CORRELATION IN THE INFLORESCENCE OF CELASTRUS 



SCANDENS. 



BY J. AUTHUR HARRIS. 



Data supplementary to those presented in the preceding 

 papers are to be found in my notes on the inflorescence of 

 Celastrus scandens. Since I have not now the time for col- 

 lecting sufficiently numerous and varied series of material 

 for a comprehensive study of variation and correlation in 

 the inflorescence it seems best to publish the rather diverse 

 notes which I have under separate titles which shall indi- 

 cate the species examined as well as the subject considered. 



In the fall of 1905, 38G clusters of the fruit of C. scan- 

 dens were collected in the woods a little north of Jefferson 

 Barracks, Mo. All of these were taken from one clump of 

 vines. Whether these were parts of the same individ- 

 ual or whether more than one plant was involved 

 was not definitely determined at the time. The fruits were 

 gathered primarily to secure a variation curve for number of 

 seeds developing per fruit, and the problems connected with 

 the inflorescence presented themselves later. At that time, 

 too, like most biologists I was unacquainted with the impor- 

 tance of knowing definitely the number of individuals from 

 which material for quantitative investigation is taken. Pre- 

 sumably only a single plant was involved, but we do not 

 know this to be true. And so far as I am able to see this 

 unfortunate omission in our observations docs not affect in 

 any way the validity of any conclusions drawn in this 

 paper, although it has necessitated the neglecting of certain 

 other points which might have been considered had our 

 knowletlge of this point been definite. 



These fruits are very easily worked with. A transversa 

 section quickly made with a razor, shows at a glance the 

 number of seeds which have developed. The normal number of 

 cells in this collection of material was three, and two seeds 

 sometimes developed in each locule. The maximum normal 

 number of seeds was assumed to be six. In sectioning 



(116) 



