REPORT ON REFRACTIONS OF SOLAR SPECTRUM. 5 



of the indices, for an increase of temperature will not hold good 

 except within very confined limits. 



In many media the violet and blue rays are absorbed ; and in 

 others the lines are very faint, or invisible : in some such cases, 

 however, their position might be estimated nearly by means of 

 coloured glasses. Some results of this kind I have stated, di- 

 stinguishing them as only rough approximations, in cases where, 

 from the nature of the substance, we can hardly hope to obtain 

 more accurate results. Among those media of this class which 

 I have recently examined, is the liquid ammonia ; the volatile 

 nature of the substance being such as to occasion so great a 

 want of homogeneity that no lines are visible, and the measures 

 are only rough estimations. 



In some cases, and those among the most interesting, not 

 even such approximate measures appear attainable. This was 

 especially the case in that highly important instance, the chro- 

 mate of lead. A good specimen of this crystal was kindly pre- 

 sented to me by H. J. Brooke, Esq., F.R.S., which was (not 

 without considerable difficulty) cut into a prism of small angle 

 by Mr. Dollond. But the appearance of the spectrum was 

 altogether confused, and the blue end wholly absorbed, so that 

 no measures could be obtained. 



In some of my first series of observations, no distinct mea- 

 sures were taken for the two bands at H, but the middle point 

 between them was taken. Such results are rejected in the pre- 

 sent report. But in the low-dispersive cases, the difference is 

 so minute, that I have thought it quite sufficient to introduce a 

 small proportional correction, to reduce the index to the exact 

 position of H. 



Some observations given from the first series, for certain 

 chemical solutions of very low dispersive power, are of little 

 value, since the differences of temperatvu-e render comparison 

 impracticable, though the whole subject of the indices of 

 chemical compounds, compared with those of their elements, is 

 one well deserving more full investigation. 



Results for a few other media are added from observations 

 now first published ', among which will be found rock-salt, so 

 interesting from its relations to heat. 



I had entertained hopes of being able before this time to 

 obtain results for a much more extended range of substances, 

 especially those of the more highly dispersive class. But there 

 are many difficulties in procuring specimens in a state suscepti- 

 ble of this mode of prismatic examination ; there are also many 

 substances, and among them the most important, to which it is 

 to be feared such examination cannot be applied. 



