464 



Prof. F. Keeble and Dr. E. F. Armstrong. 



[May 29, 



sub-epidermal layer. They must, therefore, as Baur points out, be formed by 

 the G. laburnum- component of the chimera. Examination of the flowers of 

 C. laburnum shows the presence of these red-brown or chocolate patches on 

 the standards and hence they supply Baur with a convincing argument that 

 the sub-epidermal tissues of G. Adami are derived from the laburnum 

 (C. laburnum). 



ISTow microscopic examination of transverse sections across the patches of 

 pigment cells on the standard of a flower of G. Adami shows at once that the 

 cells, of purpureas origin, which overlie the sub-epidermal pigmented cells of 

 laburnum origin, are always more deeply pigmented than are the neighbouring 

 epidermal cells, which have no anthocyan-containing cells immediately 

 beneath them. In other words the deeper pigmentation of the epidermal 

 purpureus cells coincides with the distribution of soluble (anthoeyan) pigment 

 in the underlying and contiguous laburnum cells. In other parts of the 

 petals, the laburnum cells contain only yellow carotin and there the purple 

 colour of the epidermal cells remains faint. 



Since, by the nature of the case, pigment-containing cells are impermeable 

 to their soluble pigments, it seems very probable that the heightened colour 

 of the epidermal cells which are contiguous with the red patches of sub- 

 epidermal cells is due to a migration of oxydase from the sub-epidermal, 

 laburnum cells, to the epidermal, purpureas cells. 



It might perhaps be urged that the paleness of the epidermal cells which 

 overlie the yellow tissues is due to some inhibition of oxydase on the part of 

 the latter. This, however, does not appear to be the case. For as we have 

 shown (loc. cit.) it is possible in other flowers (Primula sinensis) to destroy 

 the inhibitor of oxydase which exists in dominant white flowers by treating 

 the petals with hydrogen cyanide. This reagent, however, when applied to 

 flowers of C. Adami does not exert any influence on the strength of the 

 oxydase reactions subsequently obtained. Hence we conclude that the 

 yellow cells do not exercise an inhibitory effect on the pigment-forming 

 action of the oxydase in the epidermal cells. 



Finally, we would point out that the hypothesis of oxydase migration, not 

 only from sub-epidermal to epidermal cells, but also from the vascular 

 tissues to the epidermis, is in conformity with what is known with respect to 

 the variable behaviour of many florists' flowers. Thus, in certain seasons 

 wallflowers, and in all seasons stocks, tulips, etc., produce flowers which 

 " break." For example a pure strain of self-coloured wallflowers may 

 become blotched or flaked. There is evidence in the case of wallflowers that 

 this breaking has no relation with genetical constitution, for a strain which 

 "breaks" badly one year may behave in a perfectly normal manner and 



