PHYSICAL PROPERTIES OF WATER, ETC. 19 



23/7/87. 



25/7/87. 



The range of temperature is quite sufficient to allow a change of compressibility 

 of the water column to be noted ; but the experiments unfortunately do not enable 

 us to assert anything as to a change in that of mercury ; though, were it not for the 

 last set of experiments, there would appear to be a decided increase of compressibility 

 of mercury with rise of temperature. The experiments are only fairly consistent with 

 one another ; but this was noted at the time as the fault of the index, which, of course, 

 tells more as the quantity measured is less. It may be as well to show how to 

 deduce the compressibility of mercury from them at once, assuming the requisite data 

 for water and for glass from subsequent parts of the Report. 



Take, for instance, the first result of 25/6/86. 834 of gauge is about 305 atmospheres. Also shortening 

 of 286 mm. of water column (in glass) at 12°-3 C hy 305 atm. = 3-7 mm. nearly : — so that the compressed 

 mercury apparently loses about the content of 14-8 mm. of narrow tube = bulk of - 354 grin. Hg. 



0-354 



Apparent compressibility = = 0'00000109 



305 x 1062-3 



The average of all the normal experiments gives "000001 very nearly. 



Add compressibility of glass = 0-0000026 

 Compressibility of mercury = 0-0000036 



It is well to remember that though Grassi, working with Regnault's apparatus, 

 gave as the compressibility of mercury 



0-00000295 

 which Amaury and Descamps afterwards reduced to 



0-00000187, 

 the master ' himself had previously assigned the value 



0-00000352. 



Had Grassi's result been correct, I should have got only about half the displacements 

 observed; had that of Amaury and Descamps been correct, the apparent compres- 

 sibility would have had the opposite sign to that I obtained, so that the index would 

 not have been displaced. In such a case the construction of the instrument might 

 have been much simplified, for the index would have been placed in contact with the 

 mercury at B, and the bent part of the tube would have been unnecessary. 



1 Relation des Experiences, &c, Mem. Acad. Sci. Paris, torn. xxi. p. 461, 1847. 



