From the preceding statements, we are naturally led to inquire, in what manner these changes in the colour of tire dress are produced ? It has been supposed by some, that those quadrupeds, which, like the alpine hare and ermine, become white in winter, cast their hair twice in the course of the year ; at harvest, when they part with their summer dress, and in spring, when they throw off their winter fur. This opinion does not appear to be supported by any direct observations, nor is it countenanced by any analogical rea sonings. If we attend to the mode in which the hair on the human head becomes grey as we advance in years, it will not be difficult to perceive, that the change is not pro duced by the growth of new hair of a white colour, but by a change in the colour of the old hair. Hence there will be found some hairs pale towards the middle, and white to wards the extremity, while the base is of a dark colour. Now, in ordinary cases, the hair of the human head, un like that of the inferior animals, is always dark at the base, and still continues so during the change to grey ; hence we are disposed to conclude from analogy, that the change of colour, in those animals which become white in winter, is effected, not by a renewal of the hair, but by a change in the colour of the secretions of the rete mucosum, by ,which the hair is nourished, or perhaps by that secretion of the colouring matter being diminished, or totally sus pended.
As analogy is a dangerous instrument of investigation in those departments of knowledge which ultimately rest on experiment or observation, so we are not disposed to lay much stress on the preceding argument which it has fur nished. The appearances exhibited by a specimen of the ermine now before us are more satisfactory and convincing. It was shot on the 9th May, (1814,) in a garb intermediate netween its winter and summer dress. In the belly, and all the unde• parts, the white colour bad nearly disappear ed, in exchange for the primrose yellow, the ordinary tinge of these parts in summer. The upper parts had riot fully acquired their ordinary Slimmer colour, which is a deep yellowish brown. There were still several white spots, and not a few with a tinge of yellow. Upon exa mining those white and yellow spots, not a trace of inter spersed new short brown hair could be discerned. This would certainly not have been the case, if the change of co lour is effected by a change of fur. Besides, while some parts of the fur on the back had acquired their proper co lour, even in those parts numerous hairs could be observed of a wax yellow, and in all the intermediate stages from yellowish brown, through yellow, to white.
These observations leave little room to doubt, that the change of colour takes place in the old hair, and that the change from white to brown passes through yellow. If this conclusion is not admitted, then we must suppose that this animal casts its hair at least seven times in the year. In spring, it roust produce primrose yellow hair; then hair of a wax yellow ; and, lastly, of a yellowish brown.
The same process must he gone through in autumn, only reversed, and with the addition of a suit of white. The ab surdity of this supposition is too apparent to be farther exposed.
With respect to the opinion which we have advanced, it seems to be attended with few difficulties. We urge not in support of it, the accounts which have been published of the human hair changing its colour in the course of a single night ; but we think the particular observations on the ermine war rant us in believing that the change of co lour in the alpine hair is effected by a similar process. But how is the change accomplished in birds ? The young ptarmigans are mottled in their first plumage similar to their parents. They become white in winter, and again mottled in spring. These young birds, provid ed tire change of colour is effected by moulting, must pro duce three different coverings of leathers in the course of ten months. This is a waste of vital energy, which we do not suppose any bird in its wild state capable of sustaining; as moulting is the most debilitating process which they un dergo. In other birds of full age, two moultings must be necessary. In these changes, the range of colour is from blackish grey., through grey, to white, an arrangement so nearly resembling that which prevails in the ermine, tfiat we are disposed to consider the change of colour to take place in the old feather s, and not by the growth of new plu mage ; this change of colour being independent of the or dinary annual moultings of the birds.
Independent of the support from analogy which the er mine furnishes, we may observe that the colours of other parts of a bird vary according to the season. This is fre quently observable in the feet, legs, and bill. Now, since a change takes place in the colouring secretions of these organs, what prevents us from supposing that similar changes take place in the feathers ? But even in tire case of birds, we have before us an example as convincing as the ermine already mentioned. It is a specimen of the little auk, (?lieu alle,)which was shot in Zetland in the end of February, 1810. The chin is still in its winter dress of white, but the feather's on the lower part of the throat have assumed a dusky hue. Both the shafts and webs have become of a blackish grey colour at the base and in the centre, while the extremities of both still continue white. The change from black to white is here effected by passing through grey. If we suppose that in this bird the changes of the colour of the plumage are accomplished by moulting-, or a change of feathers, we must admit the existence of three such moultings in the course of the year—one by which the white winter dress is produced, another for the dusky spring dress, and a third for the black garb of summer. It is surely unnecessary to point out any other examples in support of our opinion on this subject. We have followed nature, and our conclusions appear to be justified by the appearances which we have described.