celain earth was made on the erroneous supposition, that the white earth which he received from a member of one of the embassies (I think, Lord Amherst) occurred naturally in this state. The second kind of material bears the name Pe-tun-tse ("white clay"). I have quite a cargo of both, and of the rocks from which either is prepared. Their examination in the laboratory may be a subject of some interest.

Near Poyang lake, the edges of these King-te-chin schists are overlain by coal-bearing strata, which are almost undisturbed. The locality (Loping) is the first in China which has become of practical value to foreign enterprise; it supplies the greater portion of the coal used by the American steamers on the Yang-tse. I found not a single fossil plant; but certain layers of limestone which overlie the coal carry an abundance of fossil shells, almost exclusively Brachiopods, in an excellent state of preservation. Though I could collect only on the waste-dumps at the pits, I gathered a large number both of specimens and species. As for beauty, the collection ranks first among all I have made in China. The fauna differs from those which I found formerly associated with the coal-measures in the northern provinces; but as I have no paleontologist to consult with, nor even paleontological books, I do not undertake to say what relation in age it bears to those other coalfields. Productus prevails in number. But the former leading fossil, P. semireticulatus, is here represented only by a few and small, though very distinct, specimens. Its place is occupied by other Producti with spirous shells, of which there are several species; some of them of the shape of P. horridus, but costate. Spirifers are very scarce. There is one Cyrtia, several Orthis, etc. But the most curious fossil is one which can only be a Siphonotreta. The natural shell has the shape of a night cap, with an opercule on the top, but no area, and is dotted with what appear to be the basal parts of spines; the folds are irreg ular. The dorsal shell is smooth. Both shells are of a horney substance. Crinoids, Orthoceras, small Porcellias, corals and sponges make up the list of the prevailing fossils.

I believe that the comparison of the various faunas which I have collected in different coal-fields will afford much of interest. No two of them are alike, yet all bear a certain resemblance in a few leading types. I have not yet found any locality that afforded an opportunity for collecting fossils at different geological levels. But every new observation supports my former conclusion, that in the vast extent of China the depositions of coal-beds continued during an extraordinarily long period, in which it shifted repeatedly to different portions of the country. I have some reason for believing that this period commenced even before the Carboniferous epoch, and, though culminating in this, continued on beyond its end.

ART. XXVIII.-Notes on Granitic Rocks; by T. STERRY HUNT, LL.D., F.R.S. SECOND PART.

(Read before the American Association for the Advancement of Science at Troy, August 20, 1870.)

(Continued from page 89.)

CONTENTS OF SECTIONS.-§ 16, Granitic veins of Maine; Brunswick; § 17, Topsham, Paris; $18, Westbrook, Lewiston; crystalline limestones; § 19, Danville, Ketchum; 20, Denuded granitic masses: $21, Banded veins; Biddeford, Sherbrooke; § 22, Veins at various New England localities; § 23, Mineral species of these veins; $24, Veins in erupted granites; $25, Geodes in granites; § 26, Veins distinguished from dykes; $27, Volger and Fournet on the origin of veins; 28, 29, Certain fissures and geodes distinguished from veins opening to the surface; § 30, 31, Temperatures of crystallization of granitic minerals.

$16. IT is in the series of micaceous schists with interstratified gneisses (§ 6) which I have elsewhere provisionally designated the Terranovan series,* that I have seen concretionary granitic veins in the greatest abundance and on the grandest scale. This stratified system, which is well seen in the White Mountains, appears to extend southward to Long Island Sound and northeastward beyond the limits of Maine. It is in this state that I have particularly studied the granitic veinstones of this system, whose history may be illustrated by a few examples from notes taken on the spot. In Brunswick the strata near the town are fine-grained, friable, dark colored, micaceous and hornblendic, passing into mica-schist on the one hand, and into well-marked gneiss on the other, and dipping to the S.E. at angles of from 15° to 40°. Very similar beds are found in the adjoining town of Topsham, and in both places they include numerous endogenous granitic veins. The course of these is generally N.W., or at right angles to the strike, though occasionally for short distances with the strike, and intercalated between the beds; the veins vary in breadth from a few inches to sixty feet, and even more. They generally consist in great part of orthoclase and quartz, with some mica and tourmaline, and offer in the associations and grouping of these minerals many peculiarities, which are met with not only in different veins, but in different parts of the same vein. In some cases, colorless vitreous quartz predominates greatly, and encloses crystals of milk-white orthoclase, often modified, and from one to several inches in diameter. At other times pure vitreous quartz forms one or both walls or the center of the vein, or else is arranged in bands parallel with the sides of the vein, and sometimes a foot or more in thickness, alternating with similar

*This Journal, July, 1870, page 83. The rocks of this White Mountain series are, in the present state of our knowledge. supposed to be newer than the Huroniau system noticed in §5, to which, with Macfarlane and Credner, I refer the crystalline schists with associated serpentines and diorites of the Green Mountains.

bands consisting wholly or in great part of orthoclase, or of an admixture of this mineral with quartz, having the peculiar structure of what is called graphic granite, or else presenting a finely granitoid mixture of the two minerals, with little or no mica, and with small crystals of deep red garnet. Prisms of black tourmaline are also met with in these veins, and more rarely beryl and even chrysoberyl. In the rock-cutting on the Lewiston railroad, just below Topsham bridge over the Androscoggin, there is a fine exhibition of these veins, which present alternate coarser and finer grained layers, traversed by long spearshaped crystals of dark mica passing from one layer to another.

§ 17. A remarkable example of a vein of considerable dimensions is seen in the feldspar-quarry in Topsham, which occurs in a dark fine-grained friable micaceous schist. At the time of my visit, in 1869, the limits of the vein were not seen, though large quantities of white orthoclase and of vitreous quartz had already been extracted. These were each nearly pure, and in alternate bands, the quartz presenting drusy cavities lined with remarkable tabular crystals. One band was made up in great part of large crystals of mica, and portions of the vein consisted of a grånular saccharoidal feldspar. The famous locality of red, green and blue tourmalines, with beryl, lepidolite, amblygonite, cassiterite, etc., at Mount Mica in Paris, is a huge granitic vein, which, with many others, is included in a dark colored very micaceous gneiss.

§ 18. In Westbrook numerous small veins of this kind, holding coarsely lamellar orthoclase with black tourmaline and red garnet, intersect strata of fine-grained whitish granitoid gneiss. In Windham the dark colored staurolite-bearing mica-schist of this series is traversed by a granitic vein holding crystals of beryl. In Lewiston a large vein of coarse graphic granite, holding black tourmaline, and showing fine-grained bands, cuts a great mass of bluish gneissoid limestone, which forms an escarpment near the railroad, about half a mile below the town. This limestone, which dips eastward about 15°, is interlaminated with thin quartzite beds, which are seen on weathered surfaces to be much contorted. The bluish crystalline limestone is mixed with grains of greenish pyroxene, and includes nodular granitic masses of white crystalline orthoclase with quartz, enclosing large plates of graphite, crystals of hornblende, and more rarely of apatite. These associations of minerals are met with in the granític veins of the Laurentian limestones, to be noticed elsewhere. The limestone of Lewiston, however, appears to be included in the great mica-schist series of the region, where similar beds, though less in extent, are met with in various places, sometimes associated with pyroxene, garnet, idocrase and sphene. A thin band of impure pyroxenic lime

stone, like that of Lewiston, occurs with the mica-schists on the Maine Central Railroad, near Danville Junction, and beds of a purer crystalline limestone were formerly quarried in the southeast part of Brunswick, where they are interstratified with thinbedded dark hornblendic and micaceous gneiss, dipping S.E. at a high angle.

§ 19. At Danville Junction strata of hornblendic and micaceous gneiss, passing into mica-schists, dip N.E. at moderate angles, and include huge veins of endogenous granite. Two of these appear in the hill just south of the railroad station, apparently running with the strike of the beds. They are seen to rest upon the mica-schist, and in one of them a mass of this rock, three feet in width, is enclosed like a tongue in the granite, which has a transverse breadth of about seventy-five feet. Notwithstanding the apparent intercalation of these granitic masses the proof of their foreign origin is evident in a transverse fracture and slight vertical dislocation of the mica-schists around the broken edges of which the granite is seen to wrap. The endogenous character of this granite is well shown by its banded structure; belts of white quartz some inches wide alternate with others of coarsely cleavable orthoclase, while other portions hold black tourmalines and garnets of considerable size.

The evidence of disturbance of the strata in connection with these endogenous granites is seen on a large scale at the falls of the Sunday River in Ketchum. These mica-schists and gneisses, similar to those already noticed, enclose great masses of endogenous granite, which are seen to be transverse to the strata. On one side of such a mass more than sixty feet wide, the schistose strata are twisted from their regular N.E. strike to the N.W., and so enclosed in the granite as to appear as if interstratified with it for short distances. The banded structure of the transverse granite veins is here very marked. Some portions present cleavage-planes of orthoclase six inches in diameter; other parts, which are less coarse, abound in mica. Similar banded granite veins abound in the adjoining towns of Newry and North Bethel, and sometimes present layers of quartz six inches or more in thickness, besides large crystals of mica, and more rarely apatite. These veins are often irregular in shape and bulging at intervals, and they sometimes run partially across the beds, which seem to have been distended and disturbed, a fact which was also observed in the thin-bedded schists in contact with some of the veins in Brunswick, and is apparently due to the expansive force of crystallization, as noticed in § 27.

$20. The locality already described at Danville offers an instructive example of a phenomenon often met with in the region now under consideration, where granitic masses, resist

ing the actions which have degraded the soft enclosing schists, stand out in relief on the surface, and seem to constitute the rock of the country. A careful search will however show that they are simply veins or endogenous masses of very limited dimensions, rising from out of the mica-schists, which are often concealed by the soil. This is well seen about the lower falls of the Presumpscott near Portland, where the mica-schists with some fine-grained gneisses, dipping S.E. at angles of from 30° to 40°, enclose large numbers of granitic veins, which, though sometimes but a few inches in breadth, often measure twenty or even fifty feet, and are usually very coarse-grained, with white mica, black tourmaline, and more rarely beryl. They are sometimes transverse to the stratification, but more often parallel, and, rising above the soil, are very conspicuous.

$21. We have already noticed the exotic granites of Biddeford, which are intruded among fine-grained bluish or grayish silicious strata. These latter are traversed by numerous veins of endogenous granite, which are very unlike in aspect to the intrusive rock. One of these veins near Saco Pool, has a diameter of about an inch and a half, and presents on either wall a layer of yellowish crystalline feldspar about one-fourth of an inch in thickness, which includes long plates of dark brown mica. These penetrate the central portion of the vein, which is a broadly crystalline bluish orthoclase, enclosing small portions of quartz after the manner of a graphic granite. The yellowish and less coarsely crystalline feldspar with its accompanying mica, had evidently lined the walls of the vein while the center yet remained open, and had moreover entirely filled a small lateral branch. The same conditions are seen in the filling of other veins in this vicinity, which are often much larger, and present upon their walls bands of an inch or two of the yellowish feldspar with mica.

The successive filling of a granitic vein is still more clearly shown in a specimen from Sherbrooke, Nova Scotia, which I owe to the kindness of Prof. H. Y. Hind. The vein, which is seen to be transverse to the adherent fine-grained mica-schist, has a breadth of nearly four inches, about two-thirds of which is symmetrical, and is included between two layers, perpendicular to the walls, consisting of a fine-grained mixture of white feldspar and quartz, each about one-fourth of an inch thick, and marked by subordinate zones, more or less quartzose. Within these two bands is a coarser aggregate, consisting of two feldspars, with some quartz and muscovite, plates of which, and crystals of pink orthoclase penetrate an irregular layer of smoky quartz varying from one-eighth to one-half an inch in diameter. This fills the center of the symmetrical portion of the vein, on one side of which is the mica-schist, while the other