462 



Prof. F. Keeble and Dr. E. F. Armstrong. [May 29, 



also in the epidermis of C. Adami — both in the buff-coloured typical 

 C. Adami flowers and in the purple flowers borne on C. Adami; but that it 

 is absent from the epidermis of C. laburnum and from that of the yellow 

 flowers of C. Adami. 



Although, as just indicated, the petals of C. purpureus and C. Adami give 

 similar epidermal oxydase reactions they differ from one another with 

 respect to the mode of reaction of their deeper tissues. For whereas the 

 petals of both G. purpureus and of purple-flowered branches of C. Adami 

 give, with benzidine alone, a well-marked reaction for bundle oxydase, the 

 vascular bundles of the buff petals of G. Adami, those of the yellow petals of 

 the laburnum-like branches borne on the graft-hybrid and those of the 

 petals of G. laburnum give no direct oxydase reaction. In the three last- 

 mentioned cases it is only after the addition of hydrogen peroxide that the 

 bundle oxydase reaction is developed. Whence it is to be concluded that 

 oxydase is present in the tissues which accompany the veins of the petals of 

 C. purpureus and purple-flowered branches of C. Adami, and that peroxydase 

 is present in the corresponding tissues of the buff and yellow flowers of 

 C. Adami, and in those of the common laburnum. 



A closer examination of the brown bundle oxydase reaction produced by 

 benzidine shows that it is most marked in C. -purpureus, fairly marked in the 

 purple-flowered branches of G. Adami, and that it does not occur in the buff 

 flowers of typically Adami branches. This is in accord with the differences in 

 pigmentation of the three flowers. For in the living flowers of G. purpureus 

 lines of darker purple (magenta) accompany the veins of the petals. Similar 

 though less well-marked magenta lines run with the veins of the purple 

 flowers of G. Adami, whereas there are no such lines of magenta colour along 

 the veins of the buff G. Adami petals. We conclude, therefore, that in 

 flowers the oxydases which accompany the veins (bundle oxydases) co-operate 

 with those contained in the epidermal cells to produce the pigment of those 

 cells. If this view be accepted, the failure of G. Adami, although its 

 epidermis is that of C. purpureus, to produce flowers of as deep a purple 

 as those of C. purpureus, is intelligible ; for the vascular system of C. Adami 

 is derived from the C. laburnum component, and in that species the vascular 

 tissues of the flower — though they contain peroxydase — possess no direct 

 oxydase, and it is therefore reasonable to suppose that they cannot contribute 

 to pigment formation in the epidermis. 



If, further, we assume, and it is an assumption for which grounds are not 

 altogether lacking, that oxydases may travel through the vascular tissues, 

 it is possible to account for the fact that the purple pigment of the flowers 

 of G. Adami is generally paler than that of the flowers of C. purpureus. 



