TRANSACTIONS OF SECTION B. 675 



burn more readily than the hydrogen, and that accordingly the luminosity of 

 ordinary flames must not be explained as is olten done by assuming the very re- 

 verse of this. The luminosity of ordinary flames is now generally admitted to be 

 due to the separation of solid carbon in the flame, and this must be ascribed to a 

 decomposition or dissociation of the hydrocarbons by heat. This heat is supplied 

 by the sheath of non-luminous combustion surrounding the flame. The author 

 considers this sheath, -which consists of tvpo parts (a barely visible outer part, and 

 a bright blue inner one) to correspond to the two cones of combustion in a non- 

 luminous flame. The author, whilst admitting from the result of experiments 

 made by Professor Lewes and by himself that acetylene was present in the interior 

 of luminous flames, thought that Professor Lewes had not established his view 

 that luminosity was mainly and essentially due to the formation and subsequent 

 dissociation of acetylene, and that he was altogether relying too much upon that 

 substance. A study of the precise cause of carbon separation was being under- 

 taken by the author with the aid of his flame-cone separator, a single hydro- 

 carbon (ethylene) being employed. 



Other experiments with the flame-cone separator relating to the spectra of 

 metallic salts were exhibited. An account of these will be found in the report 

 of the proceedings of Section A, p. 646. The chief conclusion was that the spectra 

 obtained when metallic salts were introduced into non-luminous flames were the 

 direct outcome of and were entirely dependent upon chemical changes. 



4, Report of the Oommittee on the Direct Formation of Haloids from Pure 

 Materials. — See Reports, p. 262. 



Tlie Tteaction of Hydrogen with mixtures of Oxygen and Chlorine. 

 By J. A. Hakker, B.Sc. 



6. Preliminary Note on the Action of Dry Ammonia Gas on Sulphates. 

 By Professor W. R. Hodgkinson, Ph.D., F.B.S.E., and C. C. Trench, 

 Col. B.A. 



Very little appears in chemical literature as to the action of really diy 

 ammonia gas on salts at an elevated temperature. Some chlorides are stated to 

 be reduced, others to yield nitrogen compounds when heated with ammonia, but 

 little or nothing seems to be known respecting the action with sulphates. 



The experiments were also undertaken to ascertain (1) to what e.xtent, if anj', 

 sulphates were reduced by ammonia at temperatures below that at which the 

 sulphate itself decomposes, and (2) whether nitrogen compounds (nitrides) could 

 be formed. 



All the work has been quantitative as far as possible. 



After a number of preliminary experiments, the reactions were carried on in 

 a wide glass tube, so arranged that it could be heated in an air-bath to a detinite 

 temperature when required, or by the naked flame. The substance was contained 

 in a boat of platinum or porcelain as the case demanded, which with its contents 

 was weighed before and after the action, '"^he ammonia was dried by passing first 

 over potassium hydrate and then over a considerable quantity of freshly ignited 

 lime. 



As water is, however, certainly one product of the mutual action of ammonia 

 and sulphates, this drying is perhaps not an essential precaution. To catch vola- 

 tile products the tube was drawn out to about 50 cm., and either dipped into 

 water or ended in a dry flask externally cooled. 



General Results. — The preliminary experiments showed that some metallic 

 sulphates were entirely reduced by heating with ammonia. Silver mercury and 

 bismuth behaved in this manner. Water and what at first appeared to be 



X X 2 



