A critical discussion of the conditions under which it is conceived certain eskers and sand plateaus (plains) were formed. The Auburndale district, ten miles east of Boston, presents three classes of modified drift deposits;—sand plateaus, eskers, and kames. These deposits are well exposed.

The sand plateaus have the characteristics of delta deposits of glacial streams,—even surfaces, well-bedded sands and gravels, the beds sloping outward from the "head" at an angle of 12° to 20°, and in close agreement with the slope of the plateau front, a lobate margin, deposits distinctly coarser at the head than near the front, and a series of nearly horizontal roughly cross-bedded gravels overlying the sloping beds.

The eskers are essentially of the same material as that of the plateau, often so poorly stratified as to render differentiation of the beds difficult. The interstices between the pebbles are often unfilled, although there is abundance of fine material in adjoining layers. This "open work" is taken to indicate rapid deposition, and seems to preclude the supposition that the gravels have settled down from a superglacial position, or been traversed by currents of any volume. In several instances the eskers can be followed to direct union with sand plateaus. Towards its lower end the esker frequently "gives out branches" and "the adjacent lowland surface becomes more or less encumbered with sand mounds or kames," indicating a decayed margin of the ice.

Prof. Davis' conclusions are:

"1. The eskers and sand plateaus of Auburndale and Newtonville were formed by running water just inside and outside of the ice margin in the closing stage of the last glacial epoch.

"2. The ice-sheet was a stagnant, decaying mass at the time of their formation, as is shown by the ragged outline of its margin.

"3. Eskers and sand plateaus are genetically connected; the term, feeding-esker, is fully warranted by the relation of the two in position, structure, and composition.

"4. The sand plateaus were made rapidly; this is proved by the absence of disordered beds at their heads, where space would have been opened by the backward melting of the ice had the forward growth of the plateau been slow. The eskers were also made rapidly, as is shown by their 'open-work gravels.'

"5. The diversion of the feeding streams to other outlets left the plateaus and the eskers without further energetic action as the ice melted away from them.

"6. The present form and structure of the eskers are more accordant with the supposition of a subglacial origin than of a superglacial origin; but it is not intended to imply that other eskers of more irregular form and different structure could not have been deposited in superglacial channels."

H. B. K.

The paper aims "to present some of the results of observation of the geologists of the Appalachian division during the past three years on the subject of structural geology in the Appalachian province." The structural features are all of one type but of different phases, comprised in four great districts. 1) the district of close folding, 2) a district whose chief structural characteristic is cleavage, 3) a district of open folding, 4) a district of faulting and folding. The answer to the questions, Why did the strata bend in the district of open folding, and why did they break in the district of faulting, is that the thrust affected them according to their rigidity under their respective conditions of superincumbent load. "We know that load up to a certain point restrains fracture in material under thrust." In the district of open folding the Devonian limestone is the most rigid of the strata and "the one which would most effectively transmit the compressing thrust and would control the resulting structure." In the district of open folding this limestone was prevented from breaking and faulting by a load of superincumbent strata exerting a pressure of 10,000 to 23,000 pounds per square inch, while in the faulted district a load of 5,000 to 10,000 pounds per square inch permitted the strata to break and fault.

The answer to the question, Why did the compression affect this zone, is given. "It becomes apparent on study of sections that where compression raised a great arch there previously existed a bend from a nearly horizontal to a descending position in the principal stratum transmitting the thrust. Greater anticlines and synclines originated in upward and downward convexity of initial dips, due to unequal deposits of sediments which depress underlying strata in proportion to their weight. Such folds may be called original." The Pottsville, Mahanoy, Shamokin and Wyoming coal basins of Pennsylvania belong to this class.

Experiments have recently been carried on in the office of the United States Geological Survey reproducing the different forms of folding. The experiments differed from other experiments in that 1) the materials used to simulate the stratified rocks varied in consistency from brittle to plastic, according to the depth at which deformation is supposed to take place; 2) the compression was exerted under a movable load representing the weight of superincumbent strata; 3) the strata rested on a yielding base to simulate the condition of support of any arc of the earth's crust. The following are the conclusions from the experiments:

1. "When a thrust tangentially affects a stratified mass, it is transmitted in the direction of the strata, and by each stratum according to its inflexibility. At any bend the force is resolved into components, one radial, the other tangential to the dip beyond the bend; the radial component, if directed downward, tends to depress the stratum and displace its support.

2. "A thrust so resolved can only raise an anticline or arch which is strong enough to sustain the load lifted by its development; such an arch may be called competent; and since strength is a function of the proportions of a structure, it follows that, for a given stratum, the size of a competent anticline will vary inversely as the load; or for a given load the size will vary as the thickness of the effective stratum.

3. "The superincumbent load borne by a competent anticline is transferred to the supports of the arch at the points of inflection of the limbs.

4. "When a competent arch is raised by thrust from one side, the load transferred may so depress the resulting syncline further from the force that an initial dip will be produced in otherwise undisturbed strata; this dip will rise to a bend from which a new anticline may be developed. This anticline is a result of the first, and may be called 'subsequent' in distinction to original folds. Since subsequent folds are simply competent structures, their size will be determined by conditions of thickness and load, and for like conditions they should be equal; and they must, in consequence of conditions of development, be parallel to the original fold and to each other. An example of an original fold with its subsequent anticlines is the Nittany arch and the group of parallel anticlines which lie southeast of it, extending northeast from the Broad Top basin."

C. E. P.

The post-Tertiary trenches of the Hudson and its tributaries are in the main filled with clay beds, which, covered by a thin deposit of sand, rise in terraces 130, 150, or even 180 feet above tide-water. These clays are the result of a late glacial or post-glacial submergence of the valley, but their upper surface does not indicate the amount of their submergence, as they are bottom deposits. Delta deposits made by the tributary streams, where they entered the Hudson estuary, would indicate the amount of submergence.

Such deposits are found on the Catskill a mile north of Cairo, and eroded remnants are traceable for three or four miles down stream. The surface is characterized by great numbers of water-worn stones up to fifteen or eighteen inches in diameter. The lobate margin, where present, is poorly defined. These deposits range from 290 feet (aneroid) above tide, up river, to 270 feet further down. One-tenth of a cubic mile of material seems to have been washed into the Catskill trench at the point of this delta between the time of the ice departure and the elevation of the land. Subsequent terracing has removed half that amount.

The course of the Catskill at Leeds, where it crosses a ledge of hard Corniferous limestone is probably of post-glacial superimposed origin, but the preglacial valley cannot be definitely fixed.

H. B. K.

This report covers an area of 5950 miles, two-thirds in Alabama. Topographically it falls into three divisions: 1) the Cumberland and other plateaus of the northwest; 2) in the center, anticlinal valleys—Browns and Wills, with the synclinal mountains—Sand and Lookout; 3) the monoclinal mountains, the "flatwoods" (Coosa shales) and the chert hills (Knox limestone) of the southeast. The drainage of the first is radial from the center of the plateau to the Tennessee; that of the second, once consequent upon the folded structure, is now adjusted to the strike of the soft beds.

The formations are Cambrian, Silurian, Devonian and Carboniferous. Total thickness is from 13,000 to 18,000 feet in the east, but decreases westward. Hard sandstones of the Carboniferous form the cappings of the plateaus and synclinal mountains. In the anticlinal and monoclinal valleys the Silurian and Cambrian appear. The rocks pass from the nearly horizontal beds of the plateau region, by narrow unsymmetrical anticlines with steeper dip on the northwest side, and by broad shallow synclines, to the complicated folds of the southeast. The axes of these latter folds dip more or less abruptly northward and southward, causing the ridges to assume zigzag courses. Synclines are often crossed by anticlines.

Thrust faults exist, some of great magnitude, and traceable for 200 to 300 miles. By the "Rome thrust fault" the Cambrian shales have been shoved four to five miles over upon the Carboniferous shales. Most of the overthrust strata have been worn away, but tongues of Cambrian shale still remain to all appearances lying conformably upon the Carboniferous strata. Transverse thrust faults terminate Gaylor's ridge, Dirt Seller Mountain, and Lookout Mountain on the south.

H. B. K.

References have been made in Geological literature to the beaches of the eastern portion of the Lake Erie basin, but up to the time of Mr. Leverett's work none of the beaches had been completely traced. Mr. Gilbert had discovered that several of the raised beaches do not completely encircle Lake Erie, and supposed that their eastern termini represent the successive positions of the front of the continental glacier during its retreat northeastward across the Lake Erie basin. Mr. Leverett verifies this theory by demonstrating that certain moraines are the correlatives of the beaches. They are as follows:

I. The Van Wert or upper beach and its correlative moraine, the Blanchard ridge. II. The Leipsic or second beach and its correlative moraines. III. The Belmore, or third beach and its correlative moraine.

I. The Van Wert beach extends eastward from the former southwestward outlet of Lake Erie near Fort Wayne, Indiana, to Findlay, Ohio, where it joins the Blanchard moraine. Through Indiana and Ohio its altitude is quite uniformly 210 feet above Lake Erie.

While the Van Wert beach was forming, the ice front was the northeastern shore of the lake as far east as Findlay, Ohio, its position being marked by the Blanchard moraine. East of Findlay, where the Van Wert beach joins it, the moraine is of the normal type. But west of Findlay, it presents peculiarities of topography and structure, resulting from the presence of lake water beneath the ice margin. The water was shallow and incapable of buoying up the ice-sheet, and producing icebergs. The motion of the water under the ice-sheet produced a variable structure. This is the only instance of a moraine demonstrably formed in lake water.

II. The Leipsic, or second beach, was formed after the ice had retreated from its position marked by the Blanchard moraine. Its altitude is 195 to 200 feet above Lake Erie. It has its terminus near Cleveland, where it connects with the western end of a moraine.

III. The Belmore beach and its correlative moraine. Between the Leipsic beach and the present shore of Lake Erie are several beaches. One of these, the Belmore beach, terminates near Cleveland, while the others extend into southwestern New York, and probably connect with moraines, though this connection has not been traced. The general altitude of the Belmore beach in Ohio is 160 to 170 feet above Lake Erie. Unlike the Van Wert and Leipsic beaches, it does not directly connect with a moraine at its eastern end, but a gap of ten miles intervenes. Terraces at Cleveland, Mr. Leverett thinks, make a connection between the eastern end of the beach and the western end of the moraine at Euclid, Ohio.

C. E. P.

Temperature of the Sea.—The temperature of the English Channel was similar to that where the isotherm of 32° F. is now situated. The winter temperature can scarcely have been 20° colder than at present. The Mediterranean was perhaps 5° colder than now.

Temperature of the Land (air).—It does not appear that the climate of the lowlands of southern Europe can have been 20° lower than the present mean; 10° or perhaps less appear to have been the refrigeration in the Mediterranean region. The temperature at the southern margin of the ice-sheet was about 20° colder than at present. The temperature increased rapidly towards the south. Recent observations seem to show that throughout central Europe there was a period of dry cold, causing the country to resemble the arid regions of central Asia.

J. A. B.

Geologists generally admit that there have been at least two glacial epochs, separated by one well-marked interglacial period. The closing stage of the Pleistocene period was one of cold conditions in northwestern Europe, accompanied by land depressions. After this came a genial climate with a union of the British islands among themselves and also with the continent. This was followed by a cold, humid condition.

Upham maintains that the whole of North America north of the Gulf of Mexico stood at least three thousand feet higher at the beginning of the glacial epoch than at present. Fiords were formed before glacial times and so can not be cited as evidence of high land during the glacial period. An elevation of land in the northern part of North America and Europe could not produce glaciation in their southern parts. The deflection of the Gulf Stream by the sinking of the Panama, Professor Geikie argues, could not produce the conditions which prevailed during the glacial epoch. The Earth-Movement hypothesis, he believes, accounts neither for the widespread phenomena of the ice-age, nor for the remarkable interglacial climates. Some maintain that the warm interglacial period was produced by the rise of the Panama land, the sinking of the lands to the north, and the turning of the Gulf Stream from the Pacific into the Atlantic. Why then, asks Professor Geikie, do we not have such a climate now?

J. A. B.