Prof. J. W. Mallet on the Demiti/ of Solid Merairy. 147 



The method adopted in the experiments lately made in this 

 laboratorj was the followiiii^ : — 



(1) A specific-gravity fiask was prepared from a large cylindrical 

 pipette, by closing iii, and smoothly rounding, in the flame of the 

 lamp one end of the cylinder, while the tube remaining attached 

 to the other end was cut short and united by fusion to a second 

 pipette of like shape but smaller size, the upper and open end of 

 the shortened tube of which was fitted with a small carefully 

 ground glass stopper. The neck between the larger and smaller 

 cylinders was drawn down to a small bore (about 2 millims.), and 

 at this narrowed part a fine line marked round it with a diamond. 

 The shape of the vessel is shown in the annexed sketch, on a linear 

 scale of one half the real size. The principal cylinder held about 

 53 cubic centimetres, and the small reservoir above 25 cub. 

 centims. The whole vessel weighed about 46 grammes. 



It enabled the experiments to be carried out with more than 

 half a kilogramme of frozen mercury. 



(2) This vessel having been accurately w^eighed when empty and 

 dry, its capacity np to the mark was ascertained by filling it to 

 this point with pure water at exactly 4° C, keeping it immersed 

 for some time in a large mass of water at this temperature before 

 making the final adjustment to the mark, wiping the outside dry, 

 allowing the whole to acquire the temperature of the balance-case, 

 and carefully weighing. The result of this direct caUbration, de- 

 ducting the weight of the vessel, was 59*7323 grammes or cubic 

 centimetres at 4°. 



(3) It was checked by emptying and drying the vessel, filling it 

 to the mark A\ith pure mercury at 0° C, the temperature being 

 secured by keeping the whole surrounded by melting ice long 

 enough to obtain perfect steadiness of position of the mercury, 

 and weighing after the temperature of the balance-case had been 

 regained. The mercury weighed 811°9997 grammes. 



(4) The vessel was now surrounded by steam, and the mercury 

 again brought to the mark, the temperature actually attained being 

 99°*5 C. (corrected for pressure). Allowed to cool down to the 

 temperature of the balance-case, and again weighed, the mercury 

 w^as found =799*7032 grammes. From the last two weighings 

 the coefficient of cubical expansion for 1° C. of the glass used was, 

 by the usual formula (taking absolute expansion of mercury from 0"^ 

 to 100 = -018153, as determined by Eegnault), found = -000027346. 



(5) If now the density of mercury at 0° as referred to water at 

 4° be taken at 13*596 (Eegnault), the weighing obtained in (3) 

 gives the capacity of the vessel up to the mark at 0° =59*7234 



L2 



