The results are as follows : —
1. The vascular bundles from the flower, the vegetative bud and the tendril (whether simple or branched) in all the species examined unite with bundles of the inner ring of the stem at the node at which these organs arise,
2. Three bundles enter the stem from the leaf-stalk. The median passes straight down through the internode without uniting with any other bundle. The two lateral bundles unite with the adjacent bundles of the outer ring of the stem, one on each side.
3. At each node one of the bundles of the inner ring stops and its place is taken by a bundle from the outer ring. The bundles of the inner ring anastomose at the nodes.
4. The base of the tendril in all species with branched tendrils is organised radially the bundles being arranged in a ring and there being no groove, such as is present in the petiole. This is also true of the simple tendrils of Momordica echinata.
5. The tendrils in all the species with simple tendrils except Momordica charantia are organised on the type of the petiole with horse-shoe arrangement of bundles and a groove on one side through- out their entire length. In Momordica charantia the tendril is radially organised at the base and dorsiventrally above.
6. In all species the upper part of the simple tendrils and the arms of the branched tendrils throughout their entire length are organised on the type of the petiole. The main arm of the ten- dril which is a continuation of the basal part in Luffa pentandra is, however, radially organised. The upper part of this arm was un- fortunately not examined. It would probably possess a dorsiventral structure.
The following conclusions are drawn from the above results : —
1. The vascular connections and the structure of the basal part of the branched tendrils show that they are homologous with the ordinary shoots.
2. The arms of the branched tendrils and the upper part of the simple tendrils are homologous with the ordinary leaves.
3. The simple tendrils have been derived from the branched tendrils and therefore homologous with them. Species of Momor* dica show the transition in M. echinata possessing both simple and branched tendrils and in M. charantia possessing only simple ten- drils the base of the tendril is always radially organised. In other species with simple tendrils the modifications has gone still further and even the basal part has become petiole-like.
It may be stated that conclusions one and two are in general agreement with the conclusions of Muller and Tondera, but according
1968—34
