On the Velocity of Propagation of Light in Water . 263 



their results " On the Influence of Temperature on the Refrac- 

 tion of Light " in the Philosophical Transactions*. These inves- 

 tigators made use, for their purpose, of the apparatus of Baden 

 Powell f, and made determinations (for water between 0° and 

 80°-6 C., for bisulphide of carbon between 0° and 42°* 5 C, and 

 for various alcohols and ethers between 0° and 70° C.) of the 

 alteration of the refractive index for Fraunhofer's lines A, D, H. 

 The results arrived at are collected in the following propositions. 



(1) All substances, by increase of temperature, show a dimi- 

 nution in the refractive index. The magnitude of the alteration 

 varies with various substances, being least in the case of water, 

 greatest in that of bisulphide of carbon. 



(2) The length of the spectrum diminishes rapidly with the 

 temperature in the case of substances which refract light power- 

 fully, such as bisulphide of carbon, phosphorus, &c. ; this dimi- 

 nution is scarcely perceptible in the case of water. 



(3) In the neighbourhood of the point of change of the state 

 of aggregation no remarkable change occurs in the nature of the 

 alteration of the index. 



The above conclusions have reference to the relation between 

 density and refractive index ; we shall return to this point again 

 at the end of the present paper. The extent of the investiga- 

 tions of these two physicists is certainly very great, and great 

 abundance of material is presented. But the validity of their 

 results is open to doubt, because it is absolutely essential that 

 the measurements should be exact to one unit in the fourth de- 

 cimal place, and this exactness is not attained. For water, two 

 determinations of the D line at 0° C. are given, 



0° C 1-3330 and 1*33374 (Riihlmann 1-33374). 



While the first of these numbers forms the starting-point for 

 a series of determinations from 0° to 70° C, advancing conti- 

 nually by 5° C, the second is obtained from a small number 

 of observations (which are said to be most careful) in order to 

 verify Jamin ; s result, that the index of refraction does not show 

 any singularity at 4° C. corresponding with the maximum den- 

 sity. My own observations, after elimination of accidental 

 errors, gave at 0° C. a result perfectly agreeing with Glad- 

 stone's more correct number. It hence follows that the first 

 number is too small by 7 units of the fourth decimal place. 



For 70° C. for the D line, these two physicists gave 

 70° C. . . . 1-3237 (Gl. and D.),' (1-32505 R.), 

 a value again too small, according to my experiment. The re- 

 sults moreover differ very widely from Baden Powell's and 



* Philosophical Transactions for 1858, pp. 887-894. 



f Described and figured in the Report of the British Association for 1839. 



T2 



