The Nature and Origin of Stijjules. 43 



place, though the lateral leaf bundles curve in toward the median 

 one at but a short distance from the stem. This arrangement of 

 the bundles is probably secondary in these forms for the purpose 

 of giving a firmer support to the leaf b}>- an axial concentration 

 of the vascular tissue in the sheath and a corresponding thicken- 

 ing of the surrounding tissues, a firmer support than could be 

 given by only three bundles if they did not converge till they ap- 

 proached the point of their entering the petiole. The venation of 

 the stipules is also peculiar. In Fragaria Virginiana Duchesne 

 (fig. 68), there is a single strong bundle running out into the free 

 tip of the stipule. From this are sent out one or two weak veins 

 above, and below there is a faint vascular network confined mostly 

 to the region of the tip and extending in a long curve toward the 

 outer portion of the base, where it gradually fades out without 

 forming any connection with other vascular tissue below. This 

 condition seems to indicate a former basal connection of these 

 stipular bundles, either with the lateral bundle of the leaf or pos- 

 sibly with those of the stem, forming an additional leaf-trace bun- 

 dle distributed to the stipules only. The former case is far more 

 likely. A probable explanation of this degeneration of the basal 

 stipular bundles can be found by a consideration of the conditions 

 of the environment. All the leaves being basal, the stipules are 

 clustered together and are supported by one another and by the 

 surrounding soil. They are more or less fleshy, destitute of chlo- 

 rophyll, and in their moist surroundings loss of water by evapo- 

 ration is comparatively slight. All these circumstances lessen 

 the necessity of the supply of fresh sap. The rapidly conducting 

 vascular tissue has come into disuse, and its degeneration and 

 disappearance is the natural consequence. The same arrangement 

 in forms with leafy stems is not so readily explainable except by 

 the supposition that the arrangement is ancestral. This seems 

 rather evident in the case of Agrimonia striata Michx. (figs. 53- 

 56), where the same condition of the bundles occurs, for the 

 earliest leaves representing the ancestral forms develop under the 

 same conditions as the adult leaves of Fragaria. But in Rosa it 

 would be by no means clear did we not have such intermediate 

 types as Agrimonia. Rosa humilis Marsh, (fig. 69) maybe taken 

 as typical of the genus. The venation of the tip of the stipules 

 is nearly like that in Fragaria, but with a little larger develop- 

 ment above the main bundle. The vascular network below is 



