was found to be 



2-650457 ±0-000013. 



Assuming P to be known the probable error of a single 

 determination of the specific gravity of quartz is ±0*000027. 

 I should not like to conclude this account without 

 acknowledging the kindness and consideration shown by the 

 University College of Wales, Aberystwyth, who through 

 their Registrar, Mr. Mortimer Green, provided me with a 

 suitable room and furnished it according to my requirements. 



ArrENDix. 



A. The following constants were used in the calculations : — 



Specific gravity of platinum at 0° . . 21*450 

 Specific gravity of mercury at 0° . . 13*59563 



In the scale of a hydrogen thermometer the expansion of 

 mercury for temperatures not exceeding 40° has been calcu- 

 lated from the formula 



Y t = V (l + 0-0001819 1), 



that of quartz from the formula 



V«= V [l + (33669 + 31-27^ 10- 9 /J, 

 and that of platinum from the formula 



V*=V (1 + 0-00002660- 



