March 27, 1879] 



NATURE 



483 



for9assent a avouer son emprunt ; mais la connaissance de son 

 caractere donne a I'acte qu'on lui reproche une probabilite 

 voisine de la certitude. Une lettre d'Huyghens a L'Ho5pital 

 e.st sur ce point fort instructive. 



" 'M. Leibnitz,' d it Huyghen?, 'est assurement tres habile, 

 mais il a avec cela une envie immodere'e de paroitre, comme cela 

 re voit, lorsqu'il parle de son Analyse des infinis . . . , des lois 

 harmoniques des mouvemens planetaires, oil il a suivi I'inventicn 

 de M. Newton, mais en y meslant ses p ensees qui la gastent 

 . . . Encore suis-je fort en doute, pour des raisons que je 

 pourrois alleguer, s'il n'a pas tire sa construction (de la chainette) 

 de celle de M. Bernoulli.' 



"Dans la preface de ion 'Calcul differentiel,' Euler n'attribue 



a Leibnitz que la reduction des principes de Newton en systeme. 

 Lagrange, qui cherche chez Fermat I'origine du Calcul diflerentiel, 

 ne manque pas, dans ses ' Lemons sur le Calcul des Fonctions,' de 

 signaler les concordances de I'ecrit de Leibnitz de 1684, avec la 

 theorie anteriexure de Fermat. Gauss pensait, comme on le voit 

 dans I'ecrit de Sartorius de Waltershansen, que Leibnitz, meme 

 de loin, ne doit pas etre compare a Newton." 



It is satisfactory to quote Herr Diihring through his reviewer, 

 because the introductory sentence of the extract makes evident 

 to those not already aware of M. Eertrand's hL'-torical leanings 

 that it is on Dr. Ingleby's side the eminent French mathe- 

 matician would give his voice. Thomas Muir 



High School of Glasgow, March 17 



Blue Flame from Common Salt 



Some time ago the question was raised in Nature (vol. xiii. 

 p. 287) concerning the origin of the blue flame produced when 

 common salt is thrown into a hot fire. 



Among the suggestions that were advanced, no one ofTered 

 the only explanation that is at all feasible, viz., that it is due 

 simply to hydrochloric acid. 



The blue flame is not produced by sodium chloride only, but 

 by other chlorides as well. Those I have tried are EaCL,, 

 SrClj, KCl, AmCl, HgaClj, and HCl, the last both in solution 

 and as gas. 



It would be waste of time and space to enumerate all the 

 experiments I have made on this subject ; many of them were 

 for the purpose of proving that neither carbon nor sulphur had 

 any share in the reaction. t 



Violet 



One of my methods of obtaining the flame was to bum pure 

 hydrogen from a glass jet, and allow a mingled stream of HCl 

 and N H3 from two other jets to pass into it. The best source, 

 however, is calomel, heated on wire gauze by a Bun>en burner ; 

 the next best, AmCl. 



The spectrum of the "chloride" flame is characterised by a 

 series of double bands in the green, blue, and ^olet, the least 

 refrargible of each pair being the broadest. The four pairs in 

 the violet are especially prominent. There are two red bands 

 or lines, and one orange. The least refrangible red line occu- 

 pies the place of the hydrogen line C. 



A spark between carbon points in a bottle of HCl gas yields 

 a spectrum similar in appearance to that obtained from a chlo- 

 ride, but 1 was. unable to see any violet bands, only a faint 

 continuous spectrum. 



I was able to ascertain that the red lines coincided exactly. 



amSpecti'um, oT 

 i%3ciz oncU 

 cFAmcL 



fiotinl lines 



i/Ut I cannot afi5rm with the same positiveness that all the green 

 lines and bands coincide. Some undoubtedly do, but the spec- 

 trum of the HCl was so faint, and the spectrum of the chloride so 

 transient, that measurements were very diflficult, and I do not 

 pretend to any high degree of accuracy in my delineations of the 

 lines, besides the different conditions under which the two were 

 compared might account for considerable variations. For the 

 HCl I used a coil (capable of giving a 2" spark) with Leyden 

 jar, worked by six Smees. The carbon points were ^ inch apart. 

 The spark was tried both focussed and unfocussed on the slit of 

 a 2-prism spectroscope. When comparing the two spectra side 

 by side it was diflBcult in some cases to be sure of coincidence, 

 because the flame from the HgjClj would flash out brilliantly 

 for a moment and quite overpower the more feeble lines of the 

 HCl ; it would then disappear entirely, and more chloride would 

 have to be placed on the gauze. 



I have no doubt in my own mind that HCl is the cause 

 of the blue flame. I have proved that CI is a necessary 

 constituent, and I have not been able to get it in the absence 

 of hydrogen (the spectrum of pure CI is very diflerent), 

 and besides the red H-line is present in both cases, and 

 probably the other two as well. I do not think that the 

 presence of aqueous vapour is sufficient to account for the red 

 line. 1^ , 



I subjoin the spectra as I mapped tJiein, but itTnust be borne 

 in mind that I do not vouch that they are free from error. I 

 intend to photograph them when I have sufficient leisure, and I 

 hope the results will be more definite. I may be able to find 

 violet lines in the spectrum of HCl. 



A. Percy Smith 



Temple Observatory, Rugby, 

 March 15 



