330 



PETTERSSON, SIBERIAN SEA. 



Suppose a glass vessel A measuring exactly 1 cc. to be 

 iramersed in sea-water of I.0280 sp. gravity (xjoj at 0°. The 

 weight of the vessel and likewise of the displaced volume of 

 water must be exactly 



1.0280 gr. 



If Ave substitute sea-water of I.0279 sp. gr. instead of the 

 original liquid, the weight of the displaced water (supposing 

 the temperature constant and the instrument to occup}^ its 

 former position), which tends to lift the vessel, is also I.0279 

 gr. instead of I.0280 gr. The difference 



0.0001 gr. 

 makes A sink to the next mark »b» on its scale, where it 

 attains a new state of equilibrium. 



Now we could produce ex- 

 actly the same effect by raising 

 the temperature of vessel and 

 liquid simultaneously. The di- 

 latation of 1 cc. of glass for 

 every centigrade is 



g =z 0.00002753 



(see the preceding paper). 

 Consequently the volume of A 

 at + 1° C is 



1.00002753 CC. 



The volume of sea-water, which 

 is displaced by ^ at + 1° C weighs 



-. -i 1 



1.00002753 X 1.0280 — 



if X denotes the volume of 1 cc. 

 of sea-water (of 0° C) at + 1° C. 

 This weight, which tends to lift 

 the instrument, is inferior to the 



Aveight of A (= I.0280 gr.) if a: > 1. 00002753. The difference 



1.0280 — 1. 



1.0280 



X 



1.0280 

 X 



(x — 1.00002753) 



makes the vessel A sink deeper in the fluid. By equalling 

 this quantity to the weight O.oooi gr., which is nécessary in 

 order to produce an appreciable effect, i.e. make the instru- 

 ment sink to the next mark, we may form an idea as to the 

 influence exercised by the temperature, at which the areo- 

 metric experiment is performed, upon the exactness of the 

 specific gravity 



