10 THE VOYAGE OF H.M.S. CHALLENGER. 



would compress the water no more than 44 parts in a million, and the spirit of wine no 

 less than 71 of the same parts. In making these experiments, the glass ball containing 

 the fluid to be compressed must be kept under water, that the heat of it may not be 

 altered during the operation. 



" The compression by the weight of the atmosphere, and the specific gravity of each 

 of the following fluids, (which are all I have yet tried,) were found when the barometer 

 was at 29^ inches, and the thermometer at 50 degrees. 



Millionth parts. Specific gravity. 



Compression of Spirit of Wine, 66 846 



Oil of Olives, 48 918 



Rain-Water, 40 1000 



Sea-Water, 40 1028 



Mercury, 3 13595 



These fluids are not only compressible, but also elastic : for if the weight 1 >y which they 

 are naturally compressed be diminished, they expand ; and if that by which they are 

 compressed in the condenser be removed, they take up the same room as at first. That 

 this does not arise from the elasticity of any air the fluids contain, is evident ; because 

 their expansion, by removing the weight of the atmosphere, is not greater than their 

 compression by an equal additional weight : whereas air will expand twice as much by 

 removing half the weight of the atmosphere, as it will be compressed by adding the 

 whole weight of the atmosphere. 



" It may also be worth observing, that the compression of these fluids, by the 

 same weight are not in the inverse ratio of their densities or specific gravities, as 

 might be supposed. The compression of spirit of wine, for instance, being compared 

 with that of rain-water, is greater than in this proportion, and the compression of 

 sea-water is less." 



With the exception of the mistake as to the non-effect of compressibility of glass, 

 and its consequences (a mistake into which Orsted and many others have fallen long 

 since Canton's day), the above is almost exact. The argument from the fact that thick 

 and thin vessels give the same result is unfounded ; but the discovery of the fact itself 

 shows how accurate the experiments must have been. The formula (A) below (Section 

 VII), if extended to p = 0, gives for the value of the apparent compressibility of 

 water at 10° C. (50° F.), which is what Canton really measured, the number 



0-0000461, 

 exactly the same as that given by him 126 years ago ! 



The next really great step in this inquiry was taken by Perkins in 1826. 

 He showed beyond the possibility of doubt that in water at 10° C. the compressibility 



