1895.] Decimetre of Water at its Maximum Density. 149 



air = 0*4501 A, the volume of tbe cylinder = 49*898 cubic inches, 

 and the volume of the weights = 8*156 cubic inches (sp. gr. = 8*51, 

 therefore volume of 3134 A weights = 1 cubic inch), we have for the 

 weight of the cylinder in vacuo, P 2 = 25579*5725 A* As the arbi- 

 trary unit of weight J., as we shall see later on, weighs very nearly the 

 same as a dolia of a pound, that is to say about 44*4 millirgams, and 

 as the error of Kupffer's weighings is not less than 0*7 milligram, we 

 can assume the error in P 2 to be +0*012 A, and in the mean result 

 an even smaller error will be obtained. 



Having finished all his weighings, KupfFer (loc. cit., p. 342) deter- 

 mined the relation of the weight of the arbitrary units A, by means of 

 which the weighings were made, to the weight of a standard Russian 

 pound, and after application of all necessary corrections, he came to 

 the conclusion that A was a little heavier than a dolia of true weight, 

 namely, A = 1*00000933 dolia. Therefore P 2 = 25579*81 dolias. 



Kupffer made twenty-seven weighings of the large cylinder in 

 water, which, although distilled, subsequently appeared to be not 

 quite pure. All these weighings I have re-compnted,f but I give 

 here only the final result, showing that at 13 J° R,., the weight of the 

 displaced water (reduced to weighings in vacuo) is equal to 18382*00 

 dolias, which is evidently more than the weight (18379*19 dolias) 

 of the same volume of quite pure water prepared by Hess and 

 Fritzsche. After having obtained this water, Kupffer made in it 

 twenty new full weighings of the same cylinder, arranging them in 

 five series of four weighings each, the mean result of which we give 

 here, not in the order of the weighings as made, but in the order of 

 the observed temperatures of the water, so that the change of the 

 relation between the weight and the temperature (£) and the 

 symmetry of the obtained results may be apparent. 



Corrected temperature 



on tbe mercurial Mean weight of the 

 thermometer scale cylinder in water, Kupffer, 



(Eeaumur) . in arbitrary units A. vol. ii. 



(1) 12*73° 7200*2738 A page 322 



(2) 12*79 7200*2628,, 301 



(3) 13*65 7202*4443,, „ 320 



(4) 13*84 7202*8940,, „ 316 



(5) 13*99 7203*1370,, „ 310 



These weighings were made during the intervals between the 

 weighings of the cylinder in air, and therefore for the correction for 

 the weight of air displaced by the Weights, we can take the above- 



* Kupffer finds the weight in vacuo to be 25580 *7751 or 25580 '5302, taking 

 the weight of a litre of air according to Biot and Arago, that is, higher than it 

 shoiald be. For this reason Kupffer's results are larger. 



f Vide note X P- 143. 



