328 Schively — Contributions to the Life History of 



ens for a distance of an inch or two behiud the tip, some- 

 times becoming etiolated ; tips of the secondary branches 

 were often beautifully curved downward. 



Before discussing the peculiarities of growth manifested 

 by these runners, it may be well to describe the flower 

 borne by them. The writer does not consider that there is 

 a raceme-like arrangement of the inflorescence, although 

 there seems to be ; the plan of inflorescence is somewhat 

 puzzling. Below each flower is a pair of stipules, never a 

 bract. It is easier to explain the arrangement by referring 

 first to one of the runners which we find arising from the 

 axil of the simple leaves. Now, this shoot grows to a 

 great length, producing leaves with stipules at their bases, 

 frequently each leaf is reduced to a pair of stipules. 

 Each new axil possesses the possibilities of one or more 

 buds, each of which may be similar to the parent. In this 

 part of the plant the reduction of the leaves becomes the 

 rule, and in the axil of the stipules appears a small branch 

 terminating in a flower-bud. Here may arise two— even 

 three— branches (rarely at the same time), each bearing 

 flower-buds. The main shoot may increase in length in- 

 definitely, and likewise branches may develop from the 

 apical buds hidden between the stipules at the base of the 

 flower-bearing branch. It may be said that the flowers are 

 solitary ; and when occasionally two are found terminating 

 a shoot, it may be inferred that in the axils of that special 

 pair of stipules, two flower-bearing branches arose at the 

 same time. This condition of affairs is more easily studied 

 when the fruit begins to form. It often happens in the 

 subterranean flowers that the buds are very close together. 

 This is because great reduction in length of the shoot has 

 taken place; but the stipules may usually be found even 

 then, and are always readily seen upon the longer cotyledo- 

 nary branches. 



The calyx is four-parted, but the teeth are more rounded 

 than in either of those described. As yet no evidence of 

 an original five-parted calyx has been seen. In this case 



