Whole Number 293 THE JOHNS HOPKINS IJNFVERSITY CIRCULAR 741 $ CONTRIBUTIONS TO GE-LOGY AND PLANT PHYSIOLOGY BALTIMORE, MARYLAND ISHED BY THE UNIVEBSITY ISSUED MONTHLY FROM OCTOBER TO JULY MARCH, 1917 Entered, October 21, 190C, at Baltimore, Md., as second class matter, under Act of Cengu. of July 16, 1894 BERKELEY LIBRARY UNIVERSITY Of CALIFORNIA — i I. EARTH EXCHANGE CONTRIBUTIONS TO GEOLOGY AND * PLANT PHYSIOLOGY t BALTIMORE THE JOHNS HOPKINS PRESS 1917 --EART* SCIENCI UBRARt THE JOHNS HOPKINS UNIVERSITY CIRCULAR, No. 293 MARCH, 1917 GECvLOGY : * CONTENTS PAGE ••* .'Gfeofrgicat Surveys Vith Special Reference to the Work of the Maryland Geo- •.•• • * • jlggjg^l.'Survty*. W. B. CLARK 3 The Use of Average Analyses in Defining Igneous Rocks. E. B. MATHEWS. ... 12 The Delta Character of the Tuscaloosa Formation. E. W. BERRY ......... 18 The Role of Mineralizers in Ore Segregations in Basic Igneous Rocks. J. T. SINGEWALD, JR ................................................... 24 The Environment of the Tertiary Marine Faunas of the Atlantic Coastal Plain. J. A. GARDNER ................................................... 36 The Pelecypods of the Bowden Fauna. W. P. WOODRING ................... 44 Origin of the Natural Brines of Oil Fields. F. REEVES ................... 57 An Upper Cretaceous Seacoast in Montana. W. T. THOM, JR ............. 68 A Remarkable Upper Cretaceous Fauna from Tennessee. B. WADE ......... 73 The Occurrence of the Tuscaloosa Formation as Far North as Kentucky. B. WADE ........................................................ 102 The Habitat of Belemnitella Americana and Mucronata, G. E. DORSET ...... 107 PLANT PHYSIOLOGY: The Department of Plant Physiology. B. E. LIVINGSTON ................. 133 Publications from the Laboratory of Plant Physiology, 1909-1917 ......... 154 Atmometric Units. B. E. LIVINGSTON ................................... 160 The Vapor Tension Deficit as an Index of the Moisture Condition of the Air. B. E. LIVINGSTON ............................................ ..... 170 Incipient Drying and Temporary and Permanent Wilting of Plants, as Related to External and Internal Conditions. B. E. LIVINGSTON ........ 176 The Effect of Deficient Soil Oxygen on the Roots of Higher Plants. B. E. LIVINGSTON and E. E. FREE ....................................... 182 The Experimental Determination of a Dynamic Soil-Moisture Minimum. H. E. PULLING ................................................... . . 186 Some Unusual Features of a *Sub- Artie Soil. H. E. PULLING ............. 188 The Geographical Distribution of the Citrus Diseases, Melanose and Stem-end Rot. H. S. FAWCETT ............................................. 190 Preliminary Note on the Relation of Temperature to the Growth of Certain Parasitic Fungi in Cultures. H. S. FAWCETT ....................... 193 Symptoms of Poisoning by Certain Elements, in Pelargonium and Other Plants. E. E. FREE .............................................. 195 The Effect of Aeration on the Growth of Buckwheat in Water-Cultures. E. E. FREE ..... ....................................................... 198 The Effects of Certain Mineral Poisons on Young Wheat Plants in Three-Salt Nutrient Solutions. E. E. FREE and S. F. TRELEASE ................. 199 Leaf-Product as an Index of Growth in Soy-Bean. F. M. HILDEBRANDT ..... 202 A Method for Approximating Sunshine Intensity from Ocular Observations of Cloudiness. F. M. HILDEBRANDT ................................... 205 Moisture Equilibrium in Pots of Soil Equipped with Auto-Irrigators. F. S. HOLMES ......................................................... 208 Seasonal Variations in the Growth-Rates of Buckwheat Plants under Green- house Conditions. E. S. JOHNSTON ................................. 211 On the Relation of Chlorine to Plant Growth. W. E. TOTTINGHAM ......... 217 A Study of Salt Proportions in a Nutrient Solution conaining Chloride, as Related to the Growth of Young Wheat Plants. 'S. F. TRELEASE ....... 222 The Relation of the Concentration of the Nutrient Solution to the Growth of Young Wheat Plants in Water-Cultures. S. F. TRELEASE ......... • ..... 225 The Effect of Renewal of Culture Solutions on the Growth of Young Wheat Plants in Water-Cultures. S. F. TRELEASE and E. E. FREE ........... 227 THE 1 *>^ * n •' JOHNS HOPKINS., V UNIVERSITY CIRCULAR EDITED BY THOMAS E. BALL New Series. 1917, No. 3 MARCH, 1917 Whole Number, 293 CONTRIBUTIONS TO GEOLOGY GEOLOGICAL SURVEYS WITH SPECIAL REFERENCE TO THE WORK OF THE MARYLAND GEOLOGICAL SURVEY 1 By WILLIAM BULLOCK CLARK A discussion of the organization and work of a Geological Survey would not be complete without some introductory words regarding the origin of geological surveys. Geological knowledge has been advanced by the individual working independently either in a private capacity or under university or similar auspices,, or by a group of individuals brought together in an official organization, controlled gen- erally in large measure by the lines of investigations to be followed. Cooperation is of course possible in the first case but not absolutely necessary, while team-play is an essential 1 Part of a discussion before the Scientific Association of the Johns Hopkins University. 201] 3 854788 4 Geological Surveys [202 feature of the Survey no matter how much independence may be granted in individual instances. Although;, individual effort of a sporadic sort had long not until the appointment of Werner in 1775 ast J^ofQSsor at. Freiberg in Saxony that geology can be said \tcf ^a^etSeeriv'recognized as an independent science and admitted as such into academic surroundings. The great influence of Werner in securing recognition for geology, although many of his conceptions were erroneous, has led to his being called the Father of Geology. For a half-century after his time much work was done by private initiative both in and out of the university to advance the science of geology, but it began to be recognized more and more that individual resources were inadequate to secure the vast number of facts in the field on which most lines of geology depend. It was then that public aid was solicited and secured, but secured not wholly because the legislator was impressed with the pos- sibility of advancing geology for its own sake but because the geologist was able to impress him that out of this work some- thing of a practical nature might be speedily or in the more distant future secured. It is unfortunate, perhaps, that the geologist, if he is to secure public support for such work, must be to some extent what is called a lobbyist, although the time and energy employed are not wholly lost in that his vision is broadened by frequent contact with men of affairs. Not all representatives of the people, to be sure, consider the public interest of first importance, but there are always some, often many, who do ; at least that has been my experience. It is to America that we have to look for the first recogni- tion of the part the public may play in the support of geologi- cal work through legislative appropriations, and it was North Carolina that established the first official Geological Survey. This was in 1823, when the General Assembly of the State authorized the Board of Agriculture to pay the expenses of " geological excursions " for a period of years and appointed Professor Denison Olmsted, of the State University, subse- 203] W. B. Clark 5 quently Professor at Yale, to direct the work. South Caro- lina followed the example of its sister State in 1824 with Lardner Vanuxem in charge; then Massachusetts in 1830 with Edward Hitchcock as State Geologist, the first important Survey, as the Carolina organizations were rather insignificant affairs; then Tennessee in 1831 with Dr. Gerard Troost as Geologist; then Maryland in 1833 with Jules T. Ducatel, a graduate of the Sorbonne, as State Geologist, and J. T. Alex- ander as State Topographical Engineer. Alexander has the credit of attempting the production of the first topographical maps in the country, and although they were very crude they possess much of historical interest. In 1835 the Virginia Survey was inaugurated with W. B. Eogers as Director, the New Jersey Survey with H. D. Eogers in charge, and the Con- necticut Survey with J. G. Percival and Charles U. Sheppard as Geologists. The following year, 1836, saw the inauguration of the important Survey of New York with such men as W. W. Mather, Ebenezer Emmons, Lardner Vanuxem, Thomas A. Conrad, and James Hall as Geologists, and of the Pennsyl- vania Survey which secured H. D. Eogers from New Jersey, and also the surveys of Georgia with John E. Cotting as State Geologist, and of Maine with Charles F. Jackson at the head. Following these come in succession the Surveys of Delaware, Ohio, and Michigan in 1837, Ehode Island in 1839, New Hampshire in 1840, Vermont in 1845, Alabama in 1847, Mississippi in 1850, Illinois in 1851, Wisconsin and Florida in 1853, Iowa in 1855, Arkansas in 1857, Texas in 1858, and California in 1860, so that prior to the Civil War only a few of the then existing States were without official Geological Surveys. The leading men of their time in American geology were in charge of this official work, organized at the public expense. Hitchcock, Emmons, the Eogers brothers, Vanuxem, Conrad, Hall, and the others I have named comprised the chief workers of their day. The Federal Government up to this time had done little to subsidize geological research. Some explorations of the west- 6 Geological Surveys [204 ern country had been inaugurated in which geology consti- tuted a part of the prospective plans. Among the more pro- ductive of such investigations were those made by David Dale Owen under the United States Land Office and Treasury Department in the upper Mississippi valley region in Iowa, Illinois,, Wisconsin, Minnesota, and Nebraska in various years from 1839 to 1851 and by the distinguished geologists and paleontologists Newberry, Marcou, Blake, Conrad, Hall, and others in connection with the Pacific Eailroad Surveys from 1853-55 under the War Department. Only a single foreign government had inaugurated official geological work during the early portion of this period. The Ordinance Survey of Great Britain in 1830 made a small grant to H. F. De la Beche for the survey of southwest Eng- land, but it was not until later that he was definitely appointed to make a Geological Survey. Spain and Austria established Geological Surveys in 1847 and 1849, respectively, but it was not until the next decade that similar organizations were suc- cessively established in Bavaria, Portugal, the Netherlands, Norway, Sweden, and Switzerland. Some years passed before the other foreign governments followed suit. Following our Civil War the American Government, recog- nizing the necessity of acquiring information regarding our great western country, established four exploratory geological organizations, two under the auspices of the War Department, the U. S. Geological Exploration of the 40th Parallel under Clarence King, and the U. S. Surveys west of the 100th Meridian under G. M. Wheeler, and two under the auspices of the Interior Department, the U. S. Geological Survey of the Territories under H. V. Hayden and the U. S. Geographical and Geological Survey of the Eocky Mountain Region under J. W. Powell. They were finally combined in 1879 with the title of U. S. Geological Survey, under Clarence King as first Director, and this organization with appropriations exceeding $1,000,000 annually is now conducting work in every section of the country although devoting its chief energies to the West. 205] W. B. Clark Y At the present time nearly every State in the Union is also carrying on its own official work,, generally with some form of cooperation with the Federal organization. The State Geological Survey of Maryland, as at present organized, began its operations in the spring of 1896 and is thus over twenty years old.2 Like many other similar organi- zations the Maryland Survey is carried on in conjunction with the geological department of its leading University. In States where State universities exist they are often the head- quarters of such work. The primary object of a geological survey is to determine and describe the geological formations and depict the results on maps. In order to classify these formations intelligently one must establish criteria for their discrimination based not only on their original lithological and paleontological charac- teristics but also on their often highly changed texture and structure. In such a region as Maryland, which has repre- sentatives of many types of rocks, nearly every phase of geo- logical investigation is involved. For this reason it affords a magnificent field of study for the student and has been so employed in the training of several score of graduate students at this University. The State also benefits in that it has at its command many trained or partially trained men without the expense of a permanent staff. The differentiation of geological formations and their repre- sentation on maps has passed through many interesting phases, and a few words in this place regarding the history of geo- logical maps may not be inappropriate. As far back as the end of the 17th century a scheme for depicting the mineral products of a country upon a map was submitted to the Eoyal Society of London and appears in the Philosophical Trans- actions under the quaint title of " An ingenious proposal for a 2 The Ducatel- Alexander Survey came to an end in 1842 and tho only official State geological work of any importance carried on in Maryland after that time until the organization of the present Survey was by Philip T. Tyson who as State Agricultural Chemist from 1858 to 1862, prepared the first geological map of the State. Geological Surveys [206 new sort of Maps of Countrys, together with tables of sands and clays,, such chiefly as are found in the north parts of Eng- land, drawn up about ten years since, and delivered to the Eoyal Society, March 12, 1683, by the Learned Martin Lister, M. D." 3 The first attempt at a geological map was appar- ently made by Christopher Packe in 1743 when he published with an accompanying tract "A new philosophic chorographi- cal Chart " of East Kent, England, covering an area of about thirty- two square miles. The object of the map was chiefly to delineate the topography and agricultural soils, while the geological indications are confined to notices of the position of sea beaches, gravel pits, chalk pits, etc. Much more complete maps of this character covering chiefly northern France on which the mineral substances were grouped in bands were communicated in connection with a memoir by Guettard in 1746 to the Academy of Sciences of Paris.4 Following these came maps of the same character by Fiichsel5 in 1762, by Guettard and Lavoisier about 1770, when twenty-nine uncolored geological sheets of the map of France were issued; by Guettard and Monnet in 1780, when a folio of thirty-two sheets accompanying a mineralogical de- scription of France was published, and by Desmarest in 1771, when an uncolored Geological Map of the Auvergne was prepared.6 The first colored geological map is the work of Glaser, who in 1775 depicted in colors a small district in Saxony about twenty miles long by fifteen miles broad, three tints being used : red for granite rocks with a blue dotted line to distin- guish apparently one kind of crystalline rock from another, yellow for sand, and gray for limestone. Factories, limekilns, and coal, iron, copper, silver, and gold deposits were indicated by signs. 3 Phil. Trans., vol. xiv, p. 739. *Mem. Acad. Roy. Frcmce, vol. for 1746, pp. 343-392. 5 " Historia Terrae et Maris, etc." Acta Acad. elect. Moguntinae 1762, pp. 44-209. 6 Mem. Acad. Roy. France, vol. for 1771, pp. 705-775. 207] W. B. Clark 9 In 1778 Charpentier7 published a book on the mineralogy of Saxony accompanied by a so-called petrographic map on which red is used for granite, purple for gneiss, pink for schists, blue for limestones, gray for gypsum, yellow for sand- stone, drab for river sand, and green for clay and loam. Palassou in 1781 8 published an essay on the Mineralogy of the Pyrenees in which the routes in the south of France are colored according to the rocks they cross : green for granite, yellow for schists, and red for limestone, while sands, sand- stones, and clays are indicated by signs, as are also extinct volcanoes. Much the most important of these early colored geological maps were those of William Smith, who prepared fifteen county geological maps of England between 1794 and 1821. These and his general geological map of England published in 1815 mark the beginning of modern geological cartography. Many refinements have been introduced in subsequent years and an attempt more or less successful has been made to secure cooperation on the part of geologists the world over in the use of the same colors for rocks of the same age and char- acter. At first all maps were colored by hand, but in recent years lithographic processes have been introduced, although the Geological Survey of Great Britain continued to color its geological maps by hand until a few years ago. The Mary- land Survey has made only colored lithographic geological maps. In addition to the strictly geological work carried on by the instructors and students of the Geological Department to which I shall again presently refer, the Survey has secured the cooperation of the staffs of several Federal and State Bureaus (1) in the making of the base topographic maps, necessary not only for the geological but other kinds of sur- veys; (2) in the classification and platting of the agricultural soils which are the disintegrated surface portions of the geo- Mineralogische Geog. d. Chursachsische Lande, 1778. Essai sur la Mineralogie des Monts Pyrenees, Paris, 1781. 10 Geological Surveys [208 logical formations combined with vegetable debris or humus ; (3) in determining the magnetic elements of variation, declination, and force so much affected by the underlying rocks, and (4) in the study of the surface and underground waters so largely dependent on the geological structure. In addition to these lines of work the Survey was directly con- cerned in the past in organizing the modern methods of state highway construction which were introduced in 1898, and for ten years thereafter it was the only state agency intrusted with this important service and until the transfer of its Highway Division in 1910 to the newly-organized State Roads Com- mission, on which it also had, by law, influential representa- tion until 1914. During this time the Survey built 150 miles of public highways at an expenditure of nearly $1,500,000, but more than that, developed standards of work and a trained engineering force that today largely control this important public enterprise. The Survey has also participated in the re-surveys of many of the state boundaries, including the re-survey of the Mason and Dixon Line, and also of several county boundaries. It has made extensive geological and mineral exhibits at the Buf- falo, Charleston, St. Louis, Jamestown, and San Francisco Expositions, the more important materials then secured being today maintained as a permanent exhibit in the Old Hall of Delegates in the State House at Annapolis. Returning now to the strictly geological work of the Survey, I wish to call attention to the fact that some of our official organizations, carried away by the clamor for immediately practical results, are devoting their time much more than in the past to present commercial needs, ignoring the fact that their greatest service to the public is in studying the funda- mental scientific problems furnished by the rocks even when they appear to afford no application to the industries of today. I feel that a Survey that is continually thinking of the practi- cal results it can secure should not have the name of geologi- cal, for I dislike to feel that geology has no higher public 209] W. B. Clark 11 value than to provide means for the shrewd business man who may employ its results to acquire a few more dollars. It is my belief that if the work of a Geological Survey is properly done, with one regard to the solving of the scientific problems as they arise, it will indirectly do the commercial interests of the community a greater service in the end than if the practical aspects of the work are given first place. It would not be diffi- cult to demonstrate this in the case of our Geological Surveys if I had the time. Maryland possesses three provinces : first the Coastal Plain which consists of the low-lying country extending from the ocean front to a line drawn through Elkton, Havre de Grace, Baltimore, Laurel, and Washington, which consists of nearly unaltered sediments of relatively simple structure, the basal members of which date well back in geological time, in fact, before the Eocky Mountains or the Alps were formed. They afford a series of problems of great interest but quite different from those of the other areas. Lying to the west of the landward border of the Coastal Plain and extending to the base of the Catoctin Mountain is a second area known as the Piedmont Plateau consisting of highly metamorphosed crystalline rocks cut by intrusive and extrusive eruptive rocks, the whole subjected to extensive deformation with intricate folds and faults. The rocks are very old and probably comprise the southern extension of the great Canadian shield, the oldest portion of the North Ameri- can continent. Here are problems of great interest to be solved. Beyond and extending to the western limits of the State is a third area known as the Appalachian Eegion that contains a great thickness of deposits lying in large part intermediate in position both as regards age and structure between those of the Coastal Plain and the Piedmont Plateau. Along their east- ern margin they are metamorphosed, folded, and faulted with a large admixture of eruptive rocks that become progressively less complicated westward. In this district still other prob- lems are presented. 12 Analyses of Igneous Rocks [210 I might go on and enumerate in much greater detail the innumerable questions which such an area as Maryland pre- sents to the geologist. We have been engaged, as I stated earlier, for over twenty years in trying to reach a solution of some of these problems, but our successors will, I am sure, find enough to keep them fully occupied for another genera- tion if it is not vouchsafed for us to keep actively employed in studying them during that time. The question may be asked, is often asked by the custodians of the public funds, will this work never end ? But I must answer no, not as long as there is a science of geology worthy of the name. THE USE OF AVERAGE ANALYSES IN DEFINING IGNEOUS ROCKS By EDWARD B. MATHEWS There is usually associated with the consideration of rock names and their meanings some attempt to represent the characteristic chemical composition connoted by the name. The methods employed usually /consist of the presentation either of a series of analyses of individual rocks with little or no discussion of their meaning or of an arithmetical mean of a varying number of such analyses in the form of an " aver- age " analysis also without discussion of the departures from such averages which may be shown by the analyses on which this " average " is based. Neither of these methods is very satisfactory for teaching or textbooks. The presentation in columnar form of a series of analyses each of which includes from eight to fifteen determinations bewilders the student who seldom stops to consider just what the variations amount to either absolutely or relatively, or what relations the variants bear to the general type. This method of enumerating the actual composition of individual rocks may be eminently proper in a Handbook but fails of its purpose in a textbook. 211] E. B. Mathews 13 The presentation of a single " average " analysis possesses all the charm of simplicity and ease of comprehension but fails to convey a proper conception of the complex variability underlying it. The student, with retentive memory, may hold the values assigned to the type but may gain thereby little knowledge of the real content of the term. If the rock se- lected is in itself sharply defined., or if the examples collected are sufficiently numerous the " average " analysis may be satisfactory. If, on the other hand, the individual rocks in- cluded under a given name are aggregates of minerals of vary- ing composition in various proportions such as might occur in a complex of numerous related and unrelated continuous gra- dations without any semblance of "clustering," then the " average " analysis gives nothing more than the arithmetical mean of the quantities which have been included. As Cross x remarked in his criticism of the classification proposed by Loewinson-Lessing, " the grist of this mill depends entirely upon what is put into the hopper." While even a momentary consideration shows that what has been said regarding individual concepts applies even more strongly to group concepts, the writer has considered it worth while to test quantitatively the variabilities actually involved in " average " analyses. The test is limited to anorthosites and non-feldspathic pyroxenites and peridotites and the methods employed are both graphic and arithmetic. EXAMPLES Anorthosite is composed of approximately a single mineral or at least of representatives of a single isomorphous series. The natural presumptions are that their analyses would repre- sent a continuous series and their -average an intermediate member. A graph of the analyses found in the literature shows no such evenly distributed series, at least so far as lime and soda are concerned, but three distinct types; one with 1J. Geol., vol. X, 1902, p. 481. 14 Analyses 'of Igneous Rocks [212 approximately 16% CaO, the most abundant with approxi- mately 10% and a third with about 3%%. While the usual " average " analysis of anorthosite would represent in a gen- eral way the more abundant variety the breadth of range, the discontinuity of the series and the existence of grouping is obscured. Dunite (Fig. 1) is another rock consisting largely of a single mineral. Here the analyses shows little variation except in the ferrous iron which shows an absolute range of 10% and a relative range of 156% of its mean amount. Magnesia shows an absolute range of 7% or about 15% of the average content. The isomorphism of the olivine group would suggest more uniform departures of the iron and magnesia from their aver- age values. Horriblendite (Fig. 2) is a third representative of rocks composed essentially of a single mineral or mineral group. Here the possibility of alumina in the molecule suggests wider departures from the average but the analyses show an absolute range in alumina of scarcely 9%, although nearly 80% of the average content. Ferric iron shows about the same absolute range but a much greater relative range because of its lower average content. Ferrous iron, on the other hand, with an average content similar to that of alumina, shows a relative range of 175%, while magnesia and lime show total ranges of less than 100% of the average content. From the foregoing it seems reasonable to infer that con- clusions based upon " expansions " of averages in accordance with the known isomorphism of constituent minerals are un- certain even in monomineralic rocks. Turning to rocks consisting essentially of olivine with one or more pyroxenes or hornblende two rocks were selected. Uarzburgite or Saxonite (Fig. 4) composed of olivine and an orthorhombic pyroxene shows several variant types which are entirely obscured in an average analysis. The rocks which have been named harzburgite compared with those called Saxonite generally show lower alumina and lime with higher 213] E. B. Mathews J5 FIG. 1 TO 8. Graphs showing composition of igneous rocks used in obtaining average analyses. 16 Analyses of Igneous Rocks [214 magnesia and similar silica, ferric and ferrous iron. These variants may be due to chance and their distinctness to paucity of example, but such differences as may exist are obliterated by the use of an " average " or even " typical " analysis. Lherzolite (Fig. 3) consisting of olivine, bronzite and diallage shows on the whole a wide but uniform distribution in the content of the various constituents in the analyses at hand. The graph shows the abnormality of the Iherzolite from the Protrero, San Francisco with its low magnesia and high lime which raises a question as to the applicability of its name. All of the examples thus far considered have been named by workers of different experience and training and the sugges- tion comes that the variations in composition are due in part at least to incomplete comprehension of the content of the terms used or to personal variations in usage. To illustrate range acceptable without these factors two sets of examples of unique types studied by single workers were selected. Koswite. Composed essentially of olivine and magnetite with diopside, and some hornblende, and chromite includes a series of magnetite pyroxenites described from the Urals. They are characterized chemically by high ferric and ferrous irons. While the analyses are incomplete through lack of alkalies their summation is in every instance over 100, show- ing that this lack does not vitiate them for the present purpose. The departures from the mean of values for the individual constituents is here usually only two or three per cent, or less than 50% of the value of the dominant constituents. Ariegites as defined by Lacroix are a group of pyroxenites characterized by the constant association of one or more pyroxenes and a spinel with varieties due to the presence of garnet and hornblende. In this series the absolute range is 4% to 8% and the de- partures in the case of the principal constituents is not over 10% of the mean values. 215] V. B. Mathews 17 From the foregoing it would appear that mature workers even in establishing their types believe it allowable to include rocks whose dominant constituents show departures ranging from 10% to 30% from their mean values. The use of average analyses in the description of rock groups may or may not prove more serviceable. "While there are more possible variations in the kinds and proportions of minerals the graphs may show no wider variations than those noted in discussing individual types. Thus the diagram (Fig. 7) showing pyroxenites, exclusive of the websterites (Fig. 8) although representing several kinds and many ex- amples of pyroxenites, is not much more confusing or variable than that for the websterites by themselves. The peak for lime is more marked but the predominance of the magnesia and lime with the subordination of the irons and alumina are nearly as clear. The more general diagram carries aberrant types like the pyroxenite from Rosetown, N. Y. (high alumnia and low magnesia) and the magnetite pyroxenite from Cen- tral City, Colo, (high in irons) . The former is a poor analy- sis while the latter is recognized as aberrant and their inclu- sion in any general average analysis is doubtful. Similar graphs of gabbro, dacite and camptonite show fairly well defined figures which indicate that the impressions gained from average analyses while incomplete may not be incorrect in the major essentials. Average analyses cannot be expected to bring out minor " clusterings " or many of the relationships in constituents which are disclosed by the simple serial diagrams here employed. The same is true of several of the systems of projection now employed in petrography. These, moreover, often require ex- perience and maturity beyond that of the average student of systematic petrography for their complete appreciation. 18 Tuscaloosa Formation [216 THE DELTA CHARACTER OF THE TUSCALOOSA FORMATION BY EDWARD W. BERRY During the enormous interval of time represented by ma- rine sediments in other parts of the world of late Carboni- ferous, Permian, Triassic, Jurassic and Lower Cretaceous ages the southern Appalachian region was above sea level. Physiographically the southern half of this region is segre- gated at the present time into the Piedmont Plateau, the Appalachian Mountains (which die out in northwestern Georgia), the Appalachian Valley, the Cumberland Plateau, and the Interior Lowlands. Its area south of the Ohio Eiver is over 160,000 square miles, and the actual area of this land mass during the interval from the Carboniferous to the Upper Cretaceous must have been very much greater than this, since nowhere along the margins of this massif have marine sedi- ments of these ages been deposited near enough to its present limits to be reached by deep borings near the margin of the present Coastal plain. The region of the southern Appalachians is one that has long interested physiographers. Hayes and Campbell, the chief contributors,1 have recognized three base levels or peneplains which they term in the order of their ages the Cumberland, the Highland Rim and the Coosa. They con- sider that the original Tennesee River, which they term the Appalachian Eiver, flowed southwestward by way of the valley of the Coosa Eiver throughout the Upper Cretaceous and the major portion of the Tertiary until it was diverted by stream capture due to the working back across Walden Eidge of a stream in the Sequatchie valley to the west of that ridge. 1 Hayes, C. W. and Campbell, M. R., "The Geomorphology of the Southern Appalachians," Natl. Geographic Magazine, vol. 6, pp. 63- 126, 1894; Hayes, C. W., "The Physiography of the Chattanooga District," U. S. Geol. Survey, 19th Ann. Rept., Ft. 2, pp. 1-58. 1899. 217] E. W. Berry 19 This spectacular river capture has been disputed by Johnson 2 who, it seems to me, conclusively demonstrates that the pre- sent course of the Tennessee River across Walden Ridge in a winding gorge is imposed from meanders inherited from the FIG. l. Map showing physiographic regions and areas of outcrop of Tuscaloosa, Eutaw, and Selma formations. period of earliest complete baselevelling in this area, namely from the Cumberland peneplain. The character of the Upper Cretaceous sediments of the eastern Gulf area throw considerable light on the physical 2 Johnson, D. W., " The Tertiary History of the Tennessee River..' Jour. GeoL, vol. 13, pp. 194-231, 1905. 20 Tuscaloosa Formation [218 history which has interested me chiefly in connection with the interpretation, in terms of geologic history, of the extensive fossil floras that have been found in the earliest Upper Creta- ceous or Tuscaloosa formation of this region. The Tuscaloosa formation in the area around Tuscaloosa, Alabama, and for some distance to the northwest consists of about 1,000 feet of predominantly sandy materials which give the country its present topography. These sands are usually light in color, cross-bedded and micaceous — occasion- ally there are traces of glauconitic layers. There are heavy beds of gravel made up of well rounded quartz and sub- angular chert pebbles in about equal proportions in places, especially toward the landward margin of the deposits and northward along the strike. In disconnected and interbedded lenses there is a considerable amount of argillaceous material — at times massive or heavy bedded, but generally laminated. Thin seams of lignite are present at various levels but these are generally only a few inches or less in thickness. The clays are often oxidized and mottled in color but they are as fre- quently very carbonaceous and dark in color. In some sec- tions, as in the Big Gully section southwest of the town of Tuscaloosa, there are layers filled with prostrate logs of trees of large size. Pyrite and ferruginous oxide, forming locally indurated sandstones and gravels are generally distributed, and finely disseminated gypsum crystals are very common. No fossils other than the remains of land plants have been found in the Tuscaloosa deposits. Usually the plant remains are much macerated and broken by water transportation and deposited in films of broken fragments in the laminated beds. Drift logs are common and these occasionally brought down cobbles imbedded in their roots (statement based on speci- mens collected in lignitized tree roots). There appear to have been areas of quiet waters at certain localities where the leaf remains in the clays are abundant and in a state of preservation indicating that they grew in the immediate vicinity. The outcrop of the Tuscaloosa formation, as shown in the 219] E. W. Berry 21 accompanying sketch map, is roughly lunate in outline with the southeastern horn terminating near Montgomery, Ala- bama, and the other extending as an attenuated band across western Tennessee. As will be seen, the greatest width of outcrop coincides with the maximum thickness of sediments in a belt about 125 miles in length which is at right angles to the axis of the Appalachian land mass. To the northward the deposits become thinner, are prevailingly gravels and are shown by the fossil plants to be somewhat younger than the main body of the deposits. In the interpretation of the Tuscaloosa deposits with their gravels and compound oblique cross-bedded sands, their occa- sional traces of glauconite and their abundance of driftwood, one cannot fail to be impressed with their delta-like character. We are now fairly familiar with the main features of delta deposits in different parts of the world 3 and Grabau 4 and Barrell5 have recently contributed considerable toward the interpretation of Paleozoic delta deposits. Eeturning for a moment to the physiographic history of the Tuscaloosa region we find that there are no sediments later than Pottsville age until the deposition of the Tuscaloosa in the earlier Upper Cretaceous. This long interval resulted in the nearly com- plete baselevel known as the Cumberland peneplain. There must have been some regional uplift or warping at the begin- ning of Tuscaloosa time to account for the sudden augmenta- tion in river action and the inauguration of the large delta or series of deltas along the southwestern margin of the land mass. There is no evidence in the sediments that an Appa- lachian river flower southwestward through the Coosa valley. This would also have brought the bulk of the sediments far- ther eastward than where they now occur. While I regard Johnson's evidence (op. cit.) as conclusive for the course of 3 Credner, H., " Die Delten," Petermann Geog. Mitth., Erganzungs- heft 56, pp. 1-74, pi. 3, 1878. 4 Grabau, A. W., Early Paleozoic Delta Deposits of North Amer- ica," Bull. Geol. Soc. Am., vol. 24, pp. 399-528, 113. B Barrell, J., idem., vol. 23, pp. 377-446, 1912. 22 Tuscaloosa Formation [220 the Tennessee River across Walden Ridge, I cannot help believing that the Cretaceous ancestor of this stream at the beginning of Tuscaloosa time, instead of making the sharp turn to the northwest at Guntersville, Alabama, which it does at present, continued southwestward down either Brown or Big Spring valleys and reached the sea through either the Mulberry or Locust fork of the Warrior River. This, how- ever, is not an essential part of my argument for the delta character of the Tuscaloosa formation, since there was obvi- ously at that time a stream or a series of streams draining to the southwest and engaged in removing the debris of the long weathered land mass. SW NE POTTSVILLE FIG. 2. Section showing relation of the Tuscaloosa deposits to those of the Eutaw and Selma formations. I do not wish to be understood as ignoring the fact that some of the Tuscaloosa deposits are sub-aerial and that ori- ginally the delta deposits probably continued inland up the valley or valleys for considerable distances as continental de- posits of channels, flood plains and lakes. The antecedent meanders of the present streams give clear evidence of con- ditions that prevailed on the Cumberland peneplain that were suitable for the formation of ox-bow lakes. There must have been quiet waters in the delta itself in certain bayous or possibly lakes like lakes Salvador, Ponchartrain and Borgne of the present Mississippi delta region. Certainly the leaf- bearing clays near Glen Allen and Shirley's Mill in Fayette County, Alabama, were formed in such quiet bodies of water with densely wooded shores. The relations of the Tuscaloosa formation emphasizes its delta character as is shown in the accompanying textngure. 221] E. W. Berry 23 The Tuscaloosa sands grade seaward into the glauconitic sands and thinly laminated clays of the Eutaw formation which contains a sparing representation of the marine life of the time,, which must have been in part at least contem- poraneous with the Tuscaloosa. A few plants in the near shore transgressing phase have been collected from near Havana in Hale County, Alabama. The upper Eutaw, or Tombigbee sand member I regard as a transgressing deposit and in conformity with this interpretation it contains a much better marine fauna than the earlier Eutaw deposits. Over- lying the Eutaw formation is the Selma chalk — an argilla- ceous limestone or calcareous clay which reaches its maximum thickness in the same region as does the Tuscaloosa sands, namely southwest of the axis of the Appalachian land mass. In this area the Selma chalk continues upward to the Eocene contact. Its outcrop as shown on the accompanying sketch map is almost perfectly lunate, and at its horns both to the east and the north it passes over into sands. The Selma, as shown by its abundance of Ostracea and other Mollusca is a shallow water deposit. So far as my observation goes it is entirely destitute of drift wood, lignite or any considerable sandy beds in the area of its greatest thickness and the point that I wish to make is that the southwestern drainage that explains the character of the Tuscaloosa sediments must have been reduced to a minimum or become practically non-exist- ant before the deposition of the Selma chalk. The prevailing direction of the drainage during Selma time must have been to the southeast and northwest in order to explain the Eipley sands in those regions and the absence of any except the finest terrigenous materials in the main body of the Selma chalk. Inferentially if the Cretaceous Tennessee River was a fac- tor in the building of the Tuscaloosa delta, local warping must have broken its continuity with the Warrior drainage and started it toward the northwest before the deposition of the Selma chalk. It is possible that this may have been accomplished without local warping by the simple clogging 24 Mineralizers in Ore Segregations [222 of its distributaries as a result of their own loads combined with decreased run off. This set of factors combined with the westward tilting that resulted in the Eipley Cretaceous and Midway Eocene seas penetrating up the Mississippi val- ley as far as southern Illinois is sufficient to account for the observed change, of course assuming that there has been such a change. This may be compared to the analogy of the shift- ing of the present Mississippi delta to the eastward by marine currents. The remnants of heavy gravels of Tuscaloosa age that have been traced by Wade across Tennessee and into Kentucky appear to represent the gradual migration or shifting north- ward of such a stream. That the western Highland Eim of Tennessee is a middle or late Tertiary planation of pre- vailingly siliceous rocks by the Tennessee River in its lower northward course is probably true but hardly within the scope of the present brief note. THE ROLE OF MINERALIZERS IN ORE SEGREGATIONS IN BASIC IGNEOUS ROCKS By JOSEPH T. SINGEWALD, JR. Though one of the latest groups of ore deposits to be defi- nitely recognized, the magmatic segregations were firmly es- tablished as one of the major types through the classic work of J. H. L. Vogt twenty-five years ago ; and it has been gener- ally felt by economic geologists that the mode of formation of these deposits was so clearly understood that they consti- tuted a group concerning the genesis of which there was no further question. More thorough petrographic studies of many examples of deposits classed with the magmatic segre- gations and metallographic investigation of the ores, espe- cially during the last decade, have accumulated more and more evidence to show that mineralization was not so simple and did not conform strictly to the conception of a magmatic segregation in the sense in which that term is generally 223] J. T. Sing ew aid 25 thought of. The time has arrived for a definite recognition of these discordant data and a remolding of our conceptions in harmony with them. The term magmatic segregation was borrowed by the eco- nomic geologist from the petrographer and used in the petro- graphic sense. Its application to the explanation of the genesis of certain ore deposits seemed very plausible. It is a matter of common petrographic knowledge that no large body of igneous rock is of uniform composition and that frequently the composition of a portion of the mass departs widely from the average. Consequently there were forces at work prior to or during the consolidation of the molten magma which caused local segregations of certain of its con- stituents. The nature of these forces has long been a matter of discussion and speculation but unanimity of opinion has not been attained and there is no thoroughly satisfactory explanation of the process which is known as a magmatic segregation or differentiation. The usual manifestation of the phenomenon is in the local accumulation of the more basic constituents of the magma. The basis for the application of this process as an explana- tion of ore genesis rests on certain other observations in the field of petrography. There is present in almost all igneous rocks a group of opaque minerals, occurring as accessory con- stituents and with euhedral forms, the most common repre- sentatives of which are the sulphides of iron, frequently cu- priferous, and the oxides of iron, frequently chromiferous or titaniferous. On account of their commonly euhedral forms, these minerals are regarded by petrographers as the earliest constituents of the magma to crystallize. Furthermore, these minerals are concentrated together with the basic silicates in the process of rock differentiation. There are many ore deposits world-wide in their distribu- tion possessing certain common characteristics, among which may be mentioned: (1) the ore minerals consist of one or more of the accessory opaque minerals common to igneous rocks; (2) the enclosing rock is always an igneous rock and 26 Mineralizers in Ore Segregations [224 usually basic in composition; (3) the gangue minerals of these deposits are the same as the constituent minerals of the enclosing rock; (4) the* ore body frequently passes by gradual transition into the igneous rock by a decrease in the amount of the ore minerals and increase in the amount of the silicates. These deposits appeared to be an integral part of the igneous rock in which they are found and to represent an extreme facies of the product of rock differentiation, and consequently were established as an independent group of ore deposits to which the name magmatic segregation was applied. The group was subdivided by Vogt into three divisions ac- cording as the metal occurs in the native, oxidic or sulphidic form. Segregations of native metals as primary ore deposits are of little economic importance, but the placers derived from such of native platinum in the Urals are our principal source of that metal. Segregations of oxidic ores are our only source of chrome ores, include countless deposits of titaniferous iron ores, and important deposits of non-titan- iferous iron ores. Segregations of sulphidic ores include the nickeliferous and supriferous pyrrhotites and probably a few copper sulphide deposits. The metallic content of the segregations of the native metals is usually rather sparsely disseminated through the rock, and on account of the few examples and their minor importance, the propriety of regarding the metal as a segre- gation and product of crystallization from a molten magma has been little questioned. The segregations of oxidic ores usually occur well within the igneous mass, and there are so many admirable illustrations of gradation from ore-mineral bearing rock to ore body that no particular significance has been attached to the observation in a number of instances that the ore minerals are later than the silicates, and the conception of a segregation and solidification from a molten magma has been rarely challenged. The position of the sul- phidic deposits has, however, been somewhat dubious from the start. They tend to occur on the periphery of the ig- neous mass, the sulphides often penetrate into the wallrock, 225] J. T. Singewald 27 it was early recognized that in part at least the sulphides are distinctly later than the rock-forming silicates, and the rock itself has frequently undergone considerable alteration. Many geologists have consequently insisted on .regarding them as hydrothermal deposits. A most interesting feature of the controversy over the genetic position of these sulphidic ores has been that the largest and most important example, the nickel deposits of Sudbury, Ontario, which has been cited by the advocates of the magmatic origin as a typical illustra- tion of that type, is one to which most serious objection has been raised by those contending for a hydrothermal origin. In view of the departures manifested by these deposits from the conceptions based on purely petrographic pheno- mena and concepts, it is interesting to see how the problem has been handled in four of the leading recent textbooks on ore deposits. The four selected are: R. Beck, Die Erzlager- statten, 1909 (3rd edition) ; Beyschlag-Krusch-Vogt, Die Lag erstdtten der nutzbaren Mineralien und Gesteine, 1910; W. Lindgren, Mineral Deposits, 1913; L. DeLaunay, Giles Mineraux et Metalliferes, 1913. Beck, in denning magmatic segregations, says : " In many instances there took place in the rock either before or during solidification from the molten state a concentration of the ores into irregular masses. ... In spite of the concentra- tion into a limited space, the ores of such deposits remain, what they as scattered particles in the rocks in question are, namely accessory constituents." Commenting on Vogt's ob- servation that in certain of the Swedish and Norwegian titani- ferous iron ore deposits the silicates formed first and then the titanif erous magnetite, he says, " these are departures from the rule otherwise prevailing for eruptive rocks that the iron ores belong to the earliest minerals to separate." In all chrome deposits for which he cites the sequence of crys- tallization, the chromite is the earliest constituent. Of the sulphidic deposits, on the other hand he says, " The strict proof of segregation from a molten magma cannot always be established with the same degree of sharpness. . . . For a 28 Mineralizers in Ore Segregations [226 great many occurrences, which numerous authors consider a direct segregation from eruptive rocks, one must at least con- sider probable a later secondary recrystallization of the ores by aqueous processes which brought about a partial migration and an impregnation of the wallrock." These conclusions are based largely on his own work in 1902 and 1903 on the deposits of nickeliferous pyrrhotite and chalcopyrite at Soh- land in Saxony, where he found ore deposition took place by replacement subsequent to the hydrothermal alteration of the rock, though he believes both followed immediately after its solidification. "With the exception of the treatment of the sulphidic ores, the position of Beyschlag, Krusch and Vogt is similar to that taken by Beck. Their ideas are of course largely those devel- oped by Vogt. Their conception of the genesis of these ore deposits is that, " In the same manner in which larger masses of mica and feldspar can collect out of a granite magma, segregation of ores can take place, as for example of mag- netite, titaniferous magnetite, chromite and pyrrhotite, in such igneous rocks which normally carry these ores as acces- sory constituents." They call attention to the fact that the ore minerals followed by the iron-magnesium silicates are the earliest constituents to crystallize in most eruptive rocks and they are also the constituents that migrate in magmatic differentiation. Magmatic segregation is distinguished from ore deposition in which mineralizers participate as follows: " The genetic difference consists essentially therein that the magmatic segregations result from a single differentiation process of the magma, whereas in the case of the pneumato- lytic and contact metamorphic deposits the metallic content originally belonging to the magma is transferred to an aque- ous or gaseous solution and later deposited from this through new processes." Though in most cases the two groups are considered as sharply differentiated, they admit that occa- sionally there are intermediate stages in which magmatic dif- ferentiation is accompanied by pneumatblytic or pneumato- hydatogenetic processes. They definitely state that the chrome 227] J. T. Singewald 29 ores crystallized out of a magma and that the forma- tion of the titanif erous magnetites " depends on a pure mag- matic separation, not accompanied by special pneumatolytic processes," and that the process differs from ordinary rock differentiation only in that it has proceeded much further. The characteristic association of more or less titanomagnetite with the sulphides is taken to indicate a genesis for the latter ores analogous to that of the titaniferous magnetites. In further substantiation of the magmatic origin of the sul- phides is the statement that secondary alterations such as uralitization before ore deposition or contemporaneous with it has not in general occurred. They emphasize the fluidity of the molten sulphides and the consequent power of pene- tration into minute crevices and cracks and have proposed a subdivision of injected sulphide deposits, which represents intrusions of molten sulphides into the country rock. In such an interpretation of a number of the most important examples included under this subdivision, however, they stand almost alone.' The deposits under discussion are classed by Lindgren as " Mineral deposits^Jefmed by concentration in molten mag- mas," concerning which he says : " Certain kinds of mineral deposits form integral parts of igneous rock masses and per- mit the inference that they have originated, in their present form, by processes of differentiation and cooling in molten magmas/' Of the oxidic ores he says chromite appears in all cases to be the earliest consolidated constituent, but that the titaniferous iron ores have as a rule crystallized after the silicates ; but he says further about the latter : " Petrographic research has long ago shown that ilmenite with magnetite is one of the earlier products of consolidation in magmas and is contained in almost all diabases, basalts, and gabbros. . . . The larger masses of ilmenite are simply facies of the rock itself produced by concentration from the same magma." Lindgren's position concerning the sulphides is almost identi- cal with that of Beyschlag, Krusch and Vogt, as is evidenced by such statements as, " Some of the magmatic sulphide 30 Mineralizers in Ore Segregations [228 deposits are simple basic rocks abnormal in containing much pyrrhotite, chaicopyrite and pentlandite," and, " Some de- posits in which the ore consists mainly of solid pyritic minerals present features which can hardly be explained otherwise .than by actual injection of molten sulphides," in spite of his admission that " on the whole the sulphides are the latest products crystallized." In reply to the advocates of a hydrothermal origin for certain of these deposits, % he charges them with having confused secondary changes with primary deposition. DeLaunay's treatment of these deposits differs consider- ably from any of the preceding. The first five divisions of his genetic classification of ore deposits are the following: 1. Gites d'Inclusions. 2. Gites de Segregation. 3. Gites de Depart Immediat ou de Segregation Peri- pherique Sulfuree. 4. Gites de Contact du Type Banat. 5. Impregnations Diffuses de Profundeur (includes amas pyriteux) . " Deposits of inclusions are those where a useful mineral occurs in an igneous rock in the same relation as the other constituent elements." This division is of theoretic rather than practical importance, and includes only native metals and oxides present as normal accessory constituents of an igneous rock without the intervention of mineralizers. The segregation deposits, he says, might be regarded as having been effected without intervention of volatile constituents, nevertheless the general opinion today is that water and pro- bably other mineralizers have played a role, though he con- siders them formed in a medium poor in mineralizers. The ores are native metals and oxides. The Gites de Depart Immediat, or peripheral sulphide seregations, he says, have usually been considered examples of true segregations which differ from the internal oxidic segregations by their position and nature. DeLaunay believes it necessary, however, to 229] J. T. Singeiuald 31 separate them entirely from the true segregations, for there has been a concentration of sulphides not only in the rock but at its contact, and they appear to him to be a close parent to contact metamorphic deposits which are formed when the wallrock is a limestone. They represent a type of ore deposition in which mineralizers are more abundant and active than in the preceding. The very close relation postu- lated between these deposits and typical contact metamorphic deposits indicates clearly that he does not look upon them as representing a crystallization from a molten state. The mineralization of the last group is analogous to that of the fourth, the types of deposits included under it being formed where the country rock is other than a carbonate rock. It includes most of the deposits classified by Beyschlag, Krusch and Vogt as injected sulphide deposits. There are no sharp lines of demarcation between these groups, as DeLaunay recognizes a complete transition from purely igneous deposits to hydrothermal veins and that in some instances it is difficult to decide between fusion and solution. In an attempt to settle some of the doubtful points con- cerning the mode of formation of the sulphidic ores usually classed as magmatic segregations, C. F. Tolman, Jr. and A. F. Eogers of Stanford University have just published the results of a very comprehensive petrographic and metallo- graphic investigation of those ores as a monograph entitled " A Study of the Magmatic Sulfid Ores." They formulate a number of statements which they find applicable to all of the deposits studied, the most significant of which are : 1. The first minerals to form are olivine, the pyroxenes and the feldspars. 2. Magmatic alteration of the silicates often takes place prior to the formation of the ore minerals. The most com- mon change is that of pyroxene to hornblende, but easily distinguishable from the hydrothermal process of uralitiza- tion. 3. The ores replace the silicate minerals but without re- action rims. 32 Mineralizers in Ore Segregations [230 4. The ores are introduced one after another in the fol- lowing invariable sequence: (1) magnetite and ilmenite, (2) pyrrhotite, (3) pentlandite, (4) chalcopyrite. There is a certain amount of replacement of the earlier ore minerals by the later ones. 5. Hydrothermar alteration is distinctly later than the period of ore deposition. These data of observation lead them to a theory of genesis more nearly analogous to that of DeLaunay than any of the others mentioned above. The fact that the ore minerals replace the silicates without the formation of metallic sili- cates by reaction is interpreted to mean that the ores were not introduced in a molten state but that the same agency that brought in the sulphides removed the dissolved silicates, indicating the presence of active mineralizers. The altera- tion of pyroxene to hornblende is further evidence of the presence of mineralizers. Consequently they conclude that mineralization took place at a temperature below the melting point of the ores and that they were held in solution through the agency of mineralizers. On the other hand, that ore de- position took place under conditions different from those of non-magmatic high temperature deposits is shown by the absence of the secondary silicates characteristic of ordinary pneumatolytic and hydrothermal processes, or that where present they belong to a distinctly later period. They con- clude, that the magmatic ores "have been introduced at a late magmatic stage as a result of mineralizers." The direct evidence presented by Tolman and Eogers is derived from the sulphide deposits, but the presence of a greater or less quantity of titaniferous magnetite in these, and numerous references in • the literature to the silicates pre- ceding the ores in order of crystallization in deposits of titan- iferous magnetite, led them to infer that the same observa- tions and same conclusions apply equally well to the oxidic ores. My own experience with the titaniferous magnetites corroborates the correctness of this inference. The relations between ore minerals and silicates figured and described for the 231] J. T, Singewald 33 sulphide ores are repeatedly duplicated in thin sections and polished sections from all occurrences of titaniferous iron ores in the United States. Titaniferous magnetite later than the silicates and replacing them is seen in nearly every section of the ores, though many of the contacts of the two sets of minerals show what L. C. Graton and D. H. McLaughlin have recently termed mutual boundaries, that is, boundaries that give little evidence of the sequence of the minerals. Only rarely is there unmistakable evidence of primary sili- cates distinctly later than the ore. The replacement of the silicates by the ore has not been accompanied by the forma- tion of reaction silicates and in several instances hornblendi- zatioirhas preceded the deposition of ore, phenomena in har- mony with the nature of mineralization in the case of the sulphide ores. If segregation takes place without the interven- tion of mineralizers, one might expect the deposit at Iron Mountain, Wyoming, to afford such an example. The ore body there occurs as an almost pure mass of titaniferous magnetite .cutting the anorthositic country rock as sharply as any igneous dike ever pictured. Yet the numerous oliviny crystals which occur locally in the ore are rounded and em- bayed without the formation of reaction silicates in exactly the same manner as in other occurrences. The deposit sug- gests an injection from a basic magma analogous to a peg- matite from a more acidic. In other cases, particularly at Grape Creek, Colorado, the introduction of ore has been accompanied by alteration of the feldspar so that the mag- netite is separated from it by a band of hornblende, indi- cating greater than usual activity of mineralizers. The Min- nesota deposits conform for the most part to the general rule that the ore is later than the silicates and afford some exam- ples of hornblendization preceding or contemporaneous with ore deposition, but also instances of feldspar and pyroxene later than ore. An excellent example of an iron ore deposit in a basic igneous rock giving unmistakable evidence of active partici- pation by mineralizers in the formation of the ore is afforded 3 34 Mineralizers in Ore Segregations [232 by. the Tofo deposit north of Coquimbo, Chile, being worked by the Bethlehem Steel Company. This consists of a large mass of comparatively pure magnetite forming the top of a hill on the east side of the coast range and occurring within a large area of gabbro rock. The igneous mass has undergone considerable differentiation and various rock types are repre- sented in the vicinity of the ore body from highly feldspathic to almost pure ferromagnesian silicate rocks, some of which occur as dikes. The broader relations of the ore body are such as to suggest at once a magmatic segregation; but, at the same time, there are many features that suggest pneuma- tolysis. Adjacent to the ore, there are numerous stringers of magnetite in the country rock, many of them of no greater thickness than a knife blade, which traverse it in such a way as to preclude the entrance of molten oxides and that can be explained only on the basis of the high liquidity at lower temperature that would be imparted by the presence of abundant mineralizers. The argument for the participation of mineralizers in the formation of magmatic deposits is a plausible one ,also from a general standpoint of ore genesis. Processes in nature representing different stages of a sequence from a given starting point are not usually separated by a hiatus. It is generally accepted today that igneous magmas are the pri- mary sources of the metals and modern genetic classifications group ore deposits according to their position or relation to the original sources. It has been customary, however, to- draw a sharp line between one group of deposits which it was held segregated from the molten magma and solidified with it, and such groups as represented deposition of material extracted from the magma by mineralizers and constituting the pneumatolytic and hydrothermal deposits. DeLaunay's classification recognizes no such hiatus in the sequences of mineralization, but postulates a gradually increasing partici- pation of mineralizers and hence a gradual gradation from one stage to the next. It goes even a step further and indi- 233] J. T. Singewald 35 cates that concentration of the metallic content of a magma to the extent necessary to form important ore bodies takes place only when the necessary migration of the metals is aided by the presence of mineralizers. Eor the only group recognized by him in which mineralizers did not participate in ore deposition, his gites d' inclusions, contain no deposits of economic importance; and it is only in his next group, in which mineralizers begin to play a part, that important ore deposits begin to be represented. There have been two lines of thought seeking to explain ore genesis, the one repre- sented by the French school which has always emphasized the role of mineralizers, and the other by the American and German economic geologists who have tended to draw the sharp line of demarcation between the magmatic deposits and the non-magmatic. The participation of the latter group has so greatly preponderated over that of the former during the last quarter century in the development of the science of economic geology that the views of the French school have often been completely overshadowed and have not received the attention they merit. The monograph by Tolman and Rogers will serve to establish among American economic geol- ogists the ideas embodied in the conceptions of the French school. It is hoped that this survey of the problem and the corroborative evidence contributed in the case of the iron ores will serve the same purpose. One cannot help but feel that a new study of the chrome ores with this interpretation in mind would place them in harmony with it. As their case now stands, they seem to be an exception and to represent a direct segregation as the first product of crystallization from a molten magma. 36 Environment of Tertiary Marine Faunas [234 THE ENVIRONMENT OF THE TERTIARY MARINE FAUNAS OF THE ATLANTIC COASTAL PLAIN By JULIA A. GARDNER The Miocene and Pliocene deposits of the Atlantic Coastal Plain have now been mapped in detail from New Jersey to North Carolina and the 'contained faunas, which are prolific and varied, have been rather fully described.1 The following formations have been recognized in the Middle Atlantic region : Maryland Virginia North Carolina Pliocene : "Waccamaw Yorktown Yorktown-Duplin , St. Mary's St. Mary's St. Mary's Miocene: J \ J J Cnoptank Calvert Calvert All of these formations contain extensive molluscan faunas and very large collections have been available for study and comparison. An idea of the richness of these faunas may be obtained from the following census of the faunas of the respective formations : No. of species in genera Waccamaw 325-335 130 Yorktown 364-378 143 Duplin 420-431 154 St. Mary's 326-344 129 Calvert . 80-83 50 1 The more recent literature includes the following : Whitfield, R. P., Hon. U. 8. Geol. Survey, vol. xxiv, 1894. Clark, Martin, Glenn and others, Miocene vol., Md. Geol. Survey, 1904. Gardner, J. A., "The Miocene and Pliocene Faunas of Virginia and North Carolina," Prof. Paper U. S. Geological Survey. (In press.) 235] J. A. Gardner 37 These faunas are exceedingly interesting, not only because of their diversity and the remarkable 'development of certain groups, but also because of the light they shed on the physical conditions under which they lived. In the following notes, which are based on the study of compiled tables of both recent and fossil forms, an attempt is made to summarize the proba- ble physical conditions indicated by this study. Any attempt to reconstruct bottom conditions in the ancient seas must of necessity be based upon data so meagre and so inaccurate that any hope of obtaining absolute values is vain, and yet it does' seem worth while to occasionally gather the imperfect knowledge available and to try to interpret it. Errors do, to a certain extent, neutralize one another and within certain limits general tendencies and relative values can be given with a very considerable degree of assurance. Over 800 species have been determined from the Miocene and Pliocene of Virginia and North Carolina and of these approximately 20 per cent, persist into the recent faunas. Certainly a number so large as this ought to give a fairly true line upon general temperature and bathymetric conditions in the middle and later Tertiaries. It may be well to consider the main sources of error before giving the conclusions which they modify. A. SOURCES OF ERROR IN THE DATA UPON THE TERTIARY FAUNAS. 1. Errors in determination. This is one of the least important. The greater part of the work upon the faunas in question has been done by less than half a dozen students and the same collections, for the most part, have been used for reference. Consequently the determinations, whether accurate or inaccurate, are fairly con- sistent. Furthermore, if two forms are so much alike that there is a question as to their identity, a similarity of envir- onmental conditions is implied, even though the differences may later prove to be specific. 38 Environment of Tertiary Marine Faunas [236 2. A mechanical sorting of the shells. This is a much more serious error and one which it is im- possible to eliminate. One of the most interesting phases of in-shore marine life is the dissimilarity in the dredge hauls within a limited area. The in-shore currents are quite suffi- cient to materially affect the character of the bottom and the distribution of the algal growth and thus to limit the range of a considerable number of species, particularly the vege- table feeders. Unfortunately, almost all such ecologic varia- tions have been washed away. Not only have near-by but distinct assemblages due to slight differences in environmental conditions been commingled but dead shells have been washed down from the river mouths and up from the off-shores and mixed together in a heterogeneous ensemble. The hard parts of the smaller species, many of which constitute an important item in the diet of various fishes may be carried for indefinite distances beyond their normal habitat before being laid down in their final resting place. In recent faunas extra-limital shells are usually so badly worn that their distant origin may be surmised but in the fossil forms it is much more difficult to isolate them by this method. It is, however, reasonable to suppose that forms occurring in any considerable abundance are indigenous to the fauna but inferences made from the presence of only one or two individuals should be guarded. B. SOURCES or ERROR IN THE DATA UPON THE EECENT FAUNAS. 1. Errors in determination. These are much more frequent in the Recent collections than in the fossil because the work has extended over a much greater time and the personal element is much more conspicu- ous. However, errors in the determination of the fossils are frequently parallel to those of the Recent faunas so that the final results are not always affected. The tendency in the Recent work is towards an increasingly finer distinction of species so that the ranges are becoming more and more re- 237] J* A. Gardner 39 stricted. This is especially true in certain families. In a recent zoogeographic study of the West Coast Pyrammidelli- dae 10 faunal zones were differentiated.2 The three most populous were the Oregonic with 70 species, the Californic with 164 and the Mazatlanic with 75. However, only 11 species common to the Oregonic and Californic were recog- nized and only 2 common to the Californic and Mazatlanic. No refined study such as this has ever been made upon any of the East Coast molluscs, but when it is done the number of species will doubtless be greatly increased and their ranges greatly diminished. The fossil forms can then be interpreted in terms of the Eecent with an accuracy and a detail far in advance of anything that is possible at present. The knowl- edge of the Tertiary ecology may approximate the Recent but it can never go beyond it. 2. The limited number of dredging records. Not only are the stations relatively few in number but they are so grouped that there are long stretches which have not yet been touched. The attempt has been made, however, to cover the critical areas, such as that of the Florida coast, Hatteras and Woods Hole. The New England fauna is well known in a general way and extensive collections have been made through the 'Gulf and the West Indies by the Blake and the Albatross. A very short but significant report is that of Bartsch and Henderson upon a two days' collecting trip off Chincoteague Island on the Virginia coast for the pur- pose of determining the extent of overlap of the southern fauna.3 The latest of the larger reports, that upon the Woods Hole region is by far the most satisfactory excepting that it covers an area so restricted and so little diversified. The arrangement, however, is excellent for an ecologic study, the dead shells are isolated from the living and the young 2 Bartsch, P., 1912, Proc. u. S. National Museum, vol. 42, p. 299. 3 Bartsch and Henderson, 1914, Proc. U. 8. Nat. Mus., vol. 47, pp. 411-421. 40 Environment of Tertiary Marine Faunas [238 from the adult, the number of specimens is given and the data upon depth, temperature, salinity, etc., is complete and accu- rate. The bathymetric distribution of most of the southern stations is unfortunate for those interested in determining the limits in depth of the littoral fauna. Very little shore dredging has been done and there are very few records from less than 10 fathoms. A number of unusually rich hauls were made off Hatteras between 15 and 25 fathoms. The 49 and 63 fathom stations include in addition to the native fauna a considerable number of young or more or less worn shells referrable to the more abundant species in the lesser depths but in the great majority of records these fortuitous shells have not been isolated. However, the general relationships which come out of an interpretation of the recent elements in the fossils faunas are probably true, even though the data upon which the results are based is woefully inadequate. Five formations have been recognized in the Miocene and Pliocene of Virginia and North Carolina, — the Calvert, St. Mary's and Yorktown in Virginia and the Yorktown, Duplin and Waccamaw in North Carolina. The Yorktown and Du- plin were probably, for the most part, synchronous though laid down in separate basins. Approximately 65% of the species common to the Calvert and Recent faunas have been reported from north of Hatteras, the limit of range of many of the northern and of the southern forms; 54% of the St. Mary's; 46% of the Yorktown; 35% of the Duplin, and 36% of the Waccamaw. Factors other than temperature have modified somewhat the figures for the Duplin and Waccamaw, for there is no reason to suppose that the Waccamaw sea was not quite as warm as the Duplin. The break between the late Oligocene and the early Mio- cene in the Southern Atlantic states is one of the sharpest in the stratigraphic succession of the Cenozoic. The Oligocene has not been recognized either in Virginia or North Caro- lina but the early Miocene fauna is similar in general charac- ter wherever it occurs along the East Coast. Twenty species 239] J. A. Gardner 41 from the Calvert of Virginia, approximately 12l/2% of the entire fauna, persist into the Eecent and furnish consistent evidence of environmental conditions during Calvert Times. The depth of the waters in which they live did not, in all probability, exceed 20 or 25 fathoms. The temperature was perceptibly lower than that of any other of the middle or late Tertiary faunas of that region. The bottom was prob- ably soft, dominaritly mud,, with a mixture of sand. At least a portion of the shore must have been sufficiently shel- tered to encourage the growth of kelp and ,sea lettuce and other sea weeds to which many of the smaller univalves and bivalves characteristically attach themselves. The Calvert of Maryland is unusually varied for the latitude. It is quite possible that the ancient shore line in that area was fringed with islands and sand spits similar to those along the outer margin of Virginia and North Carolina today and that dur- ing Calvert times the spits were now washed away, admitting the off-shore fauna, and now built up, protecting the waters behind them and allowing a warmer water element to creep in and establish itself. There is no evidence of any marked change in the ecology in passing from the Calvert to the St. Mary's. All of the recent species represented in the Calvert are present in the St. Mary's but the number is almost tripled. The northern element, however, is slightly less prominent and the southern element a little more so. The fauna is prolific in individuals but not greatly diversified. The outstanding differences be- tween the St. Mary's molluscs of Maryland and those of Virginia and North Carolina are mainly those of latitude, although the presence in Maryland of a considerable number of Surculas, one of the characteristically deep-water pleuroto- mids suggests deeper water in that area. The faunas in Vir- ginia and North Carolina are remarkably uniform. There are, to be sure, a few species common to Maryland and northern Virginia which are not found in North Carolina and a few of southern affinities which are restricted to North Carolina. The monotony of the assemblage indicates a long 42 Environment of Tertiary Marine Faunas [240 stretch of open shore with only an occasional bight from the vicinity of the present York River in Virginia to that of the Neuse River in North Carolina. The slope of the conti- nental shelf must have been very gentle, not more than 3' to the mile, since there is no perceptible change in the bathy- metric character of the fauna between the extreme eastern and western outcrops, a distance of 60 or 70 miles. There is no reason to believe that any part of this platform was submerged to a depth of more than 30 or 40 fathoms. The bottom was doubtless soft and, for the most part, muddy since the mud-burrowers, notably Mulinia are exceedingly prolific and widely distributed. The waters must have been sufficiently clear, however, and the bottom sufficiently shelly to furnish clutch for the numerous oyster spat and to permit them to mature. Conditions were probably not very favor- able to algal growth, since most of the groups which charac- teristically attach themselves to the sea-weeds of various kinds have a meagre representation. The elevation along the Hatteras axis at the close of the St. Mary's was apparently great enough to cut off the York- town basin in Virginia from the Duplin in southern North Carolina. The faunas of the two basins, though similar in general character, differ more in detail than one would expect in two shallow water faunas only a couple of hundred miles apart. The contemporaneity of the Yorktown and Duplin faunas was suggested by 'Dr. Dall more than fifteen years ago and even at that time he brought forward in explanantion of the conspicuous faunal differences the potency of the ocean currents, a factor which has been so emphasized of late in the distributional studies upon the West Coast. The Yorktown fauna is strikingly like that listed by Bartsch and Henderson from Chincoteague Bay, Accomac County, Virginia. The greatest break in the East Coast life from the late Tertiary on to the Recent comes at Hatteras, the point at which the Gulf stream leaves the inshore and swings out toward the open sea. Many of the sub-tropical species are able to follow along the shore as far as it is protected by the warm current, which 241] /. A. Gardner 43 also serves as an effective barrier to most of the northern forms. At Chincoteague, Bartsch and Henderson found that while along the ocean side of Chincoteague Island the fauna was consistently northern in its affinities, in the pro- tected inner bight there was an overlap of the southern faunas. Twenty-eight of the 70 species which they have listed are present in the Yorktown fauna and the number of com- mon forms will doubtless be greatly increased with further investigations. The ensemble of the fossil and Eecent faunas is conspicuously similar although the southern element is a little stronger in the former. However, much the same conditions of sandy shores and muddy bogs more or less choked with algal growth obtained in the Yorktown as along the Virginia coast today. The fauna, like those that precede and those that follow it, is characteristically shallow water and it is doubtful if any of the indigenous species lived at a depth of more than 25 fathoms. The Duplin fauna is less homogeneous. Mingled with the large pleurotomid element and a considerable number of volutoids, one of the most uniformly deep-water families, are nine species of Ilyanassa, a group that is known to occur only along inter-tidal beaches. The sediments of the Duplin are, for the most part, coarse sands. It seems on the whole reasonable to suppose that the native Duplin fauna lived near the mouth of some rather large estuary and that the streams entering the bay brought down in considerable numbers the beach-dwellers from farther up shore, while strong currents from the south sweeping along the mouth of the estuary contributed not only a southern element of living forms but also a large number of dead shells referrable to extra-limital species. One hundred and thirteen Duplin species are either identical with the Recent forms or so closely allied that they have been confused in the synonymies. Of this number 97, approximately 85%, occur between Hatteras and Florida. Of the remaining 18 only a single species, the rather uncommon Polynices heros, does not range as far south as Hatteras. Most of the characteristic Florida elements, however, are 4:4 Pelecypods of the Boiuden Fauna [212 absent, so that it seems probable that Duplin temperature conditions are more nearly duplicated between Cape Fear and Charleston, South Carolina, than along any other sec- tion of the Coast. In the succeeding Waccamaw the conditions of the Duplin were some of them intensified but not materially changed. The fauna is, on the whole, more consistent, for both the brackish and the deep water elements are rather less pro- nounced. There were, judging by the abundance of such forms as the. Olivas and Olivellas, extensive sand flats covered by from 2 to 10 fathoms of water, while the wealth of Bittiums and small Cerites and other groups of similar habits demands conditions favorable for extensive algal growth. There is a curious similarity in the general make-up of the Waccamaw and Yorktown faunas, due, doubtless to the similarity in ecology. The Waccamaw waters, however, were decidedly warmer than those of the Yorktown, in fact they were in all probability warmer than at any other period during the middle or late Tertiaries or than those off North Carolina today. The evolution toward the Recent Cape Fear fauna has been marked less by the introduction of a northern element than by the restriction of the more sensitive southern forms to the Floridian province. THE PELECYPODS OF THE BOWDEN FAUNA By WENDELL P. WOODRING 1. INTRODUCTION The marls exposed along the coast between Morant Bay and Port Morant, near Bowden, almost at the southeastern ex- tremity of the island of Jamaica, have long been known to contain a prolific and splendidly preserved molluscan fauna. In 1862 Mr. Lucas Barrett, the Director of the Jamaican Survey, deposited in the British Museum a collection ap- parently from this locality. A year later Mr. Carrick T. 243] W. P. Woodring 45 Moore 2 submitted a brief report on the mollusca. The first systematic account of the fauna was published in 1866 by the late Mr. E. J. Lechmere Guppy,3 who later made several additional contributions.4 Mr. Eobert Etheridge 5 in an ap- pendix to the report of the Jamaican Survey, published in 1869,, discussed the general aspect of the fauna. In 1896 Guppy and Ball6 issued descriptions of a number of new species. The report of Mr. Eobert T. Hill 7 on his recon- naissance of Jamaica contained a brief notice of the mollus- can elements of the fauna. In the Wagner Institute Papers Dr. Dall 8 described many new species and noted the occur- ence of previously described forms; in addition, the last fascicle contained a discussion of the correlation of the fauna and a check-list.9 2. BIOLOGICAL CHARACTER OF THE FAUNA The present study has resulted in the recognition of be- tween 190 and 200 species of pelecypods, of which almost half are new. These are segregated into 64 genera and 40 families. The superspecific groups and the number of species in each group are given in the following list: 2 Moore, C. T., Quart. Jour. Geol. 800., London, vol. 19, pp. 510-513, 1863. 3 Guppy, R. J. L., Quart. Jour. Geol. Soc., London, vol. 22, pp. 281- 295, 1866. 4 Guppy, R. J. L., Geol. Mag., decade v, vol. 4, pp. 496-501, 1867: idem, decade 2, vol. 1, pp. 404-411; 436-446, 1874; idem, vol. 2, pp. 41-42, 1875; Proc. Assoc. Trinidad. 5 Etheridge, R., Reports on the Geology of Jamaica, Part 2, West Indian Survey, Mem. Geol. Survey Great Britain, ap. 5, pp. 319-329, 1869. 6 Guppy, R. J. L., and Dall, W. H., Proc. U. S. Nat. Mus., vol. 19, no. 1110, pp. 303-331, 1896. 7 Hill, R. T., Bull. Mus. Compt. ZooL, Harvard, vol. 34 (geol. ser. 4), pp. 145-152, 1899. 8 Dall, W. H., Trans. Wagner Free Inst. fifci., Philadelphia, vol. 3, pts. 1-6, 1890-1903. 9 Idem., pt. 6, pp. 1580-1588, 1903. 4:6 Pelecypods of the Bowden Fauna [24-i Nucula 2 pp. Leda 7 spp. Yoldia 1 spp. Tindaria 1 spp. Limopsis 2 spp. Area (Area s. s.) 4 spp. Barbatta (Acar) 2 spp. (Calloarca) 4 spp. (new section) 3 spp. (Fossularca) 2 spp. Scapharca (Scapharca s. s.) 9 spp. (Argina) 1 sp. (Cunearca) 1 sp. (Bathyarca) 1 sp. (Anadara) 1 sp. Glycymeris 3 spp. Pinna 1 sp. Atrina 1 sp. Melina 1 sp. Pteria 1 sp. Ostrea 3 spp. Pecten Pecten (Pecten s. s.) 1 sp. (Euvola) 2 spp. Chlamys (Chlamys s. s.) 4 spp. (Aequipecten) 5 spp. Pseudamusium ( Pseudamusium s. s.) 1 sp. Amusium (Amusium s. s.) 1 sp. (Propeamusium) 1 sp. Spondylus 3 spp. Plica tula 1 sp. Lima (Lima s. s.) 1 sp. Lima (Mantellum) 1 sp. Limaea 1 sp. Placuanomia 1 sp Anomia 2 spp. Modiolus (Brachydontes) 1 sp. Dreissena 1 sp. Julia 1 sp. Verticordia (Trigonulina) 1 sp. (Haliris) 1 sp. Poromya 1 sp. Cuspidaria (Cardiomya) 1 sp. (Bowdenia) 1 sp. Crassatellites (Crassatellites s. s.) 2 spp. (Crassinella) 3 spp. Venericardia (Venericardia s. s.) 1 sp. (Pteromeris) 1 sp. Chama 2 spp. Echinochama 1 sp. Codakia (Codakia s. s.) 2 spp. (Jagonia) 3 spp. Myrtaea (Myrtsea s. s.) s spp. (Eulopia) 3 spp. Phacoides (Phacoides s. s.) 1 spp. Here (Here s. s.) 4 spp. (Pleurolucina) 1 sp. (Cavilucina) 1 sp. 245] W. P. Woodring 47 Pseudomiltha 1 sp. Callucina 3 spp. Parvilucina (Parvilucina s. s.) 3 spp. (Bellucina) 2 spp. Divaricella 2 spp. Diplodonta (Diplodonta s. s.) 2 spp. (Pelaniella) 1 sp. (Phlyctiderma) 1 sp. Erycina 2 spp. Anisodonta (Basterotia) 1 sp. Montacuta? 1 sp. Cardium (Cardium s. s.) 1 sp. Trachycardium 4 spp. Fragum (Fragum s. s.) 2 spp. (Trigoniocardia) 3 spp. Laevicardium 1 sp. Protocardia 2 spp. Transennella 2 'spp. Tivela 1 sp. Gafrarium (Gouldia) 1 sp. Pitaria (Hyphantosoma) 1 sp. (Lamelliconcha) 1 sp. Antigona (Ventricola) 1 sp. Cyclinella 1 sp. Chione (Chione s. s.) 3 spp. Chione (Lirophora) 1 sp. Parastarte 1 sp. Cooperella (new section) 1 sp. Tellina Arcopagia (Merisca) 4 spp. (Phyllodina) 2 spp. (Eurytellina) 1 sp. Moerella 2 spp. Angulus (Angulus s. s.) 5 spp. (Scissula) 1 sp. Strigilla 1 sp. Macoma Psammacoma (Psammacoma s. s.) 2 spp. Cymatoica 1 sp. Semele (Semele s. s.) 1 sp. Abra 2 spp. Donax 2 spp. Psammosolen 1 sp. Spisula 1 sp. Ervilia 1 sp. Corbula (Aloidis) 1 sp. (Cuneocorbula) 1 sp. (Bothrocorbula) 1 sp. Gastrochaena 1 sp. Martesia? 1 sp. Xylophaga? 1 sp. Teredina 1 sp. Teredo 1 sp.10 10 An additional form is considered the type of a new genus of doubtful affinities placed provisionally among the Isocardiacea, prob- ably near the Vesicomyacidas. 48 Pelecypods of the Bowden Fauna [246 The Prionodesmacea play an important role in the constitu- tion, being represented by 79 species, or more than 40 per cent of the fauna. The larger part of this number is contri- buted by -the taxodonts, which include 44 species. The most abundant taxodont is the genus Area, which has 28 species distributed among 10 sections. The Scapharcas are the most prolific, both individually and specifically. The section Cunearca, which usually occupies a position of importance in the mid-Tertiary faunas of the Antillean region and its peri- meters, is represented by a single small form and the sub- genus Noetia is entirely absent. Three species of Barbatia are grouped in a new section that bears a relation to Barbatia s. s. similar to the relation between Argina and Scapharca s. s. Another Barbatia of unusual type has been provisionally referred to Fossularca, although it probably represents a new section. A minute Bathyarca is abundant in one of the col- lections, but is rather rare in the other minute collections available. Among the prionodonts the Pectens are subordinate only to the Areas. They contribute 15 species representing seven sections among which are included virtually all the groups of a typical tropical fauna. The Aequipectens are the most abundant and include several species that are widely distri- buted in the Tertiary deposits of the Antillean region. With regard to specific diversification Chlamys s. s. is comparable to Aequipecten, but only one of the species is abundant. The valves of a small delicate Pseudamusium s. s. are numerous and the section Propeamusium is represented by a single valve. The oysters form a puzzling assemblage. In all the collec- tions the number of individuals is small and large forms are notably absent. The small size is probably not without sig- nificance when it is considered that one of the Bowden species reaches an imposing size in the Alum Bluff faunas and especi- ally in the Santo Domingan fauna. A similar relation obtains for an unusually large and ponderous Santo Do- mingan Spondylus. If the current synonymy for Ostrea 247] W. P. Woodring 49 megodon Hanley is accepted, this species furnishes an example of a former distribution on both sides of the Isthmus of Panama and a present restriction to the Pacific side. Another oyster probably is identical with the Eecent mangrove- oyster, 0. folium Linnaeus. Although the species may not be genetically valid, it may be assumed that the Bowden form had the peculiar habits of the oyster that is frequently found in mangrove swamps in the Antillean region. The family Limidae includes, in addition to the common Lima, the rare Limcea. Likewise among the Anomindae is found the uncommon Placunanomia, as well as the ubiquitous Anomia. The brackish-water Dreissena is not frequently encountered among American Tertiary faunas. Of greater interest is the presence of the extremely rare Julia, a genus that at the present time is confined to the Indo-Pacific region and is represented by only a few fossil species — one from the Oligocene of Florida and two from the Miocene of south- western France. A minor element in the fauna is furnished by the Anomalo- desmacea. The five species are confined to the superfamily Poromyacea and include small forms under the families Ver- ticordiidae, Poromyacidae and Cuspidariidae. One of the Cuspidarias is the type of the subgenus Bowdenia Ball. .The relative importance of the Teleodesmacea is dimin- ished by the unusually large number of prionodonts, although naturally the teleodonts include the bulk of the fauna. Among the Astartacea members of the family Astartidae are conspicuously absent, but the Crassatellitidae are represented by five species of Crassatellites, of which the most important and the most abundant belong to the subgenus Crassinella. ISTo Carditas are present, but the genus Venericardia includes a prolific Venericardia s. s. and also a small curious form that has been referred to the subgenus Pteromeris, although it is hardly typical of that group. The superfamily Lucinacea is the most diversified of the larger groups. Although only five genera are included, they are represented by 32 species. The genus Phacoides alone 50 Pelecypods of the Bowden Fauna [248 furnishes half of the species distributed among eight sections. Two phacoidean elements, Lucinisca and Miltha, as well as the genus Lucina, are absent. The Codakias and Myrtaeas are abundant and well-developed. The Divaricellas are indi- vidually numerous, whereas the Diplodontas are, as usual, represented by a small number of individuals. In contrast to the richness of the lucinoids is the meager representation of the Leptonacea. The entire superfamily includes but four species segregated into three genera and as many families. Furthermore, the four species are represented by only six valves, two of which belong to the rarely encoun- tered subgenus Basterotia of the genus Anisodonta. Among the larger groups is the genus Cardium, represented by seven sections and eleven species. The sections are such as are found in any tropical or sub-tropical mid-Tertiary American fauna, but Cerastoderma and Papyridea are not included. The Trigoniocardias, which are peculiar to the mid- American region, are a conspicuous element; indeed, a species of this section is the most abundant bivalve in the fauna. The eight veneroid genera are divided among the sub- families Meretricinae, Venerinae and Geminae. Chione is the most abundant with regard to both the number of species and individuals. Parastarte, represented by a single valve, has heretofore not been reported outside of the Floridian region either recent or fossil. The genus Tivela is not in- cluded in any of the Tertiary faunas of the North American mainland. An interesting form comparable to Cooperella in dentition is placed in a new section of that genus. Only two species of Cooperella are known, a Recent species from the west coast of North America and another from the late Mio- cene of the Atlantic Coast. The genus Tellina includes 15 species, distributed among 6 sections. Angulus has the largest number of species, but the most abundant forms are found under Merisca and Moe- rella. Among the Macomas is a typical Cymatoica. The remaining Teleodesmacea are scattered among several groups. 249] W. P. Woodring 51 A single fragmentary valve of an indeterminable Spisula is the sole representative of the Mactridae. Two of the three species of Corbula, the only non-boring Myacea, are exceed- ingly abundant. The unusually favorable conditions of pre- servation are indicated by the presence of several fragile boring Adesmacea. 3. PHYSICAL CONDITIONS The student of recent marine faunas would consider with undisguised suspicion an attempt to reconstruct environ- mental conditions on the basis of the testimony furnished by a single element in a fauna. Despite the lack of an intensive census of a restricted shallow-water West Indian area, which would be of inestimable value in projecting backward the fac- tors that determined the assemblage of an Antillean Tertiary fauna, the ensemble of Bowden pelecypods is such as to per- mit the offering of certain considerations, some of which are more or less obvious and even trite. Though it is a mere platitude to state that the fauna is tropical, yet this facies is emphasized in a striking manner by the development of certain groups and the absence of others that are prominent in the Tertiary faunas of the south- ern Atlantic Coast. The most prolific genera — Area, Pecten, Phacoides, Cardium, Tellina — are characteristically tropical or are represented only by sections or species that are con- fined to low latitudes or there reach their maximum develop- ment. According to the latest faunal lists only two of the 18 species that persist to the Recent at present range north of Cape Hatteras — the ubiquitous Anomia simplex d'Orbigny and Divaricella quadrisulcata (d'Orbigny). Eight are re- corded from Hatteras southward to the West Indies or Brazil ; seven are confined to the area south of Florida and one species is restricted to the tropical portion of the West Coast. Vir- tually the same proportions obtain for a large number of Recent species that closely resemble Bowden forms. The Areas reach their greatest importance in the warmer 52 Pelecypods of the Bowden Fauna [250 seas. The Bowden species are such as would be expected in tropical waters ; indeed a number of them are encountered in the present West Indian fauna. The genus Pecten is usually a conspicuous element in the Tertiary and Eecent faunas of all latitudes, but the large species that are characteristic of higher latitudes, are absent. Spondylus is confined to tropical or sub-tropic regions in the Eecent seas. In the middle and late Tertiary faunas of the United States the genus is restricted to rare occurrences in the Meridian region. Perhaps the most obvious indication of the temperature of the waters is fur- nished by the superfamily Astartacea, which is represented only by several Crassatellites. Even the warm-water Caloosa- hatchie and Waccamaw faunas include one or two Astartes, but in the Bowden assemblage the genus is entirely absent. An Echinochama, a genus which is preeminently Antillean, is the most ponderous bivalve in the fauna. The entire group of lucinoids is quite partial to tropical waters, although a few species, especially of the genus Divaricella, range into high latitudes. By far the greater number of the Cardiums are of the ornate type that indicates a warm-water habitat. More- over, the smooth or relatively simple forms are identical with, or closely related to, Eecent species that do not occur north of Florida. Although the distribution of the genus Tellina is almost world-wide the group is predominantly tropical. In attempting to determine the depth of the water from a consideration of the bathymetric range of identical or closely related Eecent species a rigid adherence to the evidence fur- nished by dredging records would often lead to absurd con- clusions. An example is furnished by the genus Limopsis, of which two species are present. According to available data the group as a whole is characteristic of deeper water, yet several species occur in Eocene beds of the Gulf Coast that undoubtedly were deposited in very shallow water. The fauna is essentially a shallow- water fauna. All of the Eecent species occur in water of shallow depth and many have been recorded from the intertidal zone, but the range of sev- 251] W. P. Wo'odring 53 eral is extended into considerably deeper water. The presence of apparently deeper water elements, such as Tindaria and Bathyarca, may be the result of the action of currents or other extra-limital factors. Since but a single valve of Tindaria is present it is doubtful whether the form was indigenous. It may be suggested that the depth did not exceed 30 or 40 fathoms and it is highly probable that the bulk of the fauna lived in water that was considerably shallower. The waters were clear and the bottoms free from mud. By far the larger number of the Bowden pelecypods are partial to bottoms of sand or fine gravel; even the burro wers are usually found on sandy bottoms. The absence of Mulinia and related forms that prefer a muddy bottom is not without significance. The meager representation or absence of the Leptonacea and other small forms that usually frequent muddy bottoms in sheltered near-shore positions or are attached to algae indicates an open coast and rather strong current action. Estuaries interrupted the coast line and led back to the streams that supplied the relatively coarse volcanic debris which constituted the bulk of the sediments. From the estuaries valves of Dreissena and the mangrove-oyster were carried down to the coast and mixed with the indigenous beach and off-shore dwellers. 4. EELATIONS TO THE FAUNAS OF THE NORTH AMERICAN MAINLAND The possibility of comparing an Antillean Tertiary fauna with those of the Floridian Peninsula is enhanced by the proximity of the areas and by the succession of tropical or sub-tropical faunas of the mainland. Dall11 has correlated the Bowden horizon with the top of the Alum Bluff formation which includes the Chipola, Oak Grove and Shoal Eiver mem- bers in ascending order. According to Dall 12 the Chipola "Ball, W. H., Trans. Wagner Free Inst. Sci., Philadelphia, vol. 3, pt. 6, pp. 1560, 1582, 1903; Bull. U. 8. Nat. Mus. 90, p. 8, 1916. 12 Dall, W. H., loc. tit., p. 1574, 1903. 54 Pelecypods of the Bowden Fauna [252 fauna indicates distinctly sub-tropical conditions. Berry 13 has shown that the Alum Bluff flora is sub-tropical or very warm temperate and according to Dall 14 the Oak Grove fauna indicates a slight lowering of temperature. Above the Alum Bluff formation is a sharp break that has been seized upon as a convenient location for the division between the Oliocene and Miocene. Though the succeeding Miocene faunas of Florida are imperfectly known they unquestionably indicate a more temperate f acies 15 and occupy a position near the middle of the Miocene series of Virginia and the Carolinas. The Pliocene Caloosahatchie formation of Florida has yielded a rich sub-tropical fauna. The profound hiatus in the Floridian succession is par- tially bridged by the Miocene deposits of Virginia and the Carolinas. According to Berry 16 the Calvert formation, which is the oldest, is middle Miocene, probably Tortonian. The only faunas of this region that present a warm-water f acies are those of the late Miocene Duplin and the succeeding Pliocene Waccamaw, both of which are warm temperate rather than sub-tropical. It is apparent that an attempt to compare a mid-Tertiary Antillean fauna with the faunas of Florida is seriously hampered by the absence of any tropical or sub- tropical Miocene faunas on the mainland. Furthermore, in order to make comparisons with any warm-water Miocene fauna of the Atlantic Coast it is necessary to resort to the geographically distant warm temperate Duplin fauna. In Florida the only post- Alum Bluff marine assemblage that flourished under conditions in any manner comparable to those of an Antillean fauna is the sub-tropical Pliocene Caloosahatchie fauna, which is appreciably younger than the slightly more temperate Waccamaw fauna. "Berry, E. W., U. S. Geol. Survey Prof., Paper 98-E, pp. 43-44, 1916. "Ball, W. H., loo. tit., pp. 1549, 1581, 1588-1589, 1903. 15 Dall, W. H., loc. tit., pp. 1549, 1589, 1594, 1903. "Berry, E. W., U. S. Geol. Survey Prof., Paper 98-F, p. 66, 1916. 253] W. P. Woodring 55 The Chipola marl among the Florida horizons has the largest number of species in common with the Bowden. The actual number is of little significance since the Caloosahatchie has almost the same number. It is significant, however, that the Chipolan elements are completely overshadowed by the closer affinities of a large number of groups with Duplin and even "Waccamaw and Caloosahatchie forms. Difficulties are encountered in interpreting these modern elements in terms of age relations to the faunas of the mainland, since obviously considerations of facies and geographical proximity are involved. A larger number of Bowden species are found in the present West Indian waters than in any of the Florida Tertiary faunas and more forms are common to the Caloosa- hatchie than to the Oak Grove, Duplin or Waccamaw. The taxodonts supply one-third of the total number of Recent species and all of these are found among the Arcidae, hence that family, and especially the sub-family Arcinae, has a modern aspect. Five Eecent Areas are included in the list and several others are very closely allied to Eecent forms. Four of the Areas that persist to the Recent are found in the Chipola fauna and three in the older Tampa, so that the actual number is of little weight. But a modern element is furnished by the introduction of the section Bathyarca. Two of the three oysters are believed to be identical with Recent species. Among the Pectens are several elements that are not encountered among the Oliocene faunas of Florida; these include an Euvola of modern aspect, several Acquipectens that are most closely related to Pliocene or Recent forms and a Propeamusium of decidedly modern type. Among the remain- ing Prionodesmacea the genera Limcea and Placuanomm are unrepresented in the Oligocene of the Florida section. • A Crassinella strongly suggests a Duplin and Waccamaw species. Aside from a Recent Chama, the presence of the genus EcJiinocJiama lends to the Chamidae a modern appear- ance. The Lucinacea as a whole present a modern aspect. In addition to several species that are more closely related to 56 Pelecypods of the Bowden Fauna [254 Duplin or later forms,, this relation is emphasized by the initial appearance of the section Pleurolucina of the genus Phacoides. Although the section Eulopia of Myrtaea has been reported from the Tampa fauna, it reaches its earliest development of any importance in the Bowden fauna and is not present in any of the post-Tampa Florida deposits. A peculiar Hare represents a type that has not been recognized except in the Eecent seas and typical Bellucinas have not been reported from horizons lower than the Duplin. The super- family under discussion includes a Eecent Divaricella and also a Recent Diplodonta, which is unknown from any intervening horizon. Two Eecent Cardiums, a Fragum and a Laevi- cardium, are confined to Miocene and later horizons on the mainland,, and Trachycardium includes a type unknown from beds earlier than Pliocene. The Yeneridae supply a quota of later Tertiary elements. The Bowden Tivela is the only rep- resentative of the genus recorded from American Tertiary deposits and a Recent species of Gafrarium (Gouldia)}ias not been recognized at any other Tertiary horizon. The single Cyclinella is very close to a Recent species and the most abund- ant Chione s. s. is allied to a Duplin form. The genus Paras- tarte, unrecorded from a pre-Miocene horizon, is represented by a species scarcely distinguishable from the Miocene to Recent type of the genus. Among the Tellinacea are to be noted a Recent Strigilla, a Semele that is surprisingly close to a Pliocene and Recent form and the initial appearance of the subgenus Cymatoica of the genus Macoma. Though many of the post-Chipolan elements are found among the characteristically tropical groups, yet the introduc- tion of super-specific groups, some of which are not exclu- sively tropical, can hardly be disregarded. The Bowden pele- cypods are distinctly younger than those of the Alum Bluff faunas, as those faunas are now known. It may be suggested that the Bowden fauna is Burdigalian, that is, Lower Miocene in the sense of most American stratigraphers. 255] F. Reeves 57 ORIGIN OF THE NATURAL BRINES OF OIL FIELDS By FRANK BEEVES The origin of the concentrated brines so universally found in oil-bearing strata and other porous, unmetamorphosed rocks lying at depths below the zone of active circulating ground water has never been definitely established. By some these waters are thought to be of meteoric or surface origin, t. e., they are rain waters which have in passing downward through the strata dissolved out of the rock material the salts which they now hold in solution. Others consider them to be the sea water which has remained in the pores of the strata ever since their deposition. A study of the occurrence and chemical nature of the brines found in the oil sands of southwestern Pennsylvania and West Virginia furnishes data which indicate that the waters in this area are connate or of ocean origin. This conclusion is based on the following lines of evidence : (1) The distribution of the water suggests that it is not of meteoric origin. (2) There are no adequate explanations of how the water of deposition has been removed from the strata. • (3) The association of the dry sands and "red beds5' of the area indicate that the water present accumulated with the sediments as they were being deposited. (4) The chemical nature of the brine points to it being of connate origin. THE DISTRIBUTION" OF WATER In order to consider this phase of the evidence it will be necessary to describe briefly the structural and stratigraphic features of the area under discussion. Structure of the Area. — The brines occur in the Car- boniferous and Devonian strata of the Appalachian coal basin. 58 Natural Brines of Oil Fields [256 This is a shallow geosyncline in which the surface rocks are chiefly of Pennsylvania!! age except where in the center of the basin there are from 800 to 1300 feet of Permian strata over- lying the Pennsylvanian. In this area the Mississippian and Devonian rocks underly the surface at from 1600 to 2000 feet and 2200 to 3000 feet, respectively. Around the rim of the basin these strata outcrop. On the east they reach the surface along the Alleghany front and on the west in central Ohio. The distance across the center of the basin from outcrop to outcrop of the Devonian strata is about 180 miles. The difference in elevation between the highest and lowest point which the same strata attain is about 8000 feet. Thus it is apparent that the geosyncline is to be considered as a very shallow basin. Across the basin and paralleling the Appalachian mountain folds, the strata are folded into a series of minor anticlines and synclines. Towards the east- ern outcrop these flexures become more and more pronounced. In this area the dip along the flank of the folds is about 200 feet to the mile, while in the central part of the basin it seldom exceeds 75 feet to the mile. Westward the folds die out almost entirely and the strata rise to the surface in central Ohio at the rate of about 30 feet to the mile. The strata under consideration therefore are comparatively little folded and consequently little metamorphosed. Faulting is also absent except for a minor fault of a few miles in extent in central West Virginia. Southward across Kentucky and Tennessee the basin narrows and the strata are folded and faulted to a greater degree. On this account and also because of the lack of data on the deep underground waters of the area that part of the basin is not included in this discussion. The northern end of the basin is also not considered here be- cause in that area meteoric water has entered, through old abandoned oil wells, the deeper sands and destroyed more or less the original water content of these strata. Stratigraphy of the Area. — The drill has penetrated strata from the Permian to the Lower Devonian in the search of oil in the central part of the basin. The Permian and Penn- 257] F. Reeves 59 sylvanian are the surface rocks. They comprise a series of from 2000 to 2500 feet of alternating thin-bedded shales, sandstones, limestones, clays, and coals. The Mississippian underlies the Pennsylvanian unconformably. It is made up of about 800 feet of sandstone, shales, and limestones which vary in thickness from 100 to 250 feet. The Catskill forma- tion is a non-marine facies of the Upper Devonian. It con- sists of from 500 to 800 feet of thin-bedded sandstones and red and dark-colored shales. Below the Catskill occur about 300,0 feet of compact shales. Underlying these are the Lower Devonian limestones. Occurrence of the Water. — Water is found in the sandstone and limestone members of the above stratigraphic series. In these it occurs in porous layers in which also occur oil and gas. Usually there is a structural arrangement of these materials. Generally the water occupies the synclines, the gas the anticlines, and the oil intermediate structural posi- tions. This distribution is modified by the amount of water in the sands. Where they are saturated the oil occupies the anticlinal areas. In the Appalachian oil fields, however, the most common condition encountered is where there is but sufficient water to fill up the synclines. The oil, under such conditions, occupies a belt structurally higher and the gas fills up the anticlinal areas. In sands that contain no water the oil is found in the synclines. The amount of water in a sand is usually thought to be a function of its depth. In general it may be stated that the Pennsylvanian sands are saturated, the Mississippian sands semisaturated, and the Catskill sands dry. This would appear to support the idea that the water present is meteoric in ori- gin. On examining the facts, however, this assumption does not appear to be justified for the water does not disappear with depth. Two deep wells which have penetrated the Lower Devonian strata have revealed the fact that below the dry Catskill sands are prolific water-bearing strata at depths from 5000 to 6000 feet. This occurrence, as well as the universal appearance of water at all depths in other oil fields, indicates 60 Natural Brines of Oil Fields [258 that depth is not a factor in the disappearance of water. This being so, then the usual argument that the water present has originated from descending waters loses weight. The presence of nonwater bearing sands such as the Catskill occurring between saturated strata goes to show that there has been no downward movement of meteoric water across the strata. Thus, on the assumption that the water present has a surface origin it must, then, have reached its present position by entering the strata at their outcrop. This undoubtedly explains the source of the waters occurring in strata of Pennsylvanian age, for these are saturated up to their outcrops with water obvi- ously of surface origin. But the saline waters in the Missis- sippi sandstones are not present towards their eastern outcrop. The synclines along the eastern flank of the basin contain no water and these would have to be filled before water could reach areas in the sands west of them. It is impossible that the water could have come from the westward for some of the sands, i. e., the Maxton and Hundred-foot sands, are not continuous to the western outcrop of the formations. Thus with these facts opposing the possibility of a vertical or lateral movement of the water it must be considered to be of connate origin. METHOD OF KEMOVAL OF WATER OF DEPOSITION The processes usually suggested by which the sediments have been depleted of the water deposited with them are hydration, consolidation of the sediments, expansion and evaporation of the water due to heat, and drainage resulting from elevation. A brief consideration of these hypotheses is .sufficient to prove their ineffectiveness. Hydration cannot have been a factor in the removal of the water for the few minerals of sedimentary strata capable of combining with water would more likely be hydrated while they were accumulating as water-borne sediments than while ihey were subjected to the heat and pressure incident to their condition of deeply buried strata. 259] F. Reeves 61 Consolidation of sediments though effective in lowering the amount of pore space does not remove the water from the porous area that remains after consolidation, so such an action tends to increase rather than decrease the per cent, of saturation of the total porosity of the rocks. The influence of heat resulting from the expansion and contraction following periods of burial and exposure due to erosion can be no effective agent in removing the water since water increases only 4 per cent in volume when it is raised from a temperature of 4 degrees Centigrade to 100 degrees Centigrade and this represents a much greater change in temperature of rock strata than ever occurs in a geologic cycle. Drainage in an area of the nature of the Alleghany coal basin is not possible since the basin is~ so shaped that the water cannot drain out of it. Moreover,, these sands are below sea level and hence not subject to drainage. Thus with no adequate explanation of how sea water has been removed from rock strata it is more logical to consider the water present to be of connate rather than of meteoric origin. THE ASSOCIATION OF THE DRY SANDS AND " EED BEDS " A study of the non-water-bearing strata of the Appalachian oil fields has furnished data which is to be interpreted as furnishing positive evidence that the waters present in these sands are connate in origin. As mentioned above the Cats- kill and certain areas of the Mississippian sands contain no water. This absence of water is not due to structural or porosity conditions but it is characteristic of sands which are associated with red shales. Several lines of evidence indicate that these dry areas and the " red beds " were developed when the sediments were exposed as flood-plain deposits to the ac- tion of air which oxidized the ferrous minerals present and at the same time dried out the sediments, in which condition they have remained to the present time. The acceptance of this conclusion, the arguments in support of which are given 62 Natural Brines of Oil Fields [260 elsewhere/ results in attributing a connate origin to the water present. THE CHEMICAL NATURE OF THE WATER The following is a mean analysis of 8 brines collected from strata of Mississippian age expressed in parts per million parts of water : Si02 137 Na 41585 Fe 26 K 307 Ca 12740 Br 44 Mg 2295 Hc03 19 Total 153000 parts of S04 1530 dissolved matter per million Cl 95043 parts of water The outstanding feature of the water is its high chlorine content. This makes up about the entire acidity of the water and comprises 61.12 per cent, of the total dissolved matter present. The other negative ions present, HC03 and S04, occur in small amounts, making up but about one per cent, of the salts present. Sodium is by far the most abundant of the basic ions and comprises 21.18 per cent, of the material in solution. Calcium is about one-third as abundant as so- dium and consists of 8.33 per cent of the total salts present. The other basic ions occur in unimportant amounts. In addition to their peculiar chemical nature the brines are also to be distinguished by their concentration, which is from three to seven times as great as ocean water. Another interesting feature is the similarity in content of the consti- tuents carried by the waters. Reference to the analysis on page will show that the various salts are always present in about the same relative amounts. This is the more strik- ing when it is considered that the waters were collected at points over an area of 10,000 square miles and from horizons of different geologic age occurring at depths of from 1000 to beeves, Frank: A Discussion of the Absence of Water in Certain Petroleum-bearing Strata of the Appalachian Oil Fields. Disserta- tion in Johns Hopkins University Library. 261] F. Reeves 63 p E / * / ,/ 1 / \ / f O (i d * a i i DLJ 4 i f I, f i j i ? i • \ Graphs showing the amount of each ion (expressed in per cent, of the total salts present) in the brines collected from Pennsylvania, Mississippi, Catskill and Lower Devonian strata and in the average Ocean (O) and river waters (R) of the world. 64 Natural Brines of Oil Fields [262 6000 feet. A slight change in chemical nature with depth is noticed which may be a function of stratigraphic horizon. This possibility will be discussed later. Obviously the brines are not like ocean waters yet they are more unlike surface waters and since they are to be con- sidered as originating from one of these sources it is logical to attribute them to that one which they more nearly resem- ble as this requires the explanation of fewer anomalies in the transition from the one water to the other. This comparison is shown graphically on page 63. On the horizontal lines P, M, C, D, are plotted, in percentage of total salts present, the amount of each ion in the four mean analyses of the brines from the sands of the Pennsylvanian, Mississippian, Catskill, and Lower Devonian, respectively. On line S is plotted also the percentages of the salts in the mean analyses of ocean and surface waters. With a line drawn through these points a clear idea is obtained of the similarity between the brines and their two possible sources. On the assumption that the brines are surface waters which owe their present chemical nature to changes which they have undergone as they passed downward through the rock ma- terial, it would be expected that with increase of depth there would be a progressive change at least to a point of satura- tion. Thus, for example, since there is a decrease say of calcium in the first 1300 feet of from 20.39 per cent, to 8-33 per cent., at greater depth it would be expected that the deeper brines would continue to show a decrease in the amount of this ion present. Reference to the graph shows that instead there is a decided increase of calcium with in- crease in depth below 1300 feet. Sodium shows the same anomalous change with depth. It is apparent, on the other hand, that, with the exception of magnesium, the ocean waters fall more in the general alignment of the graphs than do the surface waters. Of course it may be argued that the water would undergo a greater chemical change in the surface strata or in the zone of oxidation than at subsequent depths, 263] F. Reeves ater water Lower g Catskill Dev. 0 £ Mississippi 65 Pennsylvanian oscn O o i— ' co >£>. co SB o H*1 ^ to to to _OH- 'H-H- 'H-itotototo & o > , ^_* (& Ht» -~— •- ^-— ' -^^ ^_* Ht) -• — - — >. • ^^ *. ^ Cj? Ht» ^_^ ^_^ ^_^ ^_^ ^^ ^_^ ^^ N ., £3 *-fc 3 a, Co S^ ' s _ - o o . — I tO O CO ^C CO O O O h- rfi> tO tO CO 22S § -d to -^ co o H-I o o co to HI co co H-I H-I O C O O "h-lCOH-ICO OS CO OS 00 CO O CO Zi S' so OJ , ? i 3 t> en w. en en 80 bs to ig o°co *Q ST1 SQ" tij to H- tO H-I OS OSOSOSOS OS OSO5OSOSOS OS OSOSOSOSOSOSOSOS OS Q5 H-I p H-I tO tO H-l H-I H-I tO H-J H-l tO pi H-i H-> p tO tO tO I-* P O to co co bs co co *>• to i—i co H-I H-I co co to '\—> to b co co OH-I tOH-ios^-to ^lootoco eo tO tO tO tO H-I tO tO tO tO H-I' H-IJH-COOO p en-• ^l p to p H-I en bs co to co bs bo co co H-" bo cc " h- O H- CO O 66 Natural Brines of Oil Fields [264 but it is hardly possible that the reactions between the waters and the rock material would produce a brine so simi- lar to ocean water, since the alteration of river water to ocean water is effected largely by the action of organisms in the sea. The perpendicularity of the lines uniting the brines and the ocean water indicate the similarity between the chemical nature of the present ocean and that of former seas. The variations from the straight line,, with the exception of the graphs for calcium and sodium, is readily explained by the possible reactions which may have occurred between the rock material and the chemicals in solution. It will be noted, however, that there is a definite increase of calcium and decrease in sodium in passing downward from the Pennsyl- vanian to the Lower Devonian brines. If the waters are assumed to be of meteoric origin, then there is the anomaly of the more soluble sodium being replaced by the less soluble calcium ion as the water penetrates the earth's crust. If, however, the waters are considered to be of connate origin, then this change in depth may be a function of the geological age of the strata and hence indicate changes in the chemical content of the ocean during different geologic periods. If the salts of the ocean have been added by the rivers, then, of course, the salinity of the ocean would increa.se with the age of the earth. This increase in salinity will not affect the chemical nature of the water if the relative amount of each salt remains constant. However, since organic matter and changes in physical conditions remove some salts from solu- tion, the waters will change in nature as well as concentration. Thus the relative increase in sodium may be explained by the fact that but little of the enormous amount of this salt added continually by the rivers is lost by the ocean, while most of the other salts are constantly being removed from solution. The variation in amount of calcium in the differ- ent brines is likely due to fluctuation in the amount of C02 in the air during different geologic periods. This fluctuation 265] F. Reeves 67 is a generally accepted fact 2 and since there is an equilibrium between the amount of the C02 in the air and in the sea water it is to be expected that the sea water at times, when there is the greatest amount of C02 in the atmosphere, will hold a comparatively larger amount of calcium in solution as the solubility of calcium carbonate is a direct function of the amount of C02 present.3 The increase in concentration with depth is also an argu- ment that the brines are indigenous to the rocks in which they occur. Kichardson,4 however, suggests that this is due to the fact that there has been an upward diffusion of salts from the rock salt deposits that are known to underly the strata in question. But it has been pointed out that there are dry sands in the area under consideration intervening between these lower water-bearing horizons and the Mississippian sands and, as diffusion through dry strata is impossible, this suggestion seems untenable. Again, on this assumption it is difficult to understand why sodium would increase in relative amount with increase in distance from the salt beds. Refer- ence to the analysis given will show that the Pennsylvanian brines are richer in this salt than the Lower Devonian. It appears more likely that the concentration is due to a capil- lary migration of the water away from the sands in which it occurs and in which it accumulated. This migration would remove part of the water and most likely leave the salt behind, because fine-grained sediments have an absorptive effect on solutions passing through them 5 which would result in a 2 Chamberlain, T. C. " The influence of great epochs of limestone formation on the constitution of the atmosphere." Jour. Geol., vol. vi, 1898, pp. 609-621. 3 Johnson, John and Williamson, E. D. "The rOle of inorganic agencies in the deposition of calcium carbonate." Jour. Geol., vol. xxix, No. 8, 1916, pp. 729-750. 4 Richardson, G. B. " Note on the diffusion of sodium chloride in Appalachian oil field waters." Jour. Wash. Acad. Sci., vol. vii, no. 3, 1917, pp. 73-75. 6 Turrentine, J. W. " The occurrence of potassium salts in the saline of the United States." Bureau of Soils, Bull. 94, 1913. 68 Upper Cretdceous Seacoast in Montana [266 greater concentration of the waters. Considering this fact, then, the waters should show a decrease rather than an in- crease of mineral matter with depth, if the waters were of surface origin. Which adds more evidence to the above, that these natural brines are of connate origin. AN UPPER CRETACEOUS SEACOAST IN MONTANA1 By W. T. THOM, JR. INTRODUCTION The nonpersistence of lithologic units is the greatest ob- stacle to systematic stratigraphic work and correlation. Par- ticularly is this true when it is necessary to determine the relationship of marine and continental deposits. That this difficulty may be largely overcome by a correct " paleophysio- graphic " perspective of the period dealt with is now becoming the accepted doctrine ; and the writer has undertaken this sketch both as an illustration of the way in which ancient physiographic conditions may be deciphered and as a demon- stration of the close genetic relationship of some of the differ- ent lithologic phases of the Judith River formation. The writer studied this formation in the area lying north of the Yellowstone Biver, near Billings, Mont., while serving as an assistant to Eugene T. Hancock, of the United States Geological Survey, and it is with his very kind consent that this article is published. The accompanying sketch map will serve as a geographic guide to the reader in following the discussion. STRATIGRAPHIC GEOLOGY The later Upper Cretaceous sediments of the Montana group are exposed throughout this region, forming about 90 1 Published by permission of the Director of the United States Geological Survey. 167] W. T. Thorn 69 per cent, of the rocks outcropping within its limits. Five formational subdivisions of the Montana are present, which are, in ascending order, the Eagle sandstone, Claggett shale, Judith Eiver formation, Bearpaw shale, and Lennep sand- stone. The relationship of these units is perhaps best brought out by the accompanying diagram, but a brief description of each formtion may make the details clearer. The Eagle sandstone is developed in this area as the marine sandstone apex of a wedge of continental sediments built sea- ward at the close of Colorado time. The strand origin of the FIG. 1. Sketch map of an area in Montana. upper part of the formation seems clearly indicated by abun- dant impressions of the fossil seaweed Halymenites major, by the coarse grain of the sandstone, and by the almost universal distribution of small, flattened chert pebbles at or near the top of the formation. The Judith Eiver formation is much like the Eagle in mode of origin, but is thicker and shows the development of fresh-water and estuarine phases much farther east than the Eagle. The Lennep sandstone constitutes a third continental wedge, chiefly notable for the predominance of volcanic material among its component sediments. Separ- ating these strand and coastal-plain deposits are the marine shales of the Claggett and Bearpaw formations, which thin westward and disappear at an indeterminate distance beyond the border of the area under consideration. 70 Upper Cretaceous Seacoast in Montana [268 From this and from other evidence at hand it seems rea- sonably certain that the strand lines of the late Mesozoic lay nearly parallel to the present Eocky Mountain front, and that it migrated to and fro in response to the recurrent depressive movements of the period, which were gradually counterbalanced by intervening intervals of sedimentation. DISTRIBUTION AND CHARACTER OF THE JUDITH RIVER SEDI- MENTS The sediments of the Judith River formation are now well exposed north of Billings by reason of the extensive denuda- tion which the region has undergone; hence it is possible to form an accurate estimate of the variations of the section, both vertically and horizontally. Near Gibson, 50 miles northwest of Billings, the formation is of freshwater origin and consists of sandy .shale and nu- merous beds of hard, rather muddy, quartizitic sandstone containing reed stems and fragments of coniferous material. Both shale and sandstone are characteristically of a yellowish or tawny hue, though some of the beds, particularly among the softer units, indicate by their texture and peculiar green- ish brown color the presence of the tuffaceous material which was thrown out so copiously by the volcanoes of the Crazy Mountains during later Upper Cretaceous and early Eocene time. Farther east the volcanic material disappears and the sec- tion gradually assumes the aspect typical of the formation as it is developed along the Missouri and in the vicinity of Havre on Milk River; an alternation of light-gray to white clay shales with lignitic shale and carbonaceous sandstone giving the exposures a peculiar striped appearance. With the advent of the carbonaceous zones in the section the brackish water shell Ostrea subtrigonalis also makes its ap- pearance at different horizons in the area immediately north of Broadview, though it is conspicuously present in but a single very thick shell bed of extraordinary uniformity and 269] W. T. Thorn 71 persistence of development which lies about 50 feet above the basal sandstone of the formation. At Acton, 15 miles northwest of Billings, the formation still retains its characteristic appearance and lithology, but east of the town a very rapid lateral variation of the sedi- ments soon Incomes apparent. The thin carbonaceous sand- stones of the western section rapidly increase in magnitude and in coarseness of grain to the eastward, individual mem- bers increasing from 6 or 8 to as much as 50 feet in thickness within a distance of 5 miles. As a result the formation, as exposed about 6 miles east of Acton, consists of four massive sandstones separated by intervals of shaly sandstone or sandy shale. Abundant impressions of Halymenites major occur throughout these sandstones and indicate the inception of strand conditions, a conclusion confirmed by the discovery of marine fossils a few miles farther east, near Huntley. From where the maximum of sandstone development is attained, 6 miles east of Acton, the lower sandstone members begin to taper eastward, merging into shale lithologically indistinguishable from that of the underlying Claggett. It is further to be noted that certain surprising features of the Judith River-Bearpaw contact appear southeast of Gibson. Long, narrow ridges capped by hard andesitic sand- stone are developed for considerable distances, especially near Big Lake in the so-called Lake Basin region, their general extension being from east to west or from southeast to north- west. The cap sandstones of these ridges were probably never continuous over the intervening areas, but they lie at practically an identical horizon and are so similar lithologi- cally that they are certainly the products of the same agency. Below these upper dark sandstones the section shows great variability; at some localities sandy beds containing more or less carbonaceous shale and lignite occupy the whole of the interval down to the more typical Judith River sedi- ments, while elsewhere, even in the same ridge, the cap sandstone may overlie typical Bearpaw shale with only a 72 Upper Cretaceous Seacoast in Montana [270 few feet of thin-bedded white sandstone intervening. From this and from other corroborative evidence the conclusion therefore seems natural that the ridges represent the radiat- ing channel sandstones of an ancient delta, which was built out into the marine waters of the incipient Bearpaw sea by a river of considerable size flowing from the south or south- west. MONTANA Colorado FIG. 2. Section showing the relation of the sediments. PALEO GEOGRAPHIC INTERPRETATION Upon the basis of the foregoing data the writer draws the following picture of the physiography of Judith River time. The low coastal plain of the mainland lay to the east of the Crazy Mountains, very possibly extending thence a little beyond Gibson, and upon it a gradually increasing thickness of freshwater and subaerial deposits was laid down as regional subsidence progressed. Simultaneously a sandy barrier beach was developed and maintained in the area east of Acton, thus partially shutting off a shallow embayment or lagoon whose quiet waters, rendered brackish by the regional drainage, 271] W. T. Thorn 73 afforded a favorable habitat for multitudes of oysters,, which were later buried by the burden of fine silt accumulating in the quiet waters of the bay. Still later, as sedimentation gained upon subsidence, the site of the one-time oyste'r bed became the location of repeated coastal swamps, in which were formed the lignitic beds and carbonaceous zones now so conspicuous in the upper part of the formation. Subsequent revival of the local depressive movements of the crust temporarily restored the old embayment, into the southern end of which a large river built out a delta, even as the marine shales of the Bearpaw were being laid down in the deeper part of the bay a little farther north. As a final phase more rapid depression carried marine waters farther westward, and the Bearpaw sea covered the whole area. A REMARKABLE UPPER CRETACEOUS FAUNA FROM TENNESSEE * By BRUCE WADE During the summer of 1915 the Tennessee Geological Survey located well-preserved fossils in the Eipley formation in the northeastern part of MclSTairy County, Tennsesee. An incomplete collection was made from the locality in this region where the strata containing the fossils are best ex- posed. This collection was studied during the winter in the Geological Laboratory of the Johns Hopkins University. This partial study of the fauna has resulted in the differen- tiation of nearly 300 species, and investigations show that no single locality yet reported from the Cretaceous of North America has furnished such a large fauna made up of such well-preserved shells. The Gastropods are unusually abundant and include not only 1 Published with the permission of Dr. A. H. Purdue, State Geolo- gist of Tennessee. 74 Upper Cretaceous Fauna from Tennessee [272 a large number of new species but several forms which are regarded as new genera.2 Further collections were made at this place and in the adjoining regions during the field- season of 1916. The writer has begun, for the Tennessee Geological Survey, an investigation of all the Upper Cre- taceous deposits of the state and hopes to submit in the near future a detailed report on the Stratigraphy and Systematic Paleontology of these rocks. GENERAL GEOLOGICAL EELATIONS. The Upper Cretaceous deposits of Tennessee outcrop in a wedge-shaped area which crosses the State in a nearly north and south direction, and lies largely west of the Ten- nessee Eiver in the west-central part of the State. This area is about 67 miles wide along the southern boundary of the State, narrowing to the northward until at the Kentucky line it is only about 15 miles in width.3 Along the southern border of the State in Wayne, Hardin, McNairy and Harde- man counties these deposits may be segregated into the fol- lowing lithologic units. f Owl Creek horizon -r>. , „ ! McNairy sand member Eipley formation J _ . . ] Ferruginous clay horizon [Coon Creek horizon Selma chalk Eutaw formation Tuscaloosa formation The present discussion is limited to the lower part of the Eipley. This formation covers the western two-thirds of McNairy County, and in general is well exposed over that entire region. The four horizons or members of the Eipley 2 Some of these have been described in the Proc. Phila. Acad. Nat. 8ci., 1916, pp. 455-471, pi. 23, 24. 3 Jenkins, 0. P., Geological Map of Tennessee, State Geol. Survey, 1915. 273] B. Wade 75 named above may be traced across the county by their con- tained faunas and lithology, even though there are no sharp lines of demarcation separating the one from the other. The sediments of the Coon Creek horizon, which are described in detail below, are quite variable, ranging from local lenses of impure limestone through very fossiliferous marls to glau- conitic sands and gypsiferous clays poor in fossils. The overlying ferruginous clay horizon is sparsely fossiliferous and extends across the county in a belt about three miles wide. The McNairy sand member next above, which is typically exposed in McNairy county, has been described by Stephenson.4 This member is essentially a sand and has yielded few fossils. The so-called fucoid Halymenites major Lesq. occurs sparingly. In 1915 leaves were collected from near Selmer, Tenn., and Big Cut, Tenn., the type section for this number. These have been submitted to Prof. E. W. Berry for study. Above the McNairy sand and exposed along southwestern McNairy County and southeastern Har- din County is a fossiliferous horizon which may be traced southward into Mississippi to Owl Creek, the type locality for the Ripley formation. COON CREEK LOCALITY AND ITS STRATIGRAPHIC POSITION. The locality under immediate discussion may be known as the Dave Weeks place on Coon Creek. It is in the north- eastern part of McNairy County, 3% miles south of Enville and 7% miles north of Adams ville and % mile east of the main Henderson-Adamsville Road. The beds containing the fossils are best exposed in the valley about two hundred and fifty yards east of Dave Weeks' house along the headwaters of Coon Creek. This is a small .stream flowing northward into White Oak Creek, a tributary of the Tennessee River. Upper Cretaceous fossils have been previously collected 4 Stephenson, L. W., U. S. Geological Survey, 1914, Prof. Paper 81, p. 22. 76 Upper Cretaceous Fauna from Tennessee [274 from a few places in this general region of the Mississippi radius of 7% miles of the Dave Weeks' place. At a point % mile west of Adamsville Stephenson 5 made a collection from the Exogyra ponderosa zone of the Selma chalk. Four miles northeast of Adamsville he made another collection from the same zone at a locality referred to as " four miles southwest of Coffee Landing." Three miles west of Adamsville fossils were found by Stephenson in the Exogyra costata zone of the Selma chalk. About eleven miles southwest of the Coon Creek locality and "two and one-half miles east of Purdy," Safford6 collected in the uppermost part of the formation which he designated as the " Green Sand." The horizon from which this last collection was made probably has the same stratigraphic position as the Coon Creek beds. The best-known fossil locality in this general region is the classic Owl Creek locality 7 in Tippah County, Mississippi, often re- ferred to in the literature as Ripley, Mississippi. (See sketch map, Fig. 1.) A sketch map has been inserted on page 77 to show the location of Coon Creek and Owl Creek and to give in a general way the areal geology in the region about these localities. Big Cut and Coffee Landing, two other impor- tant localities in the Cretaceous Geology of Tennessee are shown on this map. The information given on the map south of the Tennessee-Mississippi line has been furnished by Dr. L. W. Stephenson of the United States Geological Survey. The Coon Creek horizon is stratigraphically near the base of the Eipley formation and in the Exogyra costata zone. «Idem., p. 24. 6 Safford, J. M., Geology of Tennessee, 1869, p. 416. 7 Conrad, T. A., Jowr. Acad. Nat. 8ci., Philadelphia, 1858, vol. iii, 2d ser., pp. 323-336. Conrad, T. A., Jour. Acad. Nat. Sci., Philadelphia, 1860, vol. iv, 2d ser., pp. 275-298. Stephenson, L. W., U. S. Geological Survey, 1914, Prof. Paper 81, p. 24, table 2. B. Wade 7'7 The Selma-Eipley contact is well established at Blue Cut on the Mobile and Ohio Eailroad at the state line on the south- ern extremity of McNairy County. From this point the con- FIG. 1. Map showing the areal geology of a portion of Tennessee and Mississippi. 1, McNairy Sand Member; 2, Ripley Formation; 3, Selma Chalk; 4, Eutaw; 5, Tuscaloosa. tact may be readily traced both by lithological and faunal relations to Coon Creek. It is thus evident that the Coon Creek horizon lies just above the Selma chalk and at the base of the Eipley. The Coon Creek horizon is thus strati- 78 Upper Cretaceous Fauna from Tennessee [276 graphically lower than that along Owl Creek in Mississippi. At the latter locality the fossiliferous horizon is in the upper- most beds of the Eipley and is directly overlain by Eocene limestone.8 Below the Owl Creek beds is the southern equiv- alent of the McNairy sand member of the Ripley formation.9 The McNairy sand member,, together with about 100 feet of sparsely fossiliferous, ferruginous Ripley clay, lie strati- graphically higher than the Coon Creek horizon and are exposed to the west of it. (For cartographic relations of the two localities see sketch map in Fig. 1.) Thus, it is quite evident that the Coon Creek fauna is older than the Owl Creek fauna. DESCRIPTION OF THE LOCALITY AND CHARACTER OF THE SEDIMENT A thickness of more than thirty feet of the fossil-bearing beds is exposed along the banks of Coon Creek. For one-third mile this stream flows in a narrow V-shaped channel from six to fifteen feet deep which has been cut out during the last twenty years. The stream has a steep gradient and its channel is deepened by every heavy rain. The channel fills quickly after a thundershower and its sides are kept freshly scoured by the rushing water. White shells of Crassatellites, Cucullaea, Cyprimeria, Gryphaea, Ostrea, Drilluta, Lunatia, Baculites, etc., project out of the dark greyish blue matrix and glitter in the clear water and the sunshine. In general aspect the exposure bears a striking resemblance to certain Tertiary beds. In broad physiographic relations, character of the matrix and whiteness of the shells, the Coon Creek locality resembles the well-known Upper Cretaceous exposure of Brightseat, Maryland. This locality is two miles east of Dis- trict Line and has yielded the most prolific Upper Cretaceous 8 Harris, G. D., The Midway Stage, Bull. Amer. Pal, vol. 4, no. 4, 1896, p. 24. 'Stephenson, L. W., Paper given before the Paleontological So- ciety of America, December 29, 1916. Lowe, E. N., Geology of Mississippi, Bull 12, 1915, p. 62. 277] B. Wade 79 fauna of Maryland. The shells, notably the bivalves, are probably more abundant at Brightseat, but not so well pre- served as they are at the Tennessee locality. The sediments containing the Coon Creek fauna are dark bluish green and gray clayey sands. The sand is of medium fineness and con- sists of angular and rounded grains of quartz as the major con- stituent, with glauconite, small flakes of mica, and shell frag- ments as minor constituents. Pieces of lignitic wood and small nodular masses of pyrite are common but not abundant. All of the above elastics are cemented together with a fine calcareous material, forming a compact impervious mass which varies locally in arenaceous and argillaceous content. There is locally sufficient lime for the matrix to become indu- rated into a very hard, impure and concretionary limestone. When this marl is thoroughly weathered the shells are re- moved leaving casts in a matrix which becomes yellowish brown in color, due to the oxidation of the glauconite and other ferruginous constituents. Dr. Paul C. Bowers, Chief Chemist for the Tennessee Geological Survey, has made a careful quantitative analysis of this marl and reports the following results : Si02 65.30 AloO3 8.56 Fe203 3.72 FeO 1.72 MnO 44 CaO 7.10 : 70 ff\ 2-42 P 0 trace 2 5 FeS2 45 C02 5.15 H20 5.45 Carbon 09 Total . ..101.00 80 Upper Cretaceous Fauna from Tennessee [278 HISTORICAL SKETCH The well-preserved fossils of the Eipley formation attracted the attention of the early geologists and impressed them very much. In 1856 Conrad X1 described fifty-six new species from Owl Creek and made the following observations about the fauna : "The Cretaceous strata of Mississippi have long been ob- served and partially noted by geologists and the lamented Professor Tuomey has described a number of their fossil con- tents. I now introduce quite a distinct group of shells, which are imbedded in a different matrix compared with the preva- lent cretaceous marls, green sands and limestones. The dis- covery of these beautiful organic remains is due to the inde- fatigable exertions of Dr. W. Spillman, of Columbus, who has forwarded a collection of specimens more or less perfect, consisting of nearly sixty species, all of which appear to be unpublished except Scaphites conradi. The appearance of these shells is like that of eocene species which have merely lost their animal matter, and in this respect are very unlike the condition of similar genera in the contiguous rocks of the same era. The fossils are imbedded in a sandy marl of a dark gray color, the principal constituents of which are fine scales of mica and grains of quartz mixed with fragments of small shells; and though some of the very thin species are distorted, the stronger retain their original shapes and are generally very perfect. Species of Crassatetta, Nucula and Meretrix have the valves united as in life, as well as a few of the extremely thin Inocerami, though the latter are more or less distorted by pressure. The numerical proportion of species of Cephala and Acephala is nearly equal. The external sculpture of all is as sharply defined as in existing species. Besides Scaphites and Baculites, there is only one shell in the collection which resembles a species of the green sand or lime- 11 Conrad, T. A., Jour. Acad. Nat. 8oi., Philadelphia, 2d ser., vol. iii, pp. 323-336. 279] B. Wade 81 stone, and it is quite distinct. The rare genus Pulvinites is herein for the first time introduced as an American form. The analogous species, as well as that of Gervillia, occur in the Baculite limestone of France and Normandy, which I believe is referred by d'Orbigny to his Senonian Stage, the same in which he included the Cretaceous fossils of North America. " It is interesting to find bivalves of so remote an era in sufficient preservation to exhibit generic characters as clearly defined as they are in living shells. In this condition are the hinges of Gervillia, Pulvinites, Ctenoides and Cardium. Here are also specimens of Baculites and Scaphites which exhibit the interior divested of all extraneous matter, and delight the eye with exquisite curves of the foliated septa, whilst the shells glow with brilliant iridescent tints. " This beautiful series of Cretaceous forms seems to be very limited in geographical distribution, so far as our present knowledge extends. It is probably unknown as yet beyond the limits of Tippah County, which borders on Tennessee. No account has been given of such a group by the State Geologists of Tennessee or Alabama. Dr. Spillman informs me, ' The fossils you have now under examination were found in the bluffs of Owl Creek, three miles north of the town of Ripley,' and he concurs in opinion with me that they might properly be named the ' Ripley group.' He also remarks that Ammonites placenta occurs in it with the shell preserved, and that in connection with the Ripley group, or in the same locality, are ' Exogyra costata, Gryphaea mutabilis, Ostrea plumosa, Natica petrosa, Nautilus DeKayi, etc., with the shells more or less preserved, in an argillo-calcreous marl/ but none of these species are contained in his collections sent me from Tippah County/'— Conrad, 1858. After this announcement of the discovery of well preserved Cretaceous fossils in northern Mississippi was made by Con- rad, Safford collected a few Ripley fossils from near Purdy, Tennessee and Tuomey made a large collection of unusually well preserved shells at Eufaula, Alabama, from the same 6 82 Upper Cretaceous Fauna, from Tennessee [280 horizon as Owl Creek. These collections were sent to Conrad and Gabb for study and their contributions appeared in 1860 in Volume IV of the Journal of the Philadelphia Academy of Natural Sciences. In this volume Conrad described fifty-four additional new species and Gabb four new species collected from northern Mississippi, Alabama and Tennessee. Since the work of these men, very little has been done on this un- usually prolific fauna. Geologists have often visited Owl Creek and collections have been made but nothing has been published on the systematic paleontology except for minor contributions.12 The most recent check list of the Eipley collections in the National Museum from this region is that published by Stephenson in 1914.13 STATE OF PRESERVATION OF FOSSILS A comparison of specimens from the Coon Creek collection with forms from Owl Creek in the National Museum shows that the shells from the former locality include many small and fragile individuals, and that many individuals present delicate shell parts, internal and external markings not so well defined, or entirely absent in the Tippah County Speci- mens. The hinge areas, muscle scars, buttresses, pallial lines and external sculpture are as sharp and as well defined in such genera as Cucullaea, Glycymeris, Crassatellites, Nucula, Cardium, Trigonia, Paranomia, etc., as in the shells of Ter- tiary and Eecent bivalves. Even the ligaments are occasion- ally preserved and in their natural positions in attached valves of Cardium n. sp. Cyprimeria n. sp. and Leptosolen biplicata Conrad. Many of the Gastropoda, including species of such genera as Liopeplum, Gyrodes, Ptychosyca, Voluto- morpha, Pugnellus, etc., are brilliantly glazed. The shells of Eutrephoceras, Baculites, Scaphites, and Turrilites are well 12 Ball, 1890, Trans. Wagner Free Inst. Sci., Philadelphia, vol. iii, pt. 1, p. 73. Ball, 1907, Smiths. Misc. Coll., vol. iv, pt. 1, pp. 1-23. "'Stephenson, L. W., loc. cit., p. 24, tables 1-9, 1914. 281] B. Wade 83 preserved and abundant but in many cases have been crushed by the weight of the superincumbent sediments. The proto- conchs are well defined and in a perfect state of preservation on many of the Gastropoda, especially such genera as Laxi- spira, Volutod&rma, Paladmete, Thylacus and many others. The protoconch is present and sharply differentiated in a new species of Teinostama which is strikingly like the Mio- cene form Teinostoma nana (Lea). The adult itself is only a little more than 1 mm. in its greatest dimension, yet the shell and protoconch are both brilliantly glazed and look as fresh as if they were Eecent. The small and fragile Scapho- pod Cadulus obnutus (Conrad) is abundant and perfectly preserved. Such over-specialized and projecting shell parts as the anterior calcareous tube and the fringing tubules of the genus Clavagella, and the spinose and flaring outer lips of such genera as Anchura, Aporrhais, Volutoderma, etc., occur unbroken. Fragments of non-lignitized and non-petrified wood are common arid resemble Eecent wood in state of pre- servation as shown by weight, color and woody fiber. The occurrence of so many perfect shells in unconsoli- dated sediments as old as the Cretaceous is exceedingly rare. Although these fossils have retained their original charac- ters and shell material many of them are soft and fragile so that some care is necessary in collecting and preparing them. They are easily removed from the strata with part of the matrix attached. This serves to protect the specimens in packing and shipping. When the collected material dries the sandy matrix may be readily cleared away, leaving most of the shells hard and fairly strong. The weaker specimens can be made harder and sufficiently strengthened to withstand handling and the effects of the atmosphere by a method of preparing which is used here in the Geological Laboratory. After all foreign matter has been removed from the shells they are soaked about four minutes in paraffin heated to the boiling point. The hot wax permeates the shell walls and reinforces them. The shells are darkened slightly by the wax but otherwise the method is altogether satisfactory. 84 Upper Cretaceous Fauna from Tennessee [282 At both the Owl Creek (Miss.) and Brightseat (Md.) lo- calities the fossil beds occur directly below the Cretaceous- Eocene contact. This contact represents a long interval of erosion during which the shell beds were at, or very near, the surface and were probably subjected to the action of circu- lating meteoric waters which had a disintegrating effect on the shells. The abundant springs at this horizon show that during late Pleistocene and Eecent times this uncomformable Cretaceous-Eocene contact has furnished an easy channel for ground waters which have attacked the unpetrified shells. At Coon Creek, on the other hand, the conditions are some- what different. There is no overlying uncomformable contact dircetly above the fossil beds but instead there is a great thickness of overlying impervious Eipley clays. The shells were sealed up by the Upper Cretaceous sea in compact, cal- careous sandy sediments and have been, it seems unaffected by circulating ground waters until the dawn of the present physiographic conditions. Even now these beds are so imper- vious that the ground water does not penetrate them, as is shown by the fact that well-drillers have reported the strata perfectly dry. The character of the matrix at the three locali- ties is essentially the same, so that it seems reasonable to assume that the Coon Creek shells are well preserved because they have been protected from circulating ground waters the action of which is so evident in most Cretaceous strata. OBSERVATIONS ON THE FAUNA The Coon Creek fauna is both prolific and varied. Four days collecting at this locality yielded, according to preli- minary determinations a fauna of 134 genera and 269 species, and further collecting has materially increased this number. The study is yet incomplete and some of the determinations are merely tentative but the following generalizations may be made. In the 134 genera already recognized there are, exclu- sive of the Mollusca, three genera of Vertebrata of the Class Pisces; 5 of Arthropoda of the Class Eucrustacea; 9 genera 283] B. Wade 85 of Molluscoidea of the Class Bryozoa; 1 genus of Echinoder- mata of the Class Echinoidea ; 2 of Vermes ; and 1 of Coelen- terata of the Class Anthozoa. The Mollusca, however, are by far the most abundant. In this group there are 49 genera and 110 species of Pelecypoda; 2 genera and 3 species of Scaphapoda; 60 genera and 120 species of Gastropoda; 4 genera and 7 species of Cephalopoda. It has been estimated that the Eecent east coast Molluscan fauna of the Middle Atlantic States includes more than 500 species, and there is no reason to suppose that the Upper Cre- taceous faunas were materially less prolific. On the contrary, the seas were warmer and conditions more favorable to mollus- can life, so that probably not more than one-half the entire fauna has been discovered. The Coon Creek fauna flourished near the head of the Mississippi Embayment and in about the same latitude as the Middle Atlantic States. It was probably in the same general climatic zone of the Cretaceous, so that any estimate of the east coast fauna should hold for the northern part of the Mississippi Embayment as well. The evidence afforded by the Coon Creek material shows that the above estimate is not overdrawn but probably conservative. The extent of the undescribed fauna is indicated by the fact that four days col- lecting at Coon Creek has yielded in the Mollusca alone over 100 new species, three new subgenera and eight new genera. The families and genera with the number of species in each are as follows (a preliminary list made May, 1916) : CLASS PELECYPODA Order Prionodesmacea Nfcculidae. Nucula 3 species Ledidae. Leda 2 species Yoldia 1 species Parallelodontidae. Nemodon 3 species Cucullaea 4 species Arcidae. Area 4 species Glycymeris 2 species Axinea 1 species Postligata 1 species 86 Upper Cretaceous Fauna, from Tennessee [284 Perniidae. Inoceramus 3 species Gervilliopsis 1 species Pteriidae. Pteria 2 species Ostreidae. Ostrea 8 species Exogyra 1 species Pycnodonte 1 species Gryphaeostrea 1 species Trigoniidae. Trigonia 2 species Pectinidae. Pecten 4 species Limidae. Lima 2 species Anomiidae. Paranomia 3 species Anomia 3 species Pulvinites 1 species Mytilidae. Modiolus i species Lithophaga 3 species Crenella 2 species Dreisseniidae. Dreissena 1 species Order Anomalodesmacea Anatinidae. Periplomya 1 species Anatimya 1 species Corimya 1 species Clavagellidae. Clavagella 2 species Poromyacidae. Liopistha 3 species Order Teleodesmacea Pleurophoridae. Veniella 1 species Crassatellitidae. Crassatellites 4 species Astartidae. Vetericardia 1 species Diplodontidae. Tenea 3 species Cardiidae. Ca'rdium 4 species Veneridae. Cyclina 2 species Meretrix 4 species Legumen 2 species Cyprimeria 1 species Tellinidae. Linearia 2 species Solenidae. Leptosolen 1 species Mactridae. Spisula 1 species Corbulidae. Corbula 4 species Saxicavidae. Panope 2 species Pholadidae. Pholidea 2 species Teredinidae. Teredo 4 species.14 CLASS SCAPHOPODA Dentaliidae. Dentalium 2 species Siphonodentaliidae. Cadulus 1 species 14 Besides the above named genera there are probably two others in the collection whose generic relations have not been determined on account of fragmentary material. 285] B. Wade 87 CLASS GASTROPODA Order Opisthobranchiata Suborder Tectibranchiata Acteonidae. Acteon 5 species Tornatella 2 species Ringiculidae. Ringicula 1 species Scaphandridate. Cylichna 1 species. Order Ctenobranchiata Suborder Toxoglossa Cancellariidae. Paladmete 3 species Mataxa 1 species Turritidae. Turris 1 species Surcula 6 species Volutidae. Volutoderma 3 species Volutomorpha 5 species Tectaplica 1 species Liopeplum 1 species Drilluta : 2 species Ptychosyca 2 species Mitridae. Mitra 1 species Vasidae. Xancus 3 species Fusidae. Fusus 2 species Subgenus Anomalofusus 1 species Ornopsis 2 species Fasciolaridae. Piestochilus 1 species Odontof usus 3 species Thaisidae. Sargana 2 species Busyconidae. Busycon 1 species Pyropsis 3 species Pyrif usus 3 species Buccinidae. Hemifusus 1 species Hydrotribulus 2 species Nyctilochidae. Tritonium 1 species Columbellariidae. Columbellina 1 species Strombidae. Pugnellus 1 species Rimella 1 species Aporrhaidae. Aporrhais 6 species Anchura 2 species Suborder Streptodonta Scalidae. Pseudomelania 1 species Vermetidae. Laxispira 1 species Turritellidae. Turritella 8 species Naticidae. Gyrodes 3 species Lunatia 2 species Capulidae. Thylacus 1 species Littorinidae. Littorina .1 species 88 Upper Cretaceous Fauna from Tennessee [286 Order Aspidobranchiata Eulimidae. Leiostraca 1 species Euomphalidae. Hippocampoides 1 species Turbinidae. Schizobasis 1 species Trochidae. Solariella 2 species Umboniidae. Teinostoma 1 species Delphinulidae. Urceoabrum 1 species Liotai 3 species.15 CLASS CEPHALOPODA Order Nautiloidea Nautilidae. Eutrophoceras 1 species Order Ammonoidea Lytoceratidae. Baculites 3 species Turrilites 1 species Cosmoceratidae. Genus Scaphites 1 species Probably the most significant fact revealed by the above list is that the number of univalve species is greater than the number of bivalve species. However, all three orders of the Pelecypoda are well represented. Of the Order Prionodes- macea the three families represented by the greatest number of forms are the Arcidae, Ostreidae and Mytilidae. The last two had their beginning in the Paleozoic. The Arcidae ori- ginated and suddenly became a prominent group in the latter part of the Mesozoic and developed into very great import- ance in the Tertiary. Each of these three families is repre- sented by four genera at Coon Creek. Among the Anomalo- desmacea there are three families and five genera. The Teleodesmacea are well represented. Of this order probably the individuals of the families Cardiidae, Veneridae, and Corbulidae are most abundant. A comparison of the above list with lists from the East Coast Cretaceous shows that the bivalves are relatively less abundant in the Coon Creek hori- zon than in corresponding horizons in New Jersey and the Middle Atlantic States. Several genera such as Cuspidaria, 15 Besides the above named genera there are probably thirteen genera and more than that number of species whose generic relations have not been determined on account of the fragmentary condition of the material. 287] B. Wade 89 Myrtaea, Phacoides, Docinia, Tellina, Solyma, etc., are absent from the present McNairy County collection though it is probable that further collecting may reveal some of them. The Scaphopoda are represented by the two families Den- taliidae and Siphonodentaliidae. The former originated in the Ordoviciari and are abundantly developed in the Creta- ceous and Tertiary. They are represented at Coon Creek by one genus and two species, one of which is very common. The family Siphonodentaliidae is first found in the Creta- ceous. At Coon Creek it is abundantly represented by the minute form Cadulus obnutus (Conrad). The Gastropoda are the most interesting class in the Coon Creek fauna. It will be noted from the list given above that the number of genera and species of the Gastropoda is con- siderably greater than the number of the Pelecypoda, yet, probably in every cubic yard of the Coon Creek sediments the number of bivalve individuals exceeds the univalve indi- viduals several times. In all the faunas previously reported from the Cretaceous of Eastern United States the bivalve species are more numerous than the univalve species. This majority among the latter may be due simply to the fact that in the Upper Cretaceous seas of those regions the pelecypods predominated in number of species as well as individuals, or it may be that a greater number of gastropod species existed in all the Cretaceous seas but were not preserved sufficiently to be recovered from the sediments. The chance of preser- vation of gastropod shells is not as good as it is for pelecy- pods, first, because the number of individuals per species of the Gastropoda is rarely ever as great as it is among the Pelecypoda. Second, the essential constituent of univalve shells is aragonite, and this mineral is much less stable than calcite, which is the essential constituent of the majority of pelecypod forms. Third, a gastropod shell is in greater danger of being crushed by the pressure of the inclosing sediments because of lack of support from within the shell. The body cavity of gaping bivalved shells is almost of neces- sity filled, while the sediments are intruded less readily 90 Upper Cretaceous Fauna from Tennessee [288 through the aperture in the spiral body cavities of univalves. The shells thus unsupported within become crushed by super- incumbent .sediments and are then rapidly disintegrated. In general, the Tertiary and Eecent faunas of North America contain a greater number of univalves species and it may be that about the same proportion existed in all the Cretaceous faunas. Yet it may be that the faunas, as they have been reported, represent the natural proportions in which these animals lived in the Cretaceous sea. It is possible that the gastropods became diversified in the Cretaceous and that this diversification took place only in certain provinces, where environments favored variation. Excavating and ex- tensive collecting in localities where the shells are especially well preserved will probably throw some light on this question. In the order Opisthobranchiata there are three families and four genera found in the Coon Creek collection. Of these the family Acteonidae, which had its beginning in the Devo- nian and gained great prominence in the Mesozoic is repre- sented by the genera Act eon and Tornatella. The former in- cludes probably five species and the latter two. The families Ringiculidae and Scaphandridae are each represented by one genus and one species. The order Ctenobranchiata is most abundant and is repre- sented by 48 genera, 30 of which belong to the suborder Toxoglossa and 18 -to the suborder Streptodonta. Among the Toxoglossa the family Cancellariidae is first differenti- ated in the Upper Cretaceous. It appears suddenly much diversified in that period and attains its maximum distribu- tion in the late Tertiary and Eecent. This family is repre- sented at Coon Creek by two genera. The most prolific of these is Paladmete, a genus first recognized and described by Dr. Julia A. Gardner who has recently monographed the Upper Cretaceous Mollusca of Maryland.16 The type spe- cies of Paladmete is very abundant in Maryland and north- 16 Gardner, J. A., Md. Geol. Survey, Upper Cretaceous vol., Text, p. 412, 1916. 289] B. Wade 91 ern Mississippi. The genus is represented by three species at Coon Creek. Mataxa is a form regarded by me as a new genus and is referred to the Cancellariidae. A species of Mataxa has been described from South India by Stoliczka 1T and asigned to the genus Narona of the Cancellariidae. A study of Stoliczka's description and figures together with perfect specimens from Coon Creek shows that these species belong to the same genus which is evidently not Narona, so it seems advisable to assign these two forms to Mataxa as a new genus in the Cancellariidae. The Turritidae is another family which is represented in the Cretaceous by several forms and is not found in earlier strata. There are two gen- era of this family at Coon Creek, Turris, and Surcula. The latter is especially varied, and includes probably six species. The Volutidae is the most prolific and interesting family of the Coon Creek collection. It contains six genera and sixteen species, nearly all of which are represented by abun- dant, well characterized and perfectly preserved shells. The remarkable efflorescence of the Volutes in the Upper Creta- ceous has been discussed by Dall who mongraphed the Volu- tidae in 1890 18 and in 1907.19 The Coon Creek Volutes include the genera Volutomorpha, Volutoderma, Ptychosyca, Liopeplum, two new genera, Drilluta and Tectapl^ca and two species of a form whose generic relations have not been fully determined. The genus Ptychosyca was named and described by Gabb in 1876.20 Dall did not consider this genus a Volute in his monograph of this family, but two species re- presented by well preserved material in the McNairy County collection reveal new characters that indicate that this form 17 Stoliczka, F., Geol. Surv. of India, Cret. Faunas of South India, vol. ii, p. 166, pi. xiii, figs. 15, 16. 18 Dall, 1890, Trans. Wagner Free Inst. Sci., Philadelphia, vol. iii, pt. 1, p. 72. 19 Dall, 1907, Smiths. Misc. Coll., vol. iv, pt. 1, pp. 1-23. 20 Gabb, W. M., Proc. Acad. Nat. Sci., Philadelphia, 1876, p. 295, pi. 17, figs. 2-4. 92 Upper Cretaceous Fauna from Tennessee [290 is quite probably a member of the Volutidae and near the genus Liopeplum. The Mitridae and Vasidae are each represented by one genus. The genus Xancus of the Vasidae is represented by three species which are well characterized by the manner of excavation of inner lip and number of columellar folds. This genus is well represented in the Upper Cretaceous and was first identified from the Cretaceous quite recently by Dr. Gardner.21 The Fusidae which appeared in the Jurassic and are widely developed in the Tertiary and Recent are repre- sented at Coon Creek, besides the genus Fusus, by a form which is considered of the rank of a subgenus under Fusus and given the name Anomalofusus as a new subgenus. The Fusidae are further represented by two very common well characterized species for which the new genus Ornopsis has been proposed. The family Fasciolariidae embraces the gen- era Piestochilus and Odontofusus, the latter being repre- sented by three species which are very closely related. The well-characterized genus Sargana, of the Thaisidae, is repre- sented at Coon Creek by two species, one of which is very abundant and perfectly preserved. The family Busyconidae is interesting in that it appears rather suddenly in the Cretaceous with numerous, diversified representatives. At Coon Creek it is represented by three genera and seven species. These genera are Busycon, Pyrop- sis and Pyrifusus. Pyropsis and Pyrifusus are very abundant in the Creta- ceous and have a world-wide distribution. Busycon, very commonly known as Fulgur, is rarely found in the strict sense in the Cretaceous. This extension of the range of this common East Coast Tertiary and Recent form is of particular interest. It is represented in the present Coon Creek col- lection by a single well preserved specimen, a description of which has been prepared for publication. This specimen, aside from the absence of the protoconch is perfect and pre- 21 Gardner, J. A., loo. cit., p. 434. 291] B. Wade 93 sents generic characters which cannot be mistaken. The species bears a striking resemblance to some of the medium- sized late Tertiary and Eecent species. All the typical Ful- gurs previously known have been limited to the Tertiary and Eecent of the Atlantic States. The Eocene forms are small, rather thin-shelled species, so it has been considered that the genus evolved during that period. The living Fulgurs have been very extensively studied and the life history carefully worked out. The limited geographic range has been ex- plained in a large measure by the fact that the animal is deprived of an active free-swimming larval stage by the loss of the velum before the young form emerges from the egg- capsule. This same fact might well be cited to explain the very limited distribution of Busy con in the Cretaceous. One of the earliest Tertiary species (described and referred to genus Fulgur by Harris22), occurs in the Midway group of the Eocene about 30 miles west of Coon Creek. In the family Buccinidae there is a species represented by large elegant specimens which seems to belong to a well- defined generic group for which the name Hydrotribulus has been proposed. Species of this genus have been recognized from Brightseat, Maryland and Owl Creek, Mississippi. A study of the description of a species, Tudicla monheimi (Muller) Holzapfel23 from the Aachen beds of western Ger- many shows that the European form belongs to the same genus as the American forms and it seems advisable to pro- pose a new genus for this group. Hemifusus is another genus that occurs among the Coon Creek Buccinidae and is a form which has never before been reported from the Cre- taceous of Eastern United States. In the family Nyctilochidae there is a single genus Tri- tonium. The family Columbellariidae is represented by a 22 Harris, G. D., Bull. Amer. Pal, 1896, vol. i, no. 4, p. 96, pi. 9, fig. 13. 23 Holzapfel, E. Pelaeontographica, 1888, Band xxxiv, pag. 106, Taf. xi, figs. 4-7. 94 Upper Cretaceous Fauna from Tennessee [292 single small individual which has been assigned to the genus Columbellina, a group not previously reported from the Cre- taceous of North America. One species of the genus Pugnel- lus of the Strombidae is very common in the Coon Creek beds. A species of Rimella of the same family is represented by one specimen. The family Aporrhaidae is prolific, its two gen- era Aporrhais and Anchura include probably nine species. In the suborder Streptodonta the families Scalidae, Ceri- thiidae, Trichotropidae, Vermetidae, and Turritellidae are each represented by a single genus. In the genus Turritella there are probably 8 species. Both Lunatia and Gyrodes of the Naticidae are common. The individuals of one species of Lunatia are probably more abundant than any other gas- tropod species. The family Capulidae is represented by small, fragile individuals of a single species of the genus Thylacus which was described from Owl Creek by Conrad in I860.24 The individuals of this species are small and very fragile, yet they are abundant and perfectly preserved in their natural habitat. They occur in place fitting snugly to the columellar walls in the body cavities of larger gastropods. They have the internal muscular impression produced and leaving the wall of the shell at the anterior extremities, and lack the calcareous foot-plate characteristic of the genus Hipponix of this family. The family Littoriniidae is repre- sented by the genus Littorina which is common in the Ter- tiary and Eecent of the East Coast and Gulf regions, but up to the present has not been reported from the Cretaceous of these regions. f- The genus Leiostraca of the family Eulimidae and order Aspidobranchiata is represented by abundant but often poor- ly preserved specimens, due to the fragility of the shell. The family Euomphalidae which is variously represented in both the Paleozoic and Mesozoic, includes a new genus Hip- pocampoides. This is a much depressed form with a pro- 24 Conrad, T. A., Jour. Acad. Nat. Sci., Philadelphia, 1860, vol. iv, 2d ser., p. 290, pi. 46, fig. 22. 293] B. Wade 95 duced keel and angular shoulder. Most of the specimens in this family that have been described from the Cretaceous of the Eastern United States have been referred to the genus Straparollus, and in most cases these specimens are casts and do not show any shell characters, so that it is possible that some of these casts belong to Hippocafrnpoides. The family Turbinidae is herein reported from the Cretaceous of Eastern North America for the first time. This family is abundantly developed in the Paleozoic and is common in the Eecent. The new genus Schizobasis which is character- ized by a very unique, flattened notch-like anterior canal is referred after some hesitation to the family Turbinidae. The families Trochidae and Umboniidae which have repre- sentatives from the Silurian to the Recent are each repre- sented by a genus in the Coon Creek fauna. These genera are Solariella and Teinostoma respectively and are both herein reported for the first time from the Cretaceous of North America. Another genus hitherto unknown in Ameri- can Cretaceous is Liotia of the family Delphinulidae and in this family occurs another very abundant form at Coon Creek for which the genus Urceolabrum is here proposed. This seems to be a well denned generic group near Liotia but dis- tinctly different from typical Liotiae occurring in the same strata. Besides the Coon Creek species Urceolabrum includes an undescribed species from Aufaula, Alabama, and another from the Aachen beds 25 of Vaals, Germany. In addition to the above cited genera there are probably as many as thirteen genera of Gastropoda, including that many and more species whose generic and .family relations cannot be assigned with assurance on account of their frag- mentary character. The point of greatest interest in the Coon Creek gastro- pods is the occurrence of eight new genera and one new pub- genus, many of which are represented by more than one spe- 25 Holzapfel, E., Palaeontographica, Band xxxiv, p. 170, Taf . xviii, figs. 3-7. 96 Upper Cretaceous Fauna from Tennessee [294 cies and also from more than one locality, as is shown in the literature by described species which have been questionably assigned generically. The genera Solariella, Liotia, Teinos- toma, and Columbellina have not been previously reported from North American Cretaceous, and Hemifusus and Lit- torina have not been previously reported from the Cretaceous of the Eastern United States. A typical Busycon or Fulgur 2G is for the first time found in Cretaceous sediments. The Volutes are profusely developed, being represented by six genera and eighteen species. Among the Cephalopods both the nautiloids and ammonoid orders are present and represented by abundant large well- preserved specimens. It is interesting to find the remains of the most primitive order of Cephalopoda which ranges from the Paleozoic to the Eecent associated with abundant remains of the most highly complex and degenerate cephalo- pods, indicating that both thrived under the same conditions, yet the latter became extinct and the former continues to live. The family Nautilidae is represented by one species of the genus Eutrephocems. The most common cephalopods at Coon Creek are the Baculites and Turrilites of the family Lytoceratidae. Baculites is profusely developed and proba- bly includes three species. The Cosmoceratidae include one species of the genus Scaphites. Conditions must have been especially favorable for mollus- can life in the Upper Cretaceous sea in which the members of the Coon Creek assemblage grew. A glance at a tray of specimens impressed the observer- with the fact that the shells are the remains. of once-flourishing animals. Very many of the shells are thick, stout and of imposing dimensions. Evi- dently they belonged to robust, healthy and well-fed organ- isms. The bivalve with the greatest lateral dimensions is a species of Inoceramus which was probably 15 inches in maxi- mum diameter. One species of Cardium is 5 inches in length 26 Bruce Wade, 1917, Am. Jour. Sci., vol. xliii, no. 256, p. 293, figs. 1 and 2. 295] B. Wade 97 and a Cyprimeria is 4^/2 inches. Exogyra costata and Gry- phaea vesicularis occur in their typical massiveness. The shells of Cucullaea, Crassatellites and Veniella are very abundant and evidently belonged to three very thrifty groups of Mollusca which lived under conditions especially suited to the secreting of calcium carbonate. The afflorescence of the Volutidae in the Upper Cretaceous has already been empha- sized. All the species of this family are above medium size and many of them are very large. Perfect specimens of Volutoderma in the collection attain an altitude of half a foot. There are broken specimens which when perfect must have been almost a foot long. Volutomorpha is probably the giant of the Cretaceous gastropods. There is a fragmenl of several whorls of the spire of one species of Volutomorpha in the collection which would probably be 18 inches in length were the specimen complete. The genus Ptychosyca is large and inornate while the genus Drilluta is elongate and elabo- rately sculptored. The shells of Pugnellus and Gyrodes seem to be relics of once prosperous organisms which saw no hardships in life. Lwpeplum, Lunatia, Xancus, Hydrotri- ~bulus, Ornopsis, etc., though less in dimensions than some of the above forms, evidently grew in very favorable environ- ments. Species of such genera as Columbellina, Solariella, Act eon, Cerithium, etc., are much smaller in size, yet their shells are thick and stout, and no doubt grew where condi- tions were favorable for secreting calcium carbonate. The bivalves also show various ranges in size of thick, stout shells. The Cephalopoda were the largest of the Coon Creek Mol- lusca. The genus Eutrephoceras is abundantly represented by large thick-shelled cavernous individuals more than six inches in diameter. One species of Baculites is very abundant and large. Although no complete, large individuals have been recovered from the matrix, there are several large pieces of shells and body chambers in the collection from 6 to 18 inches long and restorations of these show that some indi- viduals were five feet in length. It should be noted here that while most of the Upper Cretaceous molluscs had thick stout 7 98 Upper Cretaceous Fauna from Tennessee [296 shells with coarse, vigorous ornamentation, yet many pos- sessed small, delicate, fragile and. thin shells, but have never- theless been preserved perfectly to the present. Individuals of species of Leda, Cadulus and Teinostoma are smaller than a wheat gain. Yoldia, Anatimya, Tenea, Liopistha, Leio- straca are represented by delicate and fragile individuals. One species of Crenella is thinner than paper yet it is ele- gantly sculptured. It is impossible to postulate with assurance the depth of the water in which the Coon Creek fauna lived. Such fami- lies as the Pernidae, Volutidae and Lytoceratidae, which are very prominent in the assemblage, are usually regarded as dwellers in the open sea at a depth of about 50 fathoms. Yet the Nuculas, Corbulas, and Naticoids, etc., are for the most part dwellers in shallow water near shore. Lobsters and true crabs lived in great abundance in the Eipley sea as is shown by the remains of these forms which are very com- mon in the Coon Creek sediments. There are probably five genera of the Eucrustacea, among which is a large crab about seven inches across from .right to left and whose modern affinities live in the intertidal zone of the seas. No fora- miniefra have been found. Only two very small individuals of two species of corals have been recovered. These last two facts, together with the very abundant crab remains indicate very near-shore or intertidal waters as the habitat of the Coon Creek fauna. As regards the evidence furnished by the sediments there is no well-marked cross-bedding which would result from strong current action. However, the very presence of clastic material such as sand and clay require currents to account for transportation, and shifting of these currents to explain the intermingling of these materials. The great abundance of pelecypods which are organisms that feed for the most part on plankton is indicative of waters disturbed by cur- rents, instead of very calm seas, for plankton occurs mostly in water that is agitated by currents. No pebbles whatever have been observed in the sediments of the Coon Creek 297] B. Wade 99 horizon, wood fragments are, however, common. The totality of the evidence seems to indicate that the Coon Creek fauna lived in the agitated waters near the coast of .a low-lying land mass. A study of the distribution and variation of the faunas with reference to the character of the sediments in the Rip- ley formation of northern Mississippi and southern Tennes- see shows that the areas so favorable to molluscan life were a •'.'.*.;.',••'•! •'..',' • \ '..'/ .•••'. '•'..'• ".•.••' •'.''..' • . . '• .. • '.• • .' '. . • .' W MUv**y Group. T Owl C'rtti ffcriiiT, + N,,tA'rn fflt* 17 Mc/fitty Sin whether overcast, partly overcast or clear. The daily values for the first two of these terms vary in a regular manner throughout the year at any given place, and the ones for the third term are roughly stated in the observer's records, as just mentioned. It was desired to combine these three terms so as to get approximations of sunshine intensity for a number of different stations in Maryland for the summer of 1914, in order to make comparisons of the summer march of sun- shine intensity with that of corresponding measurements of plant growth. This has been accomplished in the manner described below. The first two terms are combined in the ordinates of the graph given by Kimball1 for the maximum possible total radiation received per day at Mount Weather, Virginia. Since this station is at about the same latitude as the stations in Maryland, the ordinate values may be taken as approximate measures of the total radiation intensity for the corresponding dates at any place in the state. These values represent the total amount of heat, expressed in gram-calories per square centimeter of horizontal surface exposed, received from the sun and sky on clear days at Mount Weather. The method 1 Kimball, Herbert H., " The total radiation received on a hori- zontal surface from the sun and sky at Mount Weather. Monthly Weather Rev. 42: 474-487. 1914. (See especially fig. 8, p. 484). 405] F. M. HiUebrandt 207 of using the graph and a weather observer's report for estimat- ing sunshine will be best shown by an example. Suppose it is desired to estimate the average daily sunshine intensity for some station in the general region of Mount Weather, for the first week of August. The average ordinate value for this week is first obtained from KimbalPs graph. For periods as short as a week or two this may be done by averaging the values for the first and last days of the period, since the curve may be taken as a straight line for such short intervals. From the report of the weather observer at the place in question, the number of clear, partly cloudy, and cloudy days is next determined for the days August 1 to August 7, inclusive, and some arbitrary weighting is given to each kind of day. We may, for instance, call clear days whole days of sunshine, partly cloudy days half days of sun- shine, and assume that cloudy days are days without any sun- shine. The scheme of weighting adopted must, of course, be adhered to in all the estimates made for different periods and stations. The system of weighting given above was used in the studies for which this method of approximating sunshine was developed. By summing these weighted daily values a num- ber is obtained which represents the equivalent number of clear days for the period considered. Suppose, in the example selected, that this equivalent number of clear days is 3.5, which is 0.5 of the total number of days in the period. The latter value may be termed " the coefficient of clear weather." By multiplying the average daily intensity value for clear days, as obtained from the curve, by this coefficient of clear weather a number is secured that is a rough approximation of the average daily sunshine intensity for the week. While it is certain that solar radiation affects plants in other ways than through its heating effect, it is no less cer- tain that by far the greater part of the energy of sunshine absorbed by plants is converted into heat (largely as latent heat of the vaporization of water), and it seems probable that the other effects produced upon the plant may be more or less proportional to the total energy equivalent of sunshine. The 208 Moisture Equilibrium [406 method of measurement of light here given, although it is only a rough approximation and depends on the heating effect of the sunshine, has been shown, as a matter of fact, to give numbers rather definitely correlated with plant growth. It has been found, for instance, that the amount of dry sub- stance produced per unit of leaf area in young soy-bean plants decreases from the beginning to the end of the growing season, in a manner which generally parallels a corresponding fall in the light intensity values as determined in the manner described above. MOISTURE EQUILIBRIUM IN POTS OF SOIL EQUIPPED WITH AUTO-IRRIGATORS By F. S. HOLMES While the auto-irrigator devised by Livingston 1 has been employed by several writers,2 for maintaining uniform mois- ture conditions in potted soils, the details of adjustment re- quired by this device, for different soils and for maintaining different moisture contents, remain still to be worked out. In order to throw some light upon this general question, a study of three different soils was undertaken to determine the relation between the equilibrium point of the soil-moisture content and the number of irrigator cups employed. One soil was a medium-fine white sand, one was a light 1 Livingston, B. E., " A method of controlling plant moisture." Plant World II: 39-40. 1908. 2 Hawkins, Lon A., " The porous clay cup for the automatic water- ing of plants." Plant World 13: 220-227. 1910. Transeau, E. N., " Apparatus for the study of comparative transpiration." Bot. Gaz. 52: 54-60. 1911. Livingston, B. E., and Lon A. Hawkins, "The water relation between plant and soil." Carnegie Inst. Wash. Pub. 204: 5-48. 1915. Hibbard, R. P., and 0. E. Harrington, "Depres- sion of the freezing-point in triturated plant tissues, and the mag- nitude of this depression as related to soil moisture." Physiol. Res. I: 441-454. 1916. 407] F. S. Holmes 209 clay loam, and the third was a mixture, of equal parts, by volume, of the other two. Pots of each kind of soil were equipped with auto-irrigators having respectively one, three and five porous cups, thus giving nine combinations. The containers were tinned sheet-metal cylinders approximately 15 cm. in diameter and 17 cm. in height. The porous cups were evenly distributed within the soil mass, when but one was used it occupied the center. A mercury tube was so ar- ranged that all water entered the soil against a pressure of from 5 to 6 cm. of a mercury column. Evaporation was pre- vented by sealing covers on the containers with plastiline. The cylinders were filled to a uniform depth of 16 cm., an at- tempt being made to secure as uniform packing as possible throughout the entire series. Weighings of the containers were made at intervals of two or three days, for the first twenty days, and thereafter at weekly intervals, to determine the rates at which water was being absorbed and to approximate the moisture content of the soil. Approximately three-fourths of the water taken up by the loam and by the sand-loam mixture occurred dur- ing the first ten days, but the sand took up only about one- half of its total amount in the same period. Approximate equilibrium of the soil moisture content was reached in about seventy-five days, in the case of the loam; in about eighty days in the case of the mixture ; and in about ninety days in the case of the sand. The number of porous clay cups em- ployed seemed to have no influence upon the length of time required for the attainment of equilibrium by either the loam or the loam-sand mixture. With the sand, however, the number of cups appeared to influence the length of this time period. With three cups equilibrium was reached sooner than with one, and with five sooner than with three. When the weighings of the cylinders and observations on the water reservoirs showed that the soil had ceased to absorb water, the cylinders were opened and samples were taken for soil-moisture determinations. Two 1-cm., full-depth cores were taken from each container, one core from as near a cup 14 210 Moisture Equilibrium [408 as possible, the other as far removed as possible. The aver- age of the two was taken to be representative of the entire soil mass. Each sample was removed and dried in eight 2-cm. sections, so that it was possible to study both the ver- tical and horizontal distribution of the soil moisture in the cylinder. There was a horizontal as well as a vertical varia- tion of small magnitude in the soil-moisture content of all the cylinders, the water content being almost always somewhat higher near the cups and at the bottom of the soil mass. The distribution of the moisture, both horizontal and vertical, was more uniform in the sand-loam mixture than in the sand, and also more uniform in the loam than in the mixture. The number of porous cups used had very little influence, if any, upon the soil moisture content of the loam; it varied as 100 : 106 : 103, for the containers having one, three and five porous cups, respectively. This influence of the number of cups was more pronounced in the case of the sand-loam mixture, the variation, with one, three and five cups, being 100 : 147 : 168. With the sand there was a still more marked effect, the moisture contents for the three cup numbers being 100 : 191 : 277 in this case. These vari- ations are all smaller than the corresponding variation? in the value of the ratio of cup number to soil mass, these values varying as 100 : 321 : 576, for all three soils. For the con- tainers with three cups the actual average soil moisture con- tent (on the basis of dry weight) was 11.0 per cent, for the loam, 5.2 per cent, for the mixture, and 1.1 per cent, for the sand. With the pressure here used (averaging 5.5 cm. of a mer- cury column) the soil moisture content at equilibrium was too low for plant cultures in the sand and perhaps also in the sand-loam mixture. In the loam, however, it was surely high enough to supply plants with the water necessary for their growth under ordinary greenhouse conditions. 409] E. 8. Johnston 211 SEASONAL VARIATIONS IN THE GROWTH-RATES OF BUCKWHEAT PLANTS UNDER GREENHOUSE CONDITIONS By EAEL S. JOHNSTON Seasonal variations in greenhouse plants are of considerable importance to plant growers as well as to experimenters in plant physiology,, but it is especially with reference to physi- ological experimentation that this study was undertaken. When it is necessary to repeat an experiment on plant growth it often occurs that the results of the second experiment are in more or less pronounced disagreement with those of the first. Since the controlled external conditions must be re- garded as the same- for both experiments, such disagreement appears to be related either to initial differences in the plants used (internal conditions) or to uncontrolled external con- ditions as these vary with the season. The first of these pos- sibilities is probably not as important as the second in most cases, for care is usually taken to select plants for the second experiment that are apparently similar to those used for the first. While this problem of similarity of internal conditions of different lots of plants is a very difficult one and is hardly susceptible of quantitative study at the present time, it is quite possible to carry out studies on the relation of growth to the usually uncontrolled (or only partially controlled) external conditions of a greenhouse., as these conditions change throughout the year. A portion of the results ob- tained from such study are here presented. A set of similar water cultures was started every two weeks and each was continued for a period of four weeks, so that the periods of successive sets overlapped. A single set con- sisted of ten plants, each suitably supported in a glass jar con- taining about 425 cc. of nutrient solution. These jars were covered, to exclude most of the light from the plant roots. The solution was renewed at the middle of each four-week 212 Variations in Growth-Rates [410 period. At the end of each week several different kinds of measurements of the plants were made, and the data thus obtained were studied to bring out the seasonal variations in growth-rates. Since the solutions were alike for all sets and the seedlings used were selected for likeness, it is fair to sup- pose that observed differences in growth-rates, between the different sets of cultures, must have been mainly due to fluc- tuations in the uncontrolled conditions of the surroundings, such as temperature, light and the evaporating power of the air. The experiments were carried out in one of the experiment greenhouses of the Laboratory of Plant Physiology. Xo artificial shade was applied to the greenhouse. Two sets of cultures were always carried out simultaneously, one under unmodified greenhouse conditions and the other in a cheese- cloth chamber in the greenhouse, but the data obtained from the chamber cultures will not be dealt with in the present paper. A continuously rotating table 76 cm. in diameter was used in each case, the jars standing near the margin of the table. Japanese buckwheat, Fagopyrum esculentum Moench., was employed, and Shive's * three-salt nutrient solution, no. R 4C2 (total osmotic value 1.75 atmospheres), was used throughout the entire series. Aside from renewing the solution at the middle of the four-week period, water was always added at the end of the third week of growth, to bring the solution back to its original volume. When the transpi- ration rates were excessive a still further addition of water was made during the fourth week of growth, in order to pre- vent the root systems from becoming unduly exposed. The first experiment began Feb. 14, 1916. Of the plant characteristics measured at the end of each four-week period of growth, only stem height, total dry weight and total area of the leaves (one surface only) are here con- 1Shive, John W., "A study of physiological balance in nutrient media." Physiol. Res. I: 327-397. 1915. 411] E. 8. Johnston 213 sidered, the values obtained being expressed as averages per plant, for each of the four-week periods. The temperature conditions, the evaporating power of the air and the intensi- ty of radiation were recorded for each of the two exposures, but these are left out of the present consideration. The results obtained from these three plant measurements are shown in the accompanying table, wherein all the values are expressed in terms of the corresponding value for the period ending May 22. In this table the dates of beginning and ending of the several culture periods are shown in the first two columns. Each value given in the table represents an average growth-rate representing a single plant, for a time period of 28 days. EXPERIMENTAL DATA Period Stem Total Total Av'ge of Beginning Ending Height. Dry Wt. Leaf Area Wt. & Area Feb. 14 Mar. 13 .73 .50 .63 .57 Feb. 28 Mar. 27 .83 .62 .81 .72 Mar. 13 Apr. 10 .85 .72 .77 .75 Mar. 27 Apr. 24 .94 .80 .76 .78 Apr. 10 May 8 .98 .89 .76 .83 Apr. 24 May 22 1.00 1.00 1.00 1.00 (67.5 cm.) (1.338 g.) (213.5 sq, , cm.) May 8 June 5 .93 .91 .93 .92 May 22 June 19 .83 .93 .98 .96 June 5 July 3 .73 .93 1.00 .97 (214.1 sq. cm.) June 19 July 17 .77 .88 .88 .88 July 3 July 31 .97 .91 .92 .92 July 17 Aug. 14 1.04 .82 .83 .83 July 31 Aug. 28 .91 .67 .77 .72 Auff 14 Sept. 11* •*••*• *-*g» -LTE Aug. 28 Sept. 25 1.07 .76 .70 .73 Sept. 11 Oct. 9 .97 .55 .58 .57 Sept. 25 Oct. 23 .78 .34 .43 .39 Oct. 9 Nov. 6 .79 .36 .51 .44 The different kinds of growth-rates are seen to vary inde- pendently, from period to period, but two of the growth Data not obtained because of insect injury to plants. 214 Variations in Growth-Rates [412 criteria,, weight and area, show variations that correspond rather closely. Both of these show high rates for the sum- mer and low ones for the spring and autumn. Judged by dry weight of plant produced the growth-rate reached its maximum (1.34 g. per plant, in 28 days) with the period ending May 22, but this value remains high until after the period ending July 31. Judged by the total leaf area, the rate does not attain its maximum (214 sq. cm, per plant, in 28 days) until later, this occurring with the period ending July 3, but this value is high for the three preceding periods and for the two following. Roughly speaking, it may be said that these two criteria give rates that are proportional, and that they agree in indicating a period of very rapid growth, extending from about May 8 to about July 17. Before the period with its middle at May 8 the rates are lower, forming a generally ascending series, from the very low values of the early spring, and after the period with its middle at July 17 they decrease rapidly (with a low secondary maximum indi- cated for the period ending Sept. 25) to very low values in the autumn. The rates of growth in height fail to show this sort of seasonal march; the maximum rate (49 cm. per plant, in 28 days) being shown for the period ending July 3, but this rate also has very low values for the periods ending March 13, Oct. 23 and JSTov. 6. By this criterion, the maximum for the period considered (72.5 cm. per plant, in 28 days) occurs with the period ending Sept. 25, but pronounced secondary maxima are shown for the periods ending May 22 and Aug. 14. This rate of growth in height appears to vary consider- ably from period to period, but in a manner entirely inde- pendent of the general advance of the season and quite inde- pendent of the variations in rates of increase in dry weight and in leaf area. As far as these data go, it therefore ap- pears that there is nothing in the usually uncontrolled ex- ternal conditions of a greenhouse in this climate, that may be expected to produce a regular march of growth-rates in 413] E. 8. Johnston 215 height, for healthy buckwheat plants, during the spring, sum- mer and autumn. McLean 2 has pointed out the approximate proportionality of the rates of production of dry weight and leaf surface, for the first four weeks of growth of soy-bean plants, and he also found that the rate of stem elongation varied quite differ- ently from the rates of production of dry weight and surface. It may be of- fundamental significance that two plants as widely different, in many other respects, as are buckwheat and soy-bean, exhibit these remarkable agreements in the manner of variation in these three growth-rates with differences in the climatic conditions of the environment. The general agreement between the seasonal variations shown by the rates of increase in dry weight and in leaf area is so marked that it appears quite permissible to combine these two criteria by averaging their relative values, to give a single value representing both together, and the averages so derived are given in the last column of the table. Of course, these two measurements of growth-rate are not directly commensu- rable, and the average values here introduced are to be re- garded merely as numerical indices of the rates of growth. This value has its maximum (1.00) for the period ending May 22, and it of course shows high value for the five fol- lowing periods. Its minimum value (0.39) occurs for the period ending Oct. 23. Of course there are many other considerations to receive attention in a study of this sort, but it already seems clear that a regular and pronounced seasonal variation in the rates of production of dry weight and leaf area may be expected in healthy buckwheat plants growing in a greenhouse in this kind of climate, although the same nutrient medium is al- ways employed. If the weight-area indices be represented 2 McLean, Forman T., "A preliminary study of climatic conditions in Maryland, as related to plant growth." Physiol. Res. 2: 129-208. 1917. 216 Variations in Growth-Rates [414 graphically they give only comparatively slight variations from a smooth curve and the actual graph may readily be smoothed to give such a curve. After this has been done the ordinates of the smoothed curve, corresponding to the various culture periods, may be measured, and the series of graphi- cally derived values thus obtained may be taken as a tentative scale to indicate approximately the relative growth-rates to be expected for this plant in these general surroundings. Of course, the seasonal march of the climatic conditions in this particular greenhouse must be expected to vary from year to year, and it surely varies from greenhouse to greenhouse; nevertheless, the tentative scale derived as just described may be of value in several ways. For the first sixteen four-week periods of the present study, beginning with Feb. 14, as given in the table pre- sented above, these relative seasonal indices of growth-rate (by either dry weight or leaf area, which appear to be propor: tional, or by their average) are respectively as follows : 61, 71, 79, 86, 91, 96, 99, 100, 99, 96, 92, 87, 81, 75, 68, 61. In this scale of growth-rate values the maximum (100) occurs for the period ending June 19, and it represents actual average growth-rates, as obtained in this study, of 1.24 g. of dry weight and 209 sq. cm. of leaf area (one surface only), per plant, per period of 28 days. While these derived results are extremely tentative and probably only very roughly ap- proximate, it is clear that we have here a new kind of descrip- tion of the climatic conditions of this greenhouse for the spring, summer and autumn of 1916, these conditions and their seasonal march being described in terms of their ability to produce dry material and leaf surface in the standard plant here employed. By such a method as this the climatic plant-producing power for any four-week period may be directly compared with that of any other similar period, no matter when or where these periods occur, the standard plant being used as an automatically integrating instrument for the measurement of 415] W. E. Tottingham 217 the effective climatic conditions. This general method for the comparative study of climatic conditions has been sug- gested by Livingston and McLean 3 and a first attempt at its employment was carried out by McLean in the paper already mentioned. ON THE RELATION OF CHLORINE TO PLANT GROWTH By W. E. TOTTINGHAM As a result of experiments conducted early in the develop- ment of the water-culture method, chlorine has been con- sidered as one of the unessential elements for the growth of plants in general. Nevertheless, all seeds contain more or less of this element and in no instance has a plant been limit- ed to this original source of chlorine through successive gen- erations, so that it may still be said that the question here raised has never been really tested. Practically all soils con- tain considerable amounts of chlorine in the form of chlorides and its occurrence in plants appears to be confined to this form. That this element may have important effects under some conditions, when applied as an agricultural fertilizer, is indicated by a common practice in some parts of Europe, of adding common salt to stimulate the growth of mangel- wurzel and of mixed meadow grasses, but the manner in which this effect is produced has not been made clear. It has been observed that unrestricted application of chlorides may lead to poisoning of the soil, and agriculturists have been advised specially against the use of potassium chloride as a source of potassium for tobacco, the potato and the sugar beet. Euro- pean investigators have reported a decreased content of starch in the potato tuber as a result of the substitution of this salt for potassium sulphate. 3 Livingston, B. E., and McLean, F. T., "A living climatological instrument." Science, n. s. 43: 362-363. 1916. 218 Chlorine and Plant Growth [416 The investigations here considered in a preliminary way were planned to supplement our knowledge of this subject. They are as yet in early stages of progress, having been begun under the auspices of the Wisconsin Agricultural Experiment Station. It was purposed to measure the responses of various plants, in form and in the weight of plant material produced, to the application of certain chlorides, and to determine any specific results brought about by this application of chlorine, upon the chemical composition of the plants. Greenhouse cultures were grown in nutrient solutions, in pure sand and in Miami silt loam, and field cultures were grown in loam. It may be said of these greehouse cultures, which were partly carried out in the winter, that, while growth is retarded by the decreased light intensities of the winter months, the partial control of climatic and soil conditions in such greenhouse cul- tures assures more reliable comparative results than are usually derived from field plots, with their natural fluctuation of climatic conditions from season to season and of fertility from plot to plot. In the water-culture experiments, in the greenhouse, the plants were grown to maturity, in either Tottingham's or Knop's nutrient solution,1 containing Ca(ISr03)2, KN03, MgS04 and KH2P04, in proper proportions, with a trace of iron as FeP04. The former had a total osmotic concentration value of about 1.75 atmospheres (0.4 per cent, of salts by weight) and the total osmotic value of the latter was about 0.9 atmospheres (0.2 per cent, of salts by weight). In some cases chlorine was introduced by replacing the MgS04 of the 4-salt solution with a molecularly equivalent quantity of MgCl2, in other cases KN03 was replaced by KC1, and in still other cases NaCl was superimposed upon the salts usual- ly present. Replacement of MgS04 by MgCl2 resulted in an increased length of roots, for pea, wheat and clover, amount- Nottingham, W. E., "A quantitative chemical and physiological study of nutrient solutions for plant cultures." Physiol. Res. I : 247-288. 1914. 417] W. E. Tottingham 219 ing to from 100 to 300 per cent. This gain in root length was correlated with somewhat smaller gains in dry weight. With wheat and clover the production of dry weight of tops was depressed by this treatment but the percentage of nitro- gen contained in the dry tops was unaffected. It will be noted that the interpretation of these effects is complicated by the fact that sulphur was absent where chlorine was pres- ent in the solution. Buckwheat was grown in Knop's solution modified by hav- ing KN03 partly or wholly replaced by KC1, thus avoiding the omission of sulphur. Such treatment led to a slightly increased production of stem and root when the replacement was only partial, but complete replacement depressed the root length and the dry weight of roots and leaves, the amount of water lost by transpiration being proportionately decreased. Total replacement of KN03 by-Nad depressed growth more than when KC1 was used and transpirational water loss was more than proportionately decreased. Comparison with the necessary control solutions indicated that this effect is to be considered specific for the NaCl molecule, an observation which adds to the accumulating evidence that molecules must be taken into consideration, and not ions only, in dealing with the relations between the plant and the solutes of a nutrient solution. The conclusion of earlier investigators, that chlorine must be added to the nutrient solution for the complete development of buckwheat, finds no support in the present work. The sand cultures of this study (also in the greenhouse) were conducted on 20-kilogram portions of sand, in open boxes with paraffined inner surfaces. The insoluble salts were incorporated with the dry sand and the others were added in successive portions of solution. The total applica- tion of salts was about 0.25 per cent, of the dry weight of the sand. With mangel-wurzel, an increase of from 40 to 120 per cent, in the dry weight of roots resulted from the applica- tion of KC1 in a complete fertilizer ration, but greater in- 220 Chlorine and Plant Growth [418 crease followed where NaCl was superimposed upon the usual complete ration. For the greenhouse cultures in Miami silt loam, fifteen or twenty kilograms of air-dry soil were employed, in cypress boxes, the salts being added as in the case of the sand cul- tures. The total application of salts approximated from 0.06 to 0.15 per cent, of the dry weight of the soil. The sugar beet produced 50 per cent, more dry substance (root) when chlorine was included with the usual salt ration than when the ration without chlorine was used. The glucose content of the root was increased somewhat, percen- tagely on the basis of dry weight, but the sucrose content was uninfluenced by this treatment. Preliminary experiments with the radish indicate that it is little affected by the chlorine supply, while the growth of the carrot is stimulated and that of the parsnip is depressed as regards content of dry matter and percentage of sugars. Similar experiments with the po- tato (" Triumph " and " Eural New Yorker " varieties) gave the same dry weights of tubers, whether potassium was sup- plied as the chloride or as the sulphate. In the field experiments, sugar beet roots showed an increase of from 10 to 30 per cent., by weight, where NaCl was ap- plied to the soil at the rate of from 260 to 520 pounds per acre, as compared with those of the unfertilized plot. The glucose content was increased, but that of sucrose was unaf- fected by this treatment.' The potato (" Triumph " variety) produced the same yield, both of total and marketable tubers, whether supplied with potassium as KC1 or as K2S04 , in the complete fertilizer ration. The addition of NaCl without other salts depressed the yield. Another experiment with potato ("Eural New Yorker" variety) showed that the starch content and cook- ing qualities of the tuber were the same whether potassium was supplied as KC1 or as K2S04, in the complete fertilizer. Fertilization with NaCl alone gave tubers of lower starch content and poor quality. It thus appears that the depress- 419] W. E. Tottingham 221 ing effect of chlorine, as reported2 for starch content and cooking quality of potato tubers, does not obtain under all conditions of culture, and fails to make itself manifest with the climatic and soil conditions of these experiments. The results outlined above leave the question of the in- fluence of the chlorine ion and chlorides upon plants still in a very complicated and unsatisfactory condition. Perhaps the most valuable general conclusion that can be drawn from a review of all the work so far reported upon this subject, is that the influence here considered appears to be impossible of any general statement. It appears that the effect of chlorine upon any given plant depends upon the nature of the plant, upon the soil conditions (aside from chloride con- tent) and upon the conditions of the surroundings generally classed as climatic. It may be that each particular case of acceleration or retardation of growth processes by chlorine presents a special problem, and that broad generalizations are not to be expected until much progress has been made toward the interpretation of environmental complexes as a whole ; for the present, we are constrained to study these con- ditions piecemeal. It seems that the promise of progress in these very complicated problems of agricultural science lies largely in more complete experimental control of the very numerous conditions that make up the environment of the plant. It is the summed or integrated effects of all of these that is registered by our plants in growth and crop produc- tion. 2 For example, see: Siichting, H., " Ueber die schadigende Wirkung der Kalirohsalze auf die Kartoffel." Lcmdw. Versuchsst. 61: 397- 449. 1905. 222 Salt Proportions [420 A STUDY OF SALT PROPORTIONS IN A NUTRIENT SOLUTION CONTAINING CHLORIDE, AS RELATED TO THE GROWTH OF YOUNG WHEAT PLANTS By S. F. TRELEASE Chlorine has been considered an unnecessary element in the nutrition of most plants,, but it seems to have produced a beneficial influence in certain cases that have been recorded. There is some practical as well as scientific interest in the question thus raised, since potassium chloride is frequently used as an agricultural fertilizer, and the influence of the chlorine thus put into the soil may not be without impor- tance. In the experiments of which this is a preliminary report the chlorine ion was introduced into nutrient solutions that already contained all the essential elements usually ab- sorbed by plant roots. These essential elements (N, S, P, Ca, Mg, K, and Fe) may be supplied to the young wheat plants as a nutrient solution containing the three salts Ca(N03)2, MgS04, and KH2P04, with a trace of iron as FeP04. To ' introduce chlorine, KC1 was added to the list just given, thus making a 4-salt solution. A solu- tion made from these four salts was used by Knop and Nobbe, and Grafe * recommends these same salts as most gen- erally useful. Detmer2 employed one set of proportions of these -four salts, and this solution has been designated by Tot- tingham 3 as Detmer^s solution. In the experiment s consid- ered in this paper the same general methods were used as were 1 Grafe, V. " Ernahrungsphysiologisches Praktikum der hoheren Pflanzen." Berlin, 1914. 2 Detmer, W., " Practical plant physiology." Translated by S. A. Moor. London, 1898. 3 Tottingham, W. E., "A quantitative chemical and physiological study of nutrient solutions for plant cultures." Physiol Res. I : 133-245. 1914. 421] 8. F. Trelease 223 employed by Tottingham and by Shive.4 The total concen- tration of the nutrient solution corresponded to an osmotic pressure of approximately 1.6 atmospheres at 25° C., and the relative proportions of the four component salts were varied in all possible ways, by increments of one-tenth of this total concentration. Eighty-four different solutions were thus in- cluded in each complete set; all of these had approximately the same total osmotic concentration, but no two had the same relative proportions of the four component salts. Six plants were grown in each culture, and the solutions were renewed every four days. The various salt proportions proved to be very different in their ability to produce growth of the young wheat plants. As has been found by other writers, the solution giving the greatest dry yield of tops is not the one giving the greatest yield of roots, and the solution producing the highest dry weight of tops and roots together has still another set of salt proportions. The highest dry yield of tops was obtained with the following partial volume-molecular concentrations of the four main constituent salts: 0.0067M KC1, 0.0138M KH2P04, 0.0047M Ca(N03)2, and 0.0081M MgS04. A trace of iron was, of course, added, as a suspension of ferric phosphate. This highest yield of wheat tops with the 4-salt solution containing chlorine was not higher, however, than was ob- tained, in these experiments, with the best salt proportions, without chlorine, of the Birner and Lucanus (Shive) 3-salt solution and of the Knop (Tottingham) 4-salt solution. If the best salt proportions are used in all three cases these three very different types of solutions give practically the same result. It therefore appears to be impossible to improve the growth of young wheat plants, as this occurs in Shive's and Tottingham's best salt proportions, by the introduction of 4 Shive, J. W., "A three-salt nutrient solution for plants." Amer. Jour. Bot. 2: 157-160. 1915. Idem, "A study of physiological bal- ance in nutrient media." Physiol. Res. I: 327-397. 1915. 224 Salt Proportions [422 chlorine into the solution. Furthermore, the best 4-salt solu- tion with chlorine contains the three essential salts in nearly the same proportions as those in which they occur in Shive's best 3-salt solution, which has the following composition: 0.0180M KH2P04, 0.0052M Ca(N03)2, and 0.0150M MgS04. The main difference in this respect lies in the Mg/Ca quoti- ent; in Shive's best solution this quotient has the value 2.88, and in the best 4-salt solution with chlorine it has the value 1.72. Both are characterized by relatively high proportions of KH2P04, and low proportions of Ca(N03)2, which is rather surprising, since many nutrient solutions heretofore proposed have a relatively high concentration of Ca(N03)2. In gen- eral, the occurrence of the morphological leaf modifications tions recognized as magnesium injury in such series as these (Tottingham, Shive) was not altered by the presence of the chlorine ion in the solution. A marked improvement over Detmer's salt proportions was obtained in the present study. The best solution gave an in- crease in dry weight of tops of 27 per cent, and 20 per cent., respectively, over the yields obtained in two solutions of the present series closely resembling Detmer's in salt proportions. An even more marked improvement over the growth obtained with Detmer's exact proportions is reported by Shive, for his best 3-salt solution, which, as has been mentioned, gave prac- tically the same yield as did the best 4-salt solution used in this study. While it seems impossible to obtain higher top yields of these plants in the 4-salt solution containing chlorine, than in the 3-salt solution without this element, it should nevertheless be remarked that the presence of chlorine may diminish to some extent the retarding effect produced by the three salts of the essential elements when these are not in the best proportions. Thus, if we start with an unbalanced. 3-salt solution, a proper addition of chlorine may sometimes accelerate the growth of the plants. The addition of a non-essential element may im- prove the physiological properties of a solution containing the essential elements in improper proportions. 423] 8. F. T release 225 Perhaps the main result of this study is, in general, that no matter whether we employ (1) the three salts KH2P04, Ca(X03)2, and MgS04, (2) the four salts KH2P04, Ca(X03)2, MgS04, and KX03, or (3) the four-salts KH2P04, Ca('N03)2, MgS04, and KC1, if we use the best proportion* of the salts in each case we may expect to obtain about the same growth. This generalization has an important bearing on the whole problem of physiological balance in nutrient solutions and furnishes what may be important suggestions bearing on our general conceptions of conditional control and conditional optima for plant activities. THE RELATION OF THE CONCENTRATION OF THE NUTRIENT SOLUTION TO THE GROWTH OF YOUNG WHEAT PLANTS IN WATER-CULTURES By S. F. TRELEASE In these experiments the salt proportions were the same in all the different solutions of each series, but the solutions differed from each other in total concentration. Three series of cultures, all carried out at the same time, are considered, each series including a concentration range of from 0.5 to 7.0 atmospheres. A different set of salt proportions was used in each series. Six plants were grown in each culture and the cultures were in duplicate, upon a rotating table. The ex- periment lasted for 32 days, from January 23 to February 24, 1917, the solutions being renewed every 4 days. In the first series the nutrient solutions contained the 4 salts KH2P04, MgS04, KC1, and Ca(X03)2 in the follow- ing relative molecular proportions: 1.000, 0.587, 0.485, 0.341. The average dry weight of tops and the average total water loss by transpiration, for six plants, are shown in the follow- ing table, whch also shows the total concentration employed in all three series. Concentration of Nutrient Solutions [424 Concentration., Dry Weight, Tops. Transpiration. atm. grants^ GO. 0.5 0.926 651 1.0 0.947 618 1.6 ' 1.152 646 2.5 1.117 554 3.5 1.030 468 4.5 0.904 386 5.5 0.821 311 7.01 0.769 246 For this particular set of salt proportions the maximum yield of tops was obtained when the nutrient solution had a total osmotic concentration of 1.6 atm. With lower con- centrations growth was considerably less, as is also true, and to a greater degree, with concentrations above the optimum. Between the concentrations 1.6 and 7.0 atm. the dry weight of tops is approximately a linear function of the concentra- tion, the dry weight decreasing as the concentration increases. The transpiration values show the same general relation to- the concentration, except that below 1.6 atm. the decrease is less clearly shown; in fact, with a concentration of 0.5 atm.. the transpiration is slightly higher than with 1.6 atm. In the second series the culture solutions were the same as those just described, except that KC1 was not included, In these cultures the relations of dry weight and transpira- tion, to total concentration, were essentially the same as in the cultures of the first series, with KC1. In the third series the salts used were the same as in the first, but in different relative molecular proportions, as fol- lows : 1.000, 1.155, 7.282, 0.699. The relation between trans- piration and concentration was the same as in the first series, but in this case there was a perfectly definite maximum of transpiration at 1.6 atm. For production of dry weight of tops, however, while the general relation to concentration was the same as in the first two series, the optimum concentra- tion was 4.5 instead of 1.6 atm. The interesting features of these results may be summar- ized as follows: (1) Transpiration and dry weight showed an approximately linear relation to the concentration of the medium above the optimum, these decreasing with an increase in concentration. (2) The optimum concentration for dry 425] 8. F. Trelease and E. E. Free weight of tops was altered from 1.6 atm. to 4.5 atm. by chang- ing the proportions of the four salts used in the first and third series. (3) With the salt proportions of the three other salts used in the first series, the omission of KC1 did not alter the relation between growth and concentration. THE EFFECT OF RENEWAL OF CULTURE SOLUTIONS ON THE GROWTH OF YOUNG WHEAT PLANTS IN WATER-CULTURES By S. P. TKELEASE and E. E. FREE One of the practical problems. in work with water-cultures is that of the frequency with which the culture solution must be renewed in order to obtain the best results. This note re- ports experiments in this connection on the growth of young wheat plants in the nutrient solution found by Shive x to be best for the production of dry weight of tops In wheat. 'Mie culture jars had a capacity of 250 cc. Six plants were grown in each jar and each culture was in triplicate. The volume of the culture solution was made up to normal by the addition of distilled water every 4 days or oftener. The details of the technique were the same as employed by SLive - Ail cultures ran 41 days, from January 6, to February 16, 1916. The results are given in the following table, in the form of dry weights of tops produced, each weight being the average of the three parallel cultures, Dry Weight, grams. Changed daily . . . , , ..... „ ..... „ .„„..».... 1.243 Changed every 3 days 1.012 Changed every week 1.020 Changed after 1 week, then every 3 days . , 0.995 Changed every 2 weeks ..»».».*. ,.;. 0.780 Changed after 2 weeks, then every 3 days 1.131 Changed after 2 weeks, then every week v ...... 0.969 Changed after 1 month 0.654 Not changed at all 0.621 1 Shive, J. W., "A study of the physiological balance in nutrient media." PhysioL Res. 1 : 327-397. 1915. 228 Renewal of Culture Solutions [42 S It is apparent that the yield is better the more frequently the solution is changed. If, after an initial period, the fre- quency of changing is increased the yield is improved. It is important, practically, that there is small difference between the cultures changed every 3 days and those changed every week. Daily change produces substantial improvement. Al- lowing the solution to remain unchanged for so long as 2 weeks is markedly injurious. The above cultures were grown on a rotating table. An additional set was grown in the same greenhouse at the same time but not on the rotating table. The results follow: Dry Weight. grams. Continuous flow of solution through culture jar at rate of about 1 liter daily 1.678 Changed every 3 days 1.222 Not changed at all 0.666 This experiment is not strictly comparable with the one done •>n the rotating table, but it seems probable that continuous flow of the solution must be regarded as more beneficial even than daily change. Parallel with the experiments on the rotating table, one set of three cultures was treated by removing the solution weekly Lnd shaking it with bone black. The solution was then filtered and restored to the culture jars. These cul- tures gave an average yield of 0.780 gram, as compared with 0.621 gram for the unchanged culture not treated with bone, black. Evidently the bone black treatment improved the solution slightly but did not correct in important degree the harmful effects of infrequent changing. It was noticed inci- dentally that the magnesium injury that is characteristic of this solution, for wheat, appeared more frequently and se- verely when the changing was frequent than when it was not. The color of the plants was greener in the more fre- quently changed solutions. 14 DAY USE RETURN TO DESK FROM WHICH BORROWED EARTH SCIENCES LIBRARY This book is due on the last date stamped below, or on the date to which renewed. Renewed books are subject to immediate recall. LD 21-40m-5,'65 (F4308slO)476 General Library University of California Berkeley I