To E. B. Tylor   19 June [1880]

My dear Sir

I have come across more than one statement that animals learn to know poisonous herbs, & one case referred to Australia. I fancy that another was by Linnæus.¹ I have no doubt that I have references, but it would take me a week of labour to find them. I believed in them because there seems to be good evidence that animals learn by seeing their comrades suffer. It is notorious that you cannot long continue poisoning rats with the same kind of poison. I have received 2 or 3 accounts from trustworthy persons (besides some published notices) that when telegraph wires are first set up in any district, many birds, especially partridges, are killed by flying against them; but that after 2 or 3 years such deaths cease. On the other hand when no harm follows animals get accustomed to what is terrific: the late W. Thompson of Belfast (an excellent observer) told me that when a R^y was first made there, which passed through some swamps, the wild birds were terrified at the trains, but that after a few months even herons remained stationary close to the train.—²

I wish that I could have aided you better & remain | Yours very sincerely | Ch. Darwin.

P.S. I have just thought of a place to look for references on cattle or sheep learning to avoid poisonous herbs

(1) Annals & Mag. of Nat. Hist 2^d series, Vol. 2. p. 364.

(2) Amœnitates Academicæ Vol. 7 p. 409. (I suppose Linnæus)

(3) Stillingfleet Tracts p. 350

(4) Youatt on Sheep p. 404.³

I do not now knows details, but these works were consulted by me,