ciated portions of both North America and Europe, by the occurrence of comparatively small areas of ice accumulation beyond the extreme boundaries of the principal icesheets. The latter condition, or ice invasion, is indicated on the outer part of the driftbearing area eastward from Salamanca, N. Y., through Staten and Long Islands, Martha's Vineyard and Nantucket, where the soft strata beneath the ice were dislocated and folded.

15. Climatic Conditions Shown by North American Interglacial Deposits. WARREN UPHAM, Somerville, Mass.

During the times both of general accumulation and growth of the ice-sheets and of their final recession, fluctuations of their borders were recorded in various districts by forest trees, peat, and molluscan shells, enclosed in beds underlain and overlain by till. Such fluctuations, while the ice accumulation was in progress, enclosed chiefly arctic or boreal species; but when the ice was being melted away, in the Champlain epoch, the remains of the flora and fauna thus occurring in interglacial beds, as at Toronto and Scarboro', Ont., may belong wholly to temperate species, such as now exist in the same district. The cold climate of the Ice age appears thus to have been followed by a temperate Champlain climate close upon the waning ice-border.

16. Glacial Lakes in Western New York and Lake Newberry, the Successor of Lake Warren. By H. L. FAIRCHILD, Rochester, N. Y.

The paper presented evidence that the finger lakes of central New York were all pre-glacial in character and that during the presence of the ice-sheet at their outlets they were backed up and discharged southward, as is abundantly shown by deltas at various heights on both sides of the present divide. Professor Fairchild cited eighteen glacial lakes from Attica on the west to the Onondaga river valley on the east. These

he has named from important towns now on the sites, as Lake Ithaca for the glacial form of Cayuga lake, which was 35 miles long, 5-10 miles broad and 1100 feet deep. It has been long known that when the ice covered western New York the great lakes discharged at Chicago to the Mississippi and the great lake formed by them is called Lake Warren, and has left a good beach. At a much later stage, when the Mohawk was uncovered, the waters ran to the Hudson, and the great lake on the site of Ontario has been called Lake Iroquois. The intermediate stage between these two, when the discharge of the water covering western New York was through the low pass at the south end of Seneca lake through Horseheads near Elmira, Professor Fairchild has called Lake Newberry. The elevations of this and the Chicago pass are such that when allowance is made for the depressed condition of the area at that time, the existence of the lake can be demonstrated.

The paper was discussed by Messrs. McGee and Gilbert, who commended the choice of the new name as felicitous and timely. J. W. Spenser also spoke, but differed with the author in some points.

Meantime, in the upper laboratory (the Williams room), the petrographic section, under the chairmanship of Professor B. K. Emerson listened to

18. The Relation of Grain to Distance from Margin in Certain Rocks. ALFRED C. LANE, Houghton, Michigan.

A description of the variation in texture and grain of some quartz diabase dikes of Upper Michigan was given, and the same compared with effusive flows of similar mineral composition. These descriptions were based on series of thin sections of known distance from the margin. Interstitial micropegmatite is primary or pneumatolytic, and the feldspar crystallization begins before that of the augite, continuing until later. The distinction between the

intrusive or dike type and the effusive type was pointed out. The main object of presenting the paper at this time is to elicit the best methods of measuring the coarseness of grain of a rock, the object being to express by some arithmetical or mathematical formula based on statistics, or in some other definite way, the relation of texture to walls and thickness in a dike. The paper elicited considerable discussion by Messrs. Hovey, Kemp, Iddings, Cross, and G. P. Merrill, in which the following points were made; the large size of the phenocrysts in some very narrow dikes; the importance of not measuring minerals of the intratelluric stage; the great variability of circumstances under which dikes cooled, as heated or cold walls, pressure, mineralizers, etc., and the difficulties of getting reliable data of the kind required by Dr. Lane.

19. Crystallized Slags from Coppersmelting. ALFRED C. LANE, Houghton, Michigan. This paper described (with exhibition of specimens) slags from the cupola furnaces used in coppersmelting, which contained large melilite crystals, between one and two centimeters square, interesting optically and in mode of occurrence. Crystallized hematite was also noted.

The specimens elicited great interest on account of the size and perfection of the crystals.

20. On the Nomenclature of the fine-grained Siliceous Rocks. L. S. GRISWOLD, Cambridge, Mass.

The writer described the difficulties met first, in his study of novaculite, and later, in connection with other siliceous rocks, such as cherts, jaspers, etc., in applying definite names. The troublesome characters of opaline, chalcedonic and quartzose silica, as regards the origin of each, presented obstacles both for mineralogic and genetic classification.

This paper elicited an interesting discussion which threatened at times to take

This paper contained a brief petrographical notice of certain dykes of diabase containing Huronite,' as the mineral was originally named by Dr. Thomson, of Glasgow, in his Mineralogy of 1836. Dr. B. J. Harrington's re-examination of this mineral in 1886 showed some very grave errors in Thomson's work and the 'huronite' must simply be regarded as an impure or altered form of anorthite, which has undergone either partial or complete 'saussuritization,' owing to metamorphic action. Certain localities were mentioned north and northeast of Lake Huron, where these dykes have been noted cutting the Huronian as well as the granitoid gneisses usually classed as Laurentian. Mr. A. P. Low, of the Canadian Geological Survey, noticed dykes containing this mineral cutting the Laurentian and Cambrian in the Labrador Peninsula. 22. The Granites of Pike's Peak, Colorado. EDWARD B. MATHEWS, Baltimore, Maryland. (Introduced by W. B. CLARK.)

This paper gave an areal and petrographical description of the granites composing the southern end of the Rampart or Colorado range and showed that great macroscopic variation may result, while the microscopic characters remain monotonously uniform. Four types in all were distinguished, based on the size of phenocrysts and coarseness of grain. The paper was discussed by Whitman Cross and J. P. Iddings, after which the section adjourned to meet again at 4:30 P. M.

the towns of Diana, Pitcairn and Wilna, but was really a review of the relations of these rocks in a wider region and was based on extended field experience. Petrographic details were presented of the several kinds of rocks, and especially of the varieties of the anorthosites, which were shown to shade into angite-syenites, and apparently into red gneiss. Many irruptive contacts of anorthosites and limestone were cited and the location of the classic mineral localities of this region was shown to be along these contacts. The same important thesis was worked out as in the preceding two papers, that the great intrusions of the Norian series were later than the gneisses and limestones.

The papers were discussed by Whitman Cross, who called attention to the close parallelism of the geology in the Pike's Peak district of Colorado; and by C. D. Walcott who referred to his own studies in the Adirondacks and similar conclusions to those advanced.

28. Lower Cambrian Rocks in Eastern California. CHARLES D. WALCOTT, Washington, D. C.

An account of the discovery of the Lower Cambrian rocks and fauna in the White Mountain range of Inyo County, Cal. See also No. 2 above. This important discovery affords a means of correllating the early Cambrian life in the remote West with those already known in the East. 29. Devonian Fossils in carboniferous strata. H. S. WILLIAMS, New Haven, Conn. The paper described the fauna of the Spring Creek limestone of Arkansas, which lies between the Keokuk-Burlington strata below and the Batesville sandstone above, and is at about the horizon of the Warsaw and Chester of the Lower Carboniferous in the Mississippi Valley. The fossils are closely related to the carboniferous fauna described by Walcott from Eureka, Nev., and by J. P. Smith from Shasta County, Cal.

But certain Devonian forms as Leiorhyncus quadricostatum and Productus lachrymosus of the New York Devonian are found with them, which are lacking in the Mississippi Valley, but are found in the Devonian of the West. The interpretation was then made, that the Arkansas fossils indicated a Devonian incursion from the westward.

During the reading of this and the succeeding titles the petrographers reconvened in the upper laboratory, as later recorded. 30. The Pottsville series along the New River, West Va. DAVID WHITE, Washington, D. C.

This paper was a careful description of the stratigraphy of the series, the determinations being based on the fossils, which evidence was presented in full.

31. The Cretaceous Deposits of the Northern Half of the Atlantic Coast Plain. Wм. B. CLARK, Baltimore, Md.

The several formations established as a result of a detailed study of the Cretaceous strata of Monmouth county, New Jersey, were shown to have a wide geographical range towards the south. They have been traced throughout the southern portion of that State, while all except the highest members of the series are found crossing Delaware and the eastern shore of Maryland. Several representatives of these formations appear on the western shore, reaching to the banks of the Potomac. 32. Stratigraphic Measurements of Cretaceous Time. G. K. GILBERT, Washington, D. C.

The writer described a great series of Cretaceous rocks, 3500-4000 ft. thick, lying in the Arkansas River Valley, west of Pueblo, Colo. They consist of layers of limestone 1 ft. to 1 ft. 6 in. thick, separated by 1 in. of shale-this alternation being uniformly repeated through the whole thickness. The writer argued that frequent continental oscillation from deep to shallow water deposits was unlikely as having caused

the beds, and hence appealed to climatic cycles.

The cycles of a year's changing seasons is too short to account for the limestone; the next longer cycle, the lunar, involves no changes of climate; hence the cycle of the precession of the equinoxes, 21,000 years long, was selected, and allowing four feet of deposit for each cycle, this portion of Cretaceous time was estimated at 21,000,000 years. There was no discussion, but a very evident feeling of solemnity at the announcement.

33. Notes on the Cretaceous of Western Texas and Coahuila, Mexico. E. T. DUMBLE, Austin, Texas.

The author being absent the paper was only read by title.

The main section then adjourned until the presidental address at 7:30 the same evening. Meantime the petrographers listened to

and

34. Spherulitic Volcanics at North Haven, Maine. W. S. BAYLEY, Waterville, Me. In the Journal of Geology a few months ago the late Dr. George H. Williams referred to the existence of old rhyolites on the coast of Maine. The author described very briefly the occurrence of these rocks, exhibited specimens of them. The specimens showed very perfect spherulites, lithophysæ and all the common features of glassy volcanics. They brought out an interesting discussion regarding the abundance of these rocks along the Altantic sea-board. J. E. Wolff spoke of their great extent near Boston, and especially at Blue Hill, where the relations with the Quincy granite are a hard problem. A. C. Lane mentioned their frequency in central Maine, as shown by the collections of L. L. Hubbard. T. G. White referred to those near Mt. Desert. J. F. Kemp spoke of recent field and petrographic work in progress on the great areas near St. John, N. B. W. S. Yeates brought up the curious phosphatic spherulites lately

found in Georgia, which closely simulate lithophysæ, and remarks were made on them by W. Cross and J. P. Iddings. 35. The Peripheral Phases of the Great Gabbro Mass of Northeastern Minnesota. W. S. BAYLEY, Waterville, Me.

On the northern border of the great gabbro mass in northeastern Minnesota are basic and granulitic rocks whose composition indicates their relationships with the gabbros with which they are associated. The basic rocks are aggregates of the basic constituents of the gabbro. They are characterized especially by the abundance of titanic iron. The granulitic rocks differ from the central gabbro mainly in structure. They consist of aggregates of rounded diallage, hypersthene and plagioclase, all of which minerals are present also in the normal rocks. The basic rocks are probably differentiated phases of the gabbro, of earlier age than the great mass of the normal rock. The granulitic phases are simply peripheral phases. Closely parallel cases were brought out in the discussion as existing in the Adirondacks (by C. H. Smyth, Jr., and J. F. Kemp), and in Quebec (F. D. Adams), where they have been been called granulites, augite-syenites and augite gneisses. H. D. Campbell mentioned the same phenomena in similar rocks in Rockbridge county, Virginia, and all the speakers commented on the peculiar development of orthoclase feldspar in the border facies of a gabbro mass.

36. The Contact Phenomena at Pigeon Point, Minn. W. S. BAYLEY, Waterville, Me.

The speaker distributed copies of his recent Bulletin U. S. Geol. Survey, No. 109, and exhibited a series of specimens which illustrate the peculiar contacts and transition rocks at Pigeon Point. Discussion followed by by J. P. Iddings and others. 37. A New Discovery of Peridotite at Dewitt, 3 miles east of Syracuse, N. Y. N. H. DARTON. Petrography of same, J. F. KEMP.