1863.] 



449 



had been employed, but some modifications were introduced in the * 

 method of observation, which are described; and the amount of 

 probable error from different sources was determined. The liquids 

 experimented on were either prepared or purified in Dr. Gladstone's 

 laboratory, or were specimens reputed to be pure, and lent for the 

 purpose of this inquiry by Prof. Hofmann, Prof. Williamson, Prof. 

 Frankland, Drs. Warren De la Rue and Hugo Miiller, Mr. Buckton, 

 Dr. Odling, Mr. A. H. Church, Mr. C. Greville Williams, and 

 Mr. Piesse. The data are collected in two long tables forming two 

 appendices: the first containing the refractive indices of the lines 

 A, D, and H, of 78 specimens at two or three different temperatures ; 

 the second, the refractive indices of all the more important lines for 

 61 of these liquids, and 10 others at the temperature of the room 

 when the observations were made. 



Five points were investigated, and the following are the results 

 arrived at with respect to each point. 



I. The relation between the change of refraction (sensitiveness) 

 and the change of volume by heat. The uniform testimony of about 

 90 different liquids examined was that both refraction and dispersion 

 diminish as the temperature increases. 



The following Table will suffice as an example, showing as it does 

 that the different rays are more sensitive in the order of their 

 refrangibility : 



This change of refraction by heat was compared with the known 

 or ascertained change of volume in bisulphide of carbon, water, 

 methylic, ethylic, and amylic alcohols, ether, acetone, acetic acid, 

 formic, acetic, and butyric ethers, methylic and ethylic iodides, 

 salicylate of methyl, bromoform, benzole, xylole, cumole, nitrobenzole, 

 hydrate of phenyl, the rectified oils of turpentine and Portugal and 

 eugenic acid, and in every case it was found that the refractive 

 index minus unity, multiplied by the volume, gave very nearly a 



