ON ABSORPTION SPECTRA AND CHEMICAL CONSTITUTION. 133 



modifications would produce an appreciable effect on the spectra 

 of compounds of low molecular weight. Corydaline, tetrahydrober- 

 berine, and their derivatives afford instances in which the replacement 

 of 2(OCH3) by CHoOa does not markedly alter the spectra, and nar- 

 cotine and hydrastine furnish an example in which the introduction 

 of an additional methoxyl group is unaccompanied by any considerable 

 effect. 



The case of cinchotenine and cinchonine may be quoted as another 

 instance. In cinchotenine the side chain -CH : CHj of cinchonine is 

 oxidised to a carboxyl group without the spectra being appreciably 

 affected. The resemblance between the two series of spectra is so close 

 that it would at once sugorest a near structural relation of the substances, 

 even if we knew nothing of their chemistry. 



The reduction of closed chain compounds is accompanied by a com- 

 plete change in the character of the spectra.' Good illustrations of this 

 are afforded by the widely different spectra of berbei-ine (fig. 9) and tetra- 

 hydroberberine (fig. 7), dehydrocorydaline (fig. 8) and coiydaline (fig. 6), 

 papaverine and tetrahydropapaverine (figs. 2 and 3). There are, 

 however, cases in which partial reduction produces very little change. 

 Hydroquinine, C20H26O2N2) is unquestionably very closely related to 

 quinine, C20H24O2N2, from which its formula only differs in containing 

 two more atoms of hydrogen. The difference between the spectra of the 

 two substances is hardly perceptible, and it is highly probable, therefore, 

 that the addition of the two atoms of hydrogen is unaccompanied by any 

 important change of structure. The change probably consists in the 

 reduction of the side chain. 



Prom the results of the examination of more than thirty alkaloids, it 

 may be laid down as a general rule that those which agree closely in 

 structure give similar absorption curves, whilst those which differ in 

 essential points of structure give dissimilar curves. 



This principle has already been recognised and applied in previous 

 investigations, particularly in the study of coloured substances and dyes, 

 and it is probably capable of extended application in the case of the alka- 

 loids, as most of these compounds have a high molecular weight, and 

 changes may be effected in their molecules without alteration of the 

 spectra which, in substances of lower molecular weight, would be attended 

 by wide differences. The essential identity of constitution subsisting 

 between two alkaloids can often be detected by the spectroscope in spite 

 of differences of structure. Cinchonine and cinchotenine give practi- 

 cally identical curves, whereas styrolene and benzoic acid, which differ in 

 the same way, give very different curves. If, therefore, an alkaloid 

 of unknown constitution is found to give spectra closely resembling 

 those of an alkaloid of known constitution, it may with great pro- 

 bability be inferred that the two only differ in the details of their 

 structure. 



The systematic study of absorption spectra is of real practical value 

 in the investigation of the alkaloids, and may often be the means of saving 

 much time and labour in their chemical investigation, especially in 

 dealing with a large number of closely related compounds. 



' Hartley, CJiem. Soc. Trans., 1885, 47, 691, and Phil Trans., 1885; also Hartley 

 and Dobbie, Chem. Soc. Trails., 1900, 77, 846. 



