134 



NATURE 



[April 12, 19 17 



the duration of these losses reduces them in proportion. 

 To obtain economy in heat, therefore, the ideal is not 

 only to insulate as well as practicable, but also to heat 

 and cast the metal in as short a time as possible, and 

 this ideal may be approached by having each lb. of 

 metal heated for the shortest jwssible time. The total 

 loss of heat per lb. of metal while it is hot is the 

 criterion. From this point of view Dr. Hering states 

 that the ideally perfect melting furnace, if such it can 

 be called, is the electric fuse, in which the intended 

 result is completed in such an exceedingly short time 

 — a fraction of a second — that the heat losses during 

 that time are vanishingly small, and hence the thermal 

 efficiency is practically loo per cent. 



With fuel heating, too great a rapidity of heating 

 generally involves high chimney losses, i.e. a lower 

 efficiency in heat transmission to the metal, and hence 

 a limit to the speed is soon reached ; but with electric 

 heating there is no chimney loss, and the possibilities 

 of rapid heating are therefore more encouraging. 

 Electric arc heating involves high radiation losses 

 from the arc itself, but in heating the metal by its 

 own resistance the heat can be generated below the 

 surface and in the metal itself, thereby eliminating all 

 heat transmission losses. Extremely rapid heating 

 then becomes possible, being limited only by the size 

 of the heat-generating capacity provided, and in the 

 case of brass or zinc by the volatilisation of the zinc 

 in the part in which the heat is set free. By the resist- 

 ance method, therefore, the ideal represented by the 

 electric fuse can be approached more closely than by 

 any other known method. Small high-speed furnaces 

 are therefore, from this pomt of view, an approach to 

 the ideal, particularly as they involve the minimum 

 of contamination of the metal being melted. In Dr. 

 Hering's opinion, it will in tim^ become possible, 

 for light castings at least, to be melted in an electric 

 furnace about as fast as the metal can be cast, in 

 which case the furnace would need to have a metal 

 capacity of only enough for about two inoulds. In 

 that case it would be so small that it could be trans- 

 ported to the moulds, thereby saving the usual large 

 heat losses in the transporting crucibles, besides the 

 heat losses in the crucibles themselves. 



Another factor, however, is involved, viz. the larger 

 the amount of metal in a furnace, the less the rate 

 of heat loss per lb., because the larger the volume, the 

 less is the surface exposed. In a large furnace with 

 a hemispherical hearth the heat loss per lb. of metal 

 through walls having uniform insulation is reduced to 

 about one-half when the capacity is increased from 

 I to 10 tons. Hence, for this reason, the larger the 

 furnace the better. 



In choosing between these two apparently conflict- 

 ing ideals the following considerations must be borne 

 in mind : — (i) When melting is the only object, then 

 the metal should be kept hot the shortest possible 

 time; hence there should be used as small a furnace 

 as is consistent with the amount of metal required 

 for one casting, (ii) When there are involved opera- 

 tions such as refining, mixing, uniformity of alloying, 

 the taking of specimens for analysis while melted, 

 or any other process requiring time, the larger the 

 furnace the better. 



LIQUID FUEL. 



«'T IQUID Fuel and its Combustion " was the title 

 •'-' of a paper read by Prof. J. S. S. Brame, on 

 February 20, before the Institution of Petroleum 

 Technologists. Attention was directed to the in- 

 creasing use of liquid fuel, and especially to its con- 

 nection with those developments of the internal com,- 

 bustion engine which have so largely determined the 



NO. 2476, VOL. 99] 



progress of aviation and submarine navigation. 

 Nevertheless he recalls the warning of Redwood 

 (1905) that no oil supplies are in sight sufficient to 

 replace anything like the bulk of solid fuel consumed. 

 The use of liquid fuel for steam raising and indus- 

 trial heating is the special subject of the paper, and 

 the following considerations are brought forward. 

 In constancy of chemical composition, whatever the 

 source, and therefore of calorific value, mineral fuel 

 oils compare very favourably with coal, and accord- 

 ingly physical considerations such as low viscosity 

 and freedom from grit may decide the choice of oil 

 fuels. Turning to our home supplies, it is gratifying 

 to note that the heavy fractions of the Scotch shale 

 oils are ideal in this respect; having been distilled 

 they are clean, while their fluidity is very satisfactory. 

 Another home product, which is deserving of the close 

 attention of liquid fuel experts, is coal-tar, the supply 

 of which must increase with the extension of 

 coal carbonisation. Its production may outgrow its 

 use? in normal channels,- and as a home-made 

 liquid fuel its rational utilisation is a matter of high 

 national importance. Nevertheless, for marine pur- 

 poses tar (and tar oils) must remain inferior to 

 petroleum, since a higher oxygen content and lower 

 calorific value are inevitable, while a capacity for giv- 

 ing off' disagreeable fumes may make it objectionable 

 in the confined space of a stokehold. Methods of 

 burning oil are surveyed historically, leading up to 

 the spray burners now almost invariably used which 

 " atomise " the oil. 



The method of spraying is varied, depending on 

 the use of compressed air or steam, or on forcing oil 

 alone under pressure through a suitable burner, a 

 method specially adapted for use in marine boilers. 

 On theoretical grounds air injection would seem to 

 be most generally efficient; steam may propel oil 

 satisfactorily into the fire, but afterwards its influence 

 on combustion can only be of negative value. The 

 general arrangements of the system for combustion 

 have more bearing on the success of a plant than the 

 choice of atomiser. It is too often overlooked that, 

 compared with solid fuel, where burning is mainly 

 confined to the fuel bed, oils require a much greater 

 volume of combustion space. 



Looking to the future. Prof. Brame points out 

 how much depends on the development of the internal 

 combustion engine; for naval purposes he believes 

 that oil firing with turbines will hold the field. 



J. W. c. 



RECENT PROGRESS IN SPECTROSCOPY.' 

 II. 



RADIATION is an electromagnetic process, 

 and must be determined by the electrical 

 state of the radiator. A molecule may be 

 neutral or for a moment charged by the loss 

 or gain of an electron. This tyf>e of ionisa- 

 tion must actually occur, as indicated by the conduction 

 of electricity through the vapour of a compound which 

 shows no evidence of chemical dissociation. What 

 causes the light emission ? It may accompany the loss 

 or gain of an electron by a neutral molecule, in which 

 case the emission centre would be charged. It may 

 be due to the shock of elastic collision with an electron 

 or ion, or to the reunion of an electron with a posi- 

 tively charged molecule, in which cases the emission 

 centre would be neutral. Luminous vapours emitting 

 band spectra usually appear to be neutral at the instant 

 of emission, so that it seems probable that band 

 emission is due either to elastic shock or to the 



1 Address delivered t-> Section B — Physics — of the American Association 

 for the Advancement of Science at the New York meeting, December, 1Q16, 

 by the chairman of the Section, Prof. E. P. Lewis. Continued from p. 118. 



