90 



NA TURE 



[November 23, 189; 



tube. I pass in air, the flame loses luminosity, and rapidly be- 

 comes an ordinary two-cone d Bun^en flame. I push the sir 

 Mipply further ; the inner cone enters the tube, and descends 

 uniil it rests on 'he end of the inner lube. The two cones of a 

 h3dr< carbon flame art- ihus widely separated ; we can aspirate a 

 sami le of the i^a-es, and see what changes have taken place in 

 the first region ol cmbu-tion. The result is one that we 

 might await wiih curi. sity, for we have now a competition. 

 There are both caibon and h\drogcn to burr, and not enough 

 oxygen to bum 1 o'h : ihe question is, which will have the pie- 

 ference ? I think I nay say that ihe ofl'-hand opinion of any 

 chemist who has not had bis attention drawn specially to this 

 point would be that the hydrogen would easily have the pre- 

 I'erence. Bu', as a matter of fact, this question was settled long 

 ago by Dalton, ar.d in the opposite sense, and in the present 

 case analysis would confiim ills conclusion. Jf we analysed 



d c 



About two-thirds of the carbon is burnt to form carbon 

 montxide, one third to form carbon dioxide, whilst rather 

 li-ss than two-thirds of the hydrogen is burnt, and more than 

 one-third remains altogether unburnt. We need not dwell on 

 the details, especially as the analysis of ihe gases was ma le 

 after th^y had cooled. The four gases — cnrtion monoxide, 

 carbon dioxide, steam, and hydrogen — act upon one, as a matter 

 of fact, while ihey are cooling do"n, and the distribution of the 

 oxyj^en that we find in our analysis of the cold gas is not 

 |)recisely that which exis's in the gases a-- they just leave the 

 inner cone. We shall only draw a general inferenc, and it is 

 one that has been recently verified in a very complete manner 

 by Prof. Dixon and his pupils. This inference is simply that 

 the carbon in the inner cone is for the most part burnt to carbon 

 monoxide, and that the hydrogen to a considerable extent is 

 set free. So much then for the inner cone. The outer cone is 



b a 



^ e f g 



Fig. 3. — ^, /■, r, d, flames with successively increasing quantities of coal-ga". d, e,/, g, flames with fixetl supply of coal-gas and siicces.'ivtly increasing 



quantities lT air. 



the gases we should find that all the carbon is burnt in the first 

 cone, whilst a coiisideiable part of the hydrogen passes through 

 unburn!. The change is not quite so simple as these words 

 might apply, as you will see from the actual figures of analysis. 

 Analysis ok Inter conal Gases from a Coal gas Air Fla.me. 



Caibon m.noxide 8"7 1 . > .•,, 



Hydr g.n 92 1 ^7 9 combustible gases. 



< arbon dioxide 4-1 



Water 16 



Nitrogen 62 



^.' [ 20'i burnt gases. 



Amount of air used 78-5 ( 9^yS^" l^'5 



' -' \ Nitro. en 620 



Amount of air still needed ... 42-9 ( Oxvgen 90 



^ ^ { Nitro.^en ..... 33'9 



NO. 1256. VOL. 49] 



due simply to the burning of the carbon monoxide and hydrogen 

 which escape from the inner C' ne. When they meet with 

 oxygen in the free air their combustion is completed. We are 

 now in possession of the explanation ol the two-coned gas air 

 flame. Applying this to the tiny gas flame to which no air has 

 been previously added, we see that the inner cor e will be formed 

 where the air has penetrated the gas .-ufhc enily to produce such 

 a gaseous mixture as we had in the lower cone ol our separator. 

 T he gases coming from this burn further out when they meet 

 with more air, and form a second cone. 



The last thing we have to explain in the ordinary gas flame is 

 the production of the yellow luminous patch, which, from the 

 illuminating point of view, is the most iin) ortant feature of all. 



Now I need scarcely remind yuu tl at the general opinion is 



