found to break out occasionally into a return to the typical form, or to be connected with it by numerous intermediates. Generally speaking, such of these aberrant races as have spread to the limits of the geographical area of the species, or have become introduced into distant countries, and have thus been adapted to a change of condition, will there be found more disposed to maintain their peculiarities, or even to diverge still more from their types. It is where the species is most at home, where it accommodates itself most readily to a variety of soils and exposures, where the stations it affects show a most ancient domicile, that the connecting links between the varieties it has produced must generally be sought for. And this is one great reason why permanence of form is so little conclusive as evidence of specific difference, unless observed in a considerable portion of the area of the species.
The investigation of the connecting links between two forms, with the view of determining whether they are distinct species or marked races of one species, is attended with great difficulty in the due appreciation of what are intermediates—of the difference between one or two definitely limited, though apparently intermediate species, and a chain of intermediate forms connecting the two extreme varieties of one species. The Allsike clover, in the colour of its flowers and mode of growth, has been looked upon as intermediate between the Dutch and the common red clover (T. repens and pratense), and some such idea suggested to Linnæus the name of T. hybridum. Yet the evidences of its specific distinctness from both are very strong. I have observed it with care in a living state over a great part of its natural area in Sweden and Central Europe. From T. pratense it is separated by characters among the most constant in the genus, without, in this instance, any tendency to variation. It is nearer to T. repens; and Professor Buckman, at the meeting of the British Association at Cheltenham, in 1856, stated that he had found it degenerate into that species. I cannot but think, however, that here there must have occurred one of those mistakes so common in botanical and experimental gardens—that a plant or its seeds have accidentally perished, and its place has been taken by someubiquitous species, so nearly allied as to escape observation when young, such as, in this instance, T. repens. I never could detect, either in those places where I have seen T. hybridum wild in the greatest abundance, nor yet in the fields where it is cultivated, any tendency to assume the creeping stems, the peculiar inflorescence, and other characters of T. repens.
Take, again, Hypericum linariifolium. The cursory inspection of a few herbarium specimens of this plant, of certain varieties of H. perforatum, and of H. humifusum, might suggest the idea that the former constituted a connecting link between the two latter. In this instance the characters are less decided, and of a less constant nature than in that of the three Clovers above quoted; yet, so far as my experience goes—and I have observed H. linariifolium living in parts of Western France, where it grows in the greatest abundance, besides numerous dried specimens from the greater portion of its area, and the two
