168 



sufficient food as one or two facts seem to indicate. In the first 

 place, branches evidently past their prime and relatively long, 

 bear as a rule undersized leaves, and, as given previously, on such 

 branches, a greater per cent, of flower buds than of leaf buds, or, 

 than of flower and leaf buds, are found to undergo development. 

 But, secondly, if for any cause the flower bud is killed, as by late 

 spring frosts, it invariably happens that one and generally more 

 of the axillary buds grow into as many branchlets. 



Finally it is conceivable that the death of a sufficient number 

 of twigs, by over-production of flowers, might in the end cause 



J 





MM? 





— C* y \ 



1M 



FlG. 4. Fruits of Rhododendron Catawbiense of the present and the past year 

 are shown ; also flower and leaf buds ; and leaves of the past and the present year. 

 The leaves are partially removed to expose the flower buds. 



the death of the entire plant. This however does not occur, as 

 will presently appear. 



When the rhododendron plant has reached a certain size, or 

 more accurately a certain stage in its life history, it begins to put 

 out branches, or suckers from the level of the ground or near it. 

 Precisely what the stage in the development of the shrub is, when 

 this takes place, is not easy to determine. It is probably asso- 

 ciated with the over-production of flowers, and the consequent 

 suppression of the development of leaf buds and is likely therefore 

 concerned with questions of nutrition. 



The course of the development of the suckers, or secondary 



