TRANSACTIONS OF THE SECTIONS. 79 



On the Composition and Valuation of Superpliosji)Iiates. 

 By Professor Gailowat. 



On an Aluminous Mineral from the Upper Chalk near Brighton. 

 By Dr. J. H. Gladstone and Mr. G. Gladstone. 



In an old clialk-pit at Hove there are many faults, and some of these are filled 

 up with a white soft miaeral that runs along- the broken layers of flint and imbeds 

 the fragments. It appears in agglomerated masses, which easily fall to powder, 

 and are porous. Sp. gr. 1-99. One piece that was analysed proved to be the hydrated 

 disilicate of almnina, that has received the name of CoUyrite, with no other im- 

 purity than one per cent, of carbonate of lime. Another piece contained 13 per 

 cent, of carbonate of lime, and 5 per cent, additional of carbonic acid, which was 

 supposed to be combined with ahmiina. As the silicic acid was proportionally 

 smaller in quantity, this piece was \-iewed as coUyrite in which about halt the silicic 

 acid had been replaced by carbonic acid. 



On the Emission and Absorption of Rays of Liglit hy certain Gases. 

 [By Dr. J. H. Gladstone, F.E.S. 



This communication arose out of an attempt to detei-mine what constituents of 

 the air give rise to the "atmospheric lines" of the solar spectrum, of which a map 

 had been exhibited by the author at the Leeds Meeting of the Association, and 

 which had been since published in a more complete form by Sir David Brewster and 

 himself A comparison of the bright rays emitted by nitrogen, oxygen, hydrogen, 

 carbonic acid, and water, when strongly heated, had shown that they do not coin- 

 cide with the absorption-bands of the atmosphere. It is possible that the three 

 bright lines of the hydrogen specti'um, as given by Angstrom and Pliicker, may be 

 in the same position as 01, F, and p of the atmospheiic spectrum. Yet the author 

 inclined to the belief that these absorption-bauds are due to two or more different 

 constituents in varying proportions, more abundant in some places than in others, 

 and probably in very mmute quantities. 



The following facts were mentioned among others : — The flame of carbonic oxide 

 burning in air gave a continuous spectnmi from about C to about k, where it ceased 

 rather abruptly: it was without either bright or dark lines. The alcohol flame 

 shows four bands — the first faint in the yellow, nearly midway between D and E ; 

 the second brighter, green, just beyond b, with the refractive index 1'6254 for glass, 

 which gives as the refractive index of b 1-6249; the third faiat and blue, about 

 half-way between F and G ; the foui-th a more luminous double line, violet, with 

 the refi'active index 1-6413, that of the line G being 1-6404. The oxyhydrogen 

 flame gave a contiuuous spectrum principally green and blue, extending to about 

 G 33, with no lines corresponding to the hydi-ogen lines of Angstrom and Pliicker, 

 The lightning- flash gave a continuous spectnmi, showing all the colom-s from red 

 to \ioIet, -with doubtfril indications of more luminous bands. That there is no 

 necessary coiTespondeuce between the lines of absorption of a gas at the ordinary 

 teniperatiu-e, and the rays emitted by it at a high temperature, is sti-ikingly proved 

 by iodine, where the absoi-ption-bands delineated by Professor Miller, the groups 

 of green caud blue bands produced when the vapour is introduced into a Bunsen's 

 flame, and the lines of the rarefied gas as observed by Pliicker, are perfectly dif- 

 ferent. By the prismatic analysis of solar light, the absence of the colom-ed gases 

 from the air can be proved, even in very minute quantity. Thus the author 

 observed that about y^th of an inch of bromine vapour interposed between the 

 eye and the object-glass of the refi-action goniometer was suflicient to exhibit the 

 absorption-bands ; and fi-om this he had reckoned that if free bromine constituted 

 one thousand millionth part of the atmosphere, it would betray its presence in the 

 solar specti-iim when the sun was on the horizon ; but there is no such indication. 

 This, however, rests on the improved assumption that a gas almost infinitely 

 difiiised along a given line will produce the same absorbent efiect as if its particles 

 were all close together at some point along that line. 



