328 A. M. Mayer — Determination of the coefficient, etc. 



The volume of water at 15° -85 equals 1-060845, and '° T V°ff 5 

 = •0000533, the coefficient of cubical expansion of the brass. 

 The previous determination with the brass cylinder gave 

 •0000536. 



The volume of water at 38°'55 is 1-00697, and'||. 6 //== 

 •0001808, the coefficient of expansion of hard rubber. Mer- 

 cury has at 30° a coefficient of '0001805. Hard robber has 

 the greatest coefficient of any solid. To show on the curve of 

 apparent expansion of water in a vessel of hard rubber it's 

 points of intersection of the axis of x, on the scale of fig. 1, the 

 axis of x would have to be extended yL of its present length. 



These determinations with the hydrometer have the advan- 

 tage' of those made by observations on the volume of water 

 contained in a vessel, because the metal of the hydrometer 

 changes its temperature with that of the water, which is not 

 the case with the vessel filled with water, whose temperature 

 lags behind that of the outside water. This makes the obser- 

 vation on the upper temperature of the water in the vessel 

 tedious and not precise unless the precautions are taken which 

 we have already mentioned. 



The accuracy of this method of determining the mean coef- 

 ficient of expansion of water, in the determined range of tem- 

 perature, depends on having the water in the vessel and the 

 substance of the hydrometer at the same temperature as the 

 water surrounding these bodies. This can be obtained to ^° C. 

 In the case of a vessel filled with water we have already stated 

 the precautions necessary. The hydrometer is floated in water 

 at 0° by surrounding with ice a vessel holding water which has 

 already been cooled to 0° and from which the floating ice has 

 been removed. 



The accuracy, however, of this method depends essentially 

 on having the true curve of the absolute expansion of water 

 between 0° and 30°. It is not possible to say to what approxi- 

 mation we have this true curve ; but that it is quite close may 

 be inferred from the following account of how Rossetti pro- 

 jected a curve which is the best we have at present. 



Prof. F. Rossetti describes his experiments and the manner 

 of obtaining the curve in the Atti dell' Instituto Veneto, vol. 

 xiii, 3 series, 1868. Kosetti's new determination of the volume 

 of water at various temperatures between —6° and 100° were 

 made with every precaution to ensure accuracy. He then 

 combines the results of his experiments in a curve projected 

 according to the method of Schiapparelli.* He then combined 

 his results, thus reduced, with those of Depretz (1839) ; Kopp 



* Schiapparelli, Sul modo di ricavare la vera expressione delle leggi della 

 natura dalle curve empiriche ; Nuovo Cimento, Tomo xxv, 1867; e Tomo xxvi, 

 1867. 



