124 



Prof. Keeble, Dr. Armstrong, and Mr. Jones. 



presence of glycine it is converted into a yellow substance which becomes 

 orange and finally orange red. A red insoluble deposit, which separates out, 

 forms an orange-red solution in alcohol. In the tintometer we find for 

 a ^-ineh cell: alcoholic extract, 3'5 red + l"5 yellow; aqueous solution, 

 2*5 red + 5 yellow. 



Finally, aesculin (from the horse-chestnut) gives a yellow precipitate, and 

 aucubin (from the red berries of Aucuba japonica), a black precipitate under 

 the conditions described. 



The property of colour formation from a glucoside and an amino-acid seems 

 to be a very general one, though we are unable to say whether the mechanism 

 is in each case the same as we have postulated for arbutin, namely, oxidation 

 of the phenol to a quinone, formation of a quinohydrone and interaction of 

 this with ammonia to form a coloured salt. 



In any case, Chodat's discovery of the resolution of amino-acid into 

 formaldehyde and ammonia is obviously of fundamental importance. The 

 ammonia may serve to provide the alkaline conditions so favourable for 

 oxidation and it may react directly to form amino-compounds. The 

 formaldehyde may take part in all manner of condensations leading to the 

 production of complex substances. 



Section 3. — The Biochemistry of Mendelian Colour Characters. 



Of the various artificial chromogens which serve to determine the presence 

 of oxydase in flowers, benzidine behaves most like the natural chromogens. 

 For example, most artificial chromogens, a-naphthol, guaiacol, etc., serve well 

 enough to indicate the presence of oxydase in the vascular tissues (veins), 

 but they do not react as a rule with the oxydase contained in the epidermal 

 cells, whereas benzidine gives uniformly good reactions with both epidermal 

 and bundle oxydase (see Part I). Again, just as the reducing agents present 

 in petals may reduce the anthocyan pigments to a colourless state, so these 

 same agents reduce and decolorise the blue oxidation product of the 

 interaction of plant oxydase and benzidine. 



Inasmuch as benzidine has proved to be of considerable value for the 

 investigation of plant oxydases, it may be useful to preface this section with 

 a brief account of what is known of the oxidation products of benzidine, 

 which are of an unusually complex character. 



Willstatter and Kalb* have shown that the first oxidation product of 

 benzidine (NHaCeH^CeH^H^) is probably the reddish-brown dipheno- 

 quinone di-imine NH:C 6 H4:C 6 H4:NH. On further more drastic oxidation, 

 two molecules of this substance unite to form the yellowish-red diaminoazo- 



* < Ber.,' 1905, vol. 38, p. 1232. 



