§3 



OF ARTS AND SCIENCES. 



2,^2^ 



§ 3. The necessary data appear to have been deter- 

 mined only for the five liquids in the table: — * 



y z= 42,000,000. T=To = 273° 



The case of water must be thrown out for reasons 

 already mentioned, which will be treated in full, later on. 



We find for alcohol, a = 2.38; for turpentine, ^^^4.23; 

 for bisulphide of carbon, a = 4-3 1, and for ether a =^ 5-04. 

 With the exception of alcohol, the numbers agree closely 

 enough with those obtained from the rough calculation of 

 § I, although the latter were derived from the ratio of the 

 cohesive forces in two widely different states of aggrega- 

 tion, while the values of a were determined by the work 

 done against a cohesion continually decreasing during the 

 change of state. 



We are not able from these data to determine exactly 

 what the law of cohesion is; but we are able to decide 

 what it is not. Since ^ is a value intermediate between 

 two values of .r, namely <^, and a^, some values of a; must be 

 equal to or greater than a, and some values equal or less. 



• With the exception of turpentine, the (internal) latent heats were from Zeuner's Tables 

 (appendix) ; the densities from Wohler's Organic Chemistry (indexed) ; the elasticity was the 

 mean from Everett, pages 52 and 53 of his Units and Physical Constants ; and the expansion is 

 the mean of Kopp and Pierre. For turpentine, see Everett, page 88; Pickering's Physical 

 Manipulation, Volume II., Appendix, Table 12; Ganot's Physics, § 334 and § 326. 



