This letter has elicited various replies and suggestions which you will perhaps allow me to answer, and I may also take the opportunity of stating the progress which my dog Van has made, although, owing greatly, no doubt, to my frequent absences from home and the little time I can devote to him, this has not been so rapid as I doubt not would otherwise have been the case. Perhaps I may just repeat that the essence of my idea was to have various words, such as "food," "bone," "water," "out," &c., printed on pieces of card-board, and, after some preliminary training, to give the dog anything for which he asked by bringing a card. I use pieces of cardboard about ten inches long and three inches high, placing a number of them on the floor side by side, so that the dog has several cards to select from, each bearing a different word. One correspondent has suggested that it would be better to use variously coloured cards. This might, no doubt, render the first steps rather more easy, but, on the other hand, any temporary advantage gained would be at the expense of subsequent difficulty, since the pupil would very likely begin by associating the object with the colour, rather than with the letters. He would, therefore, as is too often the case with our own children, have the unnecessary labour of unlearning some of his first lessons. At the same time, the experiment would have an interest as a test of the colour-sense in dogs. Another suggestion has been that, instead of words, pictorial representations should be placed on the cards. This, however, could only be done with material objects, such as "food," "bone," "water," &c., and would not be applicable to such words as "out," "pet me," &c.; nor even as regards the former class do I see that it would present any substantial advantage. Again, it has been suggested that Van is led by scent rather than by sight. He has, no doubt, an excellent nose, but in this case he is certainly guided by the eye. The cards are all handled by us, and must emit very nearly the same odour. I do not, however, rely on this, but have in use a number of cards bearing the same word. When, for instance, he has brought a card with "food" on it, we do not put down the same identical card, but another with the same word; when he has brought that, a third is put down, and so on. For a single meal, therefore, eight or ten cards will have been used, and it seems clear, therefore, that in selecting them Van must be guided by the letters. When I last wrote I had satisfied myself that he had learnt to regard the bringing of a card as a request, and that he could distinguish a card with the word "food" on it from a plain one, while I believed that he could distinguish between a card with "food" on it and one with "out" on it.