h‘ —4- -nh1‘LJAL“ A A HEAVY MINERAL ENVESTIGATION OF GLACEAL TILLS {N WESTERN NEW YORK Thesis {or ”w Degree at; M. S. MICHIGAN STATE UNIVERSITY Norman :E. Wingard 1962. rug LIBRARY Michigan State University A HEAVY MINERAL IIVESTIGATIOH OF GLACIAL TILLS IE WESTERE NEW YORK By p Norman E; Wingard A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIEECE Department of Geology 1962 AESTRACT In western New York, sand-sized heavy minerals are a far-travelled and evenly distributed constituent of glacial drift. This suite serves as a useful criterion for the inter- pretation of processes of tranSportation and deposition of tills and glacio fluvial materials. Variability of this assemblage in the tills is determined by petrographic and sedimentological examination of composite samples from selected sites in Livingston, Allegany, Ontario, Yates, Stuben and schyler and Chemung counties. Variations in size distribution in the heavy mineral suite throughout this area indicate that systematic size reduction does not result from tranSportation processes. Also, the heavy mineral suite in the tills is indicated to be more stable with reSpect to disintegration than in the related glacicxfluvial deposits. The factor which most affects the interpretation of variability in the heavy mineral fraction is basal con- tamination. Additions to the sand-sized heavy mineral assemblage are caused by pre-existing concentrations, brought about by gflgaciefluvial and/or solely fluvial action. The findings generally relate to the known glacial geology of the region. The results illustrate an effective application of heavy mineral studies for provenance ii determination and process analysis. The technique may prove of particular value in supplementing knowledge in regions where geomorphic evidence alone is inadequate for a complete understanding of the glacial history. ACKKOHLEDGELEHTS The writer wishes to eXpress his gratitude to Gordon Connally of the State University College, Kew Paltz, New York for suggesting the problem, providing assistance and guidance in the field and for reading and commenting on the manuscript. Thanks are also due to the following people, all of the Department of Geology, Michigan State University: Dr. Maynard Killer for criticism regarding content and organi- zation, whose comments are liberally incorporated throughout the paper; Dr. James Trow for helpful advice concerning the laboratory research, and for reading the manuscript; Drs. J.E. Smith, 3.3. Prouty and W.J. Hinze for their interest and advice. iv CON II7TS Abstract. . . . . . . . . . . . . . . . . Acknowledgements. , . . . . . . . . . . . ContentS. . . . . . . . . . . . . . . . . List of Illustrations , , , . . , . . . . List of TableS. . . . . . . . . . . . . . Introduction. . . . . . . . . . . . . . Purpose and scope of investigation , Previous work . . . . . . . . . . . . Field Data. . . . . . . . . . . . . . . Saripling method , , , , , . . . . . Sample localities, , , . , . . . . . Site descriptions, , , , . . . . . . Laboratory Procedure, , , . . . . . . . . Mechanical analysis, , , , , . . . Disaggregation and dispersion of Se Heavy mineral separation , , , , , . Mounting and identification, , , , . Analytical Results, , . . . . . . . . . . Mechanical analysis. 0 O 0 Identification and description of minerals Heavy mineral percentages by weight. Percentage 0:: magnetics , , , Heavy mineral percentages by count variabilitgro o o o o o o o o o o o o Attrition O O O O O O O O O O O Alteration of unstable minerals Contamination . , , , , , , , , Possible sources of contamination, , Heavy mineral ratios Summary and Conclusions . . , , , , . Suggestions for Additional Research , . , References 0 o o o o o o o o o o o o o 0 ii iv 4 Po 4 Ho Ho Ho [OFJH Ul-P-P'J-F Figures Figure 1. Figure 2. Plates Plate Plate Ilate Plate Plate Plate Plate Plate Plate Plate Plate 1. 7. 10. 11-. Flow sheet of laboratory procedure. Hlei ht percentage mean deviation. . Histograms of size frequency distribution . . . . . . . . . . . . Eistorrans of size frequency dlStrlbutlon o o o o o o o o o o o 0 Cumulative curves of size frequency distribution. . . . . . . . . . . . Cumulative curves of size frequency distribution . . . . . . . . . . . . Isoplethm ap sho wing percentage by weight of we Var minerals o o o o o o :4 Isopleti map showing percentage by weight of magnetics. . . . . . . . . Index map showing: . . . . . . O O O 1. Location of profile lines of heavy mineral percentages. 2. Location of lines along which mean deviation of weight percen- tage was calculated. 3. Rele tionsl‘.ip oetween the mean deviation of the weight per- centage and direction. 4. Terminal moraines. Percentage pr ile along A - A'. . . A Percentage profile along C - C'. . . Percentage profile along III III. 0 O O O O O O O O O O O 0 Percentage 3r file along I - I' . . $ 13 25 48 SO 51 52 54 U1 U1 Di 0\ Plate Plate Plate Plate Plate Plate Plate Plate Isopleth map of percemita:e of garnet, including opaque n7;inerals Is0pleth map showing percent hornblende, including opaque minerals . . . . . . . . . . . . . Isopleth map of percent hypersthene including opaque minerals. . . . . Isopleth map of percent garnet, excluding Opaque minerals. . . . . Is0pleth map of percent hypersthene xcluding opaque minerals. . . . . Isopleth map of percent hornblende, excluding opaque minerals. . . . . Isooleth map of clear garnet - red SEE-Ml t rabio 0/3.. 0 o o o o o o 0 lap of Western Fin; showing the range 0 glacial movement i1 I‘OCZC Striaeo o o o o o o o o o o o 57 58 6O 61 63 61+ Table Table Table Table Difference in percent composition of LIST OF PAS-ES -2, -1 grade size. Quartile measures for size distribution. - . Heavy minerals identified. Percent by weight of heavy in the tills. Distribution and frequency (+0 ient of variation of heavy mineral ages. viii frequency minerals of minerals. 18 19 2O 23 28 29 INTRODUCTION Glaciation in western Sew York state has been an important factor in develOpnent of the present tepography. This is an area in which many studies of glacial geology have been undertaken and where significant contributions have been made. Such papers include/those by T.C. Chamberlain in 883, N.J. Killer in 1914, V.E. Eonnet in 1924, H.L. Fair- child in 1932, and more recently, those by P. MacClintocn and E.A. Apfel (1944) and C. Holmes (1952). 3 Purpos .d scope of study (0 (‘3 Q 3 The purpose of this study is to examine selected samples of glacial till for the purpose of determining the nature of depositional and post depositional processes that have been Operative during the most recent glacial and intra-glacial stages of the Pleistocene. Particularly, it is intended to determine the degree to which each process has been important in producing the present distribution of the materials in glacial drift. The sand-sized fraction of the heavy mineral suite in the drift is a far-traveled and evenly distributed con- stituent. L ghter materials, dominantly composed of rock particles show greater local variation. For this reason the sand-sized heavy mineral fraction has been chosen as the subject of this analysis. Four p ssible causes of areal variability in the h avy lages are considered. (1) provenance, E; H D (D H 93 H to D c 3 (D OJ (2) distance an“ mode of tranSportation, (3) effect of p re-alacial tono repay, and (4) effects of multi’le Q A x- d P slaciation. k1 V Previous work Little research has been carried out on the distribution of heavy minerals in tills. No such investigation has been l,— previously conducted 0. the tills in the Finger Lakes region. As will be shown, it i. possible to determine general (0 variations over the area and to relate these to present knowledge of the glacial history. The relations found here may allow application of the heavy mineral technique in the interpretation of glacial geology in regions where previous geomorphological studies are not as extensive, or where adequate geomorphological evidence is lacking. Numerous attempts to . B ap the glacial deposits in central and western New York have been carried out since T.C. Chamber— lain's description of the "terminal moraine" in 1883. Put detailed laboratory analysis of the drift has not been done. A field classification or pebble count has usually been considered adequate. (KacClintoch and Apfel, 1944; Holmes, 1952). Quantitative studies of heavy minerals in drift have been made in Ontario and adjacent regions by Dreimanis, Beavely, Cook, Knox and Koretti (1957). Their purpose was to differentiate drift associated with major lobes of Pleis- tocene ice. The result was highly successful, since each lobe contained an assemblage reflecting a different pro- venance. Samples of stratified drift in the Finger Lakes region have been studied by Connally (1959) with reSpect to heavy mineral content. The results of this investigation will be noted as they relate to findings in the present treatment. -4- FIELD DATA Field sampling was accomplished by samples from a number of localities. ‘The sample sites were chosen to coincide as nearly as possible to a grid of fifteen minutes of latitude by fifteen minutes of longitude. This cor- reSponds to one location for each fifteen minute quadrangle - i.e. a sampling pattern considered appropriate for the determination of broad areal patterns in the distribution of heavy minerals in the tills. In many localities, lack of eXposure of fresh, unweathered till prevented the grid distribution from being fully realized. Where eXposures of undisturbed till were abundant, the samples were obtained at sites providing significant areal distribution. To assure that local variations in the till could not affect the randomness of sampling, four samples were taken from each location. The necessity of this procedure was borne out, in terms of size distribution, by the subsequent mechanical analyses. Safiple localities Samples of till where collected from parts of Livingston, Allegany, Ontario, Yates, Stuben, Schyler, and Chemung counties. These all lie in the western part of the Finger Lakes region, extending from Lake Ontario, southward to close to the Pennsylvania State Line (13p Plate 7.). In this region, five and possibly seven recessional sub-ages of the Wisconsin Glacial Age from Lower Cary to post Port Huron (Hankato) are represented. Throughout the region the trend of major striae and the orientation of streamlined forms, such as drumlins, is from Ngoow to N250E. (Plate 19). Connglly (1959) cites various lines of evidence to indicate that during Valley Heads (Upper Cary) glaciation and all subsequent readvances of late Glacial time, the mass transfer of ice was toward the southwest. As Opposed to this, the geomorphic evidence cited in the following pages and the related data on heavy mineral distribution suggest that the ice movement was dominantly toward the southeast. Site descriptions Descriptions of the sample sites follows. The character- istic color of the till given in these descriptions is based on an arbitrary field classification. When compared with the Hunsell Color Chart in the laboratory, the colors do not necessarily correSpond. The field colors are listed as a prime guide to identity. The color chart comparison is noted in parentheses at tho end of each description. Site 1. Stream-cut bluff formed by a small tributary to Bradner Creek, one mile north of Byersville. Seventy to 80 feet of silt-loam till exposed. The lower 35 feet of till very firm and blue-gray in color. The upper 40-45 feet oxidized brown till with no obvious stratigraphic.break with the lower portion. This till contains man pebbles and cobbles up to 8 inches in diameter. (N 5.0 . Site 2. This location not used in present analysis. A small cutbank in Springwater Creek, 2 miles north of Wayland or 5 miles northwest of Dansville. Five feet of till is exposed. The lower 2 feet is blue silt-loam till, with cobbles l to 6 inches in diameter. Some cobbles show slight stream rounding. The till becomes finer upward and is over- lain by 2 feet of lacustrine silt or silt till. The upper silty layer contains few pebbles, and incorporates cobbles at 3 up to 12 inches dimension. The till is calcareous and the cobbles exhibit the "bright lithology" described by KacClintock and Apfel (1944, pp. 1155-1156). Oxidation extends to a depth of 2.5 feet. Koderate leaching has occured in the upper 2 feet. Fifty feet downstream from his section, a drab gravel 3.5 feet thick appears to over- lie the till. The gravel is leached and oxidized. (10.0334/2). Site 4. A pebbly gray till eXposed in a 7-foot cutbank one-half mile west of Eoneoye. The clasts range in size from less than 1 inch to more than 12 inches in diameter. The till is firm and calcareous. Oxidation extends to 5 feet, leaching to 3.5. All samples taken below the oxidized zone and over a lateral distance of about 100 feet. (l0.0YR4/2). Site 5. A 10-foot bluff on a north tributary of the Canisteo River with exposed till of olive-brown appearance. Slightly alcareous and with a clay-loam matrix. Huch local fragmental material is incorporated with a great many stream-rounded pebbles and cobbles. Oxidation extends to about 4 feet, though slump cover has obscured this observation. (2.5Y6/4). Site I, Cut bank on a small tributary west of Deerlick Run, 2.5 miles north of Wheeler. This site can be seen on the map (Plate 7) about 5 miles west of Keuka Lake. The bank is 12 feet high, the upper 10.5 feet is covered with slump debris. The lower 1.5 feet had been freshly undercut, at the time of observation, exposing a pebbly chocolate- colored, calcareous till. (2.5Y5/4). Site 8. A 40-foot cutbank 1.5 miles south of Short Tract. This can be seen near the center of the map of Plate 7, about 4 miles east of the Genesee River. he lower 20 feet consists of deternatin: cross-bedded sands and gravels. The upper 20 feet is blue, calcareous, unleached till with silt matrix. Oxidation extends downward 12 inches. (5.0Y4/2). Site 9. Cutbank 1 mile west of Greenwood in Christian Hollow is about 25 miles west of Corning (Plate 7). A tan, calcareous, pebble-rich silt till is exposed in a bank 7 feet high. The upper 4 feet is oxidized till. Leaching has been effective to a depth of 2 to 5 feet. (2.5Y6/2). Site 10. An 8-foot cutbank on Trapping Brook, 1 mile east of Wellsville in the southeastern corner of the study area (Plate 7). A pebble-rich blue till, 2.5 to 5 feet thick, incorporates much local material. The bank had been freshly cut at the base but was badly slumped above, at the time of observation, therefore depth of oxidatiofiaiot determined. (5.0Y4/2). Site 11. Ten-foot cutbank on Salt Creek 1.5 miles northeast of Retsof, or 5 miles north of Geneseo on River Road (7 miles west of Conesus Lake, Plate 7). Eight feet of firm purple, calcareous and very pebbly till is exposed; the lower feet is fresh, unoxidized; the upper 4 feet oxidized but unleached. Two feet of colluvium overlies the till. (5.0‘34/2). Site 12, A 20-foot cutbank 1 mile west of Buckbee Corner (10 miles southwest of Rochester, Plate 7), eXposes firm, purple-brown, calcareous pebbly till. Lacustrine sands are intercallated toward the top. (5.0YR4/2) Site 15. Construction site for a shopping center 1 mile northwest of Pittsford, southeast of Rochester (Plate 7). A composite thickness of about 20 feet of firm, purple, pebbly sand till, with lacustrine sands, gravels and silts is eXpose .* A 12- foot embankment discloses purple till as the basal member with intercallated sand layers with some gravel. The upper 6 feet of the bank shows a firm sandy brown till. Samples were taken from the lower unit. (75YR5/4). Site 14. Thirty-foot bluff 2 miles east of Woodhull on the south Branch of Tuscarora Creek (15 miles west of Corning, Plate 7). The lower 10 feet is slump covered, above is 10 to 15 feet of drab, olive, pebble-rich clay-loam till. Above the till is a layer of banded, oxidized sand, up to 8 inches *This description was through the joint effort of G.G. Connally (personal communication) and the author. thick. Above the sand layer is another till, 15 to 50 feet thick, of which the upper 15 feet is oxidized. The till can be seen to incorporate the underlying sand layer, partially destroying the bedding. Samples were taken fron the lower till. (2.5Y5/4). Site 15. A 10-foot cutbank on heads Creek at the junction with Dry Run, 5 miles west of Coopers Plains and 7 miles north- west of Corning (Plate 7). The lower 5 feet is firm, olive-colored pebbly till with loam matrix. The drab looking pebble lithology and absence of "bright" materials (IacClintock and npfel, 1944), suggest Olean age. The till is overlain by about 5 feet of massive oxidized sand. (2.535/4). Site 16. A O-foot cutbank one-half mile southeast of Catlin (5 miles northwest of Elmira, Plate 7). The bottom 1 to 1.5 foot eXposes firm, very pebbly, olive-colored loam till, overlain by Olean-type gravel. The till contains a fairly large amount of "Binghamton-type" (fiacClintock and Apfel, 1944) pebbles, including chert and red sandstone. (5.0Y4/2). Si Located 4 miles north of HammondSport, on the west side of Keuka Lake. A 50-foot cutbank with 5 feet of fresh, gray, pebbly silt-loam till exposed near the tOp. The remainder of the bank, when observed, was covered by slump material, mostly till, indicating considerable thickness. Samples taken from top of the unit. (N5.0). Site 18. Cutbank on Flint Creek one-half mile west of Seneca Castle, 5 miles west of Seneca Lake (Plate 7). The bluff is 50 feet high with 5 till units eXposed. The lower unit is a firm, gray, pebbly silt till, extending up to 5 feet. Above is 25 feet of firm purple sandy till with "Binghamton lithology." This unit contains logs apparently of interglacial origin, although Connally (personal communication) suggests that they' may be later con- taminants. At approximately 10 feet from the top, this unit grades into a firm, sandy loam till. Sand lenses are found locally throughout the upper two units. These lenses are usually oxidized, probably as a result of recent ground water migration. Samples were taken from the middle unit. (5.0YR4/2). Site 19. A road out on the southeast side of the village of Newark (Hap, Plate 7), eXposing 20 feet of firm, red loam till that is slightly calcareous. The upper 12 inches is oxidized. The till is pebblv and contains a number of erratic boulders. (5.0YR4/4 . Site 20. Cutbank on tributary to Iud Creek 1.5 miles south of Vincent, and about 6 miles west of Canandaigua Lake (Plate 7). The bluff is 50 feet high and exposes a gray silt-loam till. The till contains many lenses or pods of sand which appear oxidized and to serve as channels of ground water migration. The till is firm, containing many pebbles. Many striated boulders are found along the bank and in the creek bed. The till is oxidized throughout the upper 8 feet. (h5.0). Site 21. Drainage ditch south of Sheldrake Creek, 4 miles southeast of Ovid, between Cayuga and Seneca Lakes (Plate 7). Six feet of firm, brown, calcareous, pebbly loam till is exposed. The till is unoxidized and unleached. (7.5YR4/4). Site 22. A 50-foot road cut 1 mile southeast of Italy, about 10 miles west of Keuka Lake (Plate 7). The lower 8 to 10 feet eXposes blue silt till, which incorporates much of the local bedrock. An 8 to 12-foot sandy layer overlies the blue till and contains many pebbles and cobbles which are more abundant toward the base. The next 10 to 15 feet of the section is not visible, but near the tOp a firm tan loam till is exposed. Samples taken from the lowest unit. (x4.5). Site 25. A freshly eroded stream channel locally reported to have resulted from flooding three months previous. It is located near Canaseraga Creek, in the west central portion of the map area shown in Plate 7. A 50-foot bluff exposes 15 feet of blue silt-loam till. The till is pebbly, including both local shale and a large number of erratics. It is over- lain by 6 feet of varved sand, silt and clay. Above the varved material lies about 10 feet of mottled gray to brown loam till. (35.0). Site 24. A O-foot stream-cut bluff located on the north side of Glen Creek, 5 miles west of Na kins Glen, near the south end of Seneca Lake (Plate 7). At the base of the section about 5 feet of lacustrine silt is exposed, followed by 10 feet of brown, pebbly, loam till and 15 to 20 feet of firm gray-brown pebbly silt-loan till (sampled). Overlyihg the till is 10 feet of coarse, oxidized gravel and 1 foot of brown, oxidized loan till. (5.0Y4/2). -10- vain” descriptions, with the V '_‘ x K.) v (0 With respect to the for exception of sites 15, 19, 21 and 22, all samples were as d. m (D I _ H) H O ’3 5 atural exposures resulting from stream erosion. Thus, most samples are from the sites of present-day valleys, with the above-noted four representing present- (D ES. day interfluvial ar -11- LABORATORY PRCCEDUEE Following is a description of the methods and procedures used in the laboratory, preparatOIy to the sample analyses. Sixteen Spot samples from four collecting sites were selected for mechanica- analysis of the sand fraction. The purpose was to determine the size characteristics of the till and to find a size grade containing the largest amount of material. Disaggregation was accomplished by means of a wooden rolling pin and a rubber pestle. The one quart Spot samples were quartered to approximately 100 grams by means of a Jones sample Splitter. The samples were then separated into Wentworth sand grades (ratio X2) by means of a nest of seives placed on a Ro-Tap shaising machine. Fifteen minutes of operation per sample was sufficient for the separation. Disarrremation and disnersion of samples In order to find a chick but adequate method to disaggregate and disperse the samples, and to determine a size grade containirg a maximum number of heavy minerals, portions of the same 16 Spot samples were separately treated (0 with N hydrochloric acid, N/lCO sodium oxalate, N/5O odium carbonate, E/2O sodium arbonate, N/2O sodium hydroxide, and water. Each treatment was subjected to 10 to 20 minutes -12- of stirring and from 10 to 18 hours of shaking in a reciprocal shaker. After drying, each sample was xamined — I under a binocular microscope for degree of disaggregation and oxide stain. The most effective and least time-con- suming procedure was 10 minutes of stirrinm followed by C, wet sieving and 10 minutes or boiling in 1:4 H31. There s) is no evidence to suggest that this procedure could -“ . u significantly alter the distributions of heavy minerals. zixure ‘4 1. low ‘:rt of laborator procedure. SpoT Spot Spot Spot 50mp|e Sample Sample Sample Discard ' r—‘———1 Check ' ‘Somplel (3’8 8°; 0 isca rd N WOSh T66 N02 0204 I Sieve 2.0-“.052mm. l Boil I34 NC” I Microsplif to appro&.309 I Brom oform Separation Li ght Sior‘e Heavy [I l I L3 fifracl’i Sievc Remove Magnetics-weigh Mag. 1 Non-M Store aunt -14- Figure l. is a flow sheet showing the procedure used for all samples. Approximately 200 gra31s of till were placed in the sodium oxalate solution and stirred. To insure that .‘ra 1118 upon 5... L.) no encrustation of clay occurred on the sand drying, and to eliminate the coarse fraction, the sodium oxalate solution was wet seived through 2.0 and .062 mm screens. The and fraction was retained and boiled in acid. Heavy mineral separation The heavy minerals were separated from the light fraction by means of bromoform (CHBrB), w i011 has a Specific gravity of 2. 89 at 200 centigra de. The method used is that outlined by Krumbein and Pettijoh (11938, p. 335). After acid treatment, each sample was dried and reduced to approximately 50 grams with an electric vibrating micro- Split. The sample was weighed on an analytical balance before bromoform separation. After separation, the heavy mineral fraction was wei ghed an ad the weigh ts recorded. From these values, percentages by weight of heavy minerals were calculated. The results are listed on Table 4. The 1M it fractions were stored for reference and the heavy fractions separated into the five Hentworti sand—size grades. The size separation of the heavies was accomplished with a nest of saall, 3-inch diaheter sieves. The sieves were agitated using the electric vibrator of the micro- split. Use of the small sieves in this way minimized loss [‘11 of the mineral grains during separation. ach size grade was weighed on an analytical ba ance and the weight recorded. In every case, the greatest frequency of heavy mineral grains occurred in the 1/8 - 1/16 mm. fraction. hagnetite was removed from the fine sand grade (1/8 - 1/16 mm.) with a small hand magnet. neic hts of the magnetics were determined and recorded (Table 4) and the weicht percentages of magnetics calculated. The remaining non-magnetic fractions were then mounted for microscopic examination. Iounting and identification Tie on-rm gneti 0 heavy minerals in the l/8 to 1/16 mm. grade size were mounted on glass slides for petrographic examination. Piperine was chosen as the mounting medium because of its relatively high index of refraction (1.68). Since most heavy minerals have hi her indices of refraction than do most of the lighter minerals, a medium with an index above th at of Canada balsam (1.52) is necessary to be of value in identification. The higher index is particularly useful for the deterzni nations of amphiboles and pyroxenes. A petrographic microscope was used for identification. Mineralogic properties considered, in addition to refractive index, were color, pleochroism, diapersion, optic Sign, birefringence, elonigation, extinction angle, crystal form, habit, and 2V. To ssure posit:ive identification, many -15.. grains were exanined in oils of known index. This allowed a more prec1se det nation of indices and permitted the I...) o (D In; grains to be freely rotated, so that more than one index could be observed. References used in conjunction with the identifications are: Larson and Berman (1954), Bloss (1961), Rogers and Kerr (1942) and Wahlstron (1955). -17- A--f.‘- ‘r'p'f'mf‘L T'|T':§TTL'“O ‘1—-'I-L‘-—I .. ¢ .1. '~J..‘ -g-J k4 sf J. \J 1 r u xecaaaical .03 I l 'he reason for carrying on a limited mechanical analysis of the tills was wo-fold: 1. To determine if different spot .amples taken from a single till exposure or horizon would have a uniform over-all size composition. 2. To determine whether or not systematic variations extend between different collecting stations. A chi-squared test was used for the first of these. The weight percentage of each phi sand class interval was compared with the mean percentape for the four spot samples. In fifteen of the sixteen samples analyzed, the hypothesis Of equal means between SDOt Samples was rejected at the 95 per cent confidence limit. Histograms and cumulative curves of tie size distri- bution are show1 in Plates 1, 2, 3 and 4. It can be seen that the middle sand grade (-1 to 2 on th ¢ scale; 3 m1. to % mm.), in every sample, contains only a very small proportion of the material (3 to 11 per cent) and is in relatively con31stent proportion between snot samples for J- H a given till. nxccptions to this consistency, even in this limited portion of the curve, can be noted. Sample lS-A is disproportionately high. as both extremes of sand size are ‘ oacned, the corresgondence becomes even less. The {0 H 'PP difference in percentage conposition between Spot samples 0 in the coarse fraction (-2, -1 ¢) is from a maximum of 42.o —18- per cent in sample 22, which is the greatest difference, to a maximum difference of 8.3 per cent in sample 10, which is the smallest (Table l). Table 1. Difference in per cent composition of -2, -1 ¢ grade size. Sample no. Greatest % Least % Differences in g , rwithin grade 10 17.45 9.13 8.32 11 15.05 1.63 15.42 15 54.48 36.83 17.65 22 55.65 12.81 42.84 A similar heterogeniety in size distribution is found in the fine grades-—the 3, 4¢ nterval. Here, however, the smallest difference between Spot samples is found in sanple 22 as opposed to the greatest difference in the coarse grade. The cumulative curves shown in Plates 3 and 4 are used to determine the first, second and third quartiles and the phi quartile deviation (QDd). These values are given in Table 2. The quartile deviations show, as do the histo- grams, the very high degree of variability both within and between the tills. The conclusion reached from he mechanical analysis is that individual Spot samples do not show adequate correlation cross mechanical analrsis of the sand fraction. be m ans of v . - ; The fact that uniform size distribution does not extend even for a few feet, within the same deposit, testifies that long- 1 distance correlation, between exposures is n t feasible or V t“ }. J. 4—‘ d (w m , n 5;) ..;e ‘JLLO O Table 2. 10—A 10-3 10-0 10-D ll-A 11-3 11-0 ll-D lB-A 15-B 15-0 lS-D 22-A 22-B 22-0 22-D -19- Qi¢ M92 083 3002 1.67 2.80 1.80 3.05 1.15 2.95 .29 2.50 -.84 1.92 O 2.48 2.77 3.10 -1.06 -0.10 "1.82 -0015 —1.60 -1.20 “-1.62 ‘0031 0.73 2.84 ~0.63 2.12 -l.70 1.30 ~2.00 ~1.30 Quartile measures for size frequency distribution. 93¢ QD¢ Average . 3 1.200 ’ 3.35 .840 3.42 .810 3.31 1.080 .982 3.42 1.855 3.35 2.095 3.24 1.620 3.45 .340 1.477 3.08 2.070 3080 2.810 0.70 1.150 3.40 2.510 2.135 3.14 1.205 3.24 1.960 3.07 2.385 2.68 1.440 1.747 -20.. Identification and description of minerals The minerals identified for all sample localities are given in Table 3. A description of the minerals identified is also included in this table. The minerals found in the till, in their approximate order of abun- dance are: hornblende, magnetite, pyrite, hypersthene, leucoxene, clear garnet, monoclinic pyroxene, titanite, red garnet, enstatite, zircon, tremolite, rutile and muscovite. Those found only as a trace in a few samples are pigeonite, sillimanite, chlorite and epi- dote. The latter are not considered adequately abundant to represent their true statistical distribution. Table 3. Heavy minerals identified. Mineral Indices yescrintion Chlorite o‘l.57 Green to light green plates with low birefringence. Y 1.68 Produced off-center optic axis figure with small 2V and incomplete extinc- tion. Epidote °<1.72 Greenish-yellow irregular grains. (pistacite) High birefringence. “.75 Biaxial negative, 2v large (n900). Most grains produced centered optic axis or typical "compass needle figure. Garnet n 1.79 Colorless, purple, pink and red varieties. Isotropic, irregular fragments with conchoidal fracture. Colorless to purple varieties were counted as a group and pink and red as another. Table 3 continued. Hornblende «1.60 X 1.70 Hypersthene«1,69 71.70 Kuscovite «(1.5 yl.6o Pigeonite e..“./..-L. IL-L .. .5XL.J VJ... CLLD-LVLIQ (n 4.. L-.. i—. \ U The tills sampled in this inve "ation contain a heavy mineral suite characteristic of the Grenville JJ aetasedimen s of eastern Ontario and the A iron ack f thus inferred }.Jo U) 0 -, .- *r 11.. A. 1. . -1 1 10.“. OJ. -185? 40111. it COthi.Ofl 'OLO'VED -JCe .L m for the heavy minerals in each of the drift sheets con- sidered. The drift of the western Finger Lakes re3ion is related to that of the Erie-Ontario Glacial Lobe, as differentiated from the Huron Glacial Lobe by Driemanis, the distribution of component materials of he tills, are sunnarized below: A. The size frequency3distribution of sand—sized F3 fwd 1i (.0 material is nor system a tie. supports previous 1 kn wledge that rapid cLange of materials in the lirht fraction t1hes place durin“ transport. B. Heavy mineral assenvlages by e1th can be evaluated in terms of direction by means of calculatin3 the mean deviation of the percenta3e by weight alon3 chosen lines. 5301 analysis reveals the greatest mean deviation in a south-south east direction. This deviation 0 {3 O *1? E3 {3 H Ci». 0 cf L.) H U) 9. H. H (D O L.) H U) m a H O 0 O I. (D 5...: becomes less as a line C. Percenta es of specific heavy ninerals, as cal- culated from the total heavy min‘ eral ass en Ml-" e, reveal -41.. an inverse relationship between garnet and hypersthene and a direct relationship between hornblende and hyper- sthene (Plates 8, 9, 10, an? 11). ‘ m1 A V‘ o _ _o . o _ ,0 A ~ -... . D. 1ne 11striout1on or nea1y m herals along (\ F}. specific map transects can be conpared with the regional distribution by usin3 Pea son's coefficient of variation The values are not sufficiently c nsistent with the areal distribution to have resulted from purely mechan— 1 ical chanses due to transport. Nor do they reveal any v ‘ relationship to anown ases of the drift. E. The ratios between different heavy minerals in undisturbed tills of Pleistocene age indicate that de- composition of unstable minerals is taking place at a much slower rate than in relatively coarse, waterlaid glacicrfluvial deposits of the same age. F. Patterns revealed by the heavy mineral isooleth ZEEE can be placed in one of four categories: 1. A general north to south decrease in weight per- centages: shown by both magnetics and non- nagnetics (Plates 5 and 6) and by percentages of hornblende to hypersthene as calculated from the entire assemblage includin3 opaque minerals (Plates 13 and 14). 2. North—northwest/south—southeast alignnent: revealed, principal y, by the garnet distribution (Plates 12, 15, and 18). 3. A more or less east to west variation: diaplayed by hypersthene. This pattern is not well defined (Plates 14 and 16). -42- 4. Random-uniform distribution: shown by the dis- tribution of the most abundant mineral, horn- blende, after eliminating the effect of dilution by Opaques (Plate 17). Although an element of subjectivity is involved in the construction of contour maps from any such data, the writer considers that the south-southeast alignment in the percentage distribution of heavy minerals is never- theless well justified. It is significant that this alignment is eSpecially revealed in the distribution of garnet, a relatively stable mineral. This indicates that fluvial control of the distribution of heavy minerals, by concentrating sand-sized material from former drift channels, in selective drainage” is probably the most important factor in the present distribution. It is also compatable with the findings to state that movement of ice has taken place largely in a southeasterly direction; in conformity with many of the geomorphic patterns, including some of the bedrock striations. On the Ontario Plain, to the north, a southwesterly trend of glacial movement is indicated by the orientation of drumlins. On the Alle3heny Plateau, however, deviations from the dominantly southeastward trend were probably minor and locally controlled. These somewhat Opposing directions of movement in the two areas appear not unreasonable when the influence of pre-glacial tOpography is considered. The ice which emerged radially from the Ontario Basin -43- (Holmes, 1952) streamed outward in a direction parallel to the main drainage channels which today form Honeoye Creek, the Cohocton and Oanisteo rivers and those streams that were deepened and widened by the passage of ice to form the present-day finger lakes of Cayuga, Seneca and Hemlock. The impermeable nature of the undisturbed tills considered in this study, has prevented extensive per- colation of ground water. The process of intrastratal solution, which appears to have produced substantial disintegration of unstable minerals in the more permeable stratified sands and gravels studied by Connally (1959), has apparently not occurred within the tills. It follows that a3e dating of the glacial drift by meals of heavy minerals must be confined to the better sorted glacio- fluvial deposits. In tills of western New York, important contributions to the heavy mineral suite have been distributed and concentrated in accordance with the main draina3e lines and the regional topography. -44- This study is considered preliminary to further I research in the distribution of materi'ls in glacial tills. It illustrates that heavy mineral analysis of tills can be useful in differentiating drift sheets of different provenance. The applicability f methods, such as the mean deviation of the weightpercentage of heavy minerals as an ndicator of direction of glacial tranSport, needs verification by further analysis of carefully located samples in areas where the d'rections of former ice movement can be, or have b en determined by other evidence. Heavy minerals from closely-Spaced, carefully chosen samples of till, obtained along lines of known glacial or interstadial drainage, should yield detailed information regarding the association of heavy minerals in tills conditioned by earlier fluvial and/or glaciofluvial action. Such knowledge can be of value in the reconstruction of glacial history in areas where subsequent advances producing later tills, have altered or destroyed other lines of evidence. “i 14 L‘ 1 LG :L-J. c5 LIJ (I) Alling, H. L., 1946, Quantitative petrolO3y of the Genesee gorge sediments, Proc. Roch. Acad. Sci., vol. 9, no. 1, pp. 5-64. Apfel, E. T., J. E. Maynard and L. W. Ploger, 1951, Possible diatrsme in Syracuse, Jew York (abs. ° Geol. Soc. Am. null., vol. 62, no. 12, pt. 2, pp. 1421, Dec. Bloss, P. D., 1961, An introduction to the methods of Optical crystallography: Eold, Rinehart and Winston, Lew York. ' Chamberlin, T. C. 1883, Termina el Lloraine of the second 5 glacial period: U. . Geol. Survey, 3rd Ann. Rept., pp. 351-360. Connally, G. G., 1962, Th e 3lacial geology of the western Fii er Lakes rsrion, lTew 1ork : Progress rept. no. II (unouolisuee paper). Connallv G. G. 1959 Heavy minerals in the filacial a, 9 ’ .1 . “.... drift of western New “or : Proc. noon. Acaa. Soi., vol. 10, no. 5, pp. 241-278. ts in the Elmira re 3ion, Denny, C. S., 1956, h’ if uuivalents in New En3land: q L 180 New Yor k and their poss Am. Jour. Sci., vol. 2 Dewitt, Walace Jr., and 3. W. Colton, 1959, Revised cor— relations of Lower Upper Devonian Rocks in western and central :ew York: Am. Assoc. Petroleum Geolo3ists, vol. 43, no. 12, pp. 2310—2823. Dixon, H. J., and I. J. Lassey Jr., 957, Introduction to statistical analysis, second ed., LcGraw Sill Book CO. , 1:10. , lgev'f .10r4‘.0 Dreima oni s,—A., G. H. Reavely, R. J. D. Cook, K. S. inox and F. J. Loretti, 1957, lea vy mineral studies in tills of Ontario and adjacent areas: Jour. Sed. Pet., vol. 27, no. 2, p). 143-161. Fa MTGlild H. L., 193 2, New York, morai;1es: Geol. Soc. Am. 3111].. , V01. 43, Pp. 627-652. 301$ es, C. D., 19: 52, Drift DiSpersion in west-central her York: Geol. Soc. 11. Bull., vol. 63, pp. 993- -1010. Krunbein, W. C. and F. J. Pettijohn, 1939, Lanual of sedi- mentary petr 3rapn y, Appleton—C 1tury- Crofts, Inc., lietr Yornz. Larsen, 3. S., and arry Berr an, 1934, Th determi11atioa of t e nonopa ue minerals Ge 01. Survey Bull. 848, U. S. Govt. Iri 1 Lacnio‘toa. Lawler, T. A., 1962, A field investi3ation of the magnetic dis turbanccs fr m 'lacial drifts in Iichigan; Lasters Tnesis, Licnigan State University (in manuscript). LacClintock, Paul and E. A. Apfel, 1944, Correlation of the drifts of the Salamanca re-entrant, Lew York: Geol. Soc. Am. Bu11., vol. 55, pp. 1143-1164. Liller, N. J., 1914, Geological history of Yew York State: Yew York State Hus. Bull. 168. Lilner, H. D., 1940, Sedimentary Petrography: hird ed., Thomas Yurby and Co., London. onnett, v. 3 1924, v ., er Lakes of central Yew York: in. Jour. Sci., vol. n Loroney, H. J., 1960, Facts from fi 3ures : Third revised edition, Fen3uin Books. Baltimore. Pettijohn, F. J., 1957, Sedinentary rocks: Harper and Bros. Hew York. Rogers, A. 3., and F. F. Kerr, 1942, Optical Lineralo3y, Second ed., LcGraw iill Book Co., Yew York. chmidt, V. 3., 1947, Boulders of inter31acia1 conglomerate in central lav Yorl:: Am. Jour. Sci., vol. 245, no. 2, pp. 127-133. a111a e features of central Sew York: Tarr, R. S., 190 r 10, V01. 16, pp. 2_:-2420 9 Geol. Soc. Am. Bu Twenhofel, W. H., and S. A. Tyler, 1941, Lethods of study 1 of sediments: LcGraw-Eill Book Co., Lew fork. -47- VonEngeln, O. D., 1929, Interglecial deposit in central few York: Geol. Soc. Am. Bull., vol. 40, pp. 469-430. 3., 955, Petrographic Lineralogy: John ns, Ino., Eew York. can ‘. _ 7‘1 wanlstrom ; 'Y‘o 3 mile; an $.11 C I? o 0 .v- Williams, A. S., R. S. Terr Description of the Nat " , Kindle, 1909 S J t t U.S. Geol. Survey, Geol. Atlas, Vol atonk distri i 2 c : o no 169, 33 p. t ‘8 88 hf percen vveiég no 0 ~48- s;ze frevuency distribution. 6 -49- Elntp 2. Histograms of size frecuency distribution. IO-A lO-B U1 9 IO- 1:. o C) E5 1% ghT perce IS-A W§i N 50 IS-C . IS-D 40 ‘ £20 IO -2-l o I 2 3 4-2-10 I 2 3 4 PHISIZE is 1.023. O ‘n\‘ t; DrlkJufl ., lO-L if y‘. L J ‘ h C lEI‘C f‘ '.-l. TV '7 .C' fl 'tive ,.‘~H" ‘ .‘, [41“ L1 L —A llitt‘: “3t. 30‘ --.—L.--_L.__.JL.. __ 1 21.3 4 O 82 DKJDF“ 90 I i — 0 0 0r 50: .0 it“; -2 -| £5! 21E PHI -51- -115 tl‘i bution. cueuoy rize frv P L p ( , U :1 1‘4"" 7. Cumll tin ilete 4. D 1‘ ‘ // A l_| / l I b b r P b r h b in! 4 i 4 q 4 i < 1 1 4 ,/,, 1 _/. . . ..I L Ah //// LI _ // LO 5 / L 5 4 no 44 ‘1? If b J} L D1! 1‘ 4‘ ‘ 4 13 ‘2 LI l5 0 IOO 90~ 80" 7O0 6 Cir 5C’ 4C- 3C~ a“ IOy/ Frownma 4 L A _ 1 D. 2 /. p>LP+bbppl 0000000000 098765432... h£m_®>> SIZCE PHI -52- ilete 5. ISOpleth map snowing percentage by weight of heavy minerals. sample sites are indicated by number. 718° 7'7 . ROCHESTER . —43° 43-— ' I T 7's“ Contour Inlerval=.OSO°/. 77° Mile. -53.. flaitt 6o Isopleth mop snowing percentage by weight of megnfitics. 780 77° I I ‘ I ROCHESTER .06. 0\ xI2 ‘77“753uiéi**-- ,.~-—e’ \\“‘\\\“ \ V03}: EWARK \V 43er- ill the - \{:&URA \ L43° W J 30 lo 20 MHQ. I0 9 ’9'! ate 'f‘. -4”, .“‘ c '1, l; ‘ f c c ‘ ‘ c 2‘ ., .. V ‘ ‘. ‘ fi-v ‘ ‘ “ *Ft ' V* * .- ‘ calcrlgt,‘ ‘ ‘ t?) r‘" * be. '1 .. n.- i to ‘ T ‘ N ‘“ . ‘ ~:' '7 ,. .. }',.,‘ a; {Jet}-.. "‘ .5 1. x ‘ -_ 1 .9 "f ‘u y 1 ‘1 v. p w -. 1 /1\ V. L J. \ ... s O "R“CH STER 3 . .,N ARK 43'» . ’- - 24 ( \ y r ’ ‘ l \ ' x I. ~ C OR NI N G -- __ -' ‘\_ E A x ., Lw\lR Ir l I 1 f N . 177° -55- 1 . n \ b 4 .. .1. ' \J P~AHWA~HMW a.“ .‘ fiJpH»v~ Uk‘ypmu‘idoblh iv.“ . ... i‘cJ‘ 1”.” M i, I‘ . fix I... r r: 12.300 ocog+mgmmh14 oncgncLoIO n mu m m. .9 w 4 .H 0—K IAV luvo mm. W mu. w .AV “aw Ham—1 labile (nu anu o P W ON! \ .mm3 In. H. m a mmr new 0 +ngo®0 nu . 1 o m 0v J w H Ifdo_T .MVOd 03 J.- ...—J a... PO + M" H .2. $me Mu... P « om. .mm... 4‘ 4 w. w mu. low m _ —56- ..M | m wzcfim sawmoku cuspsoogmn .AH madam #0:...an u ocofoiomhI a. ovcchgoIo 9 m 1 :03 n 4 w 0.1 w .me no a c 4 u . D...“ m. r 10m a mu II.- _ d l. m ,, me 9 ON fl /// . J. i // e l u. mN. _ // low a _ / u m . l _ . H m. t - 8. m. H . rmccowo ocoz+nL®1hId onco_o_c..oIo 9 . . 10V . . # J , u - J a O. x. L9/ \. rmv H 1.0 « .\\ / x o D e #l f -.\4r.lll|xlh\$. / ./, uHJu NJ m_ \.\ _ , ,. 10m a. m _ - - . _ ,/. m. mu . _ \\\ / u: _ -.‘\i& .../x U .0 4 _ xx . .mm P 9 ON fix . /, o u. m w /x i l ./ W mm m _ 4,4 / 100 u a m d t. _ B a dm 6 v m. B \ ..HHH - HHH wuoam maauouu owmwcmopmg .04 mp an apualunoH 1u9349d iueaqad -f7- Plate 12. Isopleth map of garnet percentage, including the opaque minerals. 439- - 780 ' 7J7o ' «ROCHESTER V' WARK -43° lo o la 20 80 Mile. -58- Plate 13. Isopleth map showing percent hornblende. Opaque minerals included in calculation. —45° 78° ‘ 77° 1 ' I l "ROCH ~TER x I ' IZ % 5O 1 F O X , (v Q: (o Q/ 2 Lu (9 x X IO 1 i -.. 1. o 78 Contour Interval = IO°7. 77 IFMM ——83° Nflllea -59- Plate 14. Isopleth map of percent hype rsthene. Opaque minerals included in calculation. . 78, . A 1 717 tin '0 ROC ESTER ‘43:- X X to o J . l, . 78 Contour lnTerval=2.0°7. 77 lo 9 lo 20 :0 Mile. -eo-, Plate 15. Isopleth map of g:rcent garnet, excluding opaque minerals. 78° 77° 4 1 , l \ 3R0 \ wROCHE;TER ‘ ’o r ‘ \ xlz - . O...“ I O . .‘NE ‘ R K -I 43" A 1 \ (‘0 0‘: ‘. '78 Contourplnterval =5‘70 [WM .330 Ml lea p-51- Plate 16. Isopleth map of percent hypersthene, excluding opaque minerals. 78° 7 ° 4 L - 1 7J7 , MROCHESTER X . ' W 5 43‘7 > , 1 ‘ '2. a: r. x b 2| Y.‘ x “\ 24 X CORNI .. ELgQaA LA - r Plate 17. Isopleth map of percent hornblende, excluding opaque minerals. 78° 77“ 1 a J 43'— - as x, MRNING \l ' " \‘ “43° lo 0 lo 20 80 Mile. 43L - 7-63- Plate 18. Isopleth map of clear garnet-red garnet ratio (C/R). '0 ROCHESTER l8 N 4K5 " u \’\I8 . I. COR I G x x \ [mil '0 ° " \V’ ELgu A k xld “ \ J J " l, ; I . 78 . Contour Interval =~I.O .77 lo 0 IO 20 3o Mlle. i I“; “ 1 OH 9.". 31‘; .L 11%) t; . 'ii’ 43" I ”31'! 'F (i "111* LA m if}... 91' L mWJt’...’.‘ho lIi-ii( o Lt .; ... .4, ..A W." 1‘ 1‘ iii -.:T. l :e. 7J7° 1 ROCHESTER xi \\I\a \ ‘\ \e-EANa; “~\ - __,./ ./ \ \V'Crgh’fEW/x R K "e \l— \Q' g .‘9 ‘ 5 "QfORNIN ’-. X IO G . ELgfimA «43° F157? a! 61:" "If-.1 0312.4; “'1‘: . .. . .‘.":‘£"‘n-' ‘*' ‘4? It's-”Ul, V- El WW 1 75% 41585 e '. i" .‘ I ’ W 3 51‘3“: l nll l II III I l llllllllHlH