large losses, or it may become adapted to some place of refuge where other fishes will not follow. What better refuge could a harrassed fish desire than the hiding places among stones in the shallows of a stream, where the water dashes ceaselessly by with a swiftness few fish can stem? And if, at the same time, the refugee develops a swimming power which enables it to dart like a flash against the strongest current, its safety would seem to be ensured. But what food could it find in such a place? Let us turn over the stones in such a stream, sweeping the roiled water at the same time with a small cloth net, and we shall find— larvæ of Chironomus and small Ephemerids and other such prey, and little else; food too minute and difficult of access to support a large fish, but answering very well if our immigrant can keep down his size. Here the principles of natural selection assert their power. The limited supply of food early arrests the growth of the young; while every fish which passes the allowable maximum is forced for food to brave the dangers of the deeper waters where the chances are that it falls a prey. On the other hand, the smaller the size of those which escape this alternative, the less likely will they be to attract the appetite of the small gar or other guerilla which may occasionally raid their retreat, and the more easily will they slip about under stones in search of their microscopic game.1 Like other fishes, the darters must have their periods of repose, all the more urgent because of the constant struggle with the swift current which their habitat imposes. Shut out from the deep still pools and slow eddies where the larger species float suspended in mid stream, they are forced to spend their leisure on or beneath the bottom of the stream, resting on their extended pectorals and anal, or wholly buried in the sand. Possibly this fact is correlated with the absence or rudimentary condition of the air-bladder; as it is a rule with many exceptions—but still, probably, a rule-that this organ is wanting in fishes which live chiefly at the bottom. Doubtless the search for food has much to do with this selection of a habitat. I have found that the young of nearly all species of our fresh-water fishes are competitors for food, feeding almost entirely on entomostraca and the larvæ of minute. 1 In Boleosoma, which is normally scaled in front of the dorsal fin, we often find the skin of this region bare in large specimens, and showing evident signs of rubbing. diptera.1 As a tree sends out its roots in all directions in search of nourishment, so each of the larger divisions of animals extends its various groups into every place where available food occurs, each group becoming adapted to the special features of its situation. Given this supply of certain kinds of food, nearly inaccessible to the ordinary fish, it is to be expected that some fishes would become especially fitted to its utilization. Thus the Etheostomatinæ as a group are explained, in a word, by the hypothesis of the progressive adaptation of the young of certain Percide to a peculiar place of refuge and a peculiarly situated food supply. Perhaps we may, without violence, call these the mountaineers among fishes. Forced from the populous and fertile valleys of the river beds and lake bottoms, they have taken refuge from their enemies in the rocky highlands where the free waters play in ceaseless torrents, and there they have wrested from stubborn nature a meagre living. Although diminished in size by their continual struggle with the elements, they have developed an activity and hardihood, a vigor of life and glow of high color almost unknown among the easier livers of the lower lands. The appended table will facilitate a comparison of the records of the different genera. The percentages were obtained by esti. mating carefully the ratios of each element of the food of each individual, and averaging these ratios for all the individuals of a species: DETAILS OF THE FOOD OF THE ETHEOSTOMATINÆ. 1 The Catostom[at]idæ (suckers) are an exception to this rule, feeding when young chiefly on Algae and Protozoa. ON THE FORMER EXTENT OF THE TRIASSIC FORMATION OF THE ATLANTIC STATES.1 BY ISRAEL C. RUSSELL. NEARLY two years since I read a paper before the New York Academy of Sciences, on the Physical History of the Triassic Formation in New Jersey and the Connecticut valley, in which many reasons were given for concluding that the Triassic rocks of these two regions were detached portions of one estuary formation. As several papers have been published relating to the Triassic rocks of the Atlantic States since my essay was written, increasing our knowledge of the subject, and as my interpretation of the geological records has not been accepted by some geologists, I 1 Read before the New York Academy of Sciences, March 22, 1880. take the present opportunity to briefly restate the substance of my former paper which seems to have been partially misunderstood, together with a brief review of the evidence that has since been gathered. . The broad generalization advanced in the essay mentioned above, was, that all the detached areas of Triassic rocks, from South Carolina northward to Connecticut and Massachusetts, are portions of one great estuary deposit, which has been broken up into separate areas by upheaval and denudation. The immediate aim of the paper, however, was to prove the former connection of the Triassic rocks of New Jersey with those found in the Connecticut valley. The conclusion arrived at from the study of the rocks of these two areas, was, that they are the marginal portions of an estuary deposit, the central region having been subsequently upheaved and greatly eroded. The Triassic rocks in this region thus fill a synclinal trough, the longer axis of which has been upheaved into a broad anticlinal. The facts that lead to this conclusion may be briefly stated as follows: First. The Triassic rocks in New Jersey dip to the westward at an average angle of about fifteen degrees, while the corresponding beds along the Connecticut river are inclined to the eastward at a somewhat larger angle; thus suggesting that they are portions of one great anticlinal. Second. Each area is an incomplete estuary formation, having only one line of shore deposits. This is shown in the case of the New Jersey area by the fact that all along the line of bluffs bordering the formation on the west occurs a coarse conglomerate which we have shown to be a shore deposit, derived from the bluffs of crystalline rock to the westward. In the finer sandstones and shales associated and interstratified with this conglomerate are ripple marks, sun cracks, raindrop impressions and the footprints of animals, proving beyond question that this was the shore of the basin in which the Triassic rocks were deposited. Throughout the eastern margin of the New Jersey area, which is sharply defined along the western bank of the Hudson from Jersey City northward to Stony Point, these indications of shore conditions are entirely lacking, in their stead there are sandstones, slates and shales of the character of ordinary off-shore deposits. The trap rock forming the Palisades will be noticed farther on. Crossing to the Connecticut valley we find this order reversed; on the eastern margin of this area the coarse conglomerates again occur, together with an abundance of all the other proofs of shore conditions we have mentioned; on the western margin the rocks have been formed of sand and mud deposited at a distance from the shore, and are without sun cracks, footprints, etc.; these beds correspond with the sedimentary rocks in the escarpment along the western shore of the Hudson. From these facts it seems perfectly justifiable to conclude that the variegated conglomerate bordering the New Jersey area on the west, corresponds in character and position with the coarse conglomerate occurring along the eastern margin of the Connecticut River region, thus mapping out portions of the eastern and western shores of the estuary in which the Triassic rocks were deposited. Thirdly. The occurrence of an outlying mass of Triassic beds in the towns of Southburg and Woodbury, Conn., lying between the two great areas, also favors the conclusion that the sandstones and shales of New Jersey and the Connecticut valley were once united. This little oasis in the valley of the Housatonic, is but six or seven miles long by two broad, and is separated from the Connecticut valley area by fifteen and from the Hudson by forty miles of crystalline rock.1 Fourthly. The topographical features along the western margin of the New Jersey area and the extension across the Hudson of the line of bluffs which border the formation in New Jersey, as stated on page 21 of the writer's previous essay (page 241 of the Annals), also indicates that the Triassic rocks of New Jersey at one time followed the course of this old shore line and extended eastward of the Hudson. Fifthly. The striking analogy that exists in the arrangement of the hills of trap found in these Triassic areas was also pointed out in the paper mentioned above. Nearly all the igneous rocks found in New Jersey and the Connecticut valley have been formed as sheets of molten matter intruded between the layers of sedimentary rock and have cooled and crystallized in that position. In the Connecticut valley these sheets of trap dip eastward at the same angle with the sandstones and shales, and present a bold escarpment to the westward; the ends of the long ridges are 1 Percival's Geo. Rep. of Conn., 1842, p. 410. Also Annals of N. Y. Acad. of Sci., Vol. 1, No. 8 (1878), p. 240. |