310 Sir J. Conroy. Refractive Indices and Densities of 



underground one), the temperature of the liquid in the prism could be 

 easily kept within one or two tenths of a degree of 18. 



The liquid in the prism was kept stirred by means of the platinum 

 wire, the end of which was bent round into a ring, and the bulb of the 

 thermometer was immersed in the liquid. 



The observations were made with sodium light only, as it had been 

 previously found that, owing to the brilliancy and constancy of the 

 sodium light, it was not only far easier to make observations with it, 

 but that those observations would be more accurate than observations 

 made with light of other refrangibilities. 



The readings were made by moving the telescope first from left to 

 right, then from right to left, until the intersection of the cross wires 

 coincided with the image of the sodium line ; and the micrometer was 

 read in the same way. After four readings of the micrometer had been 

 made, the prism was reversed by rotating the table of the goniometer, 

 and adjusted for minimum deviation : four readings made, and then the 

 prism again adjusted for minimum deviation, and four more readings 

 made ; the prism replaced in its original position, adjusted, and the 

 same number of readings made. 



The liquid in the prism was well stirred, and the thermometer read 

 before each adjustment of the prism for minimum deviation. Half the 

 difference between the two sets of eight readings was, of course, the 

 deviation. The readings were made thus in order, as far as possible, 

 to eliminate errors due to the position of the prism, and to slight 

 changes in temperature. 



For each solution observations were made with the prism at the 

 temperature of the room, and at 18. In the former case one set only, 

 as the temperature of the prism did not remain constant, and indeed 

 had usually risen slightly before the one set of readings could be made ; 

 in the latter case four sets were made. 



The densities of the solutions were determined by means of Dr. 

 Perkins' modification of the Sprengel density tube.* Two of these 

 tubes were used, containing respectively 34*0376 c.c. and 25*1587 c.c. 

 After being filled, the tubes were placed in a water bath at 18 for 20 

 minutes, the volumes adjusted, and weighed against counterpoises made 

 of the same kind of glass, and having about the same surface areas ; 

 the tubes and their counterpoises were wiped carefully with a dry 

 cloth before each weighing, to ensure that, as far as possible, the sur- 

 faces were in the same condition as regards moisture. After being 

 weighed, a few more drops of the solutions were introduced into the 

 tubes, which were then replaced in the bath ; four determinations being 

 made with each solution. 



The water-bath contained a Harcourt gas regulator, and a slow 

 stream of water from the main was run into it ; without this the tem- 

 * ' Ohem. Soc. Trans.,' 1884, p. 444. 



