102 Journal of the Mitchell Society. [Nov. 



We wish further to express our thanks to Prof. Young- for 

 some unpublished details of the measurements and for com- 

 ment upon the work, and we have taken the liberty of making 

 in this paper some extracts from his letters. 



SUMMARY. 



T T£ 



1. The equation, , ' = constant was deduced for 



y d — y D 



the purpose of testing- the assumption that the attraction 

 between the molecules of any particular substance obeyed the 

 law of gravitation, i. e., varied directly as the product of the 

 masses, inversely as the square of their distance apart, and 

 does not vary with the temperature. 



2. Twenty-one substances were examined. Of these, 

 eleven: ethyl oxide, di-isopropyl, isopentane, normal pentane, 

 normal hexane, normal heptane, normal octane, benzene, hex- 

 amethylene, fluo-benzene, and carbon tetrachloride were, 

 from 0° C to their critical temperature, making allowance for 

 errors of observation and the multiplication of such errors in 

 the calculation, in reasonably perfect agreement with the 

 deduced equation. The divergences in di-isobutyl, chlor-ben- 

 zene, brom-benzene, iodo-benzene, and water, all occur at such 

 points and are of such magnitude that they may easily be 

 due to the errors of observation (or the multiplication of such 

 errors in the calculation). Stannic chloride failed to agree 

 with the equation. 



3. To the associated substances water, methyl, ethyl, and 

 propyl alcohols, and acetic acid, the equation was not sup- 

 posed applicable; but within wide limits the agreement of 

 these substances, acetic acid excepted, is such as to suggest 

 the conclusion that the molecular association with which we 

 are there dealing is caused by the molecular attraction whose 

 law we are considering. 



