IV] BUDS AND BRANCHING 73 



(V) In other cases buds may be formed upon the roots: but this is less 

 common. In Platyceriiini and in Asplenium esculentum it has been shown 

 that the apex of a root may be directly transformed into a leafy bud (64, 

 p. 63, where the literature is quoted). Similar buds appear upon the roots 

 of Ophioglossum viilgatum, and other species: but these have been traced in 

 origin to one of the segments cut off from the apical cell itself (Rostowzew, 

 (46)). By such means an effective vegetative propagation is secured. 



Under these five headings diverse means of increase of the shoot in Ferns 

 have been included. Those first named appear in definite positions, and are 

 produced by primary development, distally. This would suggest that their 

 origin was primitive. Those mentioned under (IV) are less definite in position, 

 and are often late in their origin, so that they bear no direct relation to the 

 apex of the shoot. Those under (V) have evidently no relation whatever to 

 the terminal bud. There may, however, be a question in the case of (II) and 

 (III) what relation, if any, the buds so produced bear to the apex of the main 

 shoot: that is, how far they may be held to be of the nature of distal branch- 

 ings of it, diverted into a lateral position and delayed in their origin. In the 

 case of (I) the distal origin of the true dichotomies or equal forkings is clear. 

 The genesis of the second apex has been traced in Polypodmm, by Klein (41), 

 to a segment of the apical cell. The original apical cone thus gives rise to 

 two approximately equal cones, each of which acts as an equivalent apex. 

 This may take place independently of any leaf. In the adult state the axis 

 may be seen to fork at a point some distance above the insertion of the 

 nearest leaf (Fig. 61). But frequently a leaf may appear to be related to 

 the forking, and this has been designated by Velenovsky the "angular 

 leaf" (50). He states that the leaf nearest ^o the dichotomy halves its- angle, 

 and is to some extent a constant character. That this relation is not close 

 or constant is shown by Velenovsky's own drawings. Further, the orienta- 

 tion of the nearest leaf relatively to the shanks of the dichotomy is variable. 

 A group of drawings from transverse sections through the forking axes of 

 certain allied Ferns is represented in Fig. 68, i-vi. They show that some- 

 times there may be one leaf-trace related to the forking, sometimes two: 

 and it has been seen that sometimes there is none. It appears also that the 

 orientation of the leaf-traces varies relatively to the shanks of the dichotomy, 

 and that this variation may be found even in the same species (Fig. 68, ii 

 and v, are both from D.pinnatd). These facts suggest that there is no definite 

 relation of a leaf to the forking of the axis which would justify its recognition 

 as an "angular leaf" in the sense of Velenovsky. It appears rather that the 

 forking is purely an axial phenomenon, and that it may or may not be 

 associated with the insertion of one or more leaves. 



It is interesting to see that this inconstancy holds also in the case of an 

 ancient fossil. For in Rachiopteris {Botryopteris) cylindrica it has been shown 



