Maidenhair 37 



Development of a leaflet into a branch is brought about in 

 the following way. An incision forms between the two vein- 

 branches formed by the first forking of the vein that enters the 

 leaflet, and cuts the leaflet into two leaflets. One of the two thus 

 formed remains, at least temporarily, undivided. The base of 

 the other leaflet becomes elongate, attenuate, and dark-colored, 

 forming the beginning of a rachis, and this leaflet, which may be 

 called the apical one, divides into two as the original leaflet divided, 

 the incision occurring, as in the latter, between the two vein- 

 branches formed by the first forking of the vein that enters the 

 leaflet. The rest is repetition, one of every two leaflets formed 

 remaining undivided, and the other becoming transformed at 

 base into a continuation of the rachis and dividing above into 

 two leaflets. Which of the two leaflets remains undivided de- 

 pends in each case upon which of the two remained undivided in 

 the preceding one: if the left leaflet in the one case, the right 

 leaflet remains undivided in the following case, and vice versa. 

 The leaflets remaining on the sides of the branch are thus ren- 

 dered alternate.* 



The apical leaflets are cuneate. Most of the leaflets left on 

 the sides of the rachis lengthen at right angles to the rachis, be- 

 coming more or less oblong, although the lower ones, which may 

 lengthen in the opposite direction also, are apt to become more or 

 less cuneate. 



In the earliest leaves the unilateral midveins of the leaflets 

 are little developed, sometimes barely developed beyond their 

 superior basal branches. As a result, the venation at this stage 

 of development can scarcely be called anything but flabellate. 



* It will be noted that the development of leaflets into branches is here essentially 

 the same as in certain leaves with flabellate venation. See pages 21-22. 



