1895.] Decimetre of Water at its Maximum Density. 145 



those of M. Kupffer and Mr. Chaney have been carried out under all 

 possible conditions necessary for obtaining reliable results, those 

 conditions being as follows : — Direct comparisons of the weights 

 and measures used with the primary standards of weight and mea- 

 sure, accurate corrections of tbe thermometers, and particularly care 

 in obtaining entirely pure water, which before the experiment was 

 boiled so as to remove the dissolved air.* I therefore consider 

 these determinations, introducing into them all corrections in a 

 manner corresponding to our present experience, upon which I con- 

 sider it here necessary to make the following statements. I have 

 assumed : — 



1. One metre = 39*3700 + 0*0001 English inches, as it follows from 

 the combined reliable comparisons of Kater, Clark, and Tittmann. 

 It may be remembered that the Russian inch, according to the law, 

 is equal to the English inch, and that 1 cubic inch appears to be 

 equal to 16*38716 c.c. or 0*01638716 cubic decimetre. 



2. One pound avoirdupois — 453*59243 grams, as we see by the 

 recent comparisons' of Mr. Chaney ; therefore 1 grain = 0*0647989 

 gram. 



3. One Russian pound = 409*5120 + 0*0005 grams, as appears 

 from the fact that Kupffer gives 409*51156 ±0*00072, and the pre- 

 liminary weighings made at the Central Chamber of Weights and 

 Measures in 1894 give 409*51236 + 0-00022 grams. Therefore 

 1 doli (a Russian pound contains 96 zolotniks a 96 dolias) = 0*0444349 

 gram. 



4. For pure water (freed from dissolved air) I assumed that the 

 change of the specific gravity (at 4° C. — 1) as depending upon the 

 temperature t° C. is represented by the following equation : — 



s, = l- (*-*)* m 



1*9 (941 + £) (703*5 — t)' K ' 



as derived by mef from the whole of the best determinations be- 

 tween — 10° and +200°, made up to the year 1891, verified mer- 

 curial thermometers being used, as employed by Mr. Kupffer and Mr. 

 Chaney. Here I consider it necessary to mention that at ordinary 

 temperatures mercurial thermometers always give higher readings 

 (t) than those of the hydrogen thermometer (denoting its tempera- 

 ture by T) ; the difference t— T is very various, being at 20° from 



* Through the dissolved air the density of water decreases in its millionth parts, 

 as I have shown in my paper, " Investigations on the Specific Gravity of Aqueous 

 Solutions," 1887, p. 383. Maly and Marek ( £ Wiedemann's Ann.,' 1891, vol. 44, 

 p. 172) actually found this, having shown that the relative density of water con- 

 taining air at 0° C. = 0-9999975, at 10° C. = 0*9999968, at 20° C. = 0*9999996 

 and at 30° C. = 1. 



t ' Phil. Mag.,' 1892, p. 99. 



