330 Scientific Proceedings^ Royal Dublin Society. 



scope, hence the possibility has to be examined that they owe their colour to 

 the same pigment as do the red fruits, but in a greater concentration. There 

 is normally a liglit chocolate or brown-yellow tint in aqueous extracts of 

 this type, but this is too faint to have any importance in affecting the colour 

 of the fruit. 



The Peroxidase Reactions of Oopsiciim Fruits. 



At the outset it was imagined that the chocolate colour miglit be due to 

 an optical combination of red or yellow plastids with a cell-sap pigment as 

 in wall flowers (Keeble, Armstrong, and Jones, 1913), or certain forms of 

 Iris (Atkins, 1913). As the presence of antliocyan pigments has been 

 supposed to be associated with oxidases, it seemed advisable to study the 

 behaviour of sections towards benzidine and a-naphthol. These were 

 applied as before described. It was found that the nuclei were specially 

 deeply stained in cells wliich showed the reaction. Frequently the benzidine 

 reaction resulted in the deposition of groups of acicular crystals of the 

 oxidation product within the cells. With this reagent the thickened walls of 

 the epidermis became yellow, whereas the cellulose walls of the sub-epidermal 

 cells remained colourless. In some types of fruit there was no reaction at all 

 in these deeper cells. Taking this in conjunction with observations on the 

 behaviour of sclerenchyraatous walls, as in /m germanica, CatcdiM hignonioides, 

 etc. (Atkius, 1913), it appears probable that oxidases are concerned in the 

 production of modifications in the cell-wall. With a-naphthol the cells are 

 coloured purple, or lavender ; but the walls remain perfectly colourless, thus 

 permittiug the very ready examination of tlie pits as before mentioned. 



In the following table (p. 331) are recorded the results of some peroxidase 

 tests. The reactions refer to the epidermis only, as in many cases no change 

 took place in the deeper cells. This failure to react was observed both in 

 chocolate and red fruits. The times indicated are those which elapsed after 

 the addition of hydrogen peroxide. Tlie sections had previously been in 

 aqueous alcoholic solutions of the reagents till thoroughly impregnated. 

 V. R. denotes a very rapid action, li a visible reaction inside ten minutes, 

 + a reaction inside forty minutes. 



It appears from the table that the benzidine reaction is more rapidly 

 brought about than is that with a-naphthol. In addition it may be concluded 

 tiiat the peroxidase is most active in the unripe fruits, and so may be 

 connected with the changes which occur during ripening. Keeble and 

 Armstrong (1912) found that treatment of flowers, which failed to show 

 oxidase reactions, with 0'2 per cent, hydrogen cyanide was effective in 

 removing an inhibitor. For when carefully washed after treatment 

 with this reagent, the tests gave positive results. Similar procedure with 



