DISTRIBUTION OF LEAVES ON THE STEM. 127 



liko, the branchlcts take a horizontal position and the peti- 

 oles a quarter twist, which gives full exposure of the upper 

 fuce of all the leaves to the light. The and , with dimin- 

 ished divergence, increase the number of ranks ; the | and all 

 be}'ond, with mean divergence of successive leaves, effect a more 

 thorough distribution, but with less and less angular distance 

 between the vertical ranks. 



242 a . The helix or primitive spiral upon which the leaves 

 successively originate ascends, sometimes from left to right, 

 sometimes from right to left, 1 commonly without change on the 

 same axis, and prevailingly uniform in the same species ; but 

 occasionally both directions occur in the same individual. The 

 earliest leaves of a stem or branch, or the last, are often on a 

 different order from the rest ; or (as already stated) the spiral 

 change into the cyclical, or vice versa. 



243. The relation of the phyllotaxy of a branch to the leaf 

 from the axil of which the branch springs is somewhat various. 

 But in Dicotyledons, the first leaf or the first pair of the branch is 

 mostly transverse ; that is, the first leaves of the branch stand to 

 the right or left of the subtending leaf. In Monocotyledons, the 

 first branch-leaf is usually parallel to and facing the subtending 

 leaf, as shown in Fig. 304. 



244. When the internodes are considerably lengthened, the 

 normal superposition of leaves is not rarely obscured by torsion 

 of the axis : indeed, this may equally occur in short internodes, 

 sometimes irregularly or in opposite directions, sometimes uni- 

 formly in one. Thus, in Pandanus utilis, or Screw-Pine, of 

 tristichous arrangement, the three compact vertical ranks be- 

 come strongly spiral by a continuous torsion of the axis. The 

 later leaves of Baptisia perfoliata, which are normally distichous, 

 become one-ranked by an alternate twist, right and left, of the 

 successive internodes. 



245. When the internodes are short, so that the leaves approx- 

 imate or overlap, it is difficult or impossible to trace the suc- 

 cession of the leaves on the primitive spiral, but it is easy to 

 see which are superposed. The particular phyllotaxy may then 

 be determined by counting the vertical ranks, which gives the 

 denominator of the fraction. But in compact arrangements 

 these vertical ranks are commonly less manifest than certain 

 oblique ranks, which are seen to wind round the axis in oppo- 

 site directions. (See -Fig. 245, 246.) These are termed second- 

 ary spirals^ also by some parastichies. These oblique spiral 



1 That is, of the observer and as seen from without. See p. 51, foot-note. 



