169 



each of which may branch several times before terminating in the 

 ultimate twigs upon which the flowers and leaves appear. 



The flowers are arranged in umbels, they are terminal, and in 

 each case they end the growth of the twig. 



The leaves of each branchlet are in one, two, or rarely three 

 interrupted groups, from four to six leaves each. The variation 

 in the number of groups of leaves is correlated with the differ- 

 ences in the habitat of the plants, and varies with the time of 

 year. Uniformly the shrubs growing among the balsams have at 

 least one or two groups of leaves more than those away from the 

 forest, or, in other words, the leaves on such shrubs may live one 

 year or more longer than those on the open growing plant. In 

 the axils of the leaves are buds, each or all of which may develop, 

 or, as will be shown later, they may remain latent. If two or 

 more buds grow, the respective branches which they form make 

 a more or less wide angle with the parent branch ; if, on the other 

 hand as perhaps most frequently happens, one only develops, the 

 new branch turns upward and takes a direction parallel to that 

 of the parent, pushing the flower cluster, now become the fruit, 

 to one side. Sometimes it happens that none of the axillary buds 

 develop, in which case of course the branch either retains its in- 

 tegrity, or if the terminal bud is a flower-bud, the branchlet dies. 

 Now it happens that if a leaf-bud terminates a branchlet, the de- 

 velopment of the axillary buds appears to take place much less 

 frequently than when the terminal bud is destined to give rise to 

 flowers. It thus happens that profuseness in flowering brings 

 about wealth of branching, and since the plants in the open 

 blossom more than in the forest, that the former are also more 

 abundantly branched. 



A peculiarity which the rhododendron grown in the open shows 

 as regards the branching, accounts also for the rounded outline 

 characteristic of such forms. This shape is so noticeable that, as 

 suggested by a visitor to the mountain, the shrubs appear as if 

 they were cropped by grazing cattle. So far as I have observed, 

 however, they are not subject to their attack. Rather the sym- 

 metrical form is occasioned by the nearly equal and similar de- 

 velopment of its constituent branches. 



