sentence, and not until some days have elapsed will it be perfect. The mode of acquiring it seems to me exactly what I have observed in a child learning a French phrase; two or three words are constantly repeated, and then others added, until the whole is known, the pronunciation becoming more perfect as the repetition goes on. I found also on whistling a popular air to my parrot that she picked it up in the same way, taking note by note until the whole twenty-five notes were complete. Then the mode of forgetting, or the way in which phrases and airs pass from its recollection, may be worth remarking. The last words or notes are first forgotten, so that soon the sentence remains unfinished or the air only half whistled through. The first words are the best fixed in the memory; these suggest others which stand next to them, and so on till the last, which have the least hold on the brain. These, however, as I have before mentioned, can be easily revived on repetition. This is also a very usual process in the human subject: for example, an Englishman speaking French will, in his own country, if no opportunity occur for conversation, apparently forget it; he no sooner, however, crosses the Channel and hears the language than it very soon comes back to him again. In trying to recall poems learned in childhood or in school days, although at that period hundreds of lines may have been known, it is found that in manhood we remember only the two or three first lines of the 'Iliad,' the 'Æneid,' or the 'Paradise Lost.'[146] The following is communicated to me by Mr. Venn, of Cambridge, the well-known logician:-- I had a grey parrot, three or four years old, which had been taken from its nest in West Africa by those through whom I received it. It stood ordinarily by the window, where it could equally hear the front and back door bells. In the yard, by the back door, was a collie dog, who naturally barked violently at nearly all the comers that way. The parrot took to imitating the dog. After a time I was interested in observing the discriminative association between the back-door bell and the dog's bark in the parrot's mind. Even when the dog was not there, or for any other cause did not bark, the parrot would constantly bark when the back-door bell sounded, but never (that I could hear) when the front-door bell was heard. This is but a trifle in the way of intelligence, but it struck me as an interesting analogous case to a law of association often noticed by writers on human psychology.