MancJicster Memoirs, Vol. liv. (1910), No. 18- 5 



(a form of bud variation which is not infrequent in some 

 groups of plants) but that the flowers are potentially red, 

 and that a certain factor, or several factors, probably 

 dependent on the metabolic processes of the plant, 

 determine the formation or the non-formation of antho- 

 cyanin. 



That we have no segregation of characters is borne 

 out by the observation of the offspring resulting from self- 

 fertilising the flowers of different colours, while the return 

 of the yellow-coloured flowers during the late autumn 

 seems to point to an external controlling factor. 



It is interesting in this connection to remember the 

 experiments by Rawson on the effect of light of different 

 refrangibility on the colour of Nasturtium flowers.* 



A variety of T7'opccolmu viimis and inajus is now on 

 the market, called " Chamaeleon," and 1 am at present 

 engaged in comparing the change in colour of this variety 

 with the hybrid described above. I am also continuing 

 the experiments with the later generations of the plants 

 referred to, so that the present communication must be 

 taken to be of a preliminary nature only. 



* British Associatio)i Report, South Africa, 1905. 



