In general, animals migrating will always move along lines of least resistance. Some deep-sea fishes have a considerable vertical range. It is thought that some move into shallower water to deposit their eggs or place their young in warmer water, and where the peculiar kind of food they need early in life is the most abundant. To study deep sea fishes is difficult, and so little has been done that we not only know them imperfectly but also know very little concerning their life histories. In February, March and April of 1891 the United States Fish Commission steamer Albatross explored a portion of the region between the coasts of Mexico and Central America and the Galapagos Archipelago. Besides obtaining a large number of shore fishes, about nine hundred specimens of fishes were secured, ranging from a depth of one hundred to twenty-two hundred and twenty-three fathoms. This collection was carefully studied by Professor Garman, of Harvard. He found the collection to contain one hundred and eighty species, eighty-five per cent. of which were new to science. The bottoms of the oceans are far from level, and each deep basin has its own peculiar fauna. The shallower parts of the sea prevent migration of the deep water forms and no doubt living as they do in eternal darkness and in a temperature near the freezing point, there is little to induce them to much activity. The fact that they are easily captured in nets of comparatively small size would indicate that they move about slowly. Dr. Jenkins, who has lately studied the fishes of the Sandwich Islands, informs me that less than five per cent. are found on our American coast, while a large per cent. is found all the way to the Red sea. In other words, the fishes of the Sandwich Islands are East Indian rather than American. This is no doubt caused from the fact that the deep water between the islands of the American coast forms a barrier which has always prevented the two fish faunas from mingling with each other. Between Africa and the Sandwich Islands this has not been the case. A recent study of the fishes of the Galapagos Archipelago shows its fauna to be American, though in what respect its fishes differ from those of our west coast they resemble all the more the fishes of the Sandwich Islands. Two fish faunas will usually differ from each other if separated by an impassable barrier; especially is this true if the barrier be older than the two faunas.