Feb. 17, 1 88 1 J 



NATURE 



\6i 



I HAVE been much interested in reading the above, and I hope 

 that Mr. Hay's example will be followed by other observers, in 

 order to establish a fair average result of the relative cot of the 

 coke-gas grate and the ordinary coal grate. Mr. Hay's results 

 are not as favouraMe as those obtained hy myself, owing probably 

 to some imperfection in his arrangements, which are not described 

 sufficiently to form any judgment upon them. I should like to 

 know, for instance, whether the copper-back plate, which I pre- 

 sume he used, although it is not referred to, was backed by fire- 

 clay, or whether it touched the ironwork of (he fireplace, whereby 

 its heat would be conducted away. This alone might account 

 for the difference in result obtained by Mr. Hay and myself, and 

 I think this opportunity is a favourable one to send you the 

 figures resulting from my own ob;ervation of the original grate 

 described in my article in N,4TURE, vol. xxiii. p. 25. This grate 

 has now been in use froai November 8 to January 31, during 

 which period it has been alight sixty-six days, and the average 

 time during which a bright fire h.as been kept up has been eight 

 hours daily. During this period of 528 hours there has been 

 consumed — 



£ ^. '/. 



1 112 lbs. coke at l8.f. a ton 090 



581 lbs. smokeless coal at 2c.f. a ton o 52 



4100 cubic (eet of grs at p. 6d. per 1000 ... o 14 4 



I 8 6 

 Or an average of o-JiSi/. per hour, instead of o'525(j'., as restilting 

 from my first observation. The average consumption of solid 

 fuel per hour his been 3-2 lbs., and of gas 77 cubic feet. The 

 full supply of gas has generally been allowed during the first hour 

 of lighting, after which it is turned down to about a third ; this 

 I find to be a convenient mode of working. 



In comparing my results vvith Mr. Hay's, it must be borne in 

 mind that my room has a capacity of 7200 cubic feet, with 

 northern aspect, and his a capacity of 3000 cubic feet, also with 

 northern aspect ; his consumption should therefore be only 



-— - X o'JiS = o-2i6<)'. (instead of 0-3470'.), which figure would 



prove an economy in the employment of the cokegas grate over 

 the coal grate, which is o-27S5(/. by his own showing, and would 

 agree with the comparative results contained in my original 

 communication. C. W. Siemens 



February 2 



On the Spectrum of Carbon 



It is very desirable that, if possible, some definite conclusion 

 should be arrived at concerning the chemical origin of the bands, 

 which Prof. Liveing calls "hydrocarbon bands," and the im- 

 portance of the 1 oint at issue must be my excuse for again 

 addressing you on this subject. 



In my previous communications I pointed out that if it can be 

 .shown experimentally that the electric spark, in an atmosphere 

 of cyanogen /i;e/nn/i hydrogai, gives the groups in question (the 

 grou|s S and 7, wave-lengths 5165 to 5082 and 5635 to 5478 

 respectively, are here the only ones considered), they must be 

 due to carbon, and remarked that the hypothes's that they were 

 due to traces of hydrogen present as impurity is "to adopt an 

 extreme hypothesis which must be supported by cogent experi- 

 mental evidence before it can be accepted." Prof. Liveing 

 admits the ju-tice of this demand, and then goes on to say that 

 such " cogent experimental evidence, so far as the relations of 

 carbon and nitrogen are concerned, will be found in our complete 

 papers on the spectrum of carbon compounds in the Proa-ciiings 

 of (he Royal .•-ociety." This appears to me to be equivalent to 

 an admission that — as concerns carbon and liydrogen — no such 

 experimental evidence has yet been given ; which is also the 

 conclusion to which I came after perusal of the papers of Profs. 

 Liveing and Dewar referred to. 



It would seem then that the burden of proof that cyanogen 

 exists in which the spark will not give rise to the bands 5 and 7 

 rests with Prof. Liveing. Nevertheless I have repeated the 

 experiment with cyanogen, described in this journal (vol. 

 xxni. p. 197), fo as to set aside the objections raised by 

 Prof. Liveing to the former experiment. The apparatus was 

 in this case constructed of one piece of glass— a long piece 

 of hard glass tubing. This was carefully cleaned, the tube 

 was then contracted at two points, so as to separate a short por- 

 tion of the tube, into which platinum whes were fused, so as to 

 form a discharge tube. The whole tube was next heated to red- 



ness iu a furnace, while a current of oxygen passed through it 

 for some considerable time. The greater portion of the tube on 

 each side of the part containing the wires was then filled with 

 phosphoric anhydride, and a short length of the tube, separated 

 from the discharge tube by as great a length of phosphoric 

 anhydride as the length of the tube permitted, uas employed as 

 a rfctort, and filled with mercuric cyanide ; the other end of the 

 tube was drawn out and dipped beneath sulphuric acid. The 

 mercuric cyanide employed, after being finely pondered, was 

 dried for a long time in an air-bath, then transferred to a clean 

 hard-glass tube, in which it was repeatedly heated, while a 

 current of air dried by passing over calcium-chloride and phos- 

 phoric anhydride was drawn over it. From this tube it was 

 transferred immediately to the retort-tube. In uialdng the 

 experiment the mercuric cyanide was heated so as to give as slow 

 a current of cyanogen as possible, which was continued long 

 enough to expel all the air from the tube.. The tube was then 

 sealed up, leaving the discharge-tube, wuh a phosphoric anhy- 

 dride tul.e on each side of it, and put aside for a week. The 

 spectrum was then examined, with the same result as before. The 

 lube gave a brilliant carbon >pectrum, of \\\ hich 7 and 5 (the 

 positions of which were measured )_were the brightest groups. No 

 trace of the hydrogen C-line was obtained. Prof. Liveing 

 objects that this is not a sufficient proof that hydrogen is absent 

 (in which I cannot agree with him), and suggests that "a real 

 test would be to see whether, when the spark gives the line- 

 sjiectrum of carbon, the hydrogen-lines do not also appear." 

 This test is however not applicable, since (according to my 

 experience) cyanogen cannot be made to give the line-spectrum 

 of carbon. Further, in this particular case the spark could not 

 be got through the tube when the condenser was put on. 

 Giggleswick, February 11 \V. M. Watts 



"Prehistoric Europe" 



_ A.S there was no space to allow of all the auOiorilies being 

 cited in my criticism of the above work I now give those which 

 relate to the facts called in question by Dr. James Geikie in 

 Nature, vol. xxiii. p. 336. 



1. Dr. James Geikie repudiates as absurd the view attributed 

 to him, that the pala:olithic gravels "which oveilie the chalky 

 boulder clay of East Anglia were covered by an upper and 

 younger boulder clay," and denies that he ever wrote anything 

 which would justify that opinion. In " The Great Ice Age," 

 2nd edition, p. 531, he writes : "The pah-eolithic beds dovetail 

 into the glacial drifts, and are overlapped (as in Yorkshire) by 

 the deposits thrown down during the final cold period. To the 

 last interglacial period then we must refer the great bulk of the 

 palaeolithic river gravels of the south-east of England." If this 

 be true, w here are ihe glacial deposits in question to be seen ? If 

 they ever were above, or " overlapped," the palaeolithic grave's, 

 they have, so far as our present knowledge goes, been utterly 

 destroyed. Of course this view is absurd. 



2. The reindeer associated w ith the hippopotamus and hyjena in 

 the same stratum in the Victoria Cave was discovered while the 

 exploration was under my management, and was published iuBiit. 

 Assoc. Rep. 1872, Trans, p. 179, and again in Mr. Tiddeman's 

 Report, op. cit. 1876, Reports, p. 118. The animal is omitted 

 by the author where its presence would destroy his argument as 

 to climate, but he does not forget to record its subsequent dis- 

 covery at a higher level, where it falls in with his argument. It 

 maybe remarked that the association of reindeer with hippo- 

 potamus in this cave has no special theoretical value, because 

 the two animals have been found together in several other hytena 

 dens. 



3. The fossil mammalia of Mont Perrier are typical Upper 

 Pleiocene, as may be seen from the works of Croizet and Jobert, 

 Gaudry and Gervais, and as I can testify from their examina- 

 tion. The glacial origin of the overlying tuffs, which I have 

 e.xamined under the guidance of M. Julien, seems to me to be 

 open to consideral le doubt. 



4. The mammalia of Leffe, and those of the Val d'Arno with 

 which they are classified by Dr. James Geikie, characterise the 

 Upper Pleiocenes of Italy, as may be seen from the careful 

 essays published by Dr. Forsyth Majnr, and from an examina- 

 tion of the m.agnificent collection in the museum of the University 

 of F'lorence. 



5. If pages 309-318 of "Prehistoric Europe," dealing with 

 " interglacial epochs," do not imply a belief that the Neolithic 

 skull of Olmo is interglacial, I am unable to ascertain their 

 meaning. 



