traps of that region, thin veins occur of anthracitic material, which alone remains to represent the organic constituents of the altered sediments. Continuing our course southwestward the same changed condition is observed in the crystalline schists of Manhattan Island, and across the Hudson through northern New Jersey. Intrusions of trap, too, are frequent through all this region and the sole representative of the organic constituents of the sediments is anthracitic residues.

On the western slope of the Catskills, through eastern New York, the crystalline rocks which exist at varying depths below the surface are overlaid with sediments which are frequently imperfectly metamorphosed, and as one moves westward into central New York and northeastern Pennsylvania, while the coal beneath the surface is anthracite and the residues before mentioned that fill cavities in the limestone are anthracitic, still the surface rocks show less and less signs of alteration. As the summit of the Alleghanies is reached and passed, the coal beds fade by insensible stages from anthracite into unaltered splint and cannel coals. The beds of slate also become beds of pyroschists, and the formations generally assume the aspect of unaltered sediments. On the western slope of the Alleghanies the surface descends much less abruptly than it ascends on the eastern slope. The dip of the formations is much greater than that of the surface, consequently the outcropping edges of newer formations are repeatedly encountered, until in western Pennsylvania and New York metamorphism has ceased to be a problem in surface geology. These surface rocks are, however, geologically all below the coal, which in eastern Pennsylvania is metamorphosed into anthracite. There is no arbitrary line that separates the unaltered from the altered strata. The successive formations have thinned out, and in general they continue to become thinner as we go southwest; but there is no anthracite between the crest of the Alleghanies and the mountains of Arkansas. Throughout the Mississippi valley, as we pass to the west, these formations outcrop and overlie each other precisely like the shingles on a roof, with the pitch reversed.

In the Bradford oil field, in McKean county, Pa., the drill penetrates a bed of porous sandstone that lies enclosed in impervious unaltered strata. It contains a few shells and fish bones, but no other fossils. Like the surface rocks it lies sloping toward the southwest, the lower portion submerged in salt water, the middle PROC. AMER. PHILOS. SOC. XXXVII. 157. I. PRINTED JUNE 15, 1898.

portion filled with petroleum and the upper portion filled with gas; both originally under an enormous pressure. In Warren county, farther to the southwest, the drill reaches petroleum not in the McKean county sand, but in a different sand, higher in the series. Still farther southwest, in Venango county, the surface rocks are still higher in the series and the drill reaches petroleum in a pebble conglomerate that outcrops at the surface to the northeast. These pebble conglomerates, known as the "Venango Oil Sands," formed great riffles in the currents of the primeval ocean. They are several miles long and a few rods wide, level on the upper surface, and rounded on the under surface to a feather edge at the sides. One is above the other and they are covered, when they contain petroleum, with a solid, impervious shell of silica, that the drill penetrates with difficulty. The uppermost of these conglomerates consists of spherical pebbles of yellow quartz, about as large as cranberries; the lowest consists of lenticular pebbles of very white quartz. In both cases the pebbles are cemented together at their points of contact leaving large open spaces. These conglomerates are sometimes replaced by coarse, porous sandstones; neither of these contain fossils of any kind. Still farther southwest, on Slippery Rock creek in Mercer county, and at Smith's Ferry in Beaver county, another sandstone, that is barren where it occurs in Venango county, yields petroleum above the pebble conglomerate. If a line be followed farther to the left, across western Pennsylvania and into West Virginia, the outcrops of the formations would rise successively in the scale until the oil would be found in the Mahoning sandstone, which lies at the top of the Lower Productive Coal Measures. Since the development of the Lima oil fields the range of rocks holding the petroleum reaches in Ohio, Canada and Pennsylvania from the Lower Silurian, Trenton limestone, to the Lower Coal Measures. These rocks embrace nearly the entire palæozoic formations of North America. Very few wells have been sunk below the petroleum-bearing sandstone, for the obvious reason that it involved a useless expense. One of the deepest wells ever drilled in the oil region of western Pennsylvania was Jonathan Watson's deep well near Titusville. This well went down through all of the oil sands and the Devonian shales beneath them, to a depth of 3553 feet, when just as it was abandoned a hard rock was struck which was supposed to be the Corniferous limestone, which is the oil-bearing rock of Canada. The interval between the oil sands and the bottom of the

well was filled with Devonian shales, that underlie the Bradford oil sand and are supposed to extend from Allegheny county, New York, to central Kentucky; and in fact to underlie the entire petroleum region that produces Warranite-the pure paraffine petroleums. When "dry" or unproductive holes are drilled outside the productive areas, they pass, at the horizon of the oil sands, through a different rock, which is compact and incapable of holding petroleum. These underlying Devonian shales outcrop at Erie, Pa., and furnish there the material that on distillation yielded fifty gallons of distillate to the ton. Where this formation outcrops it is filled with fucoids and has yielded small petroleum and gas wells. The men who drilled Jonathan Watson's deep well told me that, "the soap stone (Devonian shale) became harder as they went down, and was redder in color, in fact, had been burnt like brick." In a comparatively few localities, petroleum has been found saturating rocks that lie one above the other. The upper rock invariably yields the most dense oil. In 1881 I saw a well in West Virginia, from which the same walking beam pumped at every stroke oil of 27 degrees from a depth of 255 feet and oil of 45 degrees from a depth of 600 to 700 feet.

The

18. I have never seen a specimen of graphite reported to have come from any locality between the crest of the Alleghanies and the Ozark uplift. This is an uplift of the paleozoic formation west of the Mississippi river, extending from central Missouri to central Texas. It resembles that of the Alleghanies, but is on a smaller scale.1 The eastern slope is more abrupt than the western. formations of the central portions, in Arkansas and the Indian Territory, are highly crystalline, graphite and anthracite are of frequent occurrence and are found on the western slope. On this slope also, but farther west, in unaltered strata immediately above the crystalline formations, bitumen occurs in enormous quantity and great variety. Over a large area in the northeastern portion of the Indian Territory heavy petroleums are found only a short distance beneath the surface, and, as I am informed, below the coal. South of the Red river, in northern Texas, bitumens occur saturating horizontal beds of sand that are intercalated between strata of more or less solid limestone. North of the Red river, in the Indian Territory, every rock formation that is at all porous appears to be

1 J. C. Branner, Former Extension of the Appalachians across Mississippi, Louisiana and Texas," Am. Jour. Sci. (4) iv, 357.

filled with bitumen. As far as I have investigated it, the bitumen is uniform in kind and quality. It has saturated beds of sand, strata of sandstone and limestone, some of which are hard and crystalline, others magnesian and almost as soft as chalk, some of them without fossils and some almost all fossils, and all of them conformable with the Upper Silurian and Lower Carboniferous rocks that enclose them. In one locality a sort of bituminous breccia occurs, of immense extent, consisting of fragments of limestone and quartzite cemented together with bitumen. In another an immense horizontal bed of sand, completely saturated with bitumen, is overlaid with thirty or forty feet of conglomerate that has been more or less penetrated with it.

Almost all the beds north of the river are in very sharp folds, that bring the strata to the surface nearly vertical, in eroded anticlinals that extend across the country in parallel lines, often many miles in length. What is of especial interest in this connection is the occurrence in the vertical limestones and sandstones of imperfectly saturated strata. The bedding varies from the thickness of paper to a few inches. The rock mass was usually most easily penetrated along the lines of the thinnest beds. Fractures which cross all these beds, including both the thin and thick ones, show the bitumen completely filling the thin beds and only partially penetrating the seams and the mass of the thicker cryptocrystalline

Nothing could more beautifully and clearly demonstrate the fact that the bitumen was not indigenous to these rocks, but had penetrated them while previously and as at present in their nearly vertical position.

19. Continuing our journey across the continent, bitumen is frequently encountered in positions contiguous to normal or local metamorphism, until we descend into the great valley of California, west of the Sierra Nevadas. Here the development of bitumen has proceeded on a scale of vast magnitude. On the western slope of the Sierras the region around Roseville, in Placer county, and the vicinity of the city of Stockton, are well known to be rich in natural gas. There are localities on these slopes that have also furnished limited supplies of petroleum, but, as before stated, the bitumen deposits of California are principally found in the Coast Ranges, including the ocean area lying between the Santa Barbara

1 W. L. Watts, The Gas and Petroleum Formations of the Central Valley of California, 1894.

islands and the main land. The richest deposits have been found in Ventura county, on the border line that separates the Cretaceous from the Lower Miocene. None of the bitumen is found in crystalline rocks; yet the evidences of both normal and local metamorphism, in strata not far distant from the bitumen-bearing rocks, are abundant. The late Eli W. Blake once visited the Santa Barbara islands and afterwards described to me the cascades of lava that had descended from the volcanic cones in the centre of the

islands over precipices into the sea. Bitumen has exuded for more than a century from the unaltered strata, whose upturned edges form the bed of the ocean, between these islands and the main land. The Tertiary formations that constitute the bluffs of the coast east and west of Santa Barbara contain deposits of bitumen of enormous extent and exhibit evidences of metamorphic action still in progress. Almost every large bluff from Point Conception to San Diego contains a solfatera, the action of which leaves the Miocene shales, originally rich in organic matter, devoid of a trace of carbon.

The best petroleum wells of Ventura county lie in the cañons of the Sulphur mountain, one of the foothills of the Coast Ranges. Other wells are similarly located with reference to these ranges.1 None of them have penetrated crystalline rocks; yet the core of the Coast Ranges only a few miles east of the wells of the Pacific Coast Oil Co., as Dr. Goodale and myself found, is granite. Fragments of crystalline rocks are washed out of many of the large cañons that head in the main Coast Range back of the foothills in which the oil wells are drilled. Deep drilling is extremely difficult in this region on account of the fragile character of the rocks. It might be impossible to carry a well down through all the bituminous strata to the crystalline rocks, but the fact that they are altered Miocene sediments and exist at a comparatively short distance below the surface does not admit of any question. The evidences of metamorphism, through the agency of hot, silicated water, are found everywhere. The formations contain abundant remains of highly organized animals; and the bitumens which they contain consist of benzoles and naphthenes, without an "appreciable amount of paraffines, if any." They also contain sulphur and nitrogen. They are evidently 1S. F. Peckham, Mineral Resources of the United States, "Petroleum in California," 1894.

2 Letter of C. F. Mabery to S. F. P.