OLD RED SANDSTONEJ of the Upper Old Red Sandstone area. Not only so, but they must have been in existence long before the formation of the thick Arran limestone, though it was only during the comparatively brief interval represented by that lime- stone that geographical changes permitted them to enter the Old Red Sandstone basin and settle for a while on its floor. Thus we see that while, on the one hand, the older parts of the Lower Old Red Sandstone were coeval with an Upper Silurian fauna which, having disappeared from the area of Britain, survived outside of that area, on the other hand, the higher parts of the Upper Old Red Sandstone were contemporaneous with a Carboniferous Limestone fauna which, having appeared beyond the British area, was ready to spread over it as soon as the conditions became favourable for the invasion. It is, of course, obvious that such an abundant and varied fauna as that of the Car- boniferous Limestone cannot have come suddenly into existence at the period marked by the base of that forma- tion. It must have had a long previous existence outside the present area of the deposits. But it is seldom that we obtain such clear evidence of the fact as in these instances fron1 the Scottish Old Red Sandstone. In the north of Scotland, on the lowlands bordering the Moray Firth, and again in the island of Hoy, one of the Orkney group, yellow and rerl sandstones, sometimes c011- taining characteristic Upper Old Red Sandstone fishes, are found lying unconformably upon the Caithness flags. In these northern tracts the san1e relation is thus traceable as in the central counties between the two divisions of the system. Turning southward across the border districts, we trace the red sandstones and conglomerates of the Upper Old Red Sandstone lying unconformably on Silurian rocks and Lower Old Red Sandstone. Some of the brecciated con- glomerates have much resemblance to glacial detritus, a11d it has been suggested that they have been connected with contemporaneous ice-action. Such are the breccias of the Lanimermuir Hills, and those which show themselves here and there from under the overlying mass of Carboniferous strata which flanks the Silurian hills of Cumberland and Westrnoreland. Red conglomerates and sandstones appear interruptedly at the base of the Carboniferous rocks eve11 as far as Flintshire and Anglesea. They are com- monly classed as Old Red Sandstone, but merely from their position and lithological character. No organic remains have been found in them. They may therefore, in part at least, be taken as the basement beds of the Carboniferous system. In Devonshire, at Barnstaple, Pilton, Marwood, a11d Baggy Point, certain sandstones, shales, and limestones (already referred to in the account of the Devonian rocks) graduate upward into the base of the Carboniferous system, and appear to represent the Upper Old Red Sandstone of the rest of Britain. They contain land plants and also nnny marine fossils, some of which are common Carboni- ferous forms. They thus indicate a transition into the geographical conditions of the Carboniferous period, as is still more clearly illustrated by the corresponding strata in Scotland. Nonru A.lEI:IC..—Tl1c Devonian system, as developed in the northern States, and eastern Canada and Nova Scotia, presents much geological interest in the union which it contains of the same two distinct petrographical and biological types found in Europe. If we trace the range of these rocks along the Allcghany chain ‘through l’cnnsylvania into New York, we find them to contain a <-haracteristic suite of marine organisms comparable with those of the Devonian system of Europe. But on the eastern side of the great range of Silurian hills in the north-eastern States, we en- counter in New Brunswick and Nova Scotia a succession of red and yellow sandstones. limestones, and shales nearly devoid of marine °T{-»"“‘_i5n15, yet full of land—plants, and with occasional traces of fish remains. GEOLOGY 345 The marine or Devonian type has been grouped in the following subdivisions by the geologists of New York:— { Catskill lied Sandstone. Chcmung group. Portage group. Genesec group. Hamilton group. Marcellus group. (‘orniferous or Upper Helderberg group. Schoharie Grit. Cauda-galli Grit. In the Lower Devonian series traces of terrestrial plants (I’sz'lo- pl1._1/ton, C'a-u-lopfcris, &c.) have been detected even as far west as Ohio. Corals (cyathophylloid forms, with Fuvositcs, Syrz'n_r;opora., &c.) abound, especially in the Cornifcrous Limestone, which is per- haps the most remarkable mass of coral-rock in the American 1’-alzcozoic series. Among the brachiopods are species of I’rn_Ia- mcrus, ;S't-)"iclt'l(uztlinia, ]t‘l¢_1/nclzonclla, and others, with the charac- teristic European form .51/z'7'ifcr culwzjzzgatus, and the world-wide Atrypa 7'cticulm'z's. The trilobites include the genera lkthnmzitcs, Proétzts, and I’hacops. The earliest known traces of American fishes occur in the Corniferous group. They consist of ichthyodorulites, and teeth of cestraciont and hybodont placoids, and plates, bones, and teeth of some peculiar ganoids (Jlacropctalichthg/s, Onyclzodus). In the Hamilton formation (embracing the Marcellus shalc, the Hamilton beds, and the Genesee shale) remains of land-plants occur, but much less abundantly than among the rocks of New Brunswick. Braehiopods are especially abundant among the sandy beds in the centre of the formation. They comprise, as in Europe, many broad-winged spirifers (S. azmcronatus, &c.), with species of 1’)-ocluctus, C/Ionctcs, Atlzy-ris, &c. The earliest American goniatitcs have been noticed in these beds. N cwberry has described a gigantic fish (Dhriclrtliys) from the Black Shale of Ohio. The Portage and Chemung groups have yielded land-plants and fucoids, also some crinoids, numerous broad-winged spirifcrs, with A'ricu—lw, and a few other lamcllibranchs. These strata consist in the New York region of shales and laminated sandstones, which attain a maximum thickness there of upwards of 2000 feet, but die out entirely towards the interior. T hey are covered by a mass of red sandstones and conglome1ates—the Catskill group, which is 2000 or 3000 feet deep in the Catskill Mountains, a11d thickens along the Appalachian region to 5000 or 6000 feet. Those rcd arcnaceous rocks bear a striking similarity in their lithological and biological characters to the Upper Old lied Sandstone of Europe. As a whole they are nnfossiliferous, but they have yielded some ferns like those of the Upper Old Red Sandstone of Ireland and Scotland (Cycloptcris), and some characteristic genera of fish, as Holoptychius and ]}0l7u'-z'oIqn's. Turning now to the eastern side of the ancient Laurentian and Silurian ridge, which, stretching southwards from Canada, sepa- rated i11 later Palzeozoic time the great interior basin from the Atlantic slopes, we find the Devonian rocks of New York, Penn- sylvania, and the interior represented in New Brunswick and Nova Scotia by a totally difl'x-rent series of deposits. The contrast strikingly recalls that presented by the Old Red Sandstone of the north of Scotland and the Devonian rocks of North Germany. On the south side of the St Lawrence the coast of Gaspé shows rocks of the Quebec group unconformably overlaid by grey limestones with green and red shales, attaining, according to Logan, a total thickness of about 2000 feet,‘ aml replete i11 some bands with Upper Silurian fossils. They are conformably followed by a vast arenaccous series of deposits termed the Gaspé Sandstones, to which the careful measurements of Logan and his colleagues of the Canadian Geological Survey assign a (lcptl1 of 7036 feet. This formation consists of grey and drab—colou1'cd sandstones, with occasional grey shales and bands of massive conglomerate. Similar rocks reappear along the southern coast of New Brunswick, where they attain a depth of 9500 feet, and again on the opposite side of the Bay of Firmly. The researches chiefly of Dr J. W. Dawson have shown that these strata contain an abundant terrestrial fiora —thc oldest of which any relics have yet been recovered, for the few Upper Silurian land—plants at present known hardly deserve to be reckoned as a known flora. In his recent census hc enumerates no fcwcl‘ than 118 species of land—plants. They are almost all ncrogens, the lycopods and ferns being largely predominant. Among the distinctive forms the following may be 1ncntioned— I‘s1'lop7z_z/ton, Arllzrostignm, Lcptophlcmn, an(l Prolotmritcs. Forty- nine ferns are given, including the genera 6'3/cloplc/‘is, £Vc2(roptcrz'3,- Splzcnoptcr-1's, and some treeferns (Psaronius, Cauloptcris). Lcpi- dodendroid and sigillaI'oid plants abound, as well as calamites. Higher forms of vegetation are represented by a few conifers (Da.clo:r3/1011., Ormoxylon, I’rototaa:z'tcs, &c.). From a locality on Lake Erie, Dr Dawson describes a fragment of dicotyledonous wood, not unlike that of some modern trees—the most ancient Upper Devonian. Lowcr Devonian. 1 Geology of Canada, p. 393. '
X. — 44