DESCRIPTION OF SPECIES. 81 



leaflets, form the pedicels by which they are attached, and which are slightly decurrent 

 on the midrib. The nerves start from the pedicel as strong bundles, composed of two 

 nerve-bundles. On passing into the limb of the leaflet the double bundles slowly diverge 

 dichotomously, each branch of the fork being composed of a nerve-bundle formed by 

 two closely approximated nerves which, unless seen under a good lens, appear as single 

 nerves. The two nerves composing each branch then slowly diverge dichotomously. 

 Each is' at first single, but in its further course becomes in turn a bundle of two closely 

 approximated nerves, which slowly diverge and produce a forking again in a similar 

 manner. The dichotomous branching occurs quite often in the lower third of the leaf- 

 lets, but in the upper portions the forking is very rare, and th e nerve-bundles run for long 

 distances parallel to each other. Between the nerves there is a fine granulation which, 

 seen under a lens, is composed of globular prominences or dots which do not seem to be 

 anything but a fine granulation of the epidermis. Sometimes, owing to distortion 

 from pressure, these dot-like elevations are drawn out into little bars which extend 

 from nerve to nerve and look like transverse nerves. 



The remarkable nervation of this plant thus described is that seen with 

 the help of a good lens on well-preserved specimens. To the unassisted 

 eye the nerves seem to issue from the pedicels as single strong nerves, which 

 then repeatedly branch in a dichotomous manner and fill the leaflet. The 

 nervation is seen to be freely branched in the lower part of the leaf, while 

 in the upper part the branches run for long distances parallel to each other. 

 The granulation also is very distinct without the help of a lens, and appears 

 either as dot like prominences or transverse bars from branch to branch. With 

 the help of a lens we find that the apparently sharply-defined single nerves 

 are really composed of two nerve-strands, so closely placed that they appear 

 as one nerve. It is seen also that what appear at the base of the leaflets to be 

 single strong nerves are really nerve-bundles made up of two double nerves, 

 so close together as to give the appearance each of a single strong nerve. In 

 other words, here the four nerve-strands are crowded together in the same 

 way as the two nerve-strands are higher up in the leaflet. Following with 

 the eye one of these strong bundles at the base of the leaflet as it passes 

 higher into the leaflet, we find that the two pairs of nerves slowly diverge, 

 giving what appears to the unaided eye to be a dichotomous forking of a 

 single nerve. Each pair, some distance higher up, forks dichotomously by 

 the very gradual separation of the two nerve-strands which compose it. 

 These two nerve-strands are at first single, so far as can be made out with 

 the lens, but they soon become double, and now in their turn fork, as before 

 6P 



