produces in the Double Refraction of Crystals. 411 



the plate of ground-glass, upon which the prism was placed 

 in my former experiments, rested on the ring of copper, which 

 having teeth upon its circumference could be put in motion 

 by a screw. By this arrangement, which allowed me to turn 

 the prism, it was as easy as formerly, when the prism re- 

 mained in the open air, to perform all the operations neces- 

 sary for the exact determination of the refraction, without 

 being obstructed by the heating apparatus, which being on 

 one side united with the boiler by a tube, was on the other 

 side attached to a rod of iron, rising from the masonry on 

 which the repeating circle rested. 



The temperature of the interior of the box remained, by 

 this means, perfectly invariable during the time occupied by 

 each experiment. From the temperature, however, of the 

 external air. of which the extremes were on different days 

 + 12° and +20°, it varied from day to day between +76° and 

 + 84°; so that there was almost a constant difference of 64°. 



With respect to the experiments themselves, I ought to re- 

 mark, that no change in the dispersion could be observed, 

 and that for this reason, I determined the ratio of refraction 

 only for the single ray F* of the spectrum. At the beginning 

 of each experiment the prism being at the temperature of the 

 external air, and being turned till the ray F was at its mini- 

 mum deviation, the variation in the deviation produced by 

 heating was measured in this position of the prism. For cry- 

 stals with one optic axis this determination was the only one 

 to be made, because the edge of the prism being parallel to 

 the axis of crystallization, the refracting angle did not change 

 with the temperature. But for crystals with two optic axes, 

 it was necessary, besides this, to determine the change which 

 the difference of temperature produced in the refracting an- 

 gle, which, as the following results will show, was very con- 

 siderable. 



With respect to the calculation of the index of refraction, 

 it must be observed that the refracting power of the air sur- 

 rounding the prism hew^ diminished, the deviation becomes 

 at first directly increased, but at the passage of the ray through 

 the last plate of mica, it is, on the contrary, somewhat dimi- 

 nished. These two corrections must be determined separately. 

 According to the experiments of MM. Biot and Arago, the 

 index of refraction of air at 0°, and at a barometrical pressure 

 = //, is 



=y 



0000588767 // 



"^ 1- + 0-00375/ "^ 6"'-76' 



• This ray or dark line is nearly the boundary between the green and 

 \.\\cbluc ^jiatc. — Edit. 



3 G 2 



