have been found at all elevations over the mountain. As they stop, at the mountain foot, and as the till on the flanks and top is angular and from the immediate neighboring formations, or consists of large bowlders of the sandstones to the north, it is evident that the lower part of the ice moved parallel to the river and mountain, while the upper part crossed the ridge and probably united with the lobe that came from the east along the enclosed valley south of Nescopeck Mountain. The glacial gravels of the Juniata valley. While studying the deposits in the valley of the Susquehanna, the writer followed a line of drainage over the low divide between the Susquehanna at Selin's Grove and the Juniata at Lewistown; but was surprised to find on the Salina bluffs south of the river, at the latter place, a series of lenticular deposits of glacial gravels and bowlders in the depressions and on the down-river side of the elevations of those bluffs, and at heights up to 80 feet above the river. Their position pointed to an origin up the Juniata; but as I was engaged in regular work, I let that matter go for the time; but looked up the authorities and found that I. C. White, in his report for Huntingdon county in the recent survey of the State, noted their appearance in that county, and advanced four theories for their origin; but without favoring either of them; while those who worked in Mifflin and Juniata counties said nothing of the occurrence of gravels at such high elevations. As Professor White* had also reported upon the counties along the north branch of the Susquehanna, he was in a position to speak regarding glacial deposits, and in the Huntingdon report (p. 31) he says, "Great heaps of bowlder trash, both rounded and angular, are often to be seen along the valleys of the principal streams; and these often very much resemble genuine drift heaps; but no striated bowlders or striated fragments of the country rock are to be found in the region." He further reports them at elevations above 100 feet from the river. The hypotheses advanced are various, and the only one of value is that of the melting snows of the glacial period that kept all the rivers of the State at a higher level than at present. Nothing is said of glacial occupation of the region. *Justice to Professor White requires the statement that his work in the recent geological survey of Pennsylvania was confined to counties widely separate from one another; so that the study of a limited portion of the Juniata valley would afford no clew to the origin of these gravels, had it not been supplemented by extended work over a large adjacent area. The writer was fortunate in having the careful and accurate work of Professor White to guide him in the study of the north branch of the Susquehanna, and has many times regretted that it was not extended to the south and west; as those regions were treated from an entirely different standpoint. It is to be noticed that nothing is said of any difference in age between these drift heaps and those of the great moraine. There is no difference, in fact, and as the writer was happy in finding glaciated stones at Lewistown, he set the deposit down as a recent glacial formation, and as recent as the glacial gravels of the upper Susquehanna. A glance at the map of Lewis and Wright shows that the terminal moraine approaches quite near the north branch of the Susquehanna at Williamsport. As the width of the extramorainic deposits will average 25 miles, the writer saw, when he reached the latter place, that Bald Eagle Mountain, rising sharply 1200 feet above the river, must have required a great thickness of ice to have been surmounted, and must have formed a glacial lake similar to that in the Lehigh valley, as has been described by the writer. In proof of this the surface deposits west of Muncy changed to slack water clays and overlaid, where visible, modified and unmodified drift. A trip through the gap in Bald Eagle Mt., to the inner summit of White Deer Mt., lying three miles to the south, and making a loop with the former where they meet the Susquehanna, showed that ice had surmounted both, and, as the glacier had crossed the main Susquehanna at Fisher's Ferry, at the northern slope of Little Mt., there was at least fifty miles of the bed on the stream covered with ice over 1200 feet thick. Previous to this a line of glacial discharge had been traced up Penn's Creek to Paddy Mountain; but its origin was unknown. Going west up the valley of the north branch of the Susquehanna, a great slack water terrace 150-170 feet high was found opposite Jersey Shore, which causes the river to make a long bend to the north and east. This was formed by a torrential discharge of glacial origin from Nippenose valley through Anti's gap in Bald Eagle Mt. Rogers has called attention to the peculiar features of the erosion of the south side of this valley, and to the "drift hill" in the gap. He also notes a similar hill in Washington gap between Sugar and Nittany valleys. The Jersey Shore terrace surprised the writer by bending to the northwest, and the stream that drains Nippenose valley was also surprising in meeting the Susquehanna in a similar way, or against the present flow of the river. Both pointed to a flow westward. A glance at the map showed that the Bald Eagle and Nittany valleys would afford an exit to deep waters, and the latter valley was first studied with the result that no signs of ice occupation were found; but it was seen that Sugar and Brush valleys had their eastern ends covered with ice, and that the northern end of Nittany valley was covered deeply with bowlder trash, as was the part adjacent to the gaps from the valleys to the east. From Bellefonte east ward to the valley of Penn's creek the surface was clear, but the glacial flow through that creek was seen to have come from Brush valley. At Mill Hall twelve feet of slack water deposits overlie what seems to be sub-glacial till, and the Bald Eagle valley soon becomes perfectly flat and filled with debris, from which the small knolls formed by varying outcrops appeared in a submerged state. These deposits were followed continuously over the divide at 1110 A. T., or 600 feet above the river at Williamsport, and down the Juniata to Huntingdon. Whereever they exist at great depths in the latter valley there is evidence of slack water, as below them there is a contraction of the channel, where it passes through the gaps in the numerous ridges. When we consider that the whole discharge from the central part of the state glacier passed through Bald Eagle and Juniata valleys, we can see at once that small ice dams must have been continually forming in these narrow gorges, from the floating bergs, and as these deposits are found from the present level of the river to an average of 80 feet above it, we can get an idea of the depth of the water. In one case, a dam sent cakes of ice from 300-400 feet up the slope of Warrior Ridge to make a small deposit of erratics. That the discharge was torrential is seen by the erosion of the rotten limestone at Barree on the up-stream side of the ridge, and its retention on the other side to great depths. We can imagine, therefore, during a portion of the earliest part of glacial times, a deep lake extending from Williamsport above the junction of the Susquehanna and Sinnemahoning; filling Nippenose, upper Nittany and Bald Eagle valleys, and discharging a deep torrent through the Juniata. A smaller side discharge from the same glacier went down Penn's creek to the Susquehanna; but was stopped by the ice at Shade Mt. for a time and sent down to Lewistown, on the Juniata; but afterwards was freed and allowed to reach the Susquehanna at Selin's Grove. This great glacial lake has been called by the writer, in honor of our State geologist, Lake Lesley. Detailed sections were photographed at various places through the Bald Eagle and Juniata valleys. At the northern entrance of the former, at Mill Hall, there is a considerable flood cone attached to the southern side of the gap in Bald Eagle Mt., where the flow was turned south from the Nittany valley. Along the Central railroad of Penn., through the latter valley, there exist thick deposits of glacial drift distributed from the gaps to the east, and the larger material lies close against the Nittany Mt. on the eastern side of the valley, while the fine clays extend over the whole valley. The flood cone at Mill Hall, therefore, does not contain gravel in its upper, slack-water part; but only clay and fine fragments of the local Devonian rocks. This lies on what seems to be subglacial till; but the exposures were so small, and the time for study so short that it may have been a hard clayey modified drift. At any rate it was dense and unstratified. One argument in favor of its being till was that it lay on rotten local rock from which the old surface had been removed. Going south along Bald Eagle valley the inequalities of the valley are covered by gravel in the place of the Mill Hall drift, and that is capped by slack water sands and clays-all unstratified. This shows that the action was continuous. On crossing the divide the modified drift continues at varying thicknesses; but always unstratified, and the slack water clays of the northern side are replaced by sands and fine stuff as the torrential nature. of the flow gradually ceased. An excellent section is seen in the creek at Vail-about 125 feet below the divide, and the same distance above the Juniata at Tyrone. At East Tyrone, in the yard of the Pennsylvania R. R., many fine sections are shown, and the succession is similar to those described. Huntingdon is built on a flood cone where the Juniata pours through a narrow gorge into a wide cross valley. ART. XVIII.-The Succession of Fossil Faunas at Springfield, Missouri; by STUART WELLER. Introduction. THE rocks under consideration in this paper, are those lying above the yellow Chouteau limestone in the region in question. They consist of beds of gray limestone, more or less crystalline in the various strata, with scattered bands of lenticular chert concretions. No results of careful stratigraphical work, based on palæontological evidence, have ever been published in connection with this region, and though the rocks have been correctly considered as equivalent to that part of the Mississippian Series known in Illinois and Iowa as the Burlington and Keokuk Groups, the subdivisions of the strata have never been properly pointed out. The object of the present paper is to dissect out the various sub-faunas, to note their order of succession and the relations which they hold to each other. During the summers of 1892 and 1893 the writer made quite extensive collections of the fossils from the quarries and other exposures in and about Springfield. From the study of these collections, the whole series of strata has been divided. into twelve zones, which can be distinguished by their faunas. AM. JOUR. SCI-THIRD SERIES, VOL. XLIX, No. 291.-MARCH, 1895. The limits of these zones, however, are not sharply defined, as lower faunas change to the ones above through gradual transitions. As the fauna of the Keokuk beds in Illinois and Iowa has been shown to be simply a later product of the same line of development which took place during the deposition of the Burlington beds, so here in Southwest Missouri, the upper faunas, which represent the Keokuk, are only a continuation of the development seen in the lower beds. Prof. H. S. Williams has suggested the name Osage Group* for this whole series of strata in Missouri, while Mr. C. R. Keyes has adopted the name Augusta Groupt for the same deposits. With the exception of the lowest beds, the Springfield limestones resemble their more eastern equivalents in Illinois and Iowa, both in lithological and in faunal characters. The great difficulty experienced in this study, has been the lack of any continuous section comprising all the beds of the group, such as is shown in the Mississippi River section. Some twenty-five localities were studied, and from them more or less complete collections of the fossils were made. Most instructive in the way of showing the succession of the faunas, is a series of sections along the Gulf railroad, from Jones Spring, four miles southeast of Springfield, to the Gulf railroad shops in the city, then turning to the south and extending along Wilson Creek to Mackey's quarry. These sections show the succession from near the base of the Burlington up into the Keokuk. Along this line the rocks dip gradually to the west, and in the successive cuts and quarries, starting from Jones Spring, successively higher beds are exposed. At one point, a short distance east of Springfield, a slight fault disturbs the regularity of the line of sections, otherwise it is continuous. Besides this series of sections, other more or less. isolated outcrops were studied. There being no continuous section, and the element of dip being uncertain, no accurate estimate of the aggregate thickness of these beds can be made, but they probably extend through one or two hundred feet. Description of Zones. Zone I.-This zone can be readily recognized wherever it occurs. Its lithological characters are, in general, distinct from any of the succeeding beds. It is a fine-grained, gray or bluish limestone, quite hard, and with a more or less marked conchoidal fracture. Some lenticular chert nodules occur *Bull. U. S. Geol. Surv., No. 80, p. 109. Washington, 1891. |