370 



NA TURE 



{August 19, 1880 



seen that a faint absorption extends as far as 650, and 

 very likely the termination of this shade is a band. 

 Fig. 6 represents the spectrum of a sample of pure abso- 

 lute alcohol. Ordinary methylated spirit gives a very 

 similar spectrum, differing only in the presence of some 

 general absorption, and with a mixture of equal parts of 

 methylated spirit and water the alcohol band was still 

 clearly visible, and only a faint indication of the water 

 band. 



On referring now to the alcoholic solution of ammonia 

 (Fig. 3), it will be seen that the probable explanation of 

 the darkening of the 630 band is owing to the coincidence 

 of the alcohol band with that of the ammonia, so that 

 really the marked difference of the two ammonia spectra 

 is in the absence of the 610 band, and this, we have seen, 

 may be accounted for by one being an aqueous and the 

 other an alcoholic solution. 



Ethylic alcohol giving this definite band, it was a matter 

 of much interest to examine other alcohols belonging to 

 the same series. We found that amylic alcohol (CjHuHO) 

 gave a single visible band (Fig. 5), which in character is 

 like the one given by ethylic alcohol, but differs in 

 position ; it extends from 63S to 630, the centre being 

 634, so that it is decidedly nearer to the red end of the 

 spectrum. 



A sample of amylene (C^Hij) gave also a band in the 

 same position as that of the alcohol, but it differs appa- 

 rently in being broader and less defined at the edges. 



The sample of meth)lic alcohol was not quite pure nor 

 free from colour, but it gave a band quite similar to that 

 of the other two alcohols. Its position is certainly very 

 nearly the same as that of the ethylic alcohol, but as far 

 as our measurements went it was a little nearer the blue, 

 but w'ith our method of measuring hardly discernible. 



It seems — pending further investigations — highly prob- 

 able that this band — and of course there may be others 

 not visible — is common to all the alcohols of the ethylic 

 series, and that its position is a function of the density of 

 the particular alcohol. Apparently however the signi- 

 ficance of this line does not stop here, for in ordinary 

 ether there is a band coincident with this alcohol-band — in 

 fact practically the visible, spectrum produced by alcohol 

 and ether are identical ; but in all cases that we have 

 seen the ether spectrum is clearer and sharper than the 

 alcohol one. We thought it of importance to examine a 

 sample of ether which should be as far as possible 

 rendered pure by ordinary means, especially that it 

 should be free from all traces of moisture : this sample 

 gave a band precisely similar to the band in the ordinary 

 commercial ether. Another sample of ether was satu- 

 rated with water : in this case the ether band was as 

 marked as ever, but the water-band was not visible. 



We have also examined two other bodies which belong 

 to the ethylic series, namely, aldehyde and acetic acid. 

 Both give bands, but they are not so clear or definite as 

 the alcohol or ether bands. Figs. 8 and 9 show these 

 bands. The aldehyde band commences sharply at 62S, 

 but on the other side it shades gradually oft' and ceases at 

 620. The band in acetic acid is very faint, in fact at 

 first, when using the 6-feet tube, we were led to think 

 there was no visible band. 



We also tried a few of the saline ethers, and, as far as 

 our investigations have gone, the ethyl compounds give a 

 band coincident with the alcohol- and ether-band. And 

 the band of the amyl compounds is coincident with that 

 of amylic alcohol. There appears, however, to be this 

 general difference between the bands in the alcohols and 

 those in the corresponding saline ethers, namely, that in 

 the latter the bands are always broader and less distinct ; 

 the saline ethers we have examined are ethyl oxalate, 

 amyl acetate, amyl iodide, and amyl nitrate. 



Passing now to the aromatic series, we find that they 

 give very marked absorption bands. Fig. 9 represents 

 the bands given by benzene ; the spectrum is remarkably 



sharp and clear, quite as clear as the ether spectrum ; the 

 figure is drawn from the spectrum produced by 8 feet of 

 the liquid. The absorption extends as far as 656 ; the 

 first band is from 707 to 698, the second from 609 to 605 ; 

 both are very dark and distinct. The third band extends 

 from 531 to 528, and is very much fainter. 



Toluene, the next higher member of this series, gave 

 also a similar spectrum, and it is equally sharp (Fig. 10). 

 As in the case of the alcohols, with increase of density 

 the bands have moved nearer the red. It will be seen 

 that the band in the red differs in position from the cor- 

 responding benzene band more than either of the other 

 two bands do. 



Cresol, unfortunately at present, we have not been able 

 to examine for want of a sufficient quantity of the pure 

 substance. 



Phenol gives a spectrum very similar to the benzene 

 spectrum ; possibly the band about 610 is somewhat 

 nearer the blue, but beyond this we could see no differ- 

 ence. In the first instance we tried melting the phenol, 

 but afterwards found it far preferable to keep it liquid by 

 the presence of a mere trace of water. 



We looked with much interest to the two following ex- 

 periments, with bodies of this series, namel}', aniline and 

 toluidene, to see how far their constitution might be indi- 

 cated by their spectrum. Figs. 11 and 12 give respec- 

 tively the spectra of these bodies. There is a band in 

 the red in the same position as the toluene band, and in 

 the case of aniline a band agreeing in part with the 606 

 Isenzene band. With toluidene, however, this band was 

 not visible, but probably this arose from its being hidden 

 by general absorption, the liquid used being slightly 

 coloured. However, besides these two bands, both of 

 these amido compounds gave a very clear band from 656 

 to 645, and it is certainly not without interest that this is 

 coincident with one of the bands given by ammonia ; 

 whether any other band coincidences occur between these 

 bodies we cannot say, as in both cases there was sufficient 

 general absorption to hide them even if present. 



Among other liquids we have tried turpentine, which 

 appears to give a definite spectrum. This is shown at 

 Fig. 13. With a thickness of 8 feet of carbon disulphide 

 and a similar thickness of carbon tetrachloride we could 

 see no bands. However, with ^the former liquid it may 

 prove that there is a band in the green, but as far >s we 

 coidd tell this is doubtful. One other experiment, which 

 has some interest, is that the benzene spectrum is 

 unaltered when the liquid is saturated with sulphur. 



Such are the principal observations which we have 

 made up to the present time. As stated at first, we look 

 upon these results as preliminary, and as having to be 

 repeated with more accurate means. Of course we have 

 only dealt with the bands visible under ordinary condi- 

 tions ; still, the above results, as far as they go, have been 

 made with much care, and we think show that most inter- 

 esting relations exist between the chemical composition 

 and constitution of a body and its absorption spectrum. 

 Obviously a far more extended series of observations 

 must be made before any general conclusions of value can 

 be deduced. William J. Russell 



Chemical Laboratory, William Lapraik 



St. Bartholomew's Hospital 



CELLULOID 



THE product of the action of strong nitric acid upon 

 cellulose has of late years met with many applica- 

 tions in the arts. 



When cotton ,wool, linen, paper, or other substance 

 largely consisting of cellulose, is immersed in strong nitric 

 acid, a mixture of two or more nitro-celluloses is pro- 

 duced ; a solution of this mixture in alcohol and ether 

 has been long known as collodion. 



About three or four \-cars ago it was shown that this 



