butte-like mass into the Eocene. If there had been a mountain chain, as Dr. Lerch maintains, with the Eocene deposited immediately thereafter, before erosion had degraded the limestone, the Cretaceous rock at this place should represent either a dome or an anticline, but such is not the case. Apparently the most logical explanation of the relation of the Cretaceous to the Eocene is that a land period followed the close of the deposition of the rocks belonging to the former series.

THE TERTIARY.

THE EOCENE, ITS CHARACTERS AND DISTRIBUTION.

The Eocene of Louisiana has been the subject of more or less study ever since Conrad, in 1834, announced the existence of strata of that age in the State. These beds are composed of lignitic clays and sands, the sands often cross bedded, with the interstratification of beds bearing a littoral fauna; the fossiliferous beds are occasionally impure limestone. Apparently beds in one place bearing marine fossils may in other places be represented by beds of the same age devoid of animal remains; therefore it is not at present possible to subdivide and correlate with accuracy all of the strata belonging to the Eocene period found in Louisiana.

The following are approximately the limits of the area occupied by the strata belonging to this period in Louisiana: On the north and west the Eocene of this State is continuous with that of Arkansas and Texas, on the south the boundary is formed by the Grand Gulf Oligocene. This Oligocene parting runs from a few miles south of Rosefield, in Catahoula Parish, by Centerville, in the same parish, crossing Little River 5 or 6 miles below Georgetown, reaching the Red River 5 miles north of Colfax, in Grant Parish, and the Sabine River near the mouth of Bayou Negrut.'

Dr. Lerch has divided the Eocene as follows:

Cocksfield Ferry beds (equivalent, in a general way, to the Claiborne sands of Alabama).

Lower Claiborne, including Ostrea sellaformis beds, Lisbon beds, and Buhrstone. Lignitic.

Some of the above data are taken from Hopkins: First Annual Report of Louisiana State Geolog ical Survey, p. 99, for 1869, published 1870.

LIGNITIC STAGE.

If this formation occurs in Louisiana, it is only in the northwestern corner of Caddo Parish and probably at Shreveport. I have seen at Port Caddo Landing, in Harrison County, Tex., which adjoins Caddo Parish, strata that I consider Lignitic. If this point is connected with the point where Mr. Harris's Lignitic-Claiborne parting, as shown on his map of southern Arkansas, reaches the Arkansas-Louisiana State line, it will be seen that the northwestern corner of Caddo Parish is most probably Lignitic. I have not been able to examine that area.

LOWER CLAIBORNE STAGE.

This stage covers by far the largest area of any of the subdivisions of the Eocene in Louisiana, extending south from the Arkansas line to Georgetown, on Little River, to St. Maurice, at the mouth of Saline Bayou, on the Red River, and to Provencal, on the Texas and Pacific Railway. Lower Claiborne fossils have been found in Bossier, Webster, Claiborne, Bienville, Jackson, Winn, Natchitoches, and Grant parishes. In Alabama the Lower Claiborne is divided into

Ostrea sellæ formis beds.

Lisbon beds.
Buhrstone.

In Mississippi Dr. Hilgard3 divides the Claiborne, below the Lignitic bed at the base of the Jackson, into

Calcareous.

Siliceous Claiborne.

Smith writes:

From the section given in Hilgard's report, it seems that the middle part of what we have called the Claiborne series, containing the great number of Ostrea sellæformis, are the beds of the Calcareous division best developed in that State."

The Siliceous Claiborne represents in large part the Buhrstone of Alabama. This formation in Alabama and Mississippi receives its name from the peculiar lithologic characters of its constituent rocks, which find no counterpart in Louisiana. As the Buhrstone in these two States is not characterized by well preserved fossil organisms, it is not possible to make any precise correlation in the Louisiana section. The region in which Lower Claiborne fossils have been found has already been indicated. The following are the general lithologic characters presented by the stage in Louisiana. In Caddo and De Soto parishes and in Natchitoches Parish" as far south as Victoria this for

1 Am. Geologist, Vol. XVI, pp. 304-309, November, 1895.

2G. D. Harris: Am. Jour. Sci., April, 1894. See also Smith and Johnson: Bull. U. S. Geol. Survey No. 43, 1887.

3 Agric. and Geol. of Miss., 1860, p. 108.

4 Bull. U. S. Geol. Survey No. 43, 1887, p. 25.

The author is not altogether certain about Smith's correlation.
This part of the Lower Claiborne is Hilgard's Mansfield group.

mation is represented by lignitic sands and clays devoid of marine fossils, but lignitiferous strata occur throughout the whole area of the Lower Claiborne. In Bossier, Claiborne, and Bienville parishes fossils are found as casts in sandstone and as ferruginous replacements. In Webster, Bienville, and Jackson parishes the fossils are usually obtained in glauconitic sands, although in the two last parishes they are occasionally present in yellowish clay. Farther south, in the extreme southern portion of Bienville Parish and in Natchitoches and Winn parishes, the fossiliferous beds are very calcareous and are rich in fossils, which are usually poorly preserved. Ostrea sellaformis is often extremely abundant. There are slight paleontologic differences to be found in the fauna contained in the calcareous clays or clayey limestones of northwestern Winn Parish and of Natchitoches Parish, as compared with the fauna found in the glauconitic sands to the northwest; but both faunas are beyond doubt Lower Claiborne, and they are very closely related. From the southeast dip of the Eocene of Louisiana it appears very probable that the calcareous beds above alluded to represent a horizon a little higher than the glauconitic beds to the north. Within the calcareous area at St. Maurice, Robertsville, and Georgetown Lower Claiborne fossils were obtained from glauconitic sands or greenish clays. These beds, from the close resemblance of their fossils to those found in the calcareous beds, in all probability represent local lithologic differences in the horizon to which the calcareous beds belong. The Lower Claiborne formation rests conformably upon the Lignitic, and passes conformably into the Cocksfield Ferry beds above.

Of Dr. Lerch's division the following belong to the Lower Claiborne:

Jackson (in part).

Arcadia clays.

Upper Lignitic (in part).

Marine Claiborne.

Lower Lignitic (at least in part).

The sections shown in figs. 1, 2, 3, 4, 5, and 6, Pl. I, indicate the general characters of the stage.

It is difficult to decide whether the strata exposed at Shreveport should be considered Lignitic or Lower Claiborne, although they are probably Lignitic.

The Mansfield group of Hilgard should in all probability be referred to the Lower Claiborne. There is in the Mansfield group, so far as we know, a complete absence of animal remains, the formation being composed of lignitiferous sands and clays. The characters of the strata, however, are lithologically the same as those of the lignitiferous beds that occur within the northern part of the Lower Claiborne area between the Ouachita and Red rivers. The marine fossils found between these two rivers are always both overlain and underlain by lignitic strata. The Mansfield area lies south, southwest, and west of

1 Fig. 1 is probably Lignític.

a part of the Lower Claiborne area, in which marine fossils have been found. Making an approximate determination of the strike of the former beds, which is either north or northeast, the fossiliferous Lower Claiborne of Bossier, Webster, Claiborne, and Bienville parishes would be along the strike line of the Mansfield. Lower Claiborne fossils have been found in Harrison County, Tex.,' northwest of the outcrops of the Mansfield; they are found north, northeast, and east of the outcrops of the formation under discussion in parishes already mentioned, and to the south the Mansfield in Natchitoches Parish dips under the Lower Claiborne at Victoria. The opinion has been expressed that the beds bearing marine fossils, found in the vicinity of Victoria, Provencal, Natchitoches, St. Maurice, etc., are stratigraphically above the fossiliferous beds found in Bossier, Webster, Claiborne, and Bienville parishes. West of the outcrops of the Mansfield, in Texas, according to L. C. Johnson, is Claiborne. As the topography of the Mansfield does not represent a basin, such as would suggest that the Lower Claiborne had been eroded off and the Lignitic exposed, it seems to the writer that the Mansfield is only a portion of the Lower Claiborne devoid of marine fossils.

4

The presence of such mollusks as Ostrea divaricata, O. sellæformis, Anomia ephippoides, Borsonia biconica, Latirus moorei, Fusus mortoni var. mortoniopsis, etc., in the Claiborne strata of Louisiana, leaves no doubt as to what the homotaxial equivalents in Mississippi, Alabama, and Texas are. The corals are, if possible, even more decisive than the mollusks; but as only a few of the species have been described, and as I wish to publish descriptions of them in my paper on the "Eocene corals of the United States," their names are not here cited.

The age of these beds has already been pointed out by Johnson,5 Lerch, and Harris, but all three have erred in making the Lower Claiborne area too small and the Jackson area too large.

Lower Claiborne fossils were collected by the Louisiana Geological Survey at Georgetown, on the railroad from Monroe to Alexandria, near Little River. This shows the existence of Lower Claiborne strata farther south than has hitherto been recorded. Some of the species

are:

Volutilithes petrosus Con.

Latirus moorei Gabb.

Lunatia eminula Con.

Venericardia planicosta Con.
Corbula sp.

1Lawrence C. Johnson. The Iron Regions of northern Louisiana and eastern Texas: Ex. Doc. No. 195, H. of R., 50th Congress, 1st session, p. 21, 1888. Map.

*The Lower Claiborne is east of the Mansfield in Natchitoches, Winn, and Grant parishes. 3Ibid., see map.

4 The average altitude along the Texas and Pacific Railway from Shreveport, after passing out of the river bottom, to Provencal, is 281 feet; along the Vicksburg, Shreveport and Pacific Railroad from Shreveport, after leaving the Red River bottom, to Calhoun, it is 247 feet. These elevations are taken from the respective railroad profiles.

Ibid., p. 20.

"A preliminary report upon the hills of Louisiana, Part II, 1893, p. 82.

"Tertiary Geology of/Southern Arkansas, 1894, p. 177, Professor Harris erred in accepting the pre. vious work of Johnson on the areal distribution.

I regard the Latirus moorei as indicative of the age of the beds. Four questions need further discussion: First, the Arcadia clays; second, the small prairies or meadows in Bienville, northern Winn, and Natchitoches parishes; third, the Upper Lignitic of Dr. Lerch; fourth, the red clays in the vicinity of Mount Lebanon and Arcadia.

The Arcadia clays.-Dr. Lerch has described a series of gray clays, typically exposed in the vicinity of Arcadia, and which he considered as resting upon the eroded surface of the Claiborne strata. To these clays in his second report the name "Arcadia clays" was given, and they were considered of Jackson age.

I was first led to doubt the existence of an unconformity for paleontologic reasons. The Lower Claiborne attains a fine development in Louisiana. In Winn and Natchitoches parishes the Lower Claiborne beds, as already noted, become very calcareous, and Ostrea sellaformis is often found in the greatest abundance. These features are characteristic of Hilgard's Calcareous Claiborne in Mississippi, which has been correlated with the Ostrea sellaformis beds of Alabama by Smith. The Ostrea sellaformis beds in Alabama are separated from the Jackson by scarcely 30 feet of strata. Whether the Calcareous beds of Winn and Natchitoches parishes are the exact equivalent of the Ostrea sellaformis beds of Alabama or not, the fact remains that they are not far from the base of the Jackson, and it seems improbable that a long period of dry land surface could have intervened. Furthermore, G. D. Harris,1 in Arkansas, discovered beds "Uppermost Claibornian, or perhaps transitional between that and the Jackson," a fact to my mind making it still more improbable that there could have been in Louisiana an erosion period between the two stages.

In order to study this supposed unconformity further, in November, 1894, I went to Arcadia to examine again some sections in that vicinity. Fig. 3, Pl. I, represents a section made in the first railroad cut west of that town. The gray or Arcadia clays were found resting conformably on the black clays. The dip of both the gray and black clays was the same in both direction and amount, and the stratification was absolutely continuous from one clay to the other. From the distribution of the color one would at first be inclined to think that there was an unconformity, but in a layer not thicker than one's finger I have seen the stratum in the length of about a foot light gray or blue, then chocolate and at last black (or almost black) where it passes into the black nucleus of the cut. Fig. 4, a section made on the Louisiana and Northwestern Railway, 6 miles south of Gibbsland, represents the same phenomena. I believe that the appearance of these sections is simply the result of weathering, and that the black clays represent an unweathered nucleus.

In some cases the difference between the gray and blackish clays may be due to different lithologic constitution, but I have never seen any evidence that an erosion period intervened between the "Arcadia clays" and the Lower Claiborne beds.

1 Tertiary Geology of Southern Arkansas, 1894, p. 93.