from progenitors more highly organized in this respect. FUMARIACEÆ.—_Corydalis claviculata_.—According to Mohl (p. 43), the extremities of the branched stem, as well as the leaves, are converted into tendrils. In the specimens examined by me all the tendrils were certainly foliar, and it is hardly credible that the same plant should produce tendrils of a widely different homological nature. Nevertheless, from this statement by Mohl, I have ranked this species amongst the tendril-bearers; if classed exclusively by its foliar tendrils, it would be doubtful whether it ought not to have been placed amongst the leaf-climbers, with its allies, _Fumaria_ and _Adlumia_. A large majority of its so-called tendrils still bear leaflets, though excessively reduced in size; but some few of them may properly be designated as tendrils, for they are completely destitute of laminæ or blades. Consequently, we here behold a plant in an actual state of transition from a leaf-climber to a tendril-bearer. Whilst the plant is rather young, only the outer leaves, but when full-grown all the leaves, have their extremities converted into more or less perfect tendrils. I have examined specimens from one locality alone, viz. Hampshire; and it is not improbable that plants growing under different conditions might have their leaves a little more or less changed into true tendrils. Whilst the plant is quite young, the first-formed leaves are not modified in any way, but those next formed have their terminal leaflets reduced in size, and soon all the leaves assume the structure represented in the following drawing. This leaf bore nine leaflets; the lower ones being much subdivided. The terminal portion of the petiole, about 1½ inch in length (above the leaflet _f_), is thinner and more elongated than the lower part, and may be considered as the tendril. The leaflets borne by this part are greatly reduced in size, being, on an average, about the tenth of an inch in length and very narrow; one small leaflet measured one-twelfth of an inch in length and one-seventy-fifth in breadth (2.116 mm. and 0.339 mm.), so that it was almost microscopically minute. All the reduced leaflets have branching nerves, and terminate in little spines, like those of the fully developed leaflets. Every gradation could be traced, until we come to branchlets (as _a_ and _d_ in the figure) which show no vestige of a lamina or blade. Occasionally all the terminal branchlets of the petiole are in this condition, and we then have a true tendril. [Picture: Fig. 8. Corydalis claviculata. Leaf-tendril of natural size]