the works of nature, these partial writers may, each in their department, be more accurate in their discoveries, and freer from errors, than more general writers; and so by degrees may pave the way to an universal correct natural history. Not that Scopoli is so circumstantial and attentive to the life and conversation of his birds as I could wish: he advances some false facts; as when he says of the hirundo urbica that "pullos extra nidum non nutrit."
This assertion I know to be wrong from repeated observation this summer; for house-martins do feed their young flying, though it must be acknowledged not so commonly as the house-swallow; and the feat is done in so quick a manner as not to be perceptible to indifferent observers. He also advances some (I was going to say) improbable facts; as when he says of the woodcock that "pullos rostra portat fugiens ab hoste." But candour forbids me to say absolutely that any fact is false, because I have never been witness to such a fact. I have only to remark that the long unwieldy bill of the woodcock is perhaps the worst adapted of any among the winged creation for such a feat of natural affection. 1
NOTE TO LETTER XXXI.
1 It is a fact that the woodcock does carry its young. The legs and beak are both employed in holding the young one to the parent's breast as it flies.
LETTER XXXII.
Dear Sir,—After an ineffectual search in Linnæus, Brisson, etc., I begin to suspect that I discern my brother's hirundo hyberna in Scopoli's new discovered hirundo rupestris, p. 167. His description of "Supra murina, subtus albida; rectrices maculâ ovali albâ in latere inferno; pedes nudi, nigri; rostrum nigrum; remiges obscuriores quam plumæ dorsales; rectrices remigibus con-
