OKDOVICIAN TRILOBITES OF THE FAMILYTELEPHIDAE AND CONCERNEDSTRATIGRAPHIC CORRELATIONSBy E. O. Ulrich ^Associate in Paleontology, United States 'National MuseumINTRODUCTIONPrevious work on Telephus.?The genus Telephus was establishedby Barrande ^ for the Bohemian species T. fractus. Of this speciesBarrande knew only the cranidium and the pygidium, and of theformer he mistook the long palpebral bands for the border of thecheeks. Consequently he expressed himself as quite unable to placethe genus in his classification of the trilobites. Two years laterAngelin ^ described three species from Ordovician formations inNorway and Sweden but added nothing toward fixing the systematicposition of Telephus. Some years later Billings * recognized thegenus on the west side of the Atlantic but, like his predecessors inthe field, failed to add anything of more than specific importanceto what had been known before. Many years later Reed ? describeda cranidium from the Girvan District in Scotland that he regardedas specifically identifiable with the Bohemian T. fractus. In 1909he described a new species, T. hihernicus, from an Ordovician forma-tion in Ireland, and five years later ^ in the supplement to his mono- 1 Tbig is one of many papers on trilobites for wbicb tbe author has worked out thefacts in the past 25 and more years but laclted the time to complete the manniscriptaand illustrations. Most of these unpublished works endeavored to present what wasknown at the time of the species of a particular genus or family. At the earnest andrepeated solicitation of friends it is now planned to bring to date and publish as manyof these old manuscripts as possible without interfering too greatly with the paramountduty of completing the long promised monographs on the Ozarkian and Canadian systems.The present installment has become possible mainly through the gratefully accepted aid ofDr. C. E. Resser, who made most of the photographs and assisted otherwise Id promotingthe effort. The originally brief discussion of the stratigraphy of the beds in whichspecies of Telephus occur has been greatly expanded and completely rewritten, so thatin the writer's opinion it has become the more important part of the paper. ' Barrande, Joachim, 1852, Syst. Sil. du centre Boheme, vol. 1, p. 890.3 Angelin, N. P., 1854, Palaeontologia Scandinavica, p. 91. * Billings, E., 1805, Paleozoic fossils: Canada Geol. Survey, vol. 1, p. 291.5 Reed, F. R. Cowper, 1903, Paleontogr. Soc, p. 44.? Reed, F. R. Cowper, 1909, Geol. Soc. Londom Quart. Journ., vol. 65, p. 149 ; 1913.Paleontogr. Soc, vol. 67, p. 16.No. 2818.?Proceedings U. S. National Museum, Vol. 76, Art. 2164441?29 1 1 2 PROCEEDINGS OF THE NATIONAL MUSEUM vol.76graph he described another new and very different species {T.salteri)^ from the Balclatchie group of Scotland.Except Barrande none of the mentioned authors even discussedthe generic relations of Telephus. The first to give us anything likea true estimate of these relations was Hadding/ who, in a paperspecially devoted to the species of Telej^hus known in 1913, suppliedmuch desired information concerning the palpebral lobes, eyes, andfree cheeks. On the basis of these new data he endeavored to showthe previously unsuspected relations of the genus to the Aeglinidae,on the one hand, and the Remopleuridae, on the other. Still, he foundsufficient differences to convince him that Telephucolumn is taken from the Mississippi and Ohio Valleys and the Ordo-vician part from the southern Appalachian Valley and centralTennessee except the Buffalo River series, which is best developed innorthern Arkansas. As will be observed, only the Chazyan part ofthe column is divided into units of formational rank, this being thepart that is mainly concerned with questions discussed in this paper.Besides, and however detailed the correlations of American Ordo-vician and Silurian formations may be, we can as yet do no betterin correlating European formations of these periods than to suggestmore or less indefinitely located positions for them in one or anotherof our series or groups.Though this column is called a time scale it should not be as-sumed that even its OrJovician part accounts for all of the timeincluded in this period. In fact it accounts only for those subordi-nate parts of the accesible depositional record that was laid down inAmerican epicontinental basins; and of these only those whosesequential relations have been established. As most if not all of thenamed minor units of the scale are separated from each other bystratigraphic breaks of undetermined time significance it followsthat these depositionally unrecorded intervals, at least, are not ac-counted for. Doubtless some of these intervals are represented,probably only in part, by deposits in other areas of the North Ameri-can Continent, but these could not be used in constructing the scalebecause their relations to those found in the southern Appalachian,Ohio, and Mississippi Valleys are insufficiently understood; and weknow even less about the correlation of the European and AmericanOrdovician deposits. Finally, as suggested previously (p. 54), theremay have been times when the continents on both sides of theAtlantic were so elevated above sea level that the basins in whichOrdovician marine deposits are now accessible were completelydrained. Obviously such times also are not accounted for. It fol-lows, then, the " generalized time scale " of the chart is incompleteto these several extents and pretends to be nothing more than atemporary standard for comparison.Perhaps I should call attention also to the fact that the correla-tions with formations in European countries differ considerably fromthose given in a similar table published by me only three years ago.^'However, the changes occur mainly in the lower two-thirds of whatI think should be included in the Ordovician system, which, as manyknow, I define differently from the original and even yet prevailingconception of that term. Briefly stated, my definition of the Ordo-vician system is based primarily on diastrophic criteria that in my ^ Relative values of criteria used in drawing the Ordovician-Silurian boundary : Geol.See. America Bull., vol. 37, p. 329, 1926. AET. 21 OEDOVICIAN TEILOBITES ULRICH 75opinion demand elimination of formations from both the top andthe base of this system as originally defined by Lapworth. Theevidence on which these views are based, particularly those facts thatconcern the upper boundary of the system, are briefly discussed aboveand much more fully in the just cited Ordovician-Silurian boundarypaper published in 1926. The evidence relating to the lower bound-ary was already rather thoroughly pointed out in my " Revision ofthe Paleozoic Systems " published in 1911, but some of it will againbe presented with local details and in generally amplified form ina work on the Paleozoic formations in Oklahoma due to appearbefore the close of the present year.Genei'alized comments on other colvAiins of table.?Regarding thechanges in correlating European and American formations thatcontinued study of the extremely difficult problems in the past threeyears has indicated, I make no apologies. The innovations are pre-sented as suggestions and not as final conclusions. They are based ontheoretical considerations and reasonable inferences and probabilitiesthat are not yet susceptible of satisfactory proof?and may never be.Still, they seem as well worth trying out as other not very dissimilarsuggestions were that have been presented in the past twenty yearsand whose merits have in the meantime been fully established.The probable bearing of the postulated differential character andslowness of the vertical movements of the surface of the earth onthe correlation of formations in the more or less widely separatedgeological provinces covered by the table is indicated by the inter-mediate placing of many of the names of the formations in theseveral columns. However, I am not at all certain that the Europeanand even some of the North American formations actually belong inthe positions assigned to them in the chart. Any of these maybelong a notch or two higher or lower in the time scale than is indi-cated by the present status and probable trend of the organic andphysical evidence studied to date. But I do feel satisfied that thetentative arrangement presented in the chart is a nearer approxima-tion to the facts in the several cases than any previous effort hasattained.Of extreme and commanding importance in working out thesequence of events and the great length of time involved in thegeologic history of the Lower Paleozic ages is the indisputable factthat so far as Icnown the least incomplete depositional record of theseages occurs in America. I venture to say further that, so far as thestratigraphic correlation of the marine deposits of these ages in theseveral largely supplementing provinces in North America is con-cerned, the record of the frequently shifting Paleozoic epicontinentalseas is also better understood than is the more epitomized and onthe whole much less completely developed record found m European 76 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 7G countries. Let me not be misunderstood here. I have no wish todeny that the local developments of the fossil faunas in Bohemia,the Baltic region, Sweden, Norway, and, perhaps in less degree,also in Great Britain have been more thoroughly exploited and theresults of their study published than in America. This admissionapplies particularly to the often extremely fossiliferous pre-Cincin-natian formations in the Appalachian Valley, central Kentucky,middle Tennessee, Missouri, Arkansas, and Oklahoma, but it shouldnot be interpreted as implying that we have neglected to collect andstudy these faunas. We have failed only in this, that publicationof the results has lagged far behind our information regarding theircharacter and stratigraphic significance.Having the maximum development of lower and middle Paleo-zoic marine deposits and also a more detailed conception of thesequence of geologic events recorded in and by them it seems notonly natural but also desirable that the American record should bethe standard for world-wide comparison. If this were concededthen even very elementary comparisons of classical north EuropeanEopaleozoic sections with American sections of the same era wouldconvince the observer that the composite European sequence is notonly inferior in completeness but also that the stratigraphic hiatusesin it are of greater chronologic significance than has been recog-nized by those most familiar with the fossil contents of the concerneddeposits.EXPLANATORY NOTES ON THE FORMATIONS IN THE SEVERAL COLUMNSOklaho77ia.?This sequence of formations is found on the flanksof the Arbuckle and Wichita Mountains in the south-central partof the State. At the base and resting on pre-Cambrian granite andporphyry are the two Upper Cambrian formations. Over thesecomes the great series of clolomitic and pure limestones to which Taffapplied the term Arbuckle limestone, but which is subdivided into anumber of formations in a work nearing completion. The lowerpart is divided into six oscillating Ozarkian formations, two of whichare confined to the Arbuckle area, two to the Wichita area, and twoare datum planes common to both. The greater upper part of theArbuckle is of Canadian age and divided into three formations, thelimestones of the Lower and Middle Canadian being provisionallyunited in one and the lithologically more varied beds of UpperCanadian age into two.The succeeding Simpson of Taff's classification comprises sevenvariously distributed and interfingering formations. Of these onlythe topmost (Bromide) has been previously named. The othersare newlv named as in the table. Each begins with a sandstone of ART. 21 ORDOVICIAN TRILOBJTES-?ULRICH 77from a few feet to more than 100 feet in thickness and is distin-guished from the others by a complete change in the character andin most cases also in the derivation of its famia. The first (Joins)and second (Oil Creek) derived their faunas from the west, whereasthe third (McLish) contains species of the Appalachian Lenoirfauna, which therefore are regarded as indicating an invasion fromthe east at this time. The fourth (Falls) contains species foundelsewhere in America only in Nevada and western Texas, whichis interpreted as showing that the sources of the invasion wasagain in the Pacific. The fauna of the fifth formation (TulipCreek) compares closely only with Stones River faunas of Tennessee,and is therefore held to be an Oklahoma recurrence of that southernfauna during Blount time that did not reach central Tennessee. Thefauna of the sixth (Criner) formation again differs radically fromthat of the next underlying formation. Genetically comparable fos-sils occur only to the east in Blount and Chambersburg formations.Finally, the faunules of the succeeding Bromide formation are essen-tially the same as those found in the Black River and early Trentonformations in Iowa and Minnesota, whose northern origin has longbeen recognized.The position of the succeeding Viola limestone in the time scalecan not as yet be fixed with precision. We know, however, that itfollows the Trenton, so that it must fall somewhere in the Cincimia-tian or into the hiatus that everywhere separates that series from theRichmond. Its graptolites compare rather well with the Upper Hart-fell of Britain, and its trilobites, among which species of Crypto-lithus predominate, agree better with British Caradoc forms thanwith any other trilobite fauna known. The Tyner and Sylvan areearly Silurian and clearly correspond to parts of the Maquoketaof Iowa. Above these come thin limestones of Upper Medinan andClinton ages that are 'better developed in eastern Missouri andnorthern Arkansas.Mississippi Valley.?At the base of this column the Ozarkian andCanadian depositional record between the top of the Upper Cam-brian and the base of the Buffalo River series in Missouri is broadlyindicated on the left side and the inferior record of the same sys-tems in Wisconsin on the right side. Throughout the valley northof Tennessee limestone of Black River age rests on the BuffaloRiver series. Evidently, about 10,000 feet of deposits?more thanhalf of this thickness consisting of limestone?that occur in eastTennessee and other parts of the Appalachian Valley are wantingin States bordering the Mississippi. Generalized, but in most casesvery detailed correlations of the formations of the Mohawkian seriesin this colimm with those in Oklahoma, Kentucky, Tennessee, or 78 PROCEEDINGS OF THE NATIONAL MUSEUM vol.76New York are too well established to require further comment here.But the case of the succeeding Galena dolomite is quite different.The more I study this formation the less I am satisfied with therej)uted Trenton age of its typical parts. Provisionally, and mainlyto emphasize my doubts regarding its precise age, I have moved itup in the scale opposite the Cincinnatian. At present it seems thatthe dolomite in the vicinity of Dubuque, Iowa, is certainly a youngerformation than the Kimmswick limestone of Missouri, with whichit has been correlated by other geologists ; and some of the beds thathave been assigned to the Galena in Wisconsin are older than Tren-ton, whereas others are younger than the Maysville of Ohio. Aspecial paper needs to be written about the Galena.Ohio Valley.?No deposits of the Buffalo River series occur inTennessee, but in Kentucky a calcareous phase of one of its sand-stones is found in deep wells as far south, at least, as Lexington.Over it are at least three of the limestone formations that make upthe Stones River group in central Tennessee. This group attainsgreater thickness in the western third of the Appalachian Valley,but pinches out completely and rapidly to the west of the Cincinnatianticline. The absence of the southern Appalachian Mosheim lime-stone in both Kentucky and central Tennessee is a notable featureof this column. Formerly the Mosheim was believed to be includedin, or to underlie, the horizon of the Murfreesboro limestone. How-ever, in August of 1928 Doctor Butts and I studied a completelyexposed section in the eastern part of Lee County, Va., inwhich both formations occur in typical development and in whichthe Mosheim overlies the Murfreesboro. In the next Orclovicianbelt to the southeast the Murfreesboro is absent and the Mosheimas usual in contact with the eroded top of the Canadian system.In Kentucky and Middle Tennessee the Lowville limestone of theBlack River group is in contact with the tdp of the Stones River,the great Blount group and also the succeeding Little Oak limestoneof east Tennessee and Alabama being absent. The Black River alsolacks some hundreds of feet of limestone beds that are present inthe section of Mulberry Valley north of Sneedville, Tenn. TheTrenton, however, is more fully represented, though its beds andfossils are very different from the beds and fossils of similar age inNew York and Pennsylvania. The Prosser of Minnesota, beingclosely akin to the New York Trenton, differs in like manner fromthe Trenton of Kentucky and Tennessee. Evidently the New York-Minnesota Trenton faunas invaded the continent from a differentsource than that which supplied the life of Trenton formations inKentucky and Tennessee. As these northern and southern facies ofthe Trenton have not yet been found interfingering or mingling ART. 21 ORDOVICIAN TRILOBITES ULRICH 79with each other we can not say what, if any, differences in age theobserved differences in their respective faunal contents may indicate.In essentials the Cincinnatian series is much the same on the Cincin-nati axis as in New York. However, both regions exhibit faunal andlithologic details that distinguish the sequence in one from that ofthe other. Most of these differences have been discussed by Ruede-mann,^" but to appreciate their full significance many as yet unpub-lished facts that have been disclosed by study of this part of thecolumn in Pennsylvania and Virginia must be taken into considera-tion. Obviously the subject is too intricate to warrant anythingmore than its mere mention on this occasion. The case is similarwith respect to the Medinan formations concerning which so muchwholly unpublished, or only partly published, information is inhand that adequate discussion of its problems constitutes the mate-rial of another of my uncompleted papers.Southern and middle Appalachian region.?It would require atleast 10 columns to present in correct and readily understandablemanner the variations in sequence of the Ozarkian, Canadian, Ordo-vician, and later formations that are known to occur in the Appa-lachian Valley region from central Pennsylvania to central Alabama.The sections would be in sets of twos and threes taken at intervalsacross the strike of the valley troughs and two, or better three, setstaken at points along the strike. But, however, interesting and il-luminating such a series of columns would be, my present purposeis particularly concerned only with the oscillations in the valleytroughs that are indicated by the distribution of the Chazyan de-posits and faunas. Accordingly this column may be characterizedas a rather unsatisfactory composite presentation of the frequentlyvarying sequence of formations in the southern and middle stretchesof the Appalachian geosyncline.Absence of the Buffalo River series emphasizes the chronologicsignificance of the break between the Ordovician and Canadian sys-tems. As a rule the value of this hiatus in the valley south ofStaunton, Va., is further increased by absence of the Murfreesborowhich, as above mentioned, has been observed to wedge in from thewest in the eastern part of Lee County, Va. Elsewhere in the valleyOrdovician sedimentation usually begins with greatly varyingthicknesses of Mosheim limestone, but there are many places in Vir-ginia, Tennessee, and Alabama at which not only the Mosheim butalso most or all of the Lenoir and the Holton are missing, so that theAthens shale is in contact with the Canadian.The facts just mentioned and the many similar variations thatoccur at the contact between the Canadian and Ordovician wherever ?"Ruedemann, Rudolf, The Utica and Lorraine formations of New York, New YorkState Mus. Bull. No. 258, 1925. 80 PROCEEDINGS OF THE NATIONAL MUSEUM vol.76 rocks of these ages were deposited in North America constitute thephysical part of the evidence on which I base my claim that this isone of the most important breaks in the Paleozoic column. Andthe organic part of the evidence is no less impressive. The strati-graphic significance of the gap between the two systems is narrowedto its observed minimum in Lawrence Count}'^, Ark., where it isreduced (1) by the insertion of tAvo formations at the top of the Cana-dian that are not present elsewhere in the Ozark region and (2) bythe downward expansion of the Buffalo River series which attains itsmaximum development in Newton County about 100 miles west ofLawrence County. To the north and west of Newton County thechronologic significance of the gap then increases rapidly to places atwhich the Powell or even the Cotter?respectively, the third andfourth formations beneath the top of the Upper Canadian?is direct-ly succeeded by deposits of Mississippian age. Surely, conclusionsbased on such data deserve respect and more general acceptance.Proceeding with the explanatory notes on the middle and southernAppalachian column, the conventions employed indicate that theMurfreesboro, Pierce, and Lebanon formations lap out eastwardlyand the Mosheim pinches out in the opposite direction. The Lenoir,however, seems to extend completely across the valley in southwesternVirginia and is supposed to have attained weak connection withthe Ridley in central Tennessee and Kentucky. But no such alter-nation of Atlantic and southern invasions of the Appalachian geo-syncline is indicated by the formations of the Blount group, allof which are confined to the eastern half or two-thirds of the valley ; and the succeeding Little Oak is found only in one or two of theearstern troughs in Alabama.In Black River time that excellent datum plane?the Lowvilleand its red sandy facies, the Moccasin or " Bays "?began a newseries of alternating tiltings in which the southern invasions pre-dominated. The Lowville, itself, extends from the Mississippi Riverfar eastward to East Tennessee and in places there overlaps theChazyan formations quite to the edge of the overthrusted LowerCambrian formations. To the north in the middle stretch of thevalley the Chambersburg, which invaded from the northeast andfollows the Lowville, extends southwardly from Pennsylvania toLexington, Va., beyond which place it has not been recognized.Usually, and perhaps throughout its extent, the Chambersburg issucceeded by the shaly lower Martinsburg facies of the Trenton.In the belt just within the Avestern side of the valley, in whichit is represented by its typical limestone facies, the Trenton issucceeded by either the Reedsburg shale phase of the Cincinnatianor by distinguishable Eden and Maysville formations. In the ART. 21 OEDOVICIAN TRILOBITES ULEICH 81southwestern corner of Virginia the Maysville is represented onlyby its lower formation (the Fairview). At Cumberland Gap andto the south in Sequatchie Valley this is succeeded directly by theSequatchie formation, which is the southern Appalachian marineequivalent of the Richmond. Following the strike northeastwardfrom Cumberland Gap up the Powell Valley the Sequatchie losesin thickness and probably disappears entirely before reaching BigStone Gap, so that younger Medinan and finally Clinton beds are incontact with the Fairview.In the Clinch River Valley belts to the southeast the Richmondis represented by the nonfossiliferous red, probably continentaldeposit known as the Juniata sandstone. This extends continuouslyfrom northeastern Tennessee to central Pennsylvania and thenceunder cover to western New York, where it is known as the Queens-ton shale or sandstone. In central Pennsylvania and New Yorkthe Juniata and Queenston are underlain by the Oswego (" GrayMedina"), sandstone, also mainly a continental deposit, that isbelieved to correspond in age with the highly fossiliferous McMillanformation of the Cincinnati section. The Brassfield and Whiteoakrepresent southern marine invasions that reached the AppalachianValley only south of Virginia and also only in places that hadbeen occupied previously by the Sequatchie. The Tuscarora andthe at least partly equivalent Clinch sandstone rest on the Juniataand like it are unfossiliferous and regarded as continental depositsthat in their case correspond in age to the fossiliferous Brassfieldand Whiteoak formations which occur in belts to the west of ClinchMountain. No indication of important movements having occurredduring the transition from the Lower to the Upper Medina or, inother words, between the Richmond and Alexandria (or "Albion"),epochs has been observed in the Appalachian Valley region betweenthe Adirondacks and central Alabama.North Appalachian Valley.?This column pertains mainly to theOrdovician and early Silurian deposits in Newfoundland, Anticosti,and the St. Lawrence and Champlain valleys. The Canadian andOzarkian formations in this region are not referred to except tostate my opinion that zones F, G, and H of the Newfoundland sec-tion are of Upper and perhaps Middle Canadian age and that theOzarkian is represented at Philipsburg, Quebec, by beds of the Upperand the Lower series. To this I may add the further opinion thatthe sections in northern Vermont and at Philipsburg, Quebec, do notinclude deposits of Middle Ozarkian age; and in the latter sectiononly one of the Champlain Valley Beekmantown formations, namely,the Cassin limestone, has been recognized. Regarding the Chazyan64441?29 6 82 PROCEEDINGS OF THE NATIONAL MUSEUM vol.76formations further comment is regarded as unnecessary because thestatement and braces on the chart give a sufficiently clear illustra-tion of my belief that the formations and zones of this age in thenortheastern St, Lawrence extension of the Appalachian Valleyregion fall between rather than opposite the formations in itssouthern part (see pp. 57 and 76). The Lowville and the in partsandy shales (Snake Hill, Canajoharie, and Schenectady), that arecorrelated with the lower half of the Martinsburg occur in theChamplain and lower Mohawk valleys. The Richmond, Alexandria,and Clinton formations at the top of the column occur on the Islandof Anticosti.In 1923 ^^ I referred the Gun River formation of the Anticostisection to the Lower Clinton. This was done on unimpeachablefossil evidence found on slabs that had been sent by Doctor Twen-hofel to Doctor Bassler for report on the Bryozoa and Ostracodaand which were marked as collected by the former in the Gun Riverformation. Twenholfel's final report ^* on the " Geology of Anti-costi Island " now being at hand it appears that these supposed GunRiver fossils were either incorrectly labeled or the beds from whichthey were collected were subsequently assigned to the lower part ofthe Jupiter River formation. Despite the elimination of these un-questionably Lower Clinton fossils review of the revised lists ofGun River fossils in Twenhofel's last report still leaves considerableand perhaps sufficient ground for my 1923 view. At present, there-fore, I will modify it only so far as to say that the Gun River ismainly and perhaps entirely of Clinton age. The underlying BecsieI regard as either contemporary with the Brassfield or slightly olderbut not as old as the Edgewood of Missouri and southern Illinois.The Ellis Bay falls somewhere between two or opposite one of thethree or four Upper Medinan formations that underlie the Brass-field in the Mississippi Valley. Partly to indicate the uncertaintyof these correlations?but mainly because alternate arrangement ofthe units accords with my view that, as a rtile, formations in theeastern part of the St. Lawrence trough are not precisely correlat-able with those of nearest dates in interior basins?the Gun Riveris placed midway between the Clinton and the Brassfield and theBecsie just beneath the space alotted to the Brassfield.Britain.?This column begins with the Tremadoc, which I believebelongs rather low in the Canadian and may be correlated in generalwith the Dictyonema fldbelliformis zone of the Bretonian of Matthewand the Schaghticoke shale of eastern New York. The Shineton "Ulrich, E. O., and Bassler, R. S., American Silurian formations: Maryland Geol. Sur-vey, Silurian vol., pp. 368-372.??Geol. Survey Canada, Mem. 154, 1928 (tiUe page reads 1927). ART. 21 OBDOVICIAN TKILOBITES ULRICH 83seems to be nearly of the same age, but the Arenig, which succeeds it,is more confidently assigned to the Upper Canadian. The Durnesson the northwest coast of Scotland belongs to another province andis closely related to American formations of the Canadian. Itslower part contains the Lecanospira fauna, which characterizes theMiddle Canadian in the Appalachian Valley, -Missouri, Oklahoma,Texas, and many areas in western North America ; and its upper halfcontains the similarly distributed Geratopea fauna which is equallycharacteristic of the Upper Canadian. So far as known the Durnesscomprises only Middle and Upper Canadian. In this respect itagrees with the Canadian as developed in Missouri and northernArkansas and in Alabama, Tennessee, and most of the valley in Vir-ginia. The lower Canadian series is differently and less widely dis-tributed. In the Mohawk Valley in New York and probably alsoin New Jersey no higher beds of the system are present. However,in the vicinity of Ticonderoga, N. Y., again in central and southernPennsylvania, and thence southward through Maryland to some un-known point in northern Virginia, and finally in the Arbuckle andWichita uplifts in Oklahoma the Middle and Upper Canadian lime-stones are underlain by varying thicknesses of the lower series. Theobserved maximum of 3,000 feet (Jonesboro limestone), is attainedat Limestone, Tenn. The section at this place is further unusualbecause the excellently exposed Jonesboro limestone is directly suc-ceeded by a 50-foot development of Lenoir limestone and this byAthens Shale.A small variety of Dldymograptus hifidus is said to occur in thebasal part of the Llandeilo. In America we have two small varietiesof this graptolite, and both occur near the boundary between theCanadian and the Ordovician. One occurs in Lawrence County,Arkansas, near the top of the Black Rock limestone, which is theyoungest of the Canadian formations in the Mississippi Valley. Theother occurs near the base of the Joins limestone with which theOrdovician (Ulrich), begins in Oklahoma. Which of the two ismost like the Llandeilo variety remains to be determined.SUPPLEMENTARY NOTES ON THE GIRVAN DISTRICT SECTIONIn deference to my belief that the Ordovician formations of theGirvan District named in the middle part of this column are withtwo exceptions not strictly correlatable with Appalachian formations,I have placed the names of most of them in midway positions withrespect to those of the latter. The following notes on the Girvansection, which became possible only since my visit to Girvan thepast summer with Prof. O. T. Jones and other British and Americanmembers of the Princeton University Summer School of Geology in 84 PROCEEDINGS OF THE NATIONAL MUSEUM VOL. TOBritain, seem more appropriate here than in preceding parts of thef)aper.The Stinchar limestone of the type locality rests on the Kirklandconglomerate which presumably represents the clastic initial depositof the Stinchar stage of submergence. The conglomerate rests ? doubtless unconformably?on Radiolarian cherts and black shalessaid to contain Arenig (UiDper Canadian) graptolites.Lithologically and to notable extent also faunally the upper partof the typical Stinchar is strikingly like the upper part of theLenoir limestone in Tennessee and Virginia, a fact observed andcommented on by Professor Jones when we studied the naturaloutcrop of the formation and its unweathered appearance in thequarry. The lower part of the Stinchar also resembles lower bedsof the Lenoir, so that as a whole the formation strongly suggestsapproximate equivalence to the mentioned American formation.The limestone exposed in the quarry at Craighead, about 3miles east of Girvan and which supplied many of the distinctive fos-sils of the district, is generally classed as Stinchar limestone. Butthis correlation is almost certainly in error. The supposed ageequivalence of the Craighead and Stinchar limestones evidently arosefrom the fact that Lapworth in describing the section at Craigheadregarded the limestone in the quarry as being succeeded normallyby a shale formation from which he had collected Glenkiln grap-tolites. In other words, Lapworth decided that the sequence atCraighead is the same as on the Stinchar where shale of Glenkilnage lies in normal sequence on the typical Stinchar limestone.I doubt that any of our 1929 party left the Craighead quarry uncon-vinced that the well displayed superposition of the shale on thelimestone in the quarry is due to overthrust faulting and not originaldeposition?namely, that the Glenkiln shale has been thrust over amuch younger, probably Medinan, limestone formation. All agi"eed,too, that the Craighead limestone is very different in lithic and faunalcharacters from the previously investigated typical Stinchar. So faras I have been able to learn not a single species of fossils is commonto the limestones of the Craighead and the Stinchar. Besides, thepublished fauna of the typical Stinchar makes but a short list,whereas a total of at least 85 species has been collected from theCraighead quarry limestone.The relations of the Craighead limestone fauna to that of theBalclatchie group is much closer. In fact, of the 85 Craighead fossilsReed's lists of Girvan fossils described in his monographs indicatethat 37 of the trilobites and brachiopods are present also in tho. Bal-clatchie beds. It should be noted, however, that with very few excep-tions Reed expresses some doubt regarding the actual specific identity AHT. 21 ORDOVICIAN TEILOBITES?ULEICH 85of the species listed as common to the Balclatchie and the Craigheadlimestone. The latter, of course, is referred by him to the Stinchar,and its fossils are listed under that heading. Comparison of Keed'slists therefore indicates a much greater similarity in the fossil con-tents of the Stinchar and Balclatchie formations than is warrantedby the facts. Indeed, I am confident that revision and correction ofthe lists will show that not a single species of the true Stinchar passesupward into the Balclatchie.Evidently the Craighead fauna comprises a large proportion?ap-proximately 40 per cent?of derivatives of Balclatchie species. Manyof these may be very close relatives that as preserved are not readilydistinguishable from their ancestor^. However, experience showsthat with good and abundant material and closer attention to detailsof structure these difficulties of discrimination will become much lessand in most instances quite ordinary. Moreover, this similarity offaunas is precisely what should be expected and what we are experi-encing over and over again in comparing faunas that invaded fromthe same sea at more or less widely different times.The Balclatchie, despite the mentioned faunal similarity to theCraighead, is unquestionably Ordovician in age. The only questionis how far beneath the top of the American development of the systemdoes it belong? In my opinion the Balclatchie, together with theBenan conglomerate which I regard as the initial deposit of its time,is not older than the Tellico of east Tennessee and both most prob-ably are entirely post-Blount age. The Ardwell Group then may boplaced near the middle of the Black River Group. I have recentlyprocured what seems good faunal and physical evidence for this viewin northern Virginia, but more field work and further study of thefossil collections is desired before I shall feel prepared to discuss theproblem.In the case of the Whitehouse group positive faunal evidencetending to show its stratigi-aphic relations to Appalachian forma-tions is as yet scant and far from conclusive. However, taking intoaccount all of the faunal similarities now suposed to have any bear-ing on the question together with probabilities suggested by the lithiccharacter of underlying and overlying formations in Girvan, we maybe safe in placing the Whitehouse within the limits of the Trentongroup or perhaps at the contact of the Trenton and Eden groups.After this disposal of the Whitehouse and the more confident refer-ence of the Drummuck to the Medinan the intervening Barren flag-stones seem to fall very naturally into the space occupied in Americaby the Upper Cincinnatian (Maysville group). The Girvan bedssupposed to be of this age agree particularly well with the upperpart of the Pulaski group and the Oswego sandstone as developedin parts of New York and Pennsylvania. 86 PKOCEEDINGS OF THE NATIONAL MUSEUM vol.76THE ASHGILLIAN OF NORTHWESTERN ENGLANDMarr's term Ashgillian does not appear on the chart, but sinceI enjoyed an opportunity the past summer to study the typical andbest known exposures of the series in northwestern England underthe able guidance of T. C. Nicholas and W. B. R. King, both ofCambridge University, some expression of my opinion concerningthe age of the beds covered by the term seems desirable. As definedby Marr ^^ and as thie beds and fossils appeared to me in the field,the Ashgillian should fall within the Medinan epoch. Whetherany part of the series is of Richmond age I am not prepared tosay, but the upper part at least I am strongly inclined to refer tothe Alexandrian.Let me say further that at no place visited by us in 1929 did wesee any convincing contact between beds admitted by our guides tobe very low Silurian and beds of Ashgillian age that my Britishfriends classify as " Upper Bala " or " Caradocian " and thereforeas " late " Ordovician. The supposed " contacts " and sometimes " passage " beds that were pointed out as marking the transitionfrom the Ordovician to the Silurian in no case presented the dias-trophic criteria and qualities that in America we insist on beingdefinitely located in the outcrop and indubitably shown to be presentat the Ordovician-Silurian contact. However, much more convinc-ing and diastrophically well marked contacts occur in the LakesDistrict and elsewhere in Britain between lower beds, but as theirfossils consist mainly of Ordovician generic types and perhaps par-ticularly because they lack monograptids they are referred by theBritish geologists to the Ordovician system.In my opinion the naturally defined base of the Silurian in theEnglish Lakes District lies at the base of the Coniston limestoneseries. This series begins with the " Stile End beds," to 50 feetthick and consisting of sandstones, grits, and as much as 10 feet ofcoarse conglomerate at the base. The Stile End beds are succeededby the Applethwaite beds?calcareous shales, banded and nodularlimestones?about 100 feet thick, with a basal zone full of pebblesderived from the underlying Borrowdale volcanic series. Here andthere the Applethwaite limestone is highly fossiliferous, the faunaconsisting mainly of corals. But these corals?among them severalspecies of Heliolites?are of kinds that viewed in the light of Ameri-can occurrences could indicate nothing older than topmost Medinaor Clinton. The Applethwaite is succeeded by Marr's Ashgill group,70 feet thick, with the Phillipsinella beds at the base and the Phacopsmucronatus beds?now admitted by Troedsson to be Silurian?in '?Marr, J. E., The Lower Paleozoic rocks of the Cautley District: Geol. Soc. LondonQuart. Journ., vol. 69, p. 5, 1913. ART. 21 ORDOVICIAN TRILOBITES ULRICH 87the middle third. The Ashgillian is succeeded by the Skelgill beds.These comprise a number of thin zones with species of Mofiograptusand at the base a thin limestone with Atrypa flexuosa and directlyover this a black mudstone with Dimovphograptus.Just why the first appearance of monograptids, unheralded as itusually is by a well-marked physical break, should determine the be-ginning of the Silurian system and the close of the Ordovician is notclear to me. It is merely an event in the course of Silurian historyand one that can hardly be expected to have been manifested at pre-cisely the same time everywhere. It is no more important than thefirst appearance of Fenestella in the Richmond or of Heirdtrypa inthe Brassfield or of Coelospira and Spirifer in the Clinton or of thesubsequent first appearances of many other generic types that becameabundant and lasted for long periods thereafter.In America we also find it troublesome to detect a satisfactoryphysical or faunal boundary between the Richmond and the Alex-andrian, and considerable difference of opinion as to the precise loca-tion of the Medinan-Clinton boundary is notable in American litera-ture. However, as regards the systemic boundary, the best informedAmerican stratigraphers?at least those who have learned their stra-tigraphy from field observations in many areas rather than from lab-oratory studies and comparisons of collections of fossils?are wellsatisfied to follow the footsteps of the geologists of the first geologicalsurvey of New York, who in the forties of the last century drew theboundary between their Ontario and Champlain divisions of the NewYork system at the generally clearly marked base of the LowerMedinan. The official survey of New York has never, so far as Iknow, receded from its position on this question except by substitutingBritish terms for New York names.After four brief but well-filled periods of field studies in Britain,Scandinavia, and Bohemia my conviction that diastrophically well-marked systemic boundaries essentially corresponding in age to thoseworked out in America may also be determined on the east side of theAtlantic is more firmly fixed than it was on my first visit to Europein 1922.Norway aoid Sweden.?This column requires little explanation.Etage 5 of the Norwegian section and the Leptaena limestone ofSweden are placed into the Silurian for practically the same reasonsas those that seemed to demand the removal of the British Keisleyand Drummuck formations from the Ordovician to the naturally de-limited Silurian system advocated hj me. The proper placing ofthe Norwegian Etage 4 and the Swedish Trinucleus {Tretaspis)and Chasmops zones I find much more difficult. Regarding theseScandinavian zones the Trinucleus zone seems at present to belong 88 PKOCEEDINGS OF THE NATIONAL MUSEUM vol.76liigh in the Ordovician?perhaps within the broader zone of theAmerican Viohi limestone., though the Trinucleidae in the two arequite different. The half dozen species in the latter are of Grypto-lithus^ whereas those in the Swedish formation are of Tretaspis.One of the latter is identified with T. hucklandi^ a Drummuck speciesin Scotland, which according to preceding argument (pp. Gl-69) isearly Silurian rather than late Ordovician. But the Viola also isnot firmly fixed in the position given it on the chart. We know onlythat it is younger than Trenton and in unconformable contact withthe Fernvale above. It may therefore correspond to the whole or tosome part of the Cincinnatian, or, if it does not belong betweenthe Eden and the Maysville, its place may be in the hiatus betweenthe Maysville and the base of the Kichmond. The fauna gives noconclusive indication whatever, and what evidence it does presentseems to favor the last interpretation rather than the others. Ifthe Trinucleus zone proves older than the Viola the position of theunderlying Chasrnops will also be lowered in the scale.Etage 3" {Orthoceras limestone) probably is older than given onthe chart. A positive statement is not yet warranted, but judgingfrom its fossils I am inclined to believe that eventually Etage 3 ' and the Kunda of the Baltic Province will be found to correspond toour Buffalo River series.Baltic region.?Above the Wesenberg no material change has beenmade from the correlations indicated in the table published in my1926 paper on the Ordovician-Silurian boundary. The Borkholm isagain correlated with the Leptaena limestone of Sweden and theKeisley of England, and all three are placed in the general horizonof the Upper Medinan. The Lyckliolm, also, is referred as before tothe Richmond. More doubt is entertained regarding the position ofthe Wesenberg and also as to the stratigraphic relations of the Kegeland the four members of the Wierland group of Raymond to Or-dovician formations of America. A somewhat lower position issuggested for the latter than in the preceding paper; but I am notcertain that the present arrangement is nearer the truth than theother. On the other hand, I can not free my mind of the suspicionthat most if not all of these east Baltic formations were not depositedat strictly the same times as those in either of the Scandinaviancountries or those in England, Scotland, and Ireland or those inNorth American areas, with which they have hitherto been more orless confidently correlated by others as well as by me. EXPLANATION OF PLATESPlate 1 PageFig, 1. Telephus reedi, new species 19Copy of Reed's figure of the cranidium of this Girvan speciesthat he had referred, probaljly erroneously, to T. fractusBarrande; X 3.Whitehouse group, Girvan District, Scotland.2. Telephus species undetermined:Copy of Hadding's figure of a small specifically undeterminedcranidium of this genus, X 5. Compare T. spiniferus andT. loegelini.Lower Dicellograptus shale near Rost&nga, Sweden.3-7. Telephus fractus Barrande 10Copies of Barrande's illustrations of this species; all naturalsize except fig. 5, which is X 2.Etage, Di, Lodewitz and Koenigshof, Bohemia.8-10. Telephus jamilandicus, new species 138. Imperfect cranidium, X 1 and X 4, that is provisionallydistinguished under this name from the other specimensfigured and referred, apparently in error by Hadding toT. biciispii Angelin. Copied from Hadding.9, 10. Side and anterior views of same, X 4.Lower part of Ogygiocaris shale, Jamtland, Sweden.11-18. Telephus haddingi, new species 1211,12. Different views of free cheeks with eyes, X 9.13, 14. Dorsal views of two cranidia, X 2.15, 16. Anterior and sideviews of a larger cranidium, X 2.17. Imperfect pygidium, with transverse and longitudinal pro-files, X 6.18. Incomplete thoracic segment, X 5.Figures copied from Hadding's paper in which these specimensare referred to T. bicuspis. For copies of Angelin's figuresof the Norwegian types of his T. bicusjns see pi. 2, figs. 20, 21.Lower part of Ogygiocaris shale, Jamtland, Sweden.19-23. Telephus granulatus Angelin 11Copies of Hadding's figures of Swedish specimens of cranidiaand a pygidium referred by him to this species. The twocranidia seem to differ slightly and neither is quite likeAngelin's figure of the Norwegian type of the species. Forcopy of the latter see pi. 2, fig. 13.Upper imrt of Ogygiocaris shale, Jamtland, Sweden.90 U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 76, ART. 21 PL. 1 ORDOVICIAN TRILOBITES OF THE FAMILY TELEPHIDAEFOR EXPLANATION OF PLATE SEE PAGE 90 U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 76, ART. 21 PL. 2 ORDOVICIAN TRILOBITES OF THE FAMILY TELEPHIDAEFOR EXPLANATION OF PLATE SEE PAGE 91 Plate 2 PageFigs. 1-9. Telephus mobergi Hadding 141. A small cranidium, X 4.2. Dorsal view of another cranidium, X 1.3-6. Anterior, dorsal and side views of same, X 3.6. Dorsal view of a large cranidium, X 3.7, 8. Free cheeks with broken eyes, X 3 and X 5.9. Half of a thoracic segment, X 3.Copies of original figures in Hadding's i)aper. Basal bedsof the Ogygiocaris shale, Jamtland, Sweden.10-12. Telephus wegelini Angelin 1310. Copy of Angelin's original figure of the cranidium ofthis species. The middle spine shown in front doubtless isa misapprehension.11. 12. Copies of Hadding's figures of two cranidia, X 3.Black Trinucleus shale, Dalarne, Sweden.13. Telephus granulatus Angelin 11Copy of Angelin's figure of the type of this species.14. Telephus ? salteri Reed _ 20Copy of Reed's dorsal view of the type of this strangespecies. As given by Reed, the greatest width of thiscranidium is 5.4 mm.Balclatchie group, Balclatchie, Girvan District, Scotland.15-17. Telephus linnarssoni, new species 15Dorsal, side and front views, X 3, of the holotype of thisspecies; after Warburg.Leptaena limestone, Boda, Dalarne, Sweden.18, 19. Telephus hihernicus Reed 17Copies of figures of two cranidia of this species, X 4, afterReed. The lower figure suggests shortening by com-pression.Tourmakeady Beds, County Mayo, Ireland.20, 21. Telephus bicuspis Angelin 12Copies of Angelin's original figures of this species. (Seealso pi. 1, figs. 11-18).22-27. Telephus americanus Billings 2122, 23. Dorsal views, X 2 and X 4, of the selected holo-type of this species. The figures given by Hadding weremade from a plaster cast of this specimen. It seemsalso to be the one from which Billings' illustration wasprepared.24, 25. Two other cranidia, the second not completelyuncovered, X 2, that agree very closely with the holo-type.26. A fourth cranidium, X 2, that difi"ers enough in theshape of the glabella to suggest a distinguishable speciesor variety.27. Probably another cranidium of the preceding variety,X 2. Probably Upper Chazyan (Div. N and V),Newfoundland. 91 Plate 3 PagttFigs. 1-10. Telephus pustulatus, new species 281. Dorsalviewof acranidium, X 3. U.S.N.M. No. 80536a.2, 3. Front and side views of the holotype cranidium, X 3,U.S.N.M. No. 80536.4. Dorsal view of same, X 4.5-7. Different views of a free cheek, all X 4, showing itssmall rim, very large eyes, and the broken base of thegenal spine. U.S.N.M., No. 805366.8. Same view as figure 7 with genal spine restored.9,10. Dorsal and end views of pygidium, X 4. U.S.N.M.,No. 80536c.Whitesburg limestone, Lexington, Va.1 1 . Telephus spiniferus, new species 2'^Dorsal view of the holotype cranidium, X 4. U.S.N.M.No. 80537. Basal 20 feet of the Athens shale, overlarge Holston limestone quarry, 3 miles southeast ofSaltville, Va.12. Telephus spiniferus calhounensis, new variety 30Dorsal view of the imperfect cranidium, X 4, on which thisvariety of species is based. Holotype, U.S.N.M. No.80538. As will be observed, it differs from the holo-type of the older typical form of the species in theshape and, apparently also, in the furrowing of theglabella. Athens shale, 75 feet beneath top bluff of Hi-wassee River, one and a half miles east of Calhoun,Tenn.1?, 14. Telephus latus, new species 26-Dorsal and anterior views of the holotype, X 3. Thespecies resembles T. pustulatus but differs in the outlineand lesser convexity of the glabella and the generalflatness of the fixed cheeks. U.S.N.M. No. 80539.Thin limestone lenses in the basal 20 feet of the Athensshale, at quarry southeast of Saltville, Va.16 'J'elephus sinuatus, new species 30Dorsal view of holotype cranidium, X 4. Glabella some-what as in T. spiniferus but fixed cheeks much narrower.U.S.N.M. No. 80540.Whitesburg limestone, Lexington, Va.16-19. Telephus prattensis, new species 3416. Dorsal view of a cranidium from Pratts Ferry, Ala.,X 3. Cotype U.S.N.M. No. 80541.17, 18. Similar views, X 3, of two cranidia from theWhitesburg limestone, 1 mile south of Bulls Gap, Tenn.Cotypes U.S.N.M. No. 80542.19. Free cheek, X 4, associated with this species andT. hipunclalur. at Pratts Ferry. Ala. U.S.N.M. No.8054 la.92 U. S. NATIONAL MUSEUM PROCEEDINGS. VOL. 76. ART. 21 PL. 3 ORDOVICIAN TRILOBITES OF THE FAMILY TELEPHIDAEFOR EXPLANATION OF PLATE SEE PAGE 92 U. S. NATIONAL MUSEUM PROCEEDINGS. VOL. 76. ART. 21 PL. 4 ORDOVICIAN TRILOBITES OF THE FAMILY TELEPHIDAEFor explanation of plate see page 93 Plate 4 PageFigs. 1-14. Telephus bicornis, new species 23I, 2. Anterior and dorsal views of a free cheek, X 4. Thelatter shows the shorter spine behind the longer one whichis regarded as the real genal spine. U.S.N.M. No.80535a.3. Dorsal view of rather small cranidium, X 3. U.S.N.M.No. 805356.4. Side view of a larger cranidium, X 3. U.S.N.M. No.80535c.5. Dorsal view of an imperfect cranidium with occipitalspine complete, X 3. U.S.N.M. No. 80535d.6. Large cranidium somewhat crushed in right anteriorthird, X 3. U.S.N.M. No. 80535e.7. Nearly complete but small cranidium, X 3. U.S.N.M.No. 80535/.8. Another cranidium, X 3. U.S.N.M. No. 80535^.9. Still another cranidium that retains about half of one ofthe glabellar spines, X 3. The specimen was tilted inphotographing so as to show the full length of the remain-ing part of the right spine. U.S.N.M. No. 80535/i.10. Anterior view, X 4, of the specimen shown in Figure 7.II. An associated hypostoma supposed to belong to thisspecies, X 3. U.S.N.M. No. 80535z.12-13. Respectively dorsal and side views of the associatedpygidium, X 3. U.S.N.M. No. 80535j.14. One of the thoracic segments, X 3. U.S.N.M. No.80535&. All these specimens are regarded as cotypes ofthe species, and all were collected from a single ledge ofWhitesburg limestone, 5 miles southwest of Bland, Va.93 Plate 5 I'ageFigs. 1-9. Telephus bipunctatus, new species 311-3. Dorsal, lateral, and anterior views of a nearly perfectcranidium, X 4. U.S.N.M. No. 80543a.4-6. Dorsal views of three other cranidia, X 4, selected toshow extremes of observed variations in shapes of parts andin surface sculpture. U.S.N.M. No. 805436.7, 8. Dorsal and posterior views of a pygidium, X 4. Onlythe broken bases of the double-headed spines on the axialrings are retained when the specimens are uncovered inbreaking the limestone matrix. U.S.N.M. No. 80543c.9. The most complete of the free cheeks, X 4. U.S.N.M.No. 80543d.Whitesburg limestone, Lexington, Va. All these specimensmay be ranked as cotypes of the species.10-15. Telephus impunctatus , new species 33^10-13. Dorsal views of four cranidia, X 3, that may be re-garded as cotypes of the species. U.S.N.M. No. 80544.The specimen of Figure 12 has been slightly distorted byobliquely transverse pressure.14. Side view of Figure 13.15. Pygidium associated with these cranidia, X 4. It iswider and the axis less convex than in T. bipunctatus whichoccurs in the same bed and place. U.S.N.M. No. 80545.Whitesburg limestone, near Albany, Tenn.16. Telephus buttsi, new species 40One of three very small cranidia, X 8. All of these have beensimilarly shortened by compression. Their general aspectsuggests T. bipunctatus and T. troedssoni, but the occipitalspine is too long, slender, and round for either of those spe-cies. They remind also of T. mobergi, but the occipitalspine is too long and the fixed cheeks too wide to warranttheir reference to that species. U.S.N.M. No. 80546.Yellow, leached shale, at base of the Athens shale, near Long-view, Ala.17-21. Telephus troedssoni Raymond 40'17. Dorsal view of a clay impression taken from the holotypeof this species, X 3. The latter is a natural mold of the ex-terior of an imperfect cranidium, somewhat distorted bycompression in slightly oblique direction. Original in Mus.Comp. Zool., Cambridge, Mass.Near base of Athens shale, near Athens, Tenn.18. An imperfect and distorted cast of the interior of a crani-dium supposed to belong to this species. The middle partof the anterior border is well preserved and clearly showsthe distinctness of the median anterior pair of spines andthe shorter spines to which the ends of the free cheeks arejoined. U.S.N.M. No. 80547.19. An associated cast of the interior of a free cheek withcompound eye, X 8. U.S.N.M. No. 80547a.94 U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 76. ART 21 PL. 5 ORDOViCIAN TRILOBITES OF THE FAMILY TELEPHIDAEFOR EXPLANATION OF PLATE SEE PAGES 94 AND 95 Fig. 20. Another cast of the interior of a cranidium found withthe preceding two specimens, X 8, more doubtfully referredto this species. This has been greatly shortened by pres-sure and is otherwise imperfect. It is figured mainly be-cause it retains most of the anterior border and shows theseparateness of the median pair of spines even in thisform, in which their bases are very close to the anteriorends of the facial sutures. U.S.N. M. No. 80547b.21. A pygidium found with the preceding, X 8. Thoughsomewhat shortened by compression, it still shows agreater width of border than is found in other species.In general it reminds most of the pygidium assigned toT. impundatus. U.S.N.M. No. 80547c.AU found with T. buttsi in the basal part of the Athens shale,1.5 miles northeast of Longview, Ala. 95 Plate 6 PageFigs. 1-7. Telephus mysticensis, new species and variety 221. Holotype cranidium, X 3. Very similar in general aspectto the cranidium of T. americanus but lacks the small surfacepustules and the glabella has shallow curved longitudinaldepressions that are not present in the types of that species.U.S.N.M. No. 80526.2. Free cheek with posterior side above, X 4. ParatypeU.S.N.M. No. 80527a. Possibly the cheek of the var.simulator3. 4. Probably anterior and dorsal views of another slightlydifferent free cheek, X 4. Paratype U.S.N.M. No. 805276.The collections contain four or five of each of these twokinds of cheeks, but to which of the two kinds of cranidiaeither belongs is, of course, uncertain. The lower side offigure 4 probably is anterior. The periphery of the eye hereoverhangs the narrow rim.5 An associated hypostoma, X 4. Paratype U.S.N.M. No.80527c.6. One of the two associated pygidia, X 4. Paratype U.S.N.M.No. 80527d.7. Cranidium, X 4. Differs in shape and contour of theglabella from the holotype and in size and shape of the fixedcheeks. Probably a distinct species that is provisionallyseparated as var. simulator, new variety. This name ischosen because it greatly resembles the southern T. prat-tensis (see pi. 3). Holotype, U.S.N.M. No. 80528. Bouldersof Chazyan limestone, near Mystic, Quebec.8-9. Telephus bilunatus, new species 398. Dorsal view of the holotype cranidium, X 4. U.S.N.M.No. 80529. The most characteristic feature of the speciesis the pair of deep lunate glabellar impressions. Whites-burg limestone, near Albany, Tenn.9. A cranidium, somewhat larger than the holotype and dis-torted by lateral compression of the soft shale matrix, X 4.U.S.N.M. No. 80530. From a bed of yellow shale nearLongview, Ala., probably of Whitesburg age but provi-sionally referred to the base of the Athens shale.10-19. Telephus tellicoensis, new species 3510-12. Three views of a free cheek of a Telephus and probablyof this species, X 4. U.S.N.M. No. 80531.13-15. Dorsal views of three cranidia, X 3. Cotypes, U.S.N.M.No. 80532. Anterior and side views of Figure 14 givenin pi. 7, figs. 10, 11.16-18. Three pygidia, X 4, the last showing the shape of theposterior marginal spine. Cotypes, U.S.N.M. No. 80533.19. Side view of specimen shown in Figure 14, for comparisonwith T. hircinus (see pi. 7, fig. 2). Basal 10 feet of Tel- . lico formation, 1.5 miles southeast of Knoxville, Tenn.20,21. Telephus transversus, new species 37Dorsal and anterior views of the holotype cranidium, X 3.Distinguished from T. tellicoensis by its greater width andbroad hollows in lateral slopes of glabella. U.S.N.M. No.80534. Associated with preceding.96 U. S. NATIONAL MUSEUM PROCEEDINGS, VOL. 76, ART. 21 PL. 6 ORDOVICIAN TRILOBITES OF THE FAMILY TELEPHIDAEFOR EXPLANATION OF PLATE SEE PAGE 96 U. S. NATIONAL MUSEUM PROCEEDINGS. VOL. 76, ART. 21 PL. 7 Ordovician Trilobites of the Family TelephidaeFor explanation of plate see page 97 Plate 7 PagoFigs. 1-9. Telephus hircinus, new species 381-3. Dorsal, side and anterior views of a small but nearlycomplete cranidium, X 4. The extraordinary strengthand height of the occipital spine and the conspicuousprominence of the anterior denticles are mainly relied onin distinguishing the species from T. tellicoensis, its near-est ally. Cotype U.S.N.M. No. 80548a.4-5. An incomplete but large and unquestioned cranidiumof the species, X 4. Cotype U.S.N.M. No. 805486.6. Another somewhat smaller cranidium, X 4.7, 8. Dorsal and side views of a pygidium supposed to belonghere, X 4. Cotype U.S.N.M. No. 80548c.9. A free cheek exposing the under side, X 4. CotvpeU.S.N.M. No. 80548d.Tellico formation east of Knoxville, Tenn.10, 11. Telephus tellicoensis, new species 35Anterior and side views of cranidium shown in Figure 14 inpi. 6. Cotype U.S.N.M. No. 80532.12-14. Telephus gelasinosus Ulrich 26Dorsal, side and anterior views of the holotype cranidium ofthis species, X 4. U.S.N.M. No. 71468.15, 16. Glaphurus pustulatus, (Walcott) 42Anterior and side views of specimen of which the dorsalview is given in pi. 8, Figure 11. U.S.N.M. No. 80551/i.17-19. Glaphurina brevicula, new species 46Dorsal, side, and anterior views, X 2, of the holotype crani-dium of this species. U.S.N.M. No. 80549.Holston limestone, 2 miles northwest of Lexington, Va.20,21. Glaphurina falcifera, new species 46Side and dorsal views, X 2, of the holotype cranidium ofthis species. U.S.N.M. No. 80550.Lower Chazyan limestone, 1 mile southeast of Bluff City,Tenn.64441?29 7 97 Plate 8 PageFiQS. 1-11. Glaphurus pustulaius (WaXcott) 421. Thorax and pygidium of rather small specimen, X 4.U.S.N.M. No. 80551a.2. Two cranidia of medium size specimens, X 4. U.S.N.M.No. 805516. Practically all the "pustules" of the surfaceare in fact only the bases of slender spines of three sizes.The length of those on the occipital ring is shown on lefthand cranidium.3. One of the largest of the cranidia so far observed, X 4.U.S.N.M. No. 80551c. The pair of large anterior spineawith two sets of three smaller spines between their basesare restored from another specimen.4. An incomplete free cheek, X4, showing the spinose genalspine and imprints of the series of very long minute spinesthat lines the under side of the outer rim. U.S.N.M. No.80551d.5. A pair of free cheeks, X 4, separated from the cranidiumbut still joined by a thin band that lies just within theseries of minute marginal spines shown in Fig. 4. Thisinner band is separated from the outer rim of the cheekby a marginal suture that seems to extend to and beneaththe genal spine. Apparently it is a separate plate similarto those found in Mesonacidae, Agnostidae, and certainother trilobites. U.S.N.M. No. 80551e.6-9. Four slightly differing hypostomas, X 4, all found in asso-ciation with abundant remains of this species and prob-ably belonging to it. U.S.N.M. No. 80551/.? 10. The largest pygidium found with positively identifiedmaterial of this species, X 3. It is relatively more trans-verse than the other specimens and differs from them alsoin details of the axis. As the same block of limestonecontained also the two cranidia on which Glaphurinalamottensis is founded it is not improbable that it belongswith that species rather than to Glaphurus pustulatus.U.S.N.M. No. 80551ff.11. Dorsal view of a specimen that consists of the cephalon,with the free cheeks in place, and five of the thoracic seg-ments, X 3. Side and anterior views of same on pi. 7.U.S.N.M. No. 80551/1.Fine grained reefy bed at base of Upper Chazy, Isle LaMotte, Vt.12, 13. Glaphurus laiior, new species 44Dorsal and anterior views, X 2, of the holotype cranidium.It is to be noted that this cranidium differs from G. pustu-latus in being larger and relatively wider, the brim nar-rower and with only two instead of three rows of pustulesin front of the glabella. U.S.N.M. No. 80552.Whiteburg limestone, 6 miles southwest of Bland, Va.14-16. Glaphurina lamottensis, new genus and species 45Anterior, lateral and dorsal views, X 2, of the holotype ofthis species. The genus, of which two other species arefigured on pi. 7, is distinguished from Glaphurus by thetotal pinching out of the brim between i\\e front of theglabella and the cephalic rim. U.S.N.M. No. 80553.Reefy basal limestone of the Upper Chazy, Isle La Motte, Vt.98 U. S. NATIONAL MUSEUM PROCEEDINGS. VOL. 76. ART. 21 PL. 8 Ordovician Trilobites of the Family TelephidaeFOR EXPLANATION OF PLATE SEE PAGE 98 INDEX PageAmphllichas, mentioned.- 65Ampyx, mentioned 52,53Amfjyxina powelli, mentioned 27Apatowiphalus, mentioned 7Arbuckle limestone, divisions of 70Arctinuras, mentioned 65Ardwell group 85Arcnig formation, Upper Canadian age of 83Ashglllian of lOugland probably Medinan inago 86Athens shale, fossils of 27northern extent of 51Balclatchic fauna, relations to Craigheadlimestone fauna 84formation, fossil evidence on ageof 53,85lower age limit of 50Barren flagstones of the Qirvan section 85Bassler, R. S., mentioned 1,61,71Becsie formation, age of 82Blount faunas 57British and American definition of Silurianat variance 63Buffalo River series absent in AppalachianValley 79unconformably over-lain in MississippiValley 77Butts, Charles, mentioned 58work of in Appalachian re-gion 71Calymene, mentioned 51Campbell, II. D., mentioned 2Canadian deposits in northern Newfound-land - 81Cause of difflculties in correlating Europeanand American formations 64Chambersburg limestone, stratigraphic posi-tion and fauna of 59Chariocephalus, mentioned 7Cincinnatian series 79 ' Close discrimination of species, need of 9Coelospira, first appearance of 87Coniston limestone series, natural base ofSilurian system 86Coral fauna of the Coniston limestone series. 86Correlation by diastrophic criteria 68inadequacy of former methods of. 67of Minnesota formations, note-- 67table 73basis of generalized time scale 73formations in Baltic region. 88formations in Britain. 82formations in MississippiValley 77 PageCorrelation table formations in Norway andSweden 87formations in Ohio Valley. 78generalized comments on _ _ 75G.rvan district, Scotland,supplementary notes on . . 83northern Appalachian for-mations 81notes on Oklahoma for-mations - -_ 67southern and middle Appa-lachian formations 79Craighead quarry limestone 84Cybeloides, mentioned 9Cybelopsis, mentioned-. 9Cyclopyge (Aeglina), mentioned 6Cyclopygidae, mentioned.- 7Decorah Bryozoa in Baltic region 60faunas, northern origin of 60Deposits of Blount age absent in Marylandand Pennsylvania 67Dicellograptus cf. complanatus in Sylvanshale-- - 66Didymograptus biflrlus, mentioned 83Dionide, mentioned 49Drummuck formation of Scotland, age of 61,68Durness limestone, relations of to AmericanMiddle and Upper Canadian formations-. 83Ellis Bay formation, position of in time scale. 82Encrinurus afl. punctatus, mentioned 61Etage 3c (Orthoceras limestone), age of 88Faunal evidence bearing on base of Silurianin Britain-.- --- 87Faunas of the Blount group 57Fcnestella, first appearance of 87Galena limestone, age of 78Glaphurina brevicula, new species 46falcifera, new species 40lamottensis, new species 46mentioned 8,9Olaphurus decipiens, mentioned 8,44latior 44mentioned 7pustulatus, described 42Graptolites, limitation of value in correlation. 48, 87Gun River fornmtion, age of 82Ilemitrypa, first appearance of 87Interprovincial correlations, present status of. 72Introduction of new (auoal elements, import-ance of in correlation.- 68Irvingclla, mentioned 7Isotelus, medium eye of, mentioned.? 27Jones, O. T., mentioned 83,84Juniata and Queenston sandstones 81Keisley limestone of England, age of 61, 04, 60Kimmswick limestone 7899 100 INDEXPageKindle, E. M., mentioned 21Lapworth, Charles, mentioned 62Leptaena limestone of Sweden, Silurian agoof 61,64,66Lichas, mentioned.. 65Little Oak limestone, stratigraphic positionand age of 59Lower and Upper Medina, relations of 81Lower Canadian deposits, distribution of 83Lyckholm and Borkholm formations ofEsthonia 61, 66Marine submergences, alternation of, onopposite sides of Atlantic 54Maximum development of Lower Paleozoicmarine deposits in America 76Middle Chazyan, Little Oak, and Chambers-burg faunas, origin of 60Mosheim and Murfreesboro limestones, rela-tions of 78Murat limestone, same as Holston 2Nemagraptus gracilis, mentioned 27North Atlantic, hypothetical faunal province. 61Odontopleuridae, mentioned 9Oklahoma, Lower Paleozoic formations in... 76Onchaspis, mentioned 52,53Ontario or Ontarian system, advantages of. . . 62Ordovician-Canadian contact usually marksgreat hiatus 80Ordovician marine deposition, relativeamounts and rates of 56Ordovician-Silurian boundary 61, 68, 87Ordovician system, need to revise Lap-worth's definition of 74Origin of Trenton and late Black Riverfaimas 60Oscillation of the surface of continents 55Oscillations in Black River time 80Persistence of indigenous faunas 68, 69Phacops mucronatus beds 86Phillipsburg, Quebec, section. Lower andUpper Ozarkian in 81Phillipsinella beds 86Physical evidence of distinctness of Canadianand Ordovician systems 80Post-Blount faunas, origin and ages of 59Raymond, P. E., mentioned-.. 7,49,53,57and Willard, Bradford, onAppalachian Brachi-oypoda, criticized. 64Recent progress in American stratigraphy. . . 69Recurrences of faunas 69Remopleurides, mentioned 6, 7 Resser, Charles E., mentioned 1,71Richmond and Clinton formations in Anti-costi 65Robergia major, mentioned 27,29marginata, mentioned.. 7Salteria oderi, mentioned... 51Schaghticoke shale, age of 82Shallowness, small size, and frequent shiftingof epicontinental seas 55Shifting of epicontinental seas indicated inEurope as in America 55 ) PageSilurian graptolites in Leptaena limestone... 66Silurian, naturally defined base of in Britain . 86Silurian system, unconformable base of inEurope and America 64Simpson formations 77Skellgill beds 87Spirifer, first appearance of 87Stetson, n. C, mentioned 50Stinchar limestone, typical 53,84Stones River-Lowville contact in Tennesseeand Kentucky, significance of 69Sylvan shale, mentioned 66Telephus americanus, described... 21mentioned 41bicornis, new species 23mentioned 26bicuspis, discussion of.. 12mentioned... 12, 13, 18, 19bilunatus, new species 39bipunctatus, new species 31mentioned 15,17,23, 29, 31, 33, 35, 36, 37, 40buttsi, new species.. 40mentioned 29character of material 2faunas, absence of in ChamplaiaValley 49fractus, comments on 10mentioned . 16, 17, 19, 23, 26, 27, 28gelasinosus, described 26mentioned 18,31granulatus, discussion of 11mentioned 23,24,25,26,27haddingi, new species 12mentioned 13Hadding's paper on 3hibernicus, described 17hircinus, new species 38mentioned .32, 35, 36, 37, 42impunctatus, new species... 33mentioned 32,37jamtlandicus, new species 13mentioned 19latus, new species 28mentioned 28,29linnarssoni, new species 15mobergi, discussion of 14mentioned 23,31,37,38,41mysticensis, new species 22mentioned.il, 22, 31, 32, 39, 49simulator, new vari-ety 23occurrence in southeastern Canada. 49original description of 1other trilobites associated with 47prattensis, new species 34mentioned 15, 23, 32, 33, 36pustulatus, new species. 28mentioned 32reedi, new species 19revised description 3salteri Reed, described. 20sinuatus, new species 30mentioned 30 INDEX 101PageTelephus spiniferus, new species _. 29mentioned 14,19,27,30spiniferus calhounensis, new vari-ety -- 30stratigraphic and geographic rangeof 5,47superstis, mentioned 15systematic position of 6tellicoensis, new species.. 35mentioned. 32, 33, 35, 36, 37, 42transversus, new species 37mentioned... 35,36troedssoni, described 40mentioned 14,29wegelini, discussion of 13, 16mentioned 15Tellico formation, variations in base of 37Tilting and warping of surface of continents. 58 PageTime required to effect modifications offaunas _ 58Tornquistia, mentioned 51,52Tremadoc formation of England, age of 82Tretaspis bucklandi, mentioned. 88mentioned 51Trilobites as factors in correlation 48Trinucleus beds, mentioned 66subradiatus Reed, mentioned 50zone of in Norway and Sweden.. 87Troedsson, Gustav T., correlation of forma-tions in northern Greenland by 67Tuscarora and Clinch sandstones 81Twenhofel, W. H., mentioned 65,82Viola linestone, age of discussed 77, 88Whitehouse group of Scotland 85Whitesburg limestone, description of 2fossils of 27 U S SOVEBNMENT PRINTING OFFICE: 1929