In 1847 we collected some wild parsnip seed from the top of the Cotteswolds, where this is among the most frequent of weeds. This seed, after having been kept carefully during the winter, was sown in a prepared bed, in the spring of 1848, in drills about eighteen inches apart. As the plants grew they were duly thinned out, leaving for the crop, as far as it could be done, the specimens that had leaves with the broadest divisions, lightest colour, and fewest hairs. As cultivated parsnips offer a curious contrast with the wild specimens in these respects, we place the following notes, side by side, on the root-leaves of plants of the same period of growth. 1st. WILD PARSNIP. | 2nd. STUDENT PARSNIP. Ft. in. | Ft. in. Whole length from the | Whole length from the base of the petiole to | base of the petiole to the apex of the leaf 0 8 | the tip of the leaf 2 0 Breadth of leaflets 0 0¾ | Breadth of leaflets 3 0¼ Length of ditto 0 1 | Length of leaflets 0 6½ Petiole and leaflets, | Petiole and leaflets hairy. Colour, dark | without hair. Colour, green. | light green. [Illustration: _Figures 1 and 2._—Roots of Wild Parsnips. Natural size.] We have before remarked that neither in size nor form are the wild roots at all comparable with the cultivated ones. Our figures 1 and 2 were taken from fine roots of the wild parsnip of the first year’s growth; that is to say, just at the same time as a crop parsnip would be at its best. They were purposely taken from specimens obtained from the same district as the seed with which our experiments were commenced. Our first crop of roots from the wild seed presented great diversities in shape, being for the most part even more forked than the originals, but still with a general tendency to fleshiness. Of these the best shaped were reserved for seeding; and having been kept the greater part of the winter in sand, some six of the best were planted in another plot for seed. The seed, then, of 1849 was sown in the spring of 1850, in a freshly-prepared bed, the plants being treated as before, the results showing a decided improvement, with tendencies in some examples in the following directions:— 1st. The round-topped long-root, having a resemblance to the Guernsey parsnip. (_Panais long_ of the French.) 2nd. The hollow-crowned long-root. “Hollow-headed” of the gardener. (_Panais Lisbonais_ type.) 3rd. The short, thick turnip-shaped root. “Turnip-rooted” of the gardener. (_Panais rond_ form.) These three forms were all of them much mis-shapen, with forked roots, that is, _fingers and toes_; but still each of them offered opportunities of procuring three original varieties from this new source.