subangular character of most of this fragmental débris and the rounded character of some, however, favors the idea that the fragments were well indurated at the time of origin of the breccia but suffered some sorting or transportation by water.

The writer confesses his inability at present to formulate a satisfactory genetic hypothesis for these cherty fragmental rocks. The thesis here proposed, however, is that ordinary terrestrial processes of erosion and stream transportation had little if any direct part in the formation of either the cherts or the fragmental cherty rocks; that both the fragmental cherty rocks and the overlying massive chert are of primary marine origin; and that the fragments and matrix of the breccias and conglomerates are essentially syngenetic. It is hoped that at some time a comparative study may be made of thin and polished sections of the cherts and chert breccias of this horizon from a number of localities in interior and northern Alaska. Specimens recently collected by the writer in northern Alaska show much vein quartz as well as chert among the pebbles of rocks that appear otherwise to be essentially the same as the rocks of this chert formation of interior Alaska. These variations may give rise to new ideas regarding the genesis of such rocks. In the meanwhile, it is hoped that field studies will continue to contribute stratigraphic and paleontologic data bearing on these questions.

STRUCTURE AND THICKNESS

The rocks of this formation exposed at the north end of Calico Bluff are relatively little deformed and constitute the lower beds of a synclinal basin that plunges gently about N. 30° W. The highest of these rocks underlying the Calico Bluff formation at the north end of the bluff are shown in Plate 8, B. The lowermost beds crop out downstream, north of Calico Bluff, for 6,000 feet in a low bench along the west bank of the river. The base of this sequence is nowhere exposed in this vicinity, and the assignment of the chert breccia and conglomerate farther downstream as the base of this formation can not be substantiated from any data obtainable along the river. This correlation is based on comparative stratigraphic studies made by the writer 58 at other localities in the Yukon-Tanana region, particularly in the Livengood district of the Fairbanks quadrangle.

Few structural observations of the chert breccia and conglomerate in the Woodchopper and Coal Creek areas are available, but the areal distribution of the two belts there present, one north and one south of the Middle Devonian Woodchopper volcanic rocks, suggests strongly the existence at this locality of an anticlinal fold whose

58 Mertie, J. B., jr., The gold placers of the Tolovana district: U. S. Geol. Survey Bull 662, pp. 239-244, 1917.

major axis of elongation trends northwestward. Unfortunately a band of Lower Cretaceous rocks, overlying unconformably these basal Mississippian rocks, lies between the Middle Devonian and the lower Mississippian rocks on both Coal and Woodchopper Creeks, completely concealing the contact between the two formations. The hills, however, are low and the exposures poor in the Woodchopper area, and it is doubtful if much information regarding the stratigraphic relations could be gained even if the Lower Cretaceous rocks were absent. The great lithologic differences between the Middle Devonian Woodchopper and the Mississippian rocks, and the evident kinship of the chert formation with Mississippian rocks, as exhibited at Calico Bluff, favor the belief that a stratigraphic and perhaps a structural break separates the base of this chert formation from the underlying Middle Devonian rocks.

The rocks of this sequence 10 or 12 miles below Thanksgiving Creek appear to dip in a general northwesterly direction and are believed to underlie the volcanic series adjoining them downstream. It is possible that these rocks form the northwest end of the Woodchopper anticline, plunging northwestward at this point under younger rocks.

Only at the north end of Calico Bluff has it been possible so far to measure the thickness of these rocks, and there, of course, only a part of the sequence is exposed. Here about 1,700 feet of strata crop out below the arbitrary line which is assumed as the base of the Calico Bluff formation, but the base of the section is not exposed. Considering the amount of chert breccia exposed in Coal and Woodchopper Creeks, particularly in Coal Creek, it is probably safe to assume that the total thickness of this formation is several thousand feet, but any more exact statement is not warranted.

AGE AND CORRELATION

Formations consisting dominantly of chert and siliceous slate and argillite have been found at several localities in interior and northern Alaska and have in general been difficult to place stratigraphically, for the following reasons: They resemble one another lithologically; they are usually incompetent, so that critical structural data are lacking; they rarely yield diagnostic fossils; and they are so far separated from the standard sections of the United States that the ordinary stratigraphic nomenclature can at best be applied only in the most general way. The formation of chert and related rocks appears to have been more than ordinarily prevalent in Devonian and Mississippian time, and where such rocks occur at two or more horizons in the same general area, as in the Eagle-Circle district, their correct differentiation becomes difficult. The chert formation.

that underlies the Calico Bluff formation illustrates some of these difficulties. At its type locality its structural relation to the Calico Bluff formation is unequivocal, but its lower stratigraphic limit has not been recognized, and fossils are absent. Across the Yukon, below the mouth of Shade Creek, a few nondiagnostic fossils have been found, but the structural relations to overlying or underlying formations can not be discerned. And 75 miles to the northwest, in the Woodchopper area, occur more chert and chert conglomerate which with no fossil evidence and little stratigraphic evidence are interpreted as the base of this formation. Obviously, the correlation of these and other chert occurrences along the Yukon as a single formation is based more upon vague geologic interpretation than upon facts, particularly as at least one other chert formation, of Middle Devonian age, is also present in this same area.

The chert formation here considered is definitely known to underlie conformably rocks of upper Mississippian age. It is believed to be no older and perhaps younger than the Rampart group and Circle volcanics, which in this paper are considered to be lower Mississippian. Although elsewhere in Alaska similar cherty rocks have been termed Mississippian or Upper Devonian, the structural and stratigraphic relations of this formation to adjoining formations along the Yukon indicate a much closer kinship with the Mississippian than with the Devonian rocks of interior Alaska; and therefore this formation along the Yukon is referred to the lower Mississippian, not with the purpose of indicating a close correlation with the Waverly or Madison epochs of the United States, but merely to show that these rocks are younger than the Woodchopper volcanics, older than the Calico Bluff formation, and possibly in part contemporaneous with the Rampart group and Circle volcanics.

Fossils were found in the siliceous slate and chert at the mouth of Shade Creek by E. M. Kindle, in 1906. This collection (No. 844) has recently been examined by G. H. Girty, of the United States Geological Survey, and has been found to contain Enchostoma sp. and Lingula cf. L. spatulata. These fossils are regarded by Girty as possibly Devonian but more probably Mississippian. Inasmuch as such forms have not been found either in the Middle Devonian or in the upper Mississippian faunas of the Yukon, it would seem quite appropriate to regard them as lower Mississippian, to conform with the age assignment which the stratigraphic and structural relations indicate.

The rocks of this formation are correlated primarily with a similar "chert formation" in the Tolovana district,50 about 225 miles to the west. At that locality the base of the formation was composed of

50 Mertie, J. B., Jr., Gold placers of the Tolovana district: U. S. Geol. Survey Bull. 662, pp. 239-244, 1918.

chert conglomerate and chert breccia, which may be correlated with the rocks seen on Coal and Woodchopper Creeks. The upper beds were mainly chert, with some siliceous slate and shale and silicified limestone; and this part of the sequence is believed to be the lithologic counterpart of the chert underlying the Calico Bluff formation. The "chert formation" of the Tolovana district was described originally as not older than Middle Devonian nor younger than Pennsylvanian, but fossils subsequently found indicate that it is Mississippian. Therefore it is correlative paleontologically also with the rocks at the north end of Calico Bluff and at the mouth of Shade Creek.

In northern Alaska a somewhat similar formation of chert conglomerate with interbedded layers of quartzite in the lower part and slate and shale in the upper part was described originally by Schrader 60 under the designation Stuver "series " and was considered pre-Devonian. Subsequently the Stuver "series" was shown by Smith and Mertie 1 to overlie the Upper Devonian rocks and to underlie the upper Mississippian rocks of northern Alaska and was therefore assigned to the basal part of the Noatak formation, of lower and upper Mississippian age. This basal part of the Noatak formation, previously called the Stuver "series," is therefore believed also to be approximately correlative with the chert formation of the Yukon. In this connection, however, it should be remembered that northern Alaska constitutes a different geologic province from interior Alaska, and it is therefore not surprising that the Noatak formation differs materially in lithology from rocks of similar age on the Yukon.

Other formations composed mainly or partly of chert are known elsewhere in Alaska, but accumulative data tend to show that rocks of several ages may be of this general lithologic character. The Upper Triassic chert of northern Alaska, described by Smith and Mertie, is a good example of one of the newly discovered chert formations that can not be correlated with the chert formation of the Yukon. The Middle Devonian cherty rocks at the head of the North Fork of Shade Creek constitute another example. Hence, in the future the burden of proof will bear heavily upon the geologist who attempts to correlate chert formations in Alaska upon lithologic evidence alone.

CALICO BLUFF FORMATION

DISTRIBUTION

The type locality of the Calico Bluff formation is at Calico Bluff, on the west bank of the Yukon River, about 8 miles due north of

"Schrader, F. C., A reconnaissance in northern Alaska: U. S. Geol. Survey Prof. Paper 20, pp. 60-62, 1904.

Smith, P. S., and Mertie, J. B., jr., Geography and geology of northwestern Alaska: U. S. Geol. Survey Bull. 815, pp. 155-157, 1930.

Eagle. (See pl. 8, B.) The rocks of this formation are also exposed in a narrow zone on the north bank of the Yukon north of Calico Bluff, and this zone apparently continues N. 30° W. into the lower valley of the Tatonduk River and beyond in the same general direction. Another narrow belt of the same rocks occurs just west of the mouth of the Seventymile River and also continues N. 30° W., cropping out again on the west bank of the Yukon about opposite the mouth of the Tatonduk,

LITHOLOGY

The rocks at Calico Bluff consist essentially of alternating beds of limestone and shale with some slate and, being only gently folded, afford an excellent opportunity for measuring a detailed section.

FIGURE 5.-Sketch of south side of Calico Bluff, Yukon River, showing folded Mississippian rocks. a, b, c, d, and e indicate some of the more prominent beds. A ragged ridge of Middle Cambrian limestone on the opposite side of the river is shown at the right

For convenience in referring to this section, a sketch of the south side of the bluff is included as Figure 5, in which some of the more prominent beds are designated by letters. The section given below was measured instrumentally from the base of zone b to the top of the bluff. The section from zone b down to the water's edge at the eastern point of the bluff was not measured instrumentally, but the thickness of beds involved was estimated by two roughly quantitative methods, which checked each other very closely. The limestone and shale are thicker at the base of the section and become progressively more thin bedded toward the top. As a result, the upper part of the section consists of a great number of thin beds of alternating limestone and shale, some of which have here been grouped together. It is doubtful whether any two sections of Calico Bluff made by different men would match, for much of the variation is more apparent