146 Royal Society : — 



Sea-salt in very beautiful colourless crystals gave* . 2* 195 

 The same, fused and rapidly cooled, was evidently in a~\ 0.004, 

 state of perfect crystallization, and gave . . . j 



that is to say, exactly the same number. 



Thus, then, there arc substances which, contrary to sulphur, 

 quartz, and the silicates, have only a very feeble tendency, or 

 none at all, to assume, even momentarily, the vitreous state. 



It might be demanded, to which of the two categories does 

 akimina belong. Natural corundum in small colourless crystals 

 gave me a density of 4'022 ; the same crystals, fused with the gas 

 blowpipe of M. Gaudin, had a density of 3'992,— an insensible 

 difference. There is not, therefore, a glass of corundum, as there 

 is of quartz, and this physical property of alumina, as all its 

 chemical properties, directly attaches aluminium to the group of 

 metallic bodies. 



[We observe with pleasure the present tendency of experimenters 

 to inquire into the intiuenee of molecular arrangement upon density. 

 The subject is one of great importance, and strongly solicits search- 

 ing examination. Our handbooks of natural philosophy and che- 

 mistry probably contain numerous erroneous statements as to the 

 influence of mechanical pressure upon density ; and he who places 

 this question on safe experimental foundations will do a good service 

 to science. We could have wished that the interesting paper before 

 us were a little more precise in the description of the mode of cool- 

 ing adopted, and of the appearance of the bodies after having been 

 cooled. — Eds.] 



XIX. Proceedings of Learned Societies. 



ROYAL SOCIETY. 

 [Continued from p. 79.] 

 May 24, 1855. — The Lord Wrottesley, President, in the Chair. 



THE following communication was read : — 

 " On the Theory of the Electric Telegraph." By Professor 

 WUliam Thomson, F.R.S. 



The following investigation was commenced in consequence of a 

 letter received by the author from Prof. Stokes, dated Oct. 16, 1854. 



11-225, which give a difference equal to 0027 of the first density, and in 

 the same sense as that of tin and bismuth. But such is the rapidity with 

 which this finely-divided lead is transformed into carbonate in the air, that 

 it was necessary to convert it into suljjhate to deduce the weight of the 

 matter employed. Does tliis complication cast some uncertainty on the 

 first number, "or ought we not rather admit it as the density of lead per- 

 fectly crystallized ? 



* In the essence of turpentine, the density of which had been previously 

 determined. 



