1895.] Decimetre of Water at its Maximum Density. 151 



The weight {in vacuo) of the water at 13§° R. being = 1 8379*19 

 dolias, and the volume at the same temperature = 49*89931 cubic 

 inches, then the weight of a cubic inch = 368'3255 dolias (Kupffer 

 found 368-341 dolias and accepted 368*361 dolias) = 16-366507 grams 

 at 13^° B,. Therefore, according to the above-mentioned relation 

 between inches and the metre, the probable weight, in vacuo, of a 

 cubic decimetre of water at its maximum density is equal to : — 



16,366507 = 999-8495 grams. 



0-998890 . 0-01638716 8 



This value is not only more reliable than those of Shuckburgh and 

 Lefevre-Gineau, becanse really pure water was here used for the 

 first time and all means employed for obtaining a correct result, but 

 this value is also more probable than all the other results obtained by 

 Kupffer* It corresponds to pure well-boiled water and to the 

 largest of the cylinders employed by him ; so that all the data here 

 show a more evident uniformity than in the other Kupffer determina- 

 tions. 



This result, from observations made forty-five years ago by the 

 Russian metrologist, obtains special importance in consequence 

 of the publication in 1892 in the 'Phil. Trans.' (vol. 183, 

 pp. 331 — 354) of investigations on the weight of a cubic inch of 

 water, made in the year 1888 by the contemporary English metro- 

 logist, Mr. H. J. Chaney, because the latter not only took care to 

 obtain quite pnre water (and by boiling removed the dissolved air), 

 not only employed apparatus and methods carefully tested, but also 

 took a cylinder and a sphere of considerably larger dimensions than 

 the largest used by Kupffer. f By its especially large dimensions (about 



Kupffer, the most probable value of P, the result being identical with the P above 

 found ; but I consider it more regular to proceed in the way I have accepted, 

 because in the same all entering values, especially 8*, have a real meaning, and 

 because I have also used this method in the discussion of Mr. Chaney's observations. 



* Thus, after a full computation of the other determinations of Kupffer, I 

 obtained for the weight in vacuo of a cubic inch of water at 13|° R. from the 

 data : 



For the small cylinder, first water 368 *389 doli. 



„ „ pure „ 368-377 „ 



„ large cylinder, first water 368 '371 „ 



„ „ pure „ 368-326 „ 



f Besides the cylinder C and the sphere S, Mr. Chaney also made measurements 

 with a quartz cylinder Q. After having received from Mr. Chaney some additional 

 data relating to these measurements, I have made a full computation, and found 

 for the cubic decimetre of water at its maximum density, the weight 999*374 

 grams. To this value I cannot attribute any importance, as the volume of Q is very 

 small, only 23"04 eubic inches, and as quartz very slowly assumes the temperature 

 of the surrounding medium. 



