610 



REPORT — 1897. 

 Atomic Weight of Cobalt. 



The preliminary series in each case is of little consequence. The second series 

 in each case represents results obtained from the -weighing of the argentic bromide, 

 while the third represents those obtained from the -weighing of the silver. It is 

 evident that the four samples of each bromide gave results essentially consistent 

 with one another, and hence that the atomic weights of cobalt and nickel cannot 

 be far from 58-99 and 58-69 respectively, if oxygen is taken as 16-000. 



5. On the Occurrence of Hydrogen in Minerals. By M. W. Travers. 



6. The Spectrographic Analysis of Minerals and Metals. By Professor 

 W. N. Hartley, F.R.S., and Hugh Ramage. 



The steps by which the authors were led to this method of analysis were 

 described and illustrated by lantern slides. After discovering the presence of 

 gallium in the crude iron smelted at Middlesbrough, and tracing it to the Cleve- 

 land ironstone, it became necessary to examine other iron ores for this rare 

 element. A combination of chemical and spectrographic methods was first used 

 on 100 grammes of sample. The results were satisfactory as far as the detection 

 of gallium was concerned, but the process occupied too much time. 



A simple method, in which 0-5 gramme of the ore was rolled in filter paper 

 and heated in the oxyhydrogen flame, the spectrum of which was meanwhile 

 photographed, was tested with very satisfactory results. Not only could gallium 

 be detected, but many other elements also at the same time. A large number of 

 minerals and meteoric bodies have been examined by the method, and tabulated 

 statements of the results were exhibited on the screen. Attention was directed to 

 the wide distribution of the elements sodium, potassium, calcium, copper, silver, 

 iron, manganese, and lead, and to the facts that every specimen of magnetite, 

 bauxite, and meteoric iron examined contained gallium, as also did many specimens 

 of blende and ironstone, and that siderite and the tin ores examined all contained 

 the metal iridium. 



Photographs of oxyhydrogen flame spectra of some of the elements were 

 exhibited, and their simple character contrasted with the complex spark spectra 

 of the same elements. One plate contained the flame spectra of the alkali metals, 

 a second plate contained those of copper, silver, and gold ; another plate those of 

 iron, cobalt, and nickel. Similarities in the spectra of similar elements were 

 indicated in these. 



SATURDAY, AVGUST 21. 

 The Section did not meet. 



