M. Iiaagen on the Refractive Indices of some Haloid Compounds. 457 



tweea 10° and 70° C. As iodide of silver melts only at a rather 

 high temperature (100° C), the effects could not be attributed 

 to the irregularities which might be produced in the neighbour- 

 hood of the temperature corresponding to the change in condi- 

 tion of the substance. 



The expansion of iodide of silver' should thus be expressed by 

 a negative coefficient — at any rate, within the above limits. 

 Moreover, in proportion as the temperature rises, the value of 

 the coefficient considerably increases, so that the contradiction 

 becomes more and more marked. 



Haagen has determined the refractive indices and specific gra- 

 vities of some liquid haloid compounds*. Two of Meyerstehr's 

 spectrometers were used for ascertaining the refractive indices. 

 By means of the larger one, which had a vernier reading to ten 

 seconds, the index could be determined with certainty to the 

 fourth place of decimals, and with a possible error of four units 

 in the fifth place. By the smaller instrument, which was used 

 for investigating fuming chlorides, the angle could be read off to 

 a minute ; the indices determined with its aid were accurate to 

 the third place; the fourth varied by three or four units. The 

 method used was that of the least deviation. A Geissler's hy- 

 drogen-tube was used as a source of light, and the refractive in- 

 dices determined in reference to the three principal lines of the 

 hydrogen-spectrum, of which the red one (a) notoriously coincides 

 with Fraunhofer's line C, the green ((S) with F, while the violet 

 (7) lies between G and H. 



The refractive indices were all made at the temperature 20°. 

 The liquid to be determined was contained in a hollow prism, 

 and its temperature indicated by a delicate thermometer with a 

 small bulb. The prism was heated in an air-bath a few degrees 

 above 20°, then placed on the spectrometer, and the least devia- 

 tion of a ray determined when the temperature had exactly sunk 

 to 20°. Repeated observations after fresh heating never varied 

 from each other more than 10, or at most 20 seconds, which 

 makes a difference in the refractive index of four to eight units 

 in the fifth decimal. 



The constants of Cauchy's formula, fi = A+ — 2 , were calculated 



from the refractive indices for the red and the violet ray. Putting 



a B , A B 



^*=A-f jjTn and ^ y = A+ ^y 



* Zeitschrift fur Chemie, January 25, 1868. 



