72 Scientific Proceedings^ Royal Dublin Society. 



Moseley investigated jantliiuin, which gives three bands in dilute 

 neutral solution, and one in an acid solution about D.' The bands were 

 drawn, described, and measured. 



Krukenburg also found a blue colour growing on moist fibrin.^ It 

 dissolved in alcohol with a violet-blue colour, and gave a band about D, 

 which, he saj's, at once resembles and yet differs from that of certain aniline 

 dyes. 



An indigo-blue colouring-matter was obtained from water bacteria by 

 Heiurich Claessen in 1890.' 



Moliseh, working with Rhodohactevium capsulatum, extracted from it 

 by alcohol a green colouring-matter, " hacteriochlorin," and from the residual 

 red-brown mass of bacteria he got a carmine-red colour " bacteriopurpiirin " 

 by extraction with carbon disulphide.* Bacteriochlorin has a band about D, 

 from X 616 to X 565 which is typical of, and due to, the bacteriochlorin ; the 

 spectrum begins at X 650 and ends at X 525. The bacteriopurpurin which 

 appears to give the colour to the organism has two absorption-bands, the first 

 about X 585-665, and the second from X 540-515, while yet a third is 

 supposed to exist. Moliseh describes the separation of the two colouring 

 matters from the growth on a microscopic scale by alcohol, and by carbon 

 disulphide, and by alcohol and olive oil. 



Similar experiments repeated with Bacillus violaceus failed to show any 

 separation of colours whatever. 



I should say that the violet colours extracted by Schneider from JBacillus 

 violaceus and Bacillus ianthinus may be, as he says, identical ; but, like Lecoq 

 de Boisbaudran's colour, they differ from the colouring-matter which is the 

 subject of this paper, in that they completely transmit the less refrangible 

 red. 



The colouring-matter janthinin is evidently different from either 

 Schneider's or mine, for it is fluorescent and has three absorption-bands. 



The measurements of the absorption-band given by bacteriochlorin 

 and also the length of its transmitted spectrum show it to be different 

 from the colouring-matter here described ; while bacteriopurpurin has two, 

 if not three, bands, the first of which resembles that of janthinin. 



The colouring-matter of Bacillus violaceus diff'ers from the dyes in the 

 absence of absorption-bands in the ultra-violet, and from the plant and 

 animal colouring-matters because of the continuous absorption of the rays 



Quart. Joiirn. Micros, xvii, pp. 1-23. 1877. 

 '■ Physiologische Studien., 5 Abtheil., pp. 43-47. 

 ' Centralblatt fiir Bakteriologie. 1890. 



Die Purpurbacterien. Moliseh. Jena, 1907. 



