to convey our ideas to them, rather than to devise any language, or code of signals, by means of which they might communicate theirs to us. No doubt, the former process is interesting and instructive, but it does not carry us very far. Under these circumstances, it has occurred to me whether some such system as that followed by deaf-mutes, and especially by Dr. Howe with Laura Bridgman, might not prove very instructive, if adapted to the case of dogs. Accordingly I prepared some pieces of stout cardboard, and printed on each in legible letters a word, such as 'food,' 'bone,' 'out,' etc. I then began training a black poodle, 'Van' by name, kindly given me by my friend Mr. Nickalls. I commenced by giving the dog food in a saucer, over which I laid the card on which was the word 'food,' placing also by the side an empty saucer, covered by a plain card. 'Van' soon learned to distinguish between the two, and the next stage was to teach him to bring me the card; this he now does, and hands it to me quite prettily, and I then give him a bone, or a little food, or take him out, according to the card brought. lie still brings sometimes a plain card, in which case I point out his error, and he then takes it back and changes it. This, however, does not often happen. Yesterday morning, for instance, he brought me the card with 'food' on it nine times in succession, selecting it from among other plain cards, though I changed the relative position every time. No one who sees him can doubt that he understands the act of bringing the card with the word 'food' on it, as a request for something to eat, and that he distinguishes between it and a plain card. I also believe that he distinguishes, for instance, between the card with the word 'food' on it and the card with 'out' on it. This, then, seems to open up a method which may be carried much further, for it is obvious that the cards may be multiplied, and the dog thus enabled to communicate freely with us. I have as yet, I know, made only a very small beginning, and hope to carry the experiment much further, but my object in troubling you with this letter is twofold. In the first place, I trust that some of your readers may be able and willing to suggest extensions or improvements of the idea. Secondly, my spare time is small, and liable to many interruptions; and animals also, we know, differ greatly from one another. Now, many of your readers have favorite dogs, and I would express a hope that some of them may be disposed to study them in the manner indicated. The observations, even though negative, would be interesting; but I confess I hope that some positive results might follow, which would enable us to obtain a more correct insight into the minds of animals than we have yet acquired."
Salts in Rivers and in the Sea.—The sea, it is well understood, is fed with salt as well as with water, by the rivers. The question then arises naturally, How is it that the rivers—admitting that they are mildly salt, although they appear to be fresh—differ from the ocean in the kind as well as in the strength of their saltness? Mr. W. Mattieu Williams answers the question by showing that, when sea-water is evaporated, sulphate of lime is the first salt to be deposited, while chloride of sodium, sulphate of magnesia, chloride of potassium, and the bromides, are deposited later. Hence, when the sea-water reaches the point of saturation with sulphate of lime, no more can be dissolved in it, but all additional supplies must be deposited. Moreover, if a soluble salt of lime were brought into the sea, its lime would combine with the sulphuric acid there combined with magnesia, or soda, or potash, which would, in obedience to a curious chemical law, leave those bases to combine with that one which would form an insoluble compound. Thus the total quantity of lime in sea-water is limited by the solubility of sulphate of lime, and this amounts to only about one part in four hundred of water.
The Caribs and the Greeks.—Mr. A. J. Van Koolnijk has published in the "Journal of the Dutch Geographical Society" an account of Carib tombs and relics which have been found in the Island of Aruba, off Dutch Guiana. Among the relics are potteries of good workmanship, elaborately ornamented and painted in a variety of colors obtained on the island. Some of the more common ornaments are figures of frogs and frogs' heads, which indicate that the Indians had