. Atkins — Oxydases and their Inhibitors in Plant Tissues. 153 



•5. Pale creamy yellow, red-brown markings much the same as in 4, Iris 

 laevigata. 



e. White with yellow at base, and extending up the central region, Iris 

 aureus. 



Perianth leaves of the above types were treated with alcoholic benzidine 

 or a-naphthol, followed by hydrogen peroxide, as advocated by Keeble and 

 Armstrong. In no case was the natural pigment, destroyed by immersion 

 in alcoliol, restored again by removal of the leaf to water. In this respect 

 Iris' differs from many flowers studied by Keeble and Armstrong, and from 

 Rosa rvgosa, as noticed earlier in this paper. Possibly this is due to the lack 

 of a sufficient supply of the chromogen, though absence of the " epidermal 

 oxydase," detected by benzidine, or presence of the characteristic reducing 

 substance of Iris, seems to be a more probable cause, as will be shown. It may 

 here be noted that while a-naphthol is turned to a lilac colour by bundle 

 peroxide, benzidine is turned first blue, then brown, by both bundle and 

 epidermal oxydase. 



After treatment with the above reagents it was seen that, in type 1, the 

 purple flower, the bundle peroxidase was absent, except for traces in the tips 

 of the veins, whereas in type 2, the pale purple, it was well marked. On the 

 other hand, in the purple type the benzidine reaction was very pronounced 

 in tlie veins, and distinct in the other cells, while in the pale purple the 

 brown in the veins and other cells was not so pronounced. The production 

 of tlie purple is accordingly seen to be due to the action of the epidermal 

 peroxidase on the chromogen. In type 3 there was only the faintest reaction 

 with benzidine in the veins, while with a-naphthol there was absolutely none. 

 Since, however, the patches of the purple colour were originally present, it 

 seems that this apparent absence of oxydase in any form is rather due to 

 the presence of the reducing substance of Iris in the white patches, and its 

 diffusion through all the tissues on the death of the protoplasm. 



On examining the yellow flowers it is seen that they all possess the 

 epidermal peroxidase, the pale yellow, type 5, having the greatest quantity, 

 and the white and yellow, type 6, the least. Accordingly, the absence of 

 purple is due to absence of chromogen, which only exists in the portions at 

 the base, which appear red-brown. The bundle peroxidase, too, is plentiful 

 in types 5 and 6, whereas in 4, the deep yellow, it is limited to the uppermost 

 extremities of the veins. Tliis is possibly due to the presence of the reducing 

 substance in their lower portions. 



' Other species of Iris at present under observation recover in part their original pigment. The 

 inhibitor in the petals maj' also be destroyed by hydrogen cyauiiie. 



