Page:Popular Science Monthly Volume 21.djvu/396

384 assort them in accordance with those characteristics; in short, to arrange or classify them. The young conchologist, for example, sees in an instant that out of a miscellaneous collection of shells some are bivalve and others univalve, and that some of them exhibit clear distinctions connected with the form of the animal to which the shell belongs. The young entomologist, with still greater ease, perceives the difference between most of the insects that come in his way, and indeed in some cases needs no instigation to look for them—the difference between a grasshopper and a house-fly, a beetle, a butterfly, and a moth, being self-evident to any one with eyes. So with the vertebrates; it requires no previous zoological instruction to enable any child to point out characters that will separate a snake from a tortoise, a rabbit from a sheep, a whale from a camel, and the rough primary division of all these creatures is at once perceptible. But with fishes this is not so. The learner, judging, as he is at first inclined to do, from outward survey, is surprised to find that the essential differences between a lamprey and an eel are deemed to be far greater than between an eel and a salmon, and that a skate is much further removed from a turbot than the latter is from a gudgeon, while a lancelet, which, when immersed in a bottle of spirit, looks so like a small smelt, differs, in the opinion of certain systematists, more from it than the smelt does from a frog, or indeed from any other existing vertebrate. All this, which the learner finds written in the first book on the subject (if he has one of the least authority) to which he has access, is so entirely in contradiction, as he thinks, to the plain evidence of his eye-sight, that he may well be staggered at the outset of his studies and discouraged from their prosecution. The classification of fishes has in truth been a task of no ordinary difficulty, and it is a subject requiring a far greater knowledge of their internal structure than can possibly be expected of a beginner.

One of the most formidable difficulties in the way of arriving at an intelligent classification has been removed by a discovery which Dr. Günther has made concerning the affinities of certain groups of fishes or fish-like animals, the relations of which to each other and to other fishes had been an inscrutable puzzle to all systematists. Among these were the ganoids, a family represented in an indefinite number of fossils, mostly of very ancient date, but few types of which survive to this day, and these restricted to the fresh waters of Eastern Asia, North America, and tropical Africa; other fossil fishes of equal antiquity, which were closely allied to the abundant sharks, dog-fishes, rays, and skates of our own seas, the "Chondropterygians" or "Elasmobranchs"; sturgeons, "Chondrosteans," possessing much of an archaic character; and besides these, there now exist two animals, commonly called "mud-fish," scientifically "Dipnoi," which have been deemed by some great authorities true fishes, by others amphibians. Furthermore, in the early days of the settlement of the Australian