Internal Prismatic Reflexion. 369 



ceded that in these six figures, but still find abundant 

 changes in the spectra by the other means which I have 

 detailed. 



I have another objection to Mr. Cooper's supposition : if 

 in the course of twenty reflexions seven emerge from the 

 posterior face, the others must emerge from the other two 

 faces of the prism, and thus form faint spectra from the hori- 

 zontal and anterior faces, as well as the principal spectrum 

 from the posterior ; but this, neither reading nor experi- 

 ment induce me to believe to be the case. 



I will again observe, sir, in conclusion, that I do not 

 now enter into the general question, whether the indices 

 of refraction be three or innumerable? but merely attempt 

 to shew, that if the former be the case, the requisite 

 filling up of the spectrum cannot be produced by multiplied 

 reflexion within the prism, because whether those reflexions 

 be five, twenty, or one hundred, the discrepancies which I 

 have described would still appear. 



I am, sir, your obedient servant, 



G. DODD. 



52, Myddelton Street, Clerkenwell, 

 August 31st, 1836. 



Article VIII. 



On some Silicates of Alumina. By R. D. Thomson, M.D. 



To the compounds of silica and alumina great interest is 

 attached, in consequence of the frequency of their occur- 

 rence, the diversity of their external characters, even 

 when closely allied by chemical composition, and their 

 utility in manufactures. 



Notwithstanding the care which Berthier has bestowed 

 on the subject, the arrangement of this class of minerals 

 labours under great disadvantages, and it is questionable, 

 whether our present knowledge is adequate to enable us to 

 introduce that precision which has been applied to other 

 bodies possessing analogy of chemical composition. 



I was led to the investigation of the composition of these 

 simple compounds, from having lately met with three species 

 of minerals, which appeared to possess distinctive external 

 characters; and could not be referred by these features 

 with certainty to uny described species. They were all 



