HYBRID RHODODENDRONS. 



241 



ancestor, retained either in the corolla as a delicate tint, such as 

 of pink, or in the anthers alone, may suddenly become very 

 pronounced. 



With regard to flowers, prepotency in the transmission 

 colour is usually, and I may say generally, recognised by florists 

 as being correlated with its intensity : so that, if they wish to 

 " improve" a flower, they select the best coloured as the male 

 parent ; but as far as the East Indian Rhododendrons herein 

 described are concerned, it sometimes happens that a paler tint, 

 such as a primrose-yellow or a shade of pink, is retained in the 

 offspring, notwithstanding the fact of one of the parents being of 

 an intense golden yellow or crimson, respectively. 



With regard to colours, it may be remarked here that they 

 are all reducible to two, yellow and rose-red. The former is pro- 

 duced by the presence of yellow granules scattered within the 

 cells of the epidermis or underlying tissue ; while the reds are 

 due to various degrees of concentration of a coloured fluid both in 

 individual cells as well as by superposition of cells containing 

 the rose-coloured fluid. The buffs or orange-colours are due to 

 combinations of the pink fluid with yellow granules, either in 

 the same cell, as occurs in some epidermides, or in adjacent cells, 

 as occurs in orange-coloured anthers examined. If there be a pink 

 throat with a yellow or orange border to the corolla, this is due 

 to the epidermal cells containing a more concentrated solution 

 of the pink fluid. 



Another point worth noticing, especially from a practical 

 point of view, is that better results as to the production of colours 

 occurred by using true species as the male parent, rather than a 

 cross. Hence it will be observed that in the following cases of 

 crosses a true species was much more often employed than a cross 

 as the staminate parent. The same rule, however, holds good 

 when a true species is used as the seed-bearing plant and a 

 cross as the male, rather than when both parents are crosses. 

 Nevertheless, some excellent results were obtained when both 

 parents were crosses. 



With regard to the forms of the corolla of the hybrid or cross, 

 this is by no means necessarily correlated to the size and shape 

 of that of either one parent rather than the other ; but it depends 

 upon some inherent prepotency in the plant. Thus the small- 

 flowered B. multicolor, and its variety Curtisii, invariably brings 



