2 4-3- + » eS ee Oe ee ee Ye Ver ea re ew eee eee ha rae tin hal oe ee ee ee POP Re eevee , Ms - "a4 ,* ’ oF F - . ‘ . * - * - * ¢ ‘ » ‘ 7 19,4 " ~ tae @ str +e oe 2 wes eee er ee ee © Pree + PAM. Fw ee ee Ae * < Ane pt + Oe > ae et-O-** - Ce ee et a ~- mt . -e ee ee o- 29 7 “ ° + = eer at eo * 1 -* ug > ome 7 ne oe ee ee fet Pine . J ote eee ee arr a Pag Aba Pads ee iby SRE he 5 a ipperi e eeeY ay a bit we whew ret nraanaliiis steer iil Jian ds 1) ' AY RTE Rt oe DPR’ nl ioe #, 9Y Dl. 64.2 jA yt 1A 4 N a , vs i ~ 5 > - ; - “ ~ QUARTERLY JOURNAL GEOLOGICAL SOCIETY OF LONDON. EDITED BY THE ASSISTANT-SECRETARY OF THE GEOLOGICAL SOCIETY. Ph e) B Be Et Quod si cui mortalium cordi et cure sit non tantum inventis herere, atque iis uti, sed ad ulteriora penetrare ; atque non disputando adversarium, sed opere naturam vincere; denique non belleet probabiliter _ @pinari, sed certo et ostensive scire; tales, tanquam veri scientiarum filii, nobis (si videbitur) se adjungant. _ -—Novum Organum, Prefatio. VOLUME THE SIXTY-SECOND. 1906. b LONDON: LONGMANS, GREEN, AND CO. fx: PaRIS: CHARLES KLINCKSIECK, 41 RUE DE LILLE. : ! SOLD ALSO AT THE APARTMENTS OF THE SOCIETY. al a f | a 8 4 MDCCCCVI. List OFFICERS OF THE GEOLOGICAL SOCIETY OF LONDON. BARRA AOI Elected February 16th, 1906. rey PrestVent, Sir Archibald Geikie, Se.D., D.C.L., LL.D., Sec.B.8. Wice-PrestVents, Robert Stansfield Herries, M.A. John Edward Marr, Sc.D., F.B.S. Aubrey Strahan, M.A., F.R.S. J.J. Harris Teall, M.A., D.Se., F.R.S. Secretaries. Prof. William Whitehead Watts, M.A., M.Sc., F.R.S. Poretqu Hecretarp. Sir John Evans, K.C.B., D.C.L., LL.D., F.R.S., F.L.S. Prof. Edmund Johnstone Garwood, M.A. Creasurer. | Horace Woollaston Monckton, Treas.L.8. COUNCIL, Henry Howe Arnold-Bemrose, M.A., J.P. Prof.Thomas George Bonney,Sc.D.,LL.D., E.R.S., F.S.A. Sir John Evans, K.C.B., D.C.L., LL.D., E.R.S. Prof. Edmund Johnstone Garwood, M.A. Sir Archibald Geikie, Sc.D., D.C.L., LL.D., Sec.R.8. Robert Stansfield Herries, M.A. ‘Finlay Lorimer Kitchin, M.A., Ph.D. Philip Lake, M.A. George William Lamplugh, F.R.S. Richard Lydekker, B.A., F.R.S. Bedford McNeill, Assoc.R.S.M. John Edward Marr, Se.D., F.R.S. Horace Woollaston Monckton, Treas.L.S. Frederick William Rudler, I.8.0. Leonard James Spencer, M.A. Aubrey Strahan, M.A., F.R.S. Charles Fox Strangways. J. J. Harris Teall, M.A., D.Se., F.R.S. Richard Hill Tiddeman, M.A. Prof. William Whitehead Watts, M.A., M.Sce., F.R.S. The Rev. Henry Hoyte Winwood, M.A. Arthur Smith Woodward, LL.D., F.R.S8. Horace Bolingbroke Woodward, F.R.S. Assistant-Secretarp, Clerk, Ltdrarian, anv Curator L. L. Belinfante, M.Sc. Assistants in Office, Library, and fMuseum. W. Rupert Jones. Clyde H. Black. Alee Field. STANDING PUBLICATION COMMITTEE. Sir Archibald Geikie, President. Prof. W. W. Watts. aration Prof. E. J. Garwood. } Secretaries. Prof. T. G. Bonney. ° Mr. R. S. Herries. Dr. F. L. Kitchin. Mr. P. Lake. Mr. G. W. Lamplugh. Mr. R. Lydekker. Dr. J. E. Marr. Prof. H. A. Miers. Mr. H. W. Monckton. Mr. L. J. Spencer. Dr. J. dy Veal Mr. H. Woods. TABLE OF CONTENTS. Bonney, Prof. Taomas Gzoreg. On the Relations of the Chalk and Boulder-Clay near Royston (Hertfordshire) ............ . BuckMAn, SyDNEY 8. Brachiopod Homceomorphy (Plate XLI) .. CANTRILL, THoMAS CrosBEzE, & H. H. THomas. On the Igneous and Associated Sedimentary Rocks of Llangynog, Caermarthen- Bure naves: OM LIE RON) esi a we eee ces Be ek _ CHAPMAN, Freperick, & D. Mawson. On the Importance of Halimeda asa Reef-forming Organism; with a Description of the Halimeda-Limestones of the New Hebrides (Plates X LIX- LD GE ean ene AU SRN RE RN s/n. AUN Re EM al aga Oa OR Cockin, GEoRGE MaRMADUKE. On the Occurrence of Limestone of the Lower Carboniferous Series in the Cannock-Chase Portion of the South Staffordshire Coalfield ................ Davison, Dr. CHartes. The Doncaster Earthquake of April 23rd, Me lanterns wae. s acc sede ne dete eel fhe ee hc ee Donan, Miss JANE, see LONGSTAFF. “Eres, Miss Gerrrupe L., & Miss I. L. Suarer. The Highest Silurian Rocks of the Ludlow District (Plate XXII) ........ Evans, Davip CLEDLYN. The Ordovician Rocks of Western Wacemarthnenshire (Plate. XLV) .u7, i ede PEST, Ob Evans, Dr. Joun Witrtiam. The Rocks of the Cataracts of the River Madeira and the Adjoining Portions of the Beni and NA Eara Nene rate eW Je oss hia ed a 2 tye Urano Mieke abo ill ei Ghat a ec uwew ove Garwoop, Prof. Epmunp JounstonE. The Tarns of the Canton Hicuo, owatzetland (Plates VII-XCRD): . cee cee ss. se Harker, ALFRED. The Geological Structure of the Sgirr of ‘Eige Rees LIRR Oe Vn aerial AEN @RNS ga SO Sid wala BM og Hick1Line, GEorGE. On Footprints from the Permian of Mansfield (Nottinghamshire) €) o)0 0) 6 6 0) 6 6 eo ae) es ssa ee 8) 6 Gian ie ee 6 6 ¢ 6 ew 6 es 8 8 Hi11, the Rev. Epwin. The Chalk and Drift in Méen 433 223 702 597 iv TABLE OF CONTENTS. Page JamriEson, THomas F. The Glacial Period in Aberdeenshire and the Southern Border of the Moray Firth .................. 13 Juxrs-BrowneE, ALFRED JoHN. The Clay-with-Flints: its Origin and Distribution (Plate Vi)iaa ec... .2 sche sees es 132 Lake, Puiirp. Trilobites from Bolivia, collected by Dr. J. W. Hyans in 1901-—PO0? (Piette: A) 44... dai. b... ee 425 Lrxpour, Prof. Gzorcz ALEXANDER Louts, & Dr. J. A. SMyTHE. On a Case of Unconformity and Thrust in the Coal-Measures of Northumberland (Plate: XU0D).\...). eka 2 su... 530 LonestaFF, Mrs. G. B. Notes on the Genera Omospira, Lopho- sptra, and Turritoma; with Descriptions of New Proterozoic Species (Plates XLT & ALVV) ocho ojare Ss syrepes tls alee ee | 652 Lorenzo, Prof. GirusEpPpE DE. The Eruption of Vesuvius in April 1906 vis) kiinccoeaeke Geutaidapetis- A saint eee 476 MarsHALL, Prof. Patrick. The Geology of Dunedin, New Zealand (Plage XXX VI XXII) a8 case pl o chesiel eter are cieaute Sire ee ocr ee 381 Mattey, Dr. CHartes ALFRED. The Carboniferous Rocks at Rush (County, Dublin)! dei. eae vrarhts 6h ke Ya neeen 4 275 Mawson, Dovetas, & F. CHapman. On the Importance of Hali- meda as a Reef-forming Organism; with a Description of the Halimeda-Limestones of the New Hebrides (Plates XLIX—LI). 702 OLpHAM, RicHarp Drxon. The Constitution of the Interior of the arth, as revealed by, Harthquakes.¢.,.:... deceit ba 10 eee 456 Psetursson, Dr. Hetei. The Crag of Iceland: an Intercalation in the Basalt-Kormation .. .. . ).!j.<4 4. 404+ sbhd-saeel eae . 712 RastaLt, Rospert Herron. The Buttermere and Ennerdale Granophyre (Plates XX VII & XXVIII ...........2 ee 253, RICHARDSON, LINSDALL. Liassic Dentaliide (Plate XLV) ...... o73: Scuwarz, Prof. Ernest H. L. The Coast-Ledges in the South- West of the Cape Colony ...\..%.02 210 1). aida ee 70 SHAKESPEAR, Mrs. G. A. On Graptolites from Bolivia, collected by Dr. .W. Mivansiin 1901902) /. kt aa be See 431 . The Tarannon Series of Tarannon (Plates XLVII & RN EY ie wak oe. Ss FERGUS th RLS: he 644 Srpty, THomas Franxkuin. On the Carboniferous Limestone (Avonian) of the Mendip Area (Somerset), with especial reference to the Paleontological Sequence (Plates XX XI- ORONO Ve el chs alee oa Sere eaten’ onic aisle ted Sorin be: Senn 324 SuaTer, Miss I. L., and Miss G. L. Eruzs. The Highest Silurian Rocks of the Ludlow District (Plate XXT1)........0....... dee 195 TABLE OF CONTENTS. v Page SmyTHE, Dr. J. A., & Prof. G. A. L. Lepour. On a Case of Unconformity and Thrust in the Coal-Measures of Northum- rae tela tem cle cea staked sal cals eee ala uaitlle “aie! oily, 01 530 Soxzas, Prof. Wint1am JoHNSoN. Recumbent Folds produced as GEG SUN Deen cps etree nea se) Soe ncce Un at oes 6 Dido ogel Ss Sow 716 Toomas, Herpert Henry, & T.C. Canrriti. On the Igneous and Associated Sedimentary Rocks of Llangynog, Caermarthen- ETUC) (CPA eis) Ou Gt EO. Gh I Da a i a eR ee 223 TREACHER, LuEWwEeLLyn, & H. J. O. Wuite. The Phosphatic Chalks of Winterbourne and Boxford (Berkshire)............ 499 VaucuHan, Dr. ArntHuR. An Account of the Faunal Succession and Correlation [of the Carboniferous Rocks at Rush, County A eatsheets| bel hes) ONIN Re ee ee te ene’ 295 Waitt, Haroup J. Osporne, & Lu. TREaAcHER. The Phosphatic Chalks of Winterbourne and Boxford (Berkshire) .......... 499 Woon, Miss HE. M. R., see SHakEspuaR, Mrs. G. A. Woopwarp, Dr. ArTHUR SmiTrH. On a New Specimen of the Chimeroid Fish, Myriacanthus paradoxus, from the Lower Lias near Lyme Regis, Dorset (Plate I) ...........0..2..065 1 PROCEEDINGS. Exoeredinps ot the Meetings) ir crabige. es «oa hs.e oe ae soles 1, CXXix Je SSTETTE Me LESEH OTe Helton tA Rigsee ner ort eae ae Po ix Enpispor Donors to the Warary 3 er egoma, Pl. xh, 86s | yiddle Bala ...| Various ........ 557-58 instabilis, sp. nov., pl.| | ie ie Ba. ee es \ Upper Bala Thraive Glen ...; 560-61 & am FAURE, ANG gs NY Wrenloclk ..02-2.5- Teenane 7-2-5 567-68 Sedgwickii, sp. nov., pl.) | Lower Llan- sel ios, WN LD cc's o } Overy news } Mulloc ose subglobosa, sp. nov., pl. : Shy, #2. 12, ons ee to ee | Llandeilo hee Craighead ...... 569 trispiralis, sp. nov., pl. \ : on, Fea; 1s os ee Ballas. cteo ace. Bec Ayrshire 2:2 559-60 Omospira orientalis, sp. nov., \ ae sik odie Pes Ly OMe ab Upper Bala...... Thraive Glen ..., 5538-54 Turritoma (¢) tenuifilosa, sp. Middle’. Milan Ww . - oodland 13. fs SUN es GOVEEY <.2..- | IPOInt laeeeeee | 570-71 ScapHopPopDA. Dentalium acutum, sp. nov.,| \ ( Gotherington, pl. xlv, figs. 10 & 11 a-11 d. BEC. 3. cctocsaed 574 —— angulatum ARES ch cicbies | Alderton, ete....| | 574-75 -— elongatum, pl. xlv, fig.| } Pliensbachian ... 4 Rai ee eared cen eeaiyradan ys celia | Cheltenham 575 figs. 18 & Wee Go. | Camerton, etc. .| (576 & cea ae } Hettangian St. Menge ...... 578-79 16 eae aes. \ Pliensbachian...| Hawsker Cliff...) 579-81 hexagonale, sp. nov., pl. operetta uly, figs. La—-lc............... Pliensbachian. } EH OUEESEES coc oe aap pl. xiv, fige. \ Pliensbachian Gameron 2... 581-82 — limatulum, text-fig. 2 .. ae \ Cleverly .......0: 582 13 SED oS Bee | Hettangian ...... Island Magee...| 582-84 oblongum, sp. nov., pl. ) (A xlv, figs. 3€-3 do... c soe | Gotherington ...| | 585 "parvulum, sp. nov., pl. age omer . xlv, figs. 9a-9e & 12 a | (aan URES ae te eae Peaes etc.| | 585-87 —— subovatum, sp. nov., pl.| |. Bill, Oe ie } Lower Jiias’...| Redear............ 587-88 FOSSILS DESCRIBED AND FIGURED. Name of Species. | Formation. Locality. ScaPHoropa (continued). Dentalium subquadratum, sp. Src nn DUPRE See cctv a ction fem sy ys subtrigonale, sp. nov., pl.| | py; . DV OS: 2/02 A. oe co scenes 2 SHRUREMOD Terquemi, sp. nov. ...... Hettangian ..... trigonale, pl. xlv, figs. : AaAB& SABO voces. El cusbucluat CHIMAROIDEL. Myriacanthus paradoxus, pl. i, Lower Lias..... AMPHIBIA (?) ming ermdetyom 86) ern Honeybourne ... Cheltenham Aiglemont Camerton, etc. . Black Ven, Lyme | Regis .. | Xi Page 588-89 589 589-90 590-94 126-30 EXPLANATION OF THE PLATES. PLATE PaGE I MYRIACANTHUS PARADOXUS, to illustrate Dr. A. 1 Smith Woodward’s paper on that fossil ......... ‘Map ILLUSTRATING THE AREA AFFECTED BY THE I DoncastER HWarRTHQUAKE oF APRIL 23RD, 1905, 5 to illustrate Dr. C. Davison’s paper on that Banthgtiake: |! scnssigene seetostece sheen cena ae saree THe Seurr or Hiae; and Bipein BomHEacn (Istz or Erea), to illustrate Mr. A. Harker’s hap paper on the Geological Structure of the Sgirr By OE TS Be eiacs Meow: coe ecto sioe Wea Cease sis he tet eee MicroscoPe-sEcTIONS OF Rocks FROM THE iy Vv Maprira, ete., to illustrate Dr. J. W. Evans’s 88 paper on the Rocks of the Cataracts of that and ObMET IVES yeahs sbuecucacse ee veisns ceeeaneee eae ( SEction across SatisBuRY PLAIn aND THE ALDER- ) BURY SYNCLINE; and SECTION FROM NEAR Braltsu- | FIELD THROUGH FartEy Down, to illustrate} 132 Mr. A. J. Jukes-Browne’s paper on the Clay- | WADI HUMES: oa eye Eel cnah aouewe nee a2 saemetetanimede ) VI | { | \ ( Map or THE Piora Laxzs; Lago Rirom sHEn From } THE West; Laco Tom viEWED FROM THE NorTH- | East; THE o1D Foos Vanuey and Laco Tom, | LOOKING SOUTH FROM THE WATERSHED ; THE VAL | Piora AND Lago CADAGNo, SEEN FROM pee | Laeo Tom; Sprin@ Issuing FROM THE RAUCH- | WACKE, BELOW Laco Tom, anp Lago | | LOOKING NORTH FROM THE WATERSHED; THE | Western Enp oF Lago Rirom, ann Foup all VII-XXI4{ = =o rue Dotomirez, sourn or Lago Tremoreio; tHE} 165 Exit or Lago TREMORGIO, SEEN FROM THE sOUTH- | west, and Lago LucenDRO sEEN FROM i Soutu ; View or Lago LucrenpDRo, SHOWING THE | Exit, anD Lago DELLA SELLA, SBEN FROM THE | SourH-west ; ConrourED MAPS OF Lago Ritom, | Lago Tom, Lago Capacno, Laco Scuro, Lago | Tanepa, and Lago pi Lucenpro, to illustrate Prof. H. J. Garwood’s paper on the Tarns of the | Canton Wicino mes vs.t sar .cssdj-cneres ete Suoarusccicne } EXPLANATION OF THE PLATES. Xili PLATE PAGE GrotocicaL Mar or THE NEIGHBOURHOOD OF XXII Luptow, to illustrate Miss G. L. Elles’s and 195 Miss I. L. Slater’s paper on the highest Silurian Rocks /Gfithat Gisthict! WIAA 6H....cd.ccocsececmencese (Gmonocrcan Map or tHE IeNnzous AND ASSOCIATED \ | Rocks or tHe Luaneyyoe District; and Micro- | XXIII-XXVI4_scorz-sncrions or THE Ienxous Rocks, to illus- + 223 | trate Messrs. T. C. Cantrill’s and H. H. Thomas’s | Ne paper omitlant tem eect eneet can -d-an ee ay aveeseoene== ) MIcROSCOPE-SECTIONS OF THE GRANITIC MARGIN oF THE BurreRMERE LaccoLitE; and oF THE XXVIT & XXVIII | ENNERDALE GRANOPHYRE, to illustrate Mr. R. H. 253 Rastall’s paper on those rocks ................0006 Avonian CoRALS AND BrAcHIOPoDS FROM RusH (County Dustin), to illustrate Dr. A. Vaughan’s 295 Notes on the Palxontology of the Carboniferous BOCK Olt Hat LOCALILY. >. eaetinie od sreeeeiis wae seaien gueienes XXIX & XXX (Avontan CorAts AND BRrAcHIOPODS FROM THE \ Menpie AREA; SKETCH-MAP SHOWING THE OUT- CROP OF THE CARBONIFEROUS LIMESTONE; and XXXI-XXXV4 = - RanGE-piacRrams or Corats anD Bracutopops 1n } 324 ‘| tHe Menpip Area (Somerset), to Leama Mr. T. F. Sibly’s paper on the Avonian of that | NP CIS Ieincas. ote creat atu ta aes seislutnane mans aide Cetidae ) ( Gmotogican Map or tHe Dunepin District (New \ | ZpaLaAND); and Mucroscore - sections oF | XXXVI-XXXIX<{ Icneovs Rocks From tHE Dunepin District, } 381 | to illustrate Prof. P. Marshall’s paper on the | We ceolopy: Of tab ameay js sieacado dans'sndchieeanatande ) TRILOBITES FROM Botivia, to illustrate Mr. P. i { ake’s paper om those fossils... s.2-+-serbences } — DeveLorpment oF Dipuyorps, to illustrate Mr. S. XLI S. Buckman’s paper on Brachiopod Homeo-} 483 TREO YO Ny ease torae he eee eee avs eah cata de anak SECTION NEAR WHITLEY (NoRTHUMBERLAND), to illus- XLIt trate Prof. G. A. Lebour’s and Dr. J. A. Smythe’s 530 paper on a Case of Unconformity and Thrust in ihe, Coal Meastres se aette sou tech imece dete dee Proterozoic Srecres or OmosrirdA, LorPHosPrra, XLII & XLIV AnD TvReirou4, to illustrate Mrs. G. B. Long- + 552 staff’s paper on those fossils ...............cceceeeee { Liassic DenTALimp2#, to illustrate Mr. L. Richard- ) any 1 son’s paper on those fossils .................... A rcte ip oi8 ( Mar or Tue QOrpovicran Rocks or WEsTERN XLVI CaERMARTHENSHIRE, to illustrate Mr. D. C.$ 597 | Hvans’s paper on those rocks ...............seeeeeee -Xiv EXPLANATION OF THE PLATES. PLATE PAGE ( GroLoeicaL SKETCH-MAP OF THE TARANNON CoUNTRY; ig and HorizonTaL SECTION ALONG THE AFoN JAEN AND GENERALIZED SECTION FROM DOLGADFAN T XLVIIL& XLVIII 4 ENERALIZ S M OLGAD AN oO Hi | TaterpDpIeé across Newypp Fynyppoeg, to illus- trate Mrs. G. A. Shakespear's paper on the J Tarannon Series of Tarannon................escesees AND Erars, to illustrate Mr. F. Chapman’s and 702 Mr. ID. Mawson’s paper on the AHalimeda- \ HTALIMEDA-LIMESTONE FROM MALEKULA, SAN Santo, XLIX-LIi | Limestones of the New Hebrides ...... «on so eee CoRRIGENDUM. Page 390, line 18 from top, ézsert a comma after the word “ stream.” PROCESS-BLOCKS AND OTHER ILLUSTRATIVE FIGURES, Fie. 10. a, Cr BESIDES THOSE IN THE PLATES. PaGE [Diagraminatic section from Kirkstone Fell to Howgill Me DM ae se cteaeta ceieercitat deere Renee easton ealors Tealnsenad uallgoa Pa vicsoe Ixix [Diagrammatic plan of the central portion of the Lake Asie ewe sce shee oe rae am came eanid 1 ondemoss lxx Diagrammatic section across Hdenside, from the north of the Lake District, to the Pennine range 2. ......2.-..<-jec sweat ease vadens-(ss 628 GPa Section West ol MlanddOWEOE: ca. 0cnsnesccss vere sees seedasarae*vaene on 628 Section along the Tarannon: River 2.6.0.1. .ds1.<.<-) 1.37 s.M. (principal earthquake). In addition to these, several disturbances of uncertain seismic origin are reported :— April 23rd, about 1.39 a.m.: Barton-on-Humber.—A very slight shock, similar to that of the principal earthquake, and accompanied by a sound like distant thunder. April 23rd, about 2.5 a.m.: Farnsfield and Normanton-on-Trent.— A slight shock, without any attendant sound, was felt at these places, both of which are more than 15 miles from the principal epicentre. At Farndon (near Newark) subterranean noises are said to have been heard on several occasions, especially on April 27th, at 11.30 p.., and on April 29th, about midnight. Il. Forr-SuHocx. (a) April 23rd, about 1.30 a.m. Number of records 2, from 2 places. A slight shock was felt at Epworth, and a rumbling sound was heard at Norton. Epworth lies between the two portions of the 1 Geol. Mag. 1904, pp. 536-37. * In the investigation of this earthquake, I have received great and welcome assistance from the following gentlemen, to whom my best thanks are offered :— Dr. Tempest Anderson, F.G.S., Mr. L. W. Bunting (of South Carlton, near Lincoln), Mr. G. Hibberd (of Laughton, near Rotherham), Mr. J. Dennis (station-master at Sutton, near Retford), the Rev. F. 8S. F. Jannings (rector of Warmsworth, near Doncaster), Mr. T. J. Moore (of Hatfield, near Doncaster), Mr. E. P. Richards (of Hathersage), and the Rev. A. P. Woodhouse (vicar of Tuxford), and the superintendents of police of the Dewsbury, Eckington, and Pocklington Divisions. The expenses of the investigation were defrayed from a grant received from the Government Research-Fund. 6 DR. CHARLES DAVISON ON THE [ Feb. 1906, isoseismal 7 of the principal earthquake, and Norton about 8 miles north of Doncaster. At about the same time, a rumbling noise was heard at Bassingham (near Lincoln), and a shock was felt at North Collingham (near Newark). In the absence of records from inter- mediate places, it is uncertain whether the last two observations refer to a fore-shock connected with the Doncaster earthquake. IT]. Principat EartrHevake. (b) April 28rd, 1.37 a.m. Intensity, 7; centre of south-western portion of isoseismal 7, lat. 53° 26°5' N., long. 1° 1:2’ W.; centre of north-eastern portion of isoseismai 7, lat. 53° 36-5’ N., long. 0° 43°7' W. Number of records, 1428, from 662 places, and 68 negative records from 66 places. Time of Occurrence. Excluding estimates that are given approximately, the total number of records of the time is 695. Of these, 110 are regarded by their observers as accurate to the nearest minute. The average of the latter is 1.37 a.m., and, as the averages for the zones included between successive isoseismals differ from this by less than half a minute, it is probable that the time of occurrence at the epicentre was almost exactly 1.37 a.m. Isoseismal Lines and Disturbed Area. The five continuous curves on the map of the earthquake (PI. IT) are isoseismal lines of intensities 7 to 3; the broken-and-dotted line represents the boundary of the area disturbed by the earthquake of April 13th, 1902. The isoseismal 7 consists of two portions, which are approximately circular in form. The south-western and larger portion, which is the most accurately drawn of the series, is 18; miles long from north-east to south-west, 172 miles wide, and 244 square miles in area. Its centre lies half a mile north of Bawtry, in lat. 53° 26:3’ N., long. 1° 1:2’ W. The course of the north-eastern portion is some- what uncertain towards the south-east, but probably does not vary by more than a mile from the true position. As drawn, it is 934 miles long from north-east to south-west, 83 miles wide, and 63 square miles in area. Its centre is about 4 miles east of Crowle, in lat. 53° 36°5' N., long. 0° 43-7' W., and lies 17 miles north-east of the former. The centre of the disturbed area of the earthquake of 1902 is in lat. 53° 33-4’ N., long. 0° 41:5’ W., or about 34 miles south-south-east of the centre of the north-eastern portion of the isoseismal 7. The next isoseismal, corresponding to intensity 6, is roughly elliptical in form, 58 miles in length, 44 miles in width, and 2050 square miles in area. The direction of its longer axis is EK. 38° N. and W. 38° 8. Its distances from the south-western portion of the isoseismal 7 are 14 miles on the north-west side and 123 miles on Wol: 62. | DONCASTER EARTHQUAKE OF 1905. t the south-east side, and, from the north-eastern portion of the same curve, 14 and 16 miles respectively. The remaining isoseismals are somewhat less accurately drawn, and, in parts, may possibly vary as much as 2 or 3 miles from their true position. The deviation cannot, however, be of much consequence. ‘The isoseismal 5 is 90 miles long from north-east to south-west, 76 miles wide, and contains 5300 square miles; its distance from the isoseismal 6 is 14 miles on the north-west, and 18 miles on the south-east side. The isoseismal 4 is 126 miles long from north-east to south-west, 108 miles wide, and about 10,700 square miles in area, and is distant 16 miles in both directions from the isoseismal 5. The isoseismal 3 is 166 miles long from north-east to» south-west, 130 miles wide, and about 17,000 square miles in area, its distance from the isoseismal 4 being 14 miles towards the north-west and 21 miles towards the south-east. I have also received records from several places out- side the isoseismal 3—from Liverpool, Lingen (in Herefordshire), Hugglecote and Winchcombe (in Gloucestershire), Soham (in Cam- bridgeshire), and Norwich; but I do not feel sure that the move- ments observed at these places were connected with the Doncaster earthquake. The area apparently disturbed by the earthquake was therefore about 17,000 square miles. Nature of the Shock. The twin-character of the shock was clearly recognized throughout a district overlapping the isoseismal 5 by a few miles in every direction, and was sensible to some observers as far as, and in three cases beyond, the isoseismal 4. Over the whole disturbed area, 32 per cent. (or roughly 1 in 3) of the observers who noticed closely the nature of the shock detected either two maxima in a continuous series of vibrations, or two detached series of vibrations separated by an interval of a few seconds. The percentage varies in different districts, being 31 within the isoseismal 7, 38 between the isoseismals 7 and 6, 32 between the isoseismals 6 and 5, and 14 between the isoseismals 5 and 4. These variations and the com- paratively-small percentage are due: (1) at places within the isoseismal 5, and especially near the epicentres, to the perception of a continuous tremor between the two parts of the shock; and (2), at places outside the isoseismal 5, to the enfeeblement of the vibra- tions composing the weaker part. ‘Thus, at Balne (9 miles north of Doncaster), the movement was continuous and contained two maxima, the first of which was the stronger; at Sheffield, the intervening tremor was less sensible, and the shock consisted of two detached parts separated by an interval of one second, the first being slightly stronger than the second ; at Thornton Curtis (9 miles north-east of Brigg), two series were felt, of which the second was the stronger, the interval between them being 3 or 4 seconds, and also two distinct sounds were heard, of which the second was the louder and separated from the first by 2 or 3 seconds; and, lastly, at 8 DR. CHARLES DAVISON ON THE [Feb. 1906, Humberstone (2 miles east of Leicester), the shock consisted of a single series of vibrations, lasting from 3 to 4 seconds, the intensity of which increased to a maximum and then died away. Owing perhaps to the variable perception of the intervening tremor, the estimates of the length of the interval between the two parts of the shock lie between rather wide limits. The average of 110 estimates is 34 seconds. There is, however, no evidence of the existence of a synkinetic band,’ within which the two parts of the shock coalesced, bordered by bands in which the shock consisted of two maxima of intensity. Thus, the interval between the occurrence of the two impulses must have been greater than the time required for the vibrations to travel from one focus to the other. That the two parts of the shock did notediffer greatly in intensity is shown by the magnitude of the area over which the twin-shock was felt, and also by the very variable testimony of observers with regard to the relative intensity of the two parts, each part being regarded as the stronger by approximately the same number of observers. If we treat all the observations as of equal value, no law in the distribution of relative intensity is apparent; and it is only by rejecting all records, except those made by observers who were ‘awake at the time and who evidently attended carefully to the phenomena, that any such law can be ascertained. It is then seen that the first part of the shock was slightly the stronger over the larger part of the disturbed area; while the second part was the stronger within a small and nearly-circular area indicated by the broken line (*) on the map of the earthquake (Pl. IL). This area, which is about 20 miles in diameter, includes the centre of the north-eastern portion of the isoseismal 7, its own centre lying about 6 miles to the north-east. Origin of,the Double Shock. It is evident that the two parts of the shock originated in two detached foci; though, in the intervening region of the fault, there must have been a slight displacement sufficient to account for the widely-felt tremor connecting the two series of vibrations. The two epicentres cannot be far distant from the centres of the two portions of the isoseismal 7, and are therefore separated by a dis- tance of about 17 miles. It follows, also, from the relative intensity of the two parts of the shock, that the impulse within the south- western focus took place a few seconds before the other and that it was slightly the stronger, the latter inference being confirmed by the larger size of the south-western portion of the isoseismai 7; further, that every point of the disturbed area was reached first by the vibrations from the south-western focus. The second part of the shock was the more intense within the circular region near the north-eastern epicentre, owing to the proximity of the corresponding fOCUS.; l Quart. Journ. Geol. Soc, vol. lxi (1905) p. 21. 2. Ibid. p. 28, fig. 3. Wols 62.1). DONCASTER EARTHQUAKE or 1905. 9 Seismographic Record. The earthquake was registered by-the Omori horizontal pendulum at Birmingham, which city is 75 miles 8. 28° W. of the principal epicentre. The record is too minute and ill-defined for reproduction, the range or double amplitude of pae largest waves (the second, third, and fourth) being less than 4) inch, and corresponding to a movement of the ground of about =55 inch. 1 The remaining vibrations are represented by mere notches on the trace, and in parts these cannot be separated. Altogether, there appear to be about thirty vibrations in about 15 seconds. The exact time of the first movement I am unable to determine. Sound-Area. The boundary of the sound-area is indicated on the map (PI. IJ) by the dotted line. Except towards the north-east, it overlaps the isoseismal 4 by a few miles, extending 11 miles beyond it towards the south-west, and falling short of it towards the north-east by 3 miles. It is 185 miles in length, 115 miles in width, and about 12,000 square miles in area. For a strong earthquake, the sound was heard by an unusually- large proportion of the observers: the percentage of audibility for the whole disturbed area being 93. The same high percentage is also maintained to a considerable distance from the epicentres : being 94 within the isoseismal 7, 94 between the isoseismals 7 and 6, 93 between the isoseismals 6 and 5, 86 between the iso- seismals 5 and 4, and 67 between the isoseismal 4 and the boundary of the sound-area. Nor is there any sensible variation with the direction from the epicentres ; for, in attempting to draw isacoustic lines by the method employed for the Derby earthquakes of 1903 and 19U4, the change in audibility from one square to another was found to be small and irregular.” There is no trace of any strength- ening of the sound in the directions at right angles to the longer axes of the isoseismals—a result which is in agreement with the absence of a synkinetic band. Nature of the Sound. The total number of descriptions of the earthquake-sound is 1061. In 43 per cent. of these, the sound is compared to passing traction- engines, motor-cars, etc., in 29 per cent. to thunder, in 12 to wind, in 4 to loads of stones falling, in 4 to the fall of a heavy body, in 6 to explosions, and in 2 per cent. to miscellaneous sounds. ‘These It is probable that the actual displacement of the ground was greater tnan this, for the deposit of soot on the paper clearly hinder red the movement of the pendulum. 2 With a very large number of observations, it would of course have been possible to draw the isacoustic lines. But when, as in this case, the number of records from places within any square is not great and the number of negative records very small, one more or one less negative record causes a considerable change in the percentage of audibility. The irregular variations in audibility are probably to be explained in this manner. 10 DR. CHARLES DAVISON ON THE [ Feb. 1906, figures agree closely with those obtained for the Derby earthquake of 1904." The variation in the nature of the sound with the distance from the epicentres is shown in the following table (1), in which the figures are percentages of comparison to the different types for each of the districts mentioned :— a | | = | @ Ea ai © | & Ss =) = | 9 o : ) a) | Ra a = e eS | e* je at S _ Taste I. Fie ess swe paliesg ne. c |) c= = a Les) e | om SS ry mM —_ = S| FS lea cea eae eS S 3 as} re) a ia “ 2 ~Lincoln OHorncastle -¢ a= do oN.Collingham Bassingham nsfield° pak o CNewark i Farndon Boston/4Z— 6 Nottingham a Sleaford Grantham, <2 == Oo : °Loughborough ; Spalding Humberstone o “oe ster © = eo Moz Po Seale of Miles ‘ 0 Se 20 2h je Cd CY Northampton THE DoncASTER HARTHQUAKE OF APRIL 23RD, 1905. o write ¢ ¥ ’ : * , - 4 e { t * 4 gene ‘ € ’ ‘ et j t eee i ‘ ‘ ‘ ret ; Yas bok ‘all h Ate ‘ ik in Ly , ft A 5 Quart. Journ. Geol. Soc. Vol. LXII, Pl, II. * Boundary of Area in which the Second Part of the Shock was the stronger, Scarboroughd Driffield© Harrogate 3 O Keighley Hornsea Beverley Rochdale Wigan Manchester Rotherham us Faye s iverpool f SN B saSitield \C; Aion des Chesterfield Le ic Normanton, 2° °Lincoln -on-Trent sfield, oN.Collingham q . Farnsfield® Bassingham o CNewark Farnd oAshbourne piteese Boston/ fo} Stoke-on-Trent Nottingham <; “Sleaford 3 7Thornham oDerby Creal Zoe" ae “Hleacham Burton-on-Trent o ©Loughborough Lichfield Humberstone Gale atey Leicester° Wolverhampton® Walsall “".. oBirmingham a Mee, € : Seale of Miles CLABES EN 5 10 15 20 26 fa) Northampton Stratford-on-Avon MAP ILLUSTRATING THE AREA AFFECTED BY THE DoNCASTER HARTHQUAKE OF APRIL 23RD, 1905. paca eee stone Vol. 62. ] THE GLACIAL PERIOD IN ABEL DEENSHIRE, ETC. 13 3. The Guactan Period 72 ABERDEENSHIRE and the Sovurmern Borver of the Moray Frets. By Tuomas F. Jamieson, F.G.S. (Read November 8th, 1905.) ConTENTS. Page TL, TEIDIELCOYC (CLC) A RRS Ie RA ates ae anc Beene rs nei ne ean oe ee 13 IT. The Red Clay of the Aberdeenshire Coast--its Origin and Deri- TWENSICINS 3S woh yaaa eee rab nen oor Geran a Me ec ee ee 13 III. Relation of the Chalk-Flints to the Red (Cle eG aren cama. aeons hte tee 20 MeO nest onsolySubmergeMmee Gece etic ne in sce cee cede gece hoes cad cvs deems 21 VY. Succession of Beds—an older Boulder-Clay discovered ............... 21 hale Wie Chalk-EPlints of Aberdeenshire “..........-..-+00.5.aeee+sese0ac00 bes. oe VII. Explanation of the Absence of Shell-Beds in the Red Clay ......... 24 ainineilteoslastelee=Slieeb. ts, tetas screen ve oe tak seeeiete non tudo cn eclossen ben tile doesent 25 IX. The Crag-Shellsand Chalk-Débris ofthe Red Clay—their Derivation. 27 The Southern Border of the Moray Firth. RewvwWarks ot Glaciation om, thre ROCKS” ..si0..6c0-co.ecs0s8 des codecsacceenee Daf XI. Transport of Boulders—Jurassic Débris and Chalk-Flints ......... 28 Mae txient anc Depthtot thewdlce (2.5.06) gcse. sare -ose eee obde sade sacmnnn oe 30 XIII. Southern Border of the last Ice-Sheet where it crossed the Spey ... 30 XIYV.. Period of Submergence—Shell-Beds ................0.cccscceeees eereccees 32 XV. The Dark-Blue Clay of the Banffshire Coast and Northern J ORRC SSIS NINES a Seeder tin 2rd le eter on ae berets e a ene ee 34 I. IyrropuctorRy. TuHat north-eastern angle of Scotland which lies between the Moray Firth and the Firth of Tay presents some special features of interest in connection with the history of the Glacial Period, inasmuch as it seems to have been less heavily covered with ice than the rest of Scotland. Nevertheless, no part of the area appears to have escaped invasion, for the ice which filled the basin of the Moray Firth extended over all the southern border of that basin, and even overflowed a considerable part of Aberdeenshire, spreading southward until it met those streams which, proceeding from the Grampian Mountains, moved eastward along the valleys of the Dee and the Don. No part of the surface, therefore, escaped abrasion by the ice ; and no remnant of pre-Glacial Tertiary deposits has hitherto been found in any part of the area, even in those spots where one would have thought that they might have been most likely to be preserved. Neither has any trace been obtained of the mammalian fauna which inhabited England during the inter-Glacial Period. No remains of the Mammoth, not even a tooth or a tusk, have been met with in any of the railway-cuttings or other ex- cavations (so far as I am aware), although such have been found in Ayrshire and in the Basin of the Forth. Il. Tur Rep Cray or tom ABERDEENSHIRE CoAstT—ITs ORIGIN AND DERIVATION. One of the most interesting features in the Glacial Geology of 14 . MR, T, F, JAMIESON ON THE [Feb. 1906, Aberdeenshire is the Red Clay which occurs along the eastern coast of that county. Some account of it will be found in vol. xxxviii of this Journal (1882) p. 160, where I endeavoured to show that it had been formed during a time when there was a flow of ice along the coast from south to north, caused apparently by the approach of the Scandinavian Glacier, which had the effect of blocking the natural outlet of the Scottish ice to the east, and compelling it to turn along the coast nearly at right-angles to its natural course. This northward flow brought with it red sediment and multitudes of stones from the sandstone-rocks and associated volcanic beds of Forfar and Kincardine. A subsequent visit to that district enabled me to ascertain that the mineral character of the volcanic fragments which occur in this Aberdeenshire Clay, is identical with that of the rocks along the coast between Montrose and Lunan Bay. Here I may mention that so long ago as December 2nd, 1840, Sir Charles Lyell? remarked that the observations of Mr. Blackadder and himself led them to infer, that a great flow of glacier-ice had at one time gone along the Valley of Strathmore from Dunkeld north-eastward to the sea at Lunan Bay; a circumstance which harmonizes very well with the further transport from Lunan Bay northward along the coast to Aberdeen. If this diversion of the Scottish ice from its natural direction was caused by the approach of the Scandinavian Glacier, it enables us to connect the Glacial history of Scotland with that of the Norwegian ice when it attained its great spread to the west. I shall therefore give some further particulars regarding these Red-Clay beds, as new opportunities of studying them have arisen since | wrote my paper on the subject. A line of railway has been formed from Ellon to Cruden, right through the district where they are best developed, and in the cuttings of this line some interesting sections were brought to view. Large excavations have also been made near Peterhead, in connection with a convict-prison and a harbour of refuge; while near Aberdeen fresh diggings have been opened up in the clay-pits, as well as excavations for some deep sewers Close beside them. These red beds exhibit a great variety of character. Sometimes they consist of thick masses of pure clay, which is often finely laminated, a proof of tranquil deposition in water of some depth. Sometimes thick beds of fine sand and gravel are associated or inter- stratified with this clay. In other places the clay is of coarser quality, being mixed with stones, and showing no appearance of stratification; in which cases it often presents a hard firm texture, quite like a Boulder-Clay. All these varieties, however, pass occasionally into each other, as if they belonged to one and the same series. The character of the sections changes frequently and abruptly, sometimes within a few yards. The arrangement is not often in straight regular beds, but is generally undulating, often rapidly so. Finally, there are in some places esker-like mounds of gravel, such as those beside the Loch of Slains, which contain the * Proce. Geol. Soe. vol. iii (1842) p. 342. Vol, 62.] | GLACIAL PERIOD IN ABERDEENSHIRE, BTC. 15 débris of Crag-shells and fragments of Secondary limestone. These mounds, I find, form part and parcel of this Red-Clay Series, for the Red Clay occurs both above and below them. From the foregoing account it will be seen that this Red-Clay Series presents features somewhat resembling the Drift-deposits of Lancashire and Cheshire, and like them it occasionally contains sea- shells, generally more or less broken. In certain localities, as at Aberdeen and Peterhead, clay of a different colour is interstratified or mixed with the red, show’ ing that there has been in such places a mingling of sediment from different sources. The esker- like mounds of gravel occur along what seems to have been the western border of the mass of ice which caine northward along the coast. These mounds may have been formed by streams of water tumbling from the margin of the glacier, and washing off the mineral débris embedded in or lying on the ice. The purer masses of clay seem to have resulted from ciouds of muddy sediment subsiding in a sheet of water lying in front of the ice, between it and the land. In connection with these Red-Clay beds one noteworthy circum- stance is the evidence which they afford that, at the time when this northward flow took place along our coast, the Aberdeenshire ice was shrinking, or had already shrunk, much within its former limits. During the preceding period, when the subjacent grey Boulder-Clay was deposited, the Aberdeenshire ice came down to the coast, and even advanced beyond it; but, during the deposition of the Red Clay, this local ice had made a decided retreat, and the coastal district was under water. Specially was such the case in the Valley of the Dee. This river takes its rise among the mountains of Braemar, the highest group of hills in Britain, and flows along a well-defined valley terminating in the sea at Aberdeen. It might therefore be thought that here, if anywhere, we should have had a very large and persistent glacier. Nevertheless it is quite clear, that during the deposition of the Red Clay the Glacier of the Dee had receded some distance up the valley ; forthe Red Clay lies right across the mouth of the stream at Aberdeen, and wasted patches of it have been noted for about a couple of miles up the bed of the river. This shows that the Aberdeenshire ice must have been comparatively thin, so that it melted away long before the heavy ice which covered the Western Highlands and the valleys of Perthshire. There is nothing unintelligible in this, for the Tay and its main tributaries take their rise in the rainy district of the West, where the fall of snow during the Glacial Period seems to have been very heavy ; whereas Aberdeenshire and its rivers lie on the drier side of the country, where the fall of snow was doubtless much less. Therefore, from meteorological considerations alone, we might expect the local ice there to have been much thinner. Consequently, the. glacier that filled the Valley of the Dee would shrink long before the glacier which came down the Valley of the lay, and along Strathmore to Lunan Bay. - 16 MR. T, F. JAMIESON ON THE [Feb. 1906, In Strathspey, which has a comparatively slight rainfall, asimilarly- early shrinkage took place, as we know from the phenomena of the Parallel Roads of Glen Roy, and other facts which J shall mention farther on. In proof of the state of matters at the mouth of the Dee, we have excellent evidence in the clay-pits at Aberdeen. These were formerly worked on both sides of the river, but of late only on the south side at Torry, where the clay has long been dug for the manufacture of drainpipes and bricks. JI have visited the excavations at intervals for the last forty years or more, and am consequently well acquainted with the varying nature of the sections displayed there. . The river at Torry is bordered on the south side by a ridge about 150 feet high, which consists of granite and gneiss covered by Glacial Drift. The bed of fine clay, used for industrial purposes, occupies a sort of hollow nestling in the northern flank of this ridge, and thins out at an altitude of about 50 or 60 feet above the sea-level. It rests upon a grey Boulder-Clay which is rarely exposed to view, owing to the rise of water in the bottom of the pits. The lowermost portion lying immediately on the top of this Boulder-Clay is of a similar grey colour, and contains some irregular masses of sand and gravel of a like hue. The thickness of this grey portion, when best displayed, I found to be about 14 feet. This grey clay is delicately laminated by the interposition of films of mica between the leaves of clay. Above it comes a mass of red clay of a stiffer and more waxy nature than the grey, owing to the much smaller proportion of fine sand which it contains. This red clay at Torry attains in some places a depth of 14 feet, thinning out towards the ridge and thickening towards the river. Where it meets the subjacent grey clay, the two are generally interstratified for some distance. Above the red there is a few feet of grey clay, similar in colour to that which hes beneath, but of coarser texture and without appearance of lamination. his upper grey clay is often altogether absent, owing to the denudation which the top of the beds has undergone. A mass of ferruginous pebbly gravel of later origin, from 4 to 7 feet thick, lies upon the denuded top of the clay. The red and grey clays at the bottom are remarkably free from stones ; often not a pebble of any kind is to be seen, but now and then a stone or two does occur, generally of no great size. No shells have ever been found in this Torry Clay, so far as | know. From the red clay I have, by washing and sifting, extracted one or two specimens of foraminifera, but they seem to be rare, and I did not meet with any ostracoda. In the grey clay I found no organisms of any kind. These clays at Torry have all the appearance of having been deposited in very tranquil water, after the Glacier of the Dee had retreated some distance up the valley. The stillness of the water may have been caused by a thick covering of ice, which would account also for the absence of stones. The grey bottom-clay was probably formed from the fine mud proceeding from the end of the Dee Glacier, but the red clay marks the incoming of sediment from a Nols 62..| GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 17 totally-different quarter, when the Glacier of the Dee had apparently retreated still farther back. In the old clay-pit on the north side of the river opposite Torry, the section was very similar,’ but the beds were deeper, the Red Clay being 17 feet thick. This pit is now closed, and those at Torry also. The Red Clay and sand, however, attain their greatest development to the north of Aberdeen, in the neighbourhood of Ellon, and in the coast-parishes of Slains, Cruden, Peterhead, and St. Fergus, which lie to the north of the River Ythan. On the lands of Ardiffery, near the parish-church of Cruden, the depth reaches 100 or 150 feet, and may be even more. Here a large proportion of the mass consists of sand, much of it very fine-grained and free from stones ; varying in colour from reddish to pale-grey. Other portions are coarser and more pebbly, with some small débris of yellow and grey limestone. Crumbs and dust of Crag-sheils” can generally be detected where this limestone-débris occurs in the sand and gravel. Beds of Red Clay are met with in any part, both above and below the sand, while seams of the two are often interstratified. The character of the surface here is very moundy—large swelling mounds, but these do not run in long narrow lines like eskers. The stratification in their interior is generally undulating and irregular. The sand and gravel looks as if it had been shot down in great heaps. Much of it is well washed and free from muddy sediment, and there is a marked absence of big stones and boulders in the interior of these sand-beds; but a few big stones seem to occur on the surface in all parts of the district. It seems to me that the explanation of this heavy mass of material being lodged here, may be that great streams of water flowed off from the glacier at this place, washing out the mud, sand, and mineral-débris embedded in the ice. Farther off from where the border of the glacier seems to have existed, we find the deposit to be more of a clayey nature, lying in flatter widespread sheets ; whereas, along what had been the margin of the ice, the surface is moundy, the accumulations much deeper, and the stuff consists mostly of washed gravel and sand, sometimes of very pebbly gravel, with occasional masses of coarse mud or Boulder-Clay. In one of the railway-cuttings at Port Errol, in the parish of Oruden, there was a bed of Red Clay from 30 to 40 feet thick, having much the same composition from top to bottom. It showed no distinct stratification or lamination, and contained some small stones dispersed through it. At Belscamphie, which forms the north-western boundary of the pebbly gravel that contains the débris of Crag-shells (Joc. cit.) and Secondary limestone, the section was as follows. At the bottom was a mass of Boulder-Clay of a very dark indigo colour, containing small fragments of crystalline schist and granite. This rests upon the schistose rocks of the district, which crop out a little farther on in the * Quart. Journ. Geol. Soc. vol. xiv (1858) p. 510, * Ibid. vol. xxxviii (1882) p. 145. Q.J.G.8. No, 245.. c 18 MR. T. F. JAMIESON ON THE [Feb. 1906, exposure. The Boulder-Clay has an irregular undulating surface, and varies in thickness, but is not more than a few feet deep where the bottom is exposed. Above it lay a bed of Red Clay, followed by a bed of fine clean-washed sand. Then, on the top of the sand, came the rough gravel containing pebbles of yellow limestone, sandstone,. and crumbs of (Crag) shells. In some places this stuff reached nearly to the top of the cutting; in others it was covered by a few feet of Red Clay and gravel. The total depth of the section was about 16 feet. The chief interest lay in its showing that the pebbly gravel containing débris of Crag-shells forms a member of the Red Clay Series, and that the whole lies above a bed of Boulder-Clay of quite. a different hue and character, containing not a vestige of the limestone, sandstone, or shelly débris. This section is near Pitlurg: railway-station. In some of the projecting parts of the coast the Red Clay is often quite like a Boulder-Clay, being very firm and unstratified, with small stones dispersed through it. This is well seen at present along Peterhead Bay. Close to the south side of the town the shore is formed of low rocks of red granite, which are covered by a mass of hard grey Boulder-Clay, 10 or 12 feet thick, containing stones of gneiss, quartz, and granite. The surface of the rock does not show marks of glaciation, so far as [could see. In passing along the Bay southward, the grey Boulder-Clay is very soon observed to become covered by a mass of Red Clay, which gradually attains. a thickness of about 10 feet. Itis nearly as hard and firm as the grey clay beneath, and contains a good many stones, some of which seem to have been derived from the Old-Red-Sandstone Conglo- merate. It has all the appearance of a Boulder-Clay, such as would be found in a red-sandstone district. On advancing still farther along- the Bay southward, one sees the granite gradually sink out of sight, and the grey Boulder-Clay thin nearly quite away into a mass of coarse granite-rubble. The Red Clay, on the other hand, grows. thicker; and at Invernettie, about a mile south of Peterhead, where- the bay curves more inland, the granite and its covering of grey Boulder-Clay have both sunk out of sight, but the Red Clay attains. a much greater development. The section is not now exposed as it was some years ago, when. the tile-work was in full operation, but when best displayed I found it to be as follows. At the base was a mass of fine stratified sand, which seemed to descend below sea-level, but was not much exposed. On the top of this sand was a bed of fine laminated brown clay (what the workmen called leaf-clay), which varied’ in thickness from 1 to 4 feet. Above this was a mass of clay, attaining in some places a thickness of 15 feet, of a decided red colour; generally of fairly-pure quality, but showing no clear stratification, and having some stones dispersed through it. The- top of this bed exhibited an undulating outline, and was covered by an irregular seam of coarser stuff from 6 inches to 3 feet thick, composed chiefly of small débris of gneiss and mica-schist. Above: Vol. 62.] GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 19 this was a deep mass—20 feet thick or more—composed of darkish mottled brown clay, varying much in colour, as if a red and a blue had been jumbled together. It was quite unstratified, and contained more stones than the Red Clay beneath. These stones are of a great variety of kinds, and among them are some large ice-worn boulders, from 2 to 4 feet long. Above this came from 3 to 4 feet of coarse pebbly clay, of a more ferruginous tint, with a foot or so of arable soil at the surface. The total height of the bank is about 65 feet above the sea. The relative thickness of the beds varies in different places, but the section is now quite obscured by slips. In some places here the Red Clay is curiously streaked with clay of a dark bluish colour, derived apparently from a different source. In one spot, near the base of the section, I found a boulder of granite 2 feet in diameter, sticking in the fine sand and reaching partly up into the laminated clay above it. This stone was encased in a skin of bluish-grey gritty mud, 1 or 2 inches thick, quite different from the clay surrounding it, as if the stone had been covered with this mud when it dropped out of the ice, and fell down into the clay at the bottom of the water. In the excavations for the convict-prison, which adjoins Inver- nettie on the south side, an irregular undulating band of blue clay was found in the midst of the red, accompanied occasionally by a bed of gravel, which contained a few broken shells. Many blocks of red sandstone were met with in the cuttings for the prison ; likewise a bit of hard Chalk, with a cylindrical fossil in it about the size and thickness of a cigarette (? Belemnitella), and stones of many different kinds. Among the broken shells, I observed Cyprina islandica, Pecten islandicus, Astarte arctica, Panopea, Mytilus, Cardium, Fusus, and Balanus.* I have noticed a similar bed of blue clay among the red near Cruden Bay. Itis generally very undulating, of irregular thickness, and contains ice-scratched boulders. ‘To the north of Peterhead, in the parish of St. Fergus, dark-blue clays are more developed, and are found embedded in the red, which shows that they belong to approximately the same period. North of St. Fergus the Red ~ Clay disappears, and is replaced by clay of a dark indigo colour, sometimes nearly black. ‘This dark clay extends along the coast westward, past Banff, on to Portsoy and Cullen, and indicates a transport from north-west or west. Since writing my former paper on the Red Clay I have found a few more instances of glacial markings on the rocks along the coast of Cruden, running in a general direction from south to north, or parallel to the coast; but they were exceedingly slight and of no great extent. Those at Murdoch Head, pointing towards the Buchan-Ness Lighthouse, are still the best that I have been able to discover, and are unequivocal. The 1 Further details regarding the clay-beds near Peterhead will be found in Quart. Journ. Geol. Soc. vol. xiv (1858) p. 518. c 2 20 MR, T. F, JAMIESON ON THE [Teb. 1906, rock there, however, is being quarried away, so that ere long they may disappear. ‘hese markings are at the top of the cliff, and are covered by red Boulder-Clay, at an altitude of 100 feet above the sea. The glaciation of the rocks along the coast of Slains, Cruden, and Peterhead seems to have been so slight as to be scarcely observable. Perhaps the ice was not very thick, or was buoyed up by the water which submerged the coast at that time, or it may have had little motion. The granite below the grey Boulder-Clay at Peterhead shows a like absence of glaciation. At the Govern- ment Quarry, on the west side of Stirling Hill, the rock, when newly uncovered, does show indications of abrasion by an agency coming from south or south-east, but I could find no undoubted scratches or scoring by the ice. Ill. Retarion or turn CHxartk-F irs to run Rep Cray. In the patches of Red Clay found sporadically on the top of Stirling Hill and the adjoining eminences, fragments are common of the crystalline schists, which form the rocky coast of Slains and Cruden lying to the south-east at a lower level. This shows that there had been an upward transport over the hill from that quarter during the time of the Red Clay. All this harmonizes with the general bearing of the evidence formerly detailed. More interesting evidence, however, is derived from the distri- bution of the Chalk-flints which cover the ridge running inland from the sea at Buchan Ness. These flints lie generally thickest on the top of that ridge, but extend also down both sides of it. I find, however, that for about 2 miles inland from the coast they have been swept off, apparently by an agency moving across the ridge from south to north. ‘The end of the ridge next the sea is called - Stirling Hill, the adjoining summit to the west is called the Hill of Longhayen, and the next Coldwells. They are all composed of red granite, and rise to a height of about 300 feet above the sea, forming a continuous ridge: the last-mentioned portion attaining an altitude of 329 feet, which is the highest summit. All over the southern slope of these heights the flints are entirely absent, and on the top and eastern end of Stirling Hill the same is nearly the case. On the top of Longhaven Hill a stray flint or two may be found, and they become more numerous as we approach the northern brow of the ridge, but on descending the north side they immediately become plentiful; while, in the Den of Boddam, which is a little, narrow, winding ravine running along the north side of | Longhaven Hill, they lie in vast profusion. The same is the case on the farm of Sandford Hill, on the north side of the Den. All over this space the flints lie thicker than I have seen them in any other part of Aberdeenshire, or, indeed, of Scotland, forming in some places large mounds known as the Saddle Hills. Blocks of granite, some of them 4 to 6 feet long, lie on the top of the flints along the northern shoulder of the Longhaven ridge, as if Vol. 62.] GLACIAL PERIOD IN ABERDEENSHIRE, Ere. 21 they had been swept down from above by the same agency as that which carried off the flints. At the granite-quarry on the west side of Stirling Hill, there is a like absence of flints on the crest of the ridge, but they are found plentifully in the hollow to the north of the quarry, where they are mixed with much coarse granitic sand, which seems to have been scoured off the rock by the passage of the ice. At the Hill of Coldwells, although there are no flints on the south side, these make their appearance as soon as we reach the top, where they are in many places quite plentiful. The bare granite crops out all over the summit of Longhaven and Stirling Hills, especially along the seaward front of the latter. At Coldwells, which lies farther inland and attains a somewhat greater height, the denudation has been less complete, and the rock is scarcely exposed. Another proof of the northward movement of ice along the east side of Aberdeenshire may be mentioned. When the railway- cutting at New Machar Station was in progress, I observed that the edges of the nearly-vertical beds of gneiss were bent over at the surface from south or south-west to north or north-east, as if by the passage of ice coming from the southward. ‘This station is about 9 miles north-north-west of Aberdeen. Esker-like mounds of gravel may be traced a little inland along the coast from Belhelvie northward, throughout the parish of Foveran, on to the village of Newburgh on the River Ythan. These mounds occa- sionally have some Red Clay on the top, and contain fragments derived from the red sandstones. They probably mark the border of the stream of ice which came along the coast from the south, during the time of the deposition of the Red Clay. LV. Question or SUBMERGENCE. It is quite clear to me that this Red Clay has been brought to Aberdeenshire by a drift of ice from the south, at a time when the coast was submerged beneath water to a level exceeding 300 feet above the present coast-line. Whether this submergence was caused by a depression of the land beneath the present sea-level, or by a sheet of water hemmed in between the ice and the land, is not quite clear. The evidence afforded by the Red-Clay beds in Aberdeenshire is not conclusive on this point, for as yet they have failed to yield any instance of a bed of undisturbed sea-shells in what would seem to be their original habitat; although such instances do occur to the southward at Montrose, Errol, Elie, and probably some other localities. In Aberdeenshire, the few shells met with in the Red Clay are generally more or less broken, and may have been transported from a distance; it is true that a few whole ones have also been found. V. Succrsston or Beps—awn Orprr Bourprer-Cray DIscovERED. The whole of our Glacial Beds in Scotland are, no doubt, later than the Forest-Bed Series of the Norfolk coast, perhaps later even 2 MR. T. F. JAMIESON ON THE [ Feb. 1906, than the Chalky Boulder-Clay of that region. In the Ellon district of Aberdeenshire the regular succession, beginning at the surface, is—(1) Upper Boulder-Clay and gravel; (2) the Red-Clay Series, consisting of clay, sand, and gravel; and (3) grey Boulder- Clay, which usually rests upon the old rocks of the district, and seems to have been lodged by a sheet of ice coming from about west-north-west. Recently, however, I have found remains of a still older Boulder- Clay of a very distinct character. ‘These instances occurred near Ellon railway-station, between the line of railway and a low rocky eminence called the Craigs of Auchterellon. This eminence consists of a mass of gneiss jutting out across the valley on the north side of the Ythan, and rising to a height of nearly 100 feet above that stream. Some wells have been sunk at the east side of these rocks, and it was in digging them that this older clay was discovered. After getting through the Red Clay and the underlying grey Boulder-Clay, a mass of very dark indigo-coloured clay was reached, of a finer quality, with fewer and smaller stones; but what was most remarkable, was the circumstance that it contained small fragments of sea-shells, generally mere crumbs, but some few of them in such condition as enabled me to recognize Cyprina islandica, Astarte arctica, and a bit of a Balanus. Moreover, some of them showed clear marks of glacial rubbing, like those met with in the Caithness Drift. Now, the grey Boulder-Clay here has never yielded any trace of shells or other organisms. It has always proved barren, with no remains of animal or vegetable life. This dark blackish clay is, therefore, evidently a distinct and older deposit, and its occurrence in this particular spot has no doubt been due to its lying on the east and lee side of the rocky eminence, which sheltered it from the demolishing action of the ice-stream that lodged the grey Boulder-Clay, whose flow was here from west to east. Many other wells have been sunk at Ellon, but it is only in those close to the east side of the Craigs that this older black clay has been found; and it is thickest in those wells which are nearest to the east side of the rocks. It is also the case that the Red Clay is thickest and best preserved on the east side of similar eminences, and for the very same reason : these having sheltered it, to some extent, from the denuding action of the later ice-sheet, which lodged the Upper Boulder-Clay and gravel. In the well nearest the Craigs the black clay was 20 feet thick, and then the bottom was not reached. In another, a little farther off, about the same thickness was got, but here the bottom was apparently touched. It was covered by 20 feet of the grey Boulder-Clay, above which came 6 feet of waxy Red Clay, and 6 inches of arable soil at the surface. In both wells the black clay contained remains of shells, and was of similar character. The discovery of this old dark Boulder-Clay, so distinct in its features, and lying below the grey (which is the bed usually found next the rock in this quarter), brought to my recollection the fact Vol. 62.] GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 23 that I had noticed a somewhat similar dark clay in a like position in other places, but not showing any shell-fragments. Thus, in a railway-cutting a few miles south of Ellon, I had observed that the lowermost part of the section was composed of a very dark indigo- coloured clay, in which I found a piece of fossiliterous shale con- taining the impression of a small ammonite, and what seemed to be fish-scales. The bottom-clay in the railway-cutting at Belscamphie, on the Cruden line, mentioned on p. 17, is also very like the dark bed at Craigs of Auchterellon. I found, however, nothing remarkable in it, although I did get a crumb or two of what might have been shell, which effervesced on being treated with acid. I could also mention other localities in this quarter, where a very dark bluish clay occurs beneath a Boulder-Clay of greyer hue. I therefore think that there may have been a fairly-widespread dark clay of older date than the grey, and formed by a previous phase of the ice. The action of a thick sheet of glacier-ice moving over the surface of a country would be something like that exerted by a carpet dragged over a floor, and,if long-continued, would leave very little if any older loose stuff behind it. Its plastic nature and heavy pressure adapt it so closely to all irregularities of the floor on which it rests, that hardly any place can escape its friction; and this accounts for the entire absence of all remains of pre-Glacial Tertiary deposits in Scotland. ach recurrence of an ice-sheet would tend to wipe out the material left by its predecessor, especially if the later one was heavy and leng-continued. Although, then, we have some evidence of more than one re- currence of an ice-sheet in this part of Scotland, no evidence has hitherto been obtained of warm intervals, further than what may be inferred from the melting-away of the vast mass of ice which preceded and followed the deposition of the Red Clay and the shell- beds at Clava and elsewhere. It must have taken a great deal of heat to melt these enormous masses. The deposition of the Red Clay seems to have followed close upon the retreat of the ice which lodged the grey Boulder-Clay. There is no sign of any interval there; but a long interval, and possibly a warm one, may have occurred between the time of the Red Clay and that of the later ice-sheet which lodged the Upper Boulder-Clay and gravel. The blackish shelly clay at Ellon must, I suspect, have come down the Ythan Valley from the Moray-Firth direction, and may be of the same age as the dark clay at Maud, to be afterwards mentioned (p. 29), which contains Oolitic fragments and pieces of serpulite- grit from the North-Western Highlands. ‘These latter probably came down the valley of Loch Shin in the great ice-streqam which seems to have descended along that route into the Moray Firth, and pressed on over the southern border of that basin. Their occur- rence in the northern part of Aberdeenshire is a very curious and interesting fact, which we owe to the observation of Mr. John Milne, formerly tenant of the farm of Atherb. 24+ MR. T. F, JAMIESON ON THE [ Feb. 1906, VI. Tur Cuatx-Fiints or ABERDEENSHIRE. In regard to the beds of Chalk-flints which occur in Aberdeenshire, when I first saw them, more than forty years ago, I fancied that they might have been derived from a bed of Chalk that had formerly existed in the district. They are highly waterworn, and always accompanied by a quantity of quartz-pebbles which are likewise intensely waterworn. These I supposed might have been derived from the quartz-rocks that occur in the district. The flints and quartz-pebbles, however, extend for many miles along the top of a ridge of granite; and there [remarked that the granitic débris had none of this water-rolled character. The staurolite- and mica-schists (which form another part of the range of heights on which the flints lie) also show a similar absence of water-rolling. This feature, in short, is confined to the flints and the quartz-pebbles associated with them. It would, therefore, seem that both have . been transported from a distance in company, and that the water- worn features had been impressed upon them before they were brought to the district where they now lie. They do not occur on any part of the large spreading hill of Mormond in the northern extremity of Aberdeenshire, so far as my observation goes; but are mostly confined to a narrow helt of country running in a general east-and-west direction to the coast at Buchan Ness, a little to the south of the town of Peterhead. More than one phase of the Glacial Period has occurred since the flints were brought to Aberdeenshire, for this took place before the epoch of the Red Clay. They have accordingly been swept away, and re-arranged in many places by later developments of the ice and other denuding agents. VII. Expranation oF tHE ABSENCE OF SHELL- Bens IN THE Rep Cray. The general absence of marine fossils in the Red Clay of Aberdeenshire may, perhaps, be accounted for in the following way. During the preceding stage of the Glacial Period, when the subjacent Boulder-Clay was laid down, the Scandinavian ice (as James Croll first poited out) appears to have occupied the shallow bed of the North Sea and coalesced with the Scottish ice, so that the sea-water was excluded from all the East Coast of Scotland, and nothing but ice prevailed. Now, when the thaw began to set in, the thinner ice which lay over the eastern and northern parts of Aberdeenshire would no doubt melt first, and the water which took its place would be fresh or brackish, so long as free communication with the ocean was shut off by the Scandinavian Glacier. No migration of marine life into it would therefore be possibile, until the latter glacier began to recede. It might have been a considerable time before that took place; consequently, the RedClay would be deposited in a sheet of water hemmed in by the ice, and it would only be at a later stage that any marine forms of life would be able to gain admittance. This later stage is probably represented by the clay- Vol. 62.] GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 25 beds at Montrose, Errol, Elie, etc., where we find remains of Arctic mollusca and asteroidea, apparently in the place where the animals lived and died. ‘The clay-beds at Aberdeen, as I have shown, afford good evidence that the glacier of the Dee Valley retreated, before the ice which came along the coast from the south gave way. Thus we actually have some facts in support of the explanation which I have suggested, while the absence of ordinary marine conditions presented by these beds seems to render some such explanation necessary. VIII. Tue wast Icu-Saeet, Subsequent to the Red-Clay epoch, the inland ice again made a great advance, and apparently for the last time. The most satis- factory evidence of this in the Aberdeenshire district, is to be found at the seaward end of the Dee Valley and the coast immediately to the north of it. Im this area there is clear proof, to one who is intimately acquainted with the ground, that the Dee Glacier came down to the coast with a breadth of at least 5 or 6 miles. In doing so, it destroyed the Red Clay along the greater part of that width, leaving some patches at the side of its track, but clearing it out completely along the central portion of its route. The River Dee, at the lower end of its course, is bordered on the south by a continuous ridge of higher ground than it is on the north. The effect of this was to shunt the end of the glacier off to the north-east, so that, although the river now terminates at the city of Aberdeen, the central current of the ice-stream was diverted a mile or two farther north, crossing the present mouth of the River Don. This is well shown by the clear marks of the ice on the surface of the granite at the quarries of Cairncry and Persley, and also by the fact that the Red Clay has been wiped out most completely in that direction. Some patches have been left beside the present mouth of the Dee, as at Torry, Ruthrieston, and the Duthie Park. The large mounds of coarse gravel and stones, on which much of the city of Aberdeen is built, are moraines of this stream of ice which came down the valley of the Dee.. The Broadhill, the Castle- Hill (on which the barracks are built), Ferryhill, and many other eminences now covered with houses, are of this nature. In some of these mounds, masses of the Red Clay are to be found which have been dislodged by the ice and mixed with the moraine-gravel, as, for example, on the west side of the Broadhill. Some large ice-worn boulders occur in these mounds, generally near the top. One of these may be seen in the quadrangle of Marischal College, where it was placed a good many years ago, I believe by Prof. Nicol; but most of these big stones have been used up for building- purposes long ago. The country around Aberdeen, before its reclamation, was so very rugged and encumbered with stones, that it is difficult now to realize its former appearance. Dr. James Anderson, who wrote a good sketch of the district in 1794 for the then Board of Agriculture, tells us that. it was of the most barren nature that could anywhere 26 MR. I. F, JAMIESON ON THE [Feb. 1906, be seen; but that, about 50 years before the date of his report, a commencement had been made towards improving it. The expense of trenching and clearing the ground of stones and afterwards manuring it, often amounted to £50 or £60 an acre before a crop could be put in. Some fields, indeed, he says, cost £100 an acre. Yet, strange to say, the operation proved a profitable one. The demand for granite-stones to London about that time was so great as to help materially to make the enterprise pay. On the lands of Nigg and Loirston, on the south side of the Dee, where the right flank of the glacier rested, the big stones cleared off the ground are so numerous that they have to be piled into what are locally termed consumption-dykes. These consist of two parallel stone-walls, 4 to 6 feet high, with a space of 10 to 12 feet between them, into which the stones are piled. These stones are of all sizes, up to 3 or 4 feet in length, and some larger. Asan instance Imay give the following: from 12 acres of land on the farm of North Loirston, the quantity of stones taken off formed a consumption-dyke (as I was told by the late Prof. Dickie) 300 yards long, 30 to 35 feet broad, and 6 feet high; in addition to which smaller stones were got to fill the furrow-drains, which were made 24 feet apart. This last glacier of the Dee Valley must have protruded beyond the present coast-line. Its northern flank lay in the parish of Belheivie, about 5 or 6 miles north of Aberdeen, where there is a great accumulation of gravel-mounds running inland, from a place on the coast called the Black Dog, to Parkhill in the parish of New Machar. These gravel-mounds rest upon the wasted top of the Red Clay at Milden and the Black Dog. At the latter place the clay is now dug for the manufacture of bricks and drainpipes. The right flank of the glacier lay on the hills of Nigg, which form the southern border of the Dee. In the intermediate space between Aberdeen and the Black Dog, the Red Clay has been swept clean away, causing a gap in it 4 or 5 miles wide. liven in the little valley of the Ythan, there is evidence of Glacial action subsequent to the Red-Clay epoch. This valley is about half way between Aberdeen and Peterhead, and takes its rise among hills only a few hundred feet in height; consequently, it is difficult to believe that it could have originated anything like a true glacier. Perhaps the facts can be explained by the melting of deep masses of snow and ice, or it may be that a lobe or offshoot of the Moray-Firth ice came down this way. Anyhow, the Red Clay has been much wasted, and covered in places by masses of coarse gravel, and sometimes by a profusion of large stones 2 to 4 feet in length. At Michael Moor, a few miles to the west of Ellon, there is a sheet of this gravel in which the stones are so large and plentiful, that the agriculturist in clearing the ground has been obliged to form them into broad consumption-dykes, and farther up the river something like miniature lateral moraines occur. At Ellon, in some places big stones 3 to 4 feet long are found in great number, just at the bottom of the graveland lodged on the top of the fine Red Clay. af Vol. 62.] GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 27 In other parts of the country, the result of the last ice-sheet has generally been to form at the surface a thin covering of coarse mud, charged more or less with stones. In the lower coast-district this lies on the top of the Red Clay, and derives its colour partly from it; but, along the course of the rivers and minor streams, there is always more or less washed gravel, formed by the currents generated by the final melting of the snow and ice. IX. THe Crac-SHELLs AND CHaLK-DEBRIS oF THE RED CLhay— THEIR DERIVATION. The Crag-shells, and the débris of Secondary limestone accom- panying them, which are found in some of the gravel-beds and clay of the Aberdeenshire coast, have evidently been brought thither by the ice-stream which produced the Red Clay, and may therefore have been transported from a very great distance. It has to be kept in view that pieces of chalk, sometimes iceworn and scratched, likewise occur in this Red Clay. They are met with at Montrose, Belhelvie, and also, though rarely, near Peterhead. Small pieces of coal have also been got in the same clay. The Cretaceous fragments have probably been brought to the east coast of Scotland by the Scandinavian ice, which passed over strata of Chalk in part of its route; and possibly the débris of the Crag may also have been brought by the same agency. The bits of chalk are not uncommon in the Belhelvie clay and also at Montrose. It is not likely that they could have come from England, as the movement of ice to the south of the Firth of Forth was such as would have carried them in the opposite direction. Their occurrence at Montrose and Belhelvie tends to prove that this clay in Forfar and Aberdeen is all one and the same deposit. THe SouTtHERN BorpvDER OF THE MoRAY FIRTH. In this district, the valleys have a general trend from south-west to north-east, as will be seen from a glance at any good map. The Ness, the Nairn, the Findhorn, the Spey, the Deveron, are the principal rivers, and they all flowin this direction. It might therefore be supposed that, during the Glacial Period, the ice would move down to the coast along these lines; but there is ample evidence to show that in some part of that period a great transport of boulders and mineral matter took place along the coast from west to east, right across the lower ends of these river-valleys, while marks left by the ice upon the rocky framework of the country, together with the drift of transported stones, prove that in some places the movement was from north-west to south-east. * X. Marks oF GLACIATION ON THE Rocks. The finest display of this sort that I have seen in the district, is on the top of a low hill called the Carden Moor, not far from 28 MR. T. F. JAMIESON ON THE [Feb. 1906, Alves railway-station, about 6 miles west of Elgin. The late Dr. Gordon (of Birnie) took me to see it many years ago. The ridge lies east and west, and rises to a height of 250 or 300 feet above the level of the sea. The rock along the top is bare, and composed of a hard, fine-grained, grey sandstone of Devonian age, which has been strongly rubbed by the ice. The scratches and furrows point about 10° or 15° north of west, so that the line of movement crossed the ridge obliquely. Although the surface is unprotected by any covering of earth or clay, the markings are in many places as fresh and clear as if they had been made but a few years ago. That the movement came from the west-north-west is shown by the transport of numerous fragments of the stone for many miles to the south-east. On the sandstone of Quarrywood Hill, near the manse of New Spynie, Dr. Gordon showed me other instances pointing west 15° north, and on a hill near Burghead, west 30° to 35° north. The surface of the Cornstone beneath the Boulder-Clay at Links- field Quarry, near Elgin, is also strongly glaciated, and even polished by the ice, the scores and scratches running from north- west to south-east. In one place, I found the direction almost due north and south. Along the coast from Speymouth eastward to Fraserburgh, I have not met with any good display of glaciation on the rocks, but merely a few traces here and there, near Melrose Head, Crovie, Troup, and Aberdour. On the cliff at Troup Head I found some markings running east and west. At Kinnaird Head beside the town of Fraserburgh, the indications are more decided, showing a movement of ice coming from the north-west. The same direction is maintained round the corner of the coast to the fishing-village of St. Colms. XI. Transport oF BovtpErs—J uRAssic DHBRIS AND CHALK-FLINTs. The eastward movement of boulders along the southern shore of the Moray Firth has long been known to observers in that quarter. At Linksfield! a huge mass of Oolitic beds, 40 feet thick and several hundred yards in length, has been transported bodily, and lodged on the top of the iceworn surface of a Devonian limestone, locally known as Cornstone, there being a layer of red Boulder- Clay 1 to 4 feet thick between the two. Other large transported masses of a similar nature have been found in the Elgin district at Lhanbryde, Spynie, and elsewhere. At Plaidy, on the Turriff railway-line in Aberdeenshire, a great mass of greenish-blue Oolitic clay, brought probably by ice from the shores of the Moray Firth, has been stranded on the top of the slate- rocks of the district. It is so big that a tile-work has been established in it, and worked for many years. It was first discovered in making 1 See the late Major L. Brickenden’s account of it, Quart. Journ. Geol. Soe. vol. vii (1851) p. 289. a” ol. 62. | GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 29 the railway in 1858. At that time I sent a notice of it to the Geological Society," in which I termed it an outlier of Lias, but it is neither more nor less than a huge transported mass of clay, and is enveloped in Glacial Drift of a different colour. Judging from the character of the ammonites and other fossils which it contains plentifully, it bas probably been derived from some bed belonging to the horizon of the Oxford Clay. It is much wasted on the outside, and partly mixed with the Glacial Drift surrounding it, but its full extent in some directions is hardly yet known. Fossiliferous fragments from the Lias, the Oolite, and other Secondary rocks are found here and there over the country eastward, from Elgin and Speymouth across Banffshire, on to the coast of Aberdeen, beside Peterhead. Near Maud railway-station, in the north of Aberdeenshire, there is a small farm called Atherb. Mr. John Milne, the former tenant of that farm, is a man with the eye of a hawk for all manner of curious stones. He made a wonderful collection from the fields around it. Many of the pieces found by him are of a fine-grained grey sandstone containing Oolitic fossils ; but what is still more interesting, he got several fragments of the Pipe-Rock, one of the characteristic Cambrian beds of the North- Western Highlands. I have myself found a small piece of shale, containing the impression of an ammonite and other fossils, em- bedded in the Boulder-Clay of a railway-cutting a mile or two south of Ellon, which is the farthest point in a south-easterly direction where such have been obtained. Fragments of the Gamrie sandstone and conglomerate have been carried eastward along the coast towards Fraserburgh, and may be seen at various places along that line, and at the northern base of Mormond Hill. The fossiliferous Greensand débris, which occurs at Moreseat in Aberdeenshire, proves to be also transported, probably from the _ north-west, and is embedded in Glacial Drift.2 The Chalk-flints found so plentifully in that quarter have also, I suspect, been brought from the Moray Firth by the same glacial agency. These flints, as I have said, are found chiefly along a belt of country running across the north of Aberdeenshire in an east-and- west direction, terminating at the coast near Peterhead, and may have been shed off along the southern border of one of the streams of ice which brought so great a quantity of other débris from the Moray Basin. They are generally most plentiful at the surface, except where covered by peat. Frequently they are mixed up with some gritty earth or glacial mud, or have some of it beneath them ; but, at the western extremity of the district where they occur, they are covered in some places by Boulder-Clay of a later date, containing ice-scratched stones. This I observed near Delgaty many years ago, in 1858.: it shows that there has been a recurrence of glaciation after the fiints were laid down. 1 Quart. Journ. Geol. Soe. vol. xv (1859) p. 181. -2 See the Report of the Committee on Cretaceous Fossils in Aberdeenshire, Rep. Brit. Assoc. 1897 (Toronto) pp. 333 ef segq. 30 MR. 't.. F. JAMIESON ON THE [ Feb. 1906, No chalk ever occurs with these flints, but there is occasionally some whitish earthy matter of a non-calcareous nature, consisting of silicate of alumina, derived no doubt from the weathering and waste of the flint-pebbles. This has sometimes been mistaken for chalk. The north-eastern extremity of Aberdeenshire consists of low undulating ground, in the midst of which rises a large lumpy hill called Mormond, 769 feet high, composed of quartz-rock and staurolite-schist. At the western base of this hill is a low tract of coarse-grained grey granite. Blocks of this granite are scattered over the hill and far up it, to a height much exceeding that of the rock in place, while some of the smaller granitic débris is to be found on to the very top. There Lalso found stones of hard white quartz showing marks of strong glaciation, one of them being ground down to a smooth edge and streaked with fine scratches, showing that the abrasion had been most intense; but I found no flints on the hill. XII. Exrenr anp Depra oF tHE Ics. All this seems to indicate pretty clearly that the ice had gone over the top of Mormond. From this hill westward to Inverness, at the head of the Moray Firth, is a distance of 80 miles in a straight line. If the glacier extended all the way, and was high enough to pass over the top of Mormond, as it seems to have done, it would be at least 800 feet thick along the coast opposite that point, and probably more. Then, if we allow an average slope of only half a degree, or 46 feet per mile, along the surface of the ice to Inverness, this would amount to 3680 feet on the 80 miles, to which add 800 feet for the height at Mormond, making 4480 feet for the altitude at the head of the Moray Firth. But the average surface-slope of the Greenland ice, where Nansen crossed it, is a good deal more than this, so that, in all probability, the height at Inverness was greater than I have stated. XIII. Sournern Borper or tar Last Ice-SHEET WHERE IT CROSSED THE SPEY. There is some interesting evidence to show at what point the southern edge of this great ice-stream crossed the Spey at the last — time of its extension thither. Along that river, above Fochabers, the sandstone-rocks are of a deep red, and the Boulder-Clay derived from them is of the same hue. This red clay has been carried 1 The ice-sheet which filled the basin of the Moray Firth was at one time so extensive that its right flank flowed round the northern parts of Banff and Aberdeenshire, while its left turned round over Caithness, originating the shelly Boulder-Clay of that quarter. [Since this paper was written I have seen an article by Dr. W. Mackie, of Elgin, in Trans, Edin. Geol. Soc. vol. viii (1901) p. 91, on the distribution of erratics over Eastern Moray. It contains many interesting observations on the subject. At Ben Rinnes and some other places on the east side of the Spey to the south of Rothes, boulders are found which Dr. Mackie believes to have come across the valley from the west during some part of the Glacial Period. probably by the agency of land-ice.—T. F. J., Nov. 25th, 1905. | Wol..62;|| GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 31 eastward up the hollow traversed by the railway between Mulben and Keith. ‘The rocks along this hollow are not of sandstone, but consist of quartz-rock and clay-slate, which give rise to Boulder- Clay of a grey colour, quite distinct from the red. These rocks are covered therefore by grey Drift, above which the red is some- times visible. J traced this red clay as far on at least as Keith, so that the Moray ice must have moved eastward up this hollow. At Rothes, a little farther south, and about 12 miles from the coast, we get to the outer edge of the red clay, which has been driven past the boundary of the sandstone. Here a hill called Ben Aigan, composed of quartz-rock and schist, 1500 feet high, flanks the Spey on the east. Lying against the northern base of this hill, there is a high bank bordering the stream and rising about 180 feet above it. This bank affords, or did afford many years ago, a most instructive section. The top consists of well-washed sand and gravel of a greyish colour, much of the sand being so fine as. to be blown about by the wind. Beneath that is a great depth of fine stratified silt and sand, of a general greyish colour, in the midst of which, but well down, a mass of red Boulder-Clay makes. its appearance, stretching horizontally through the silt from the north, and thinning ont to the south in tongues and ribbons, which are more or less interstratified with the grey silt. North- ward, that is to say down the river-side, the red pebbly mud grows rapidly thicker, and occupies the whole depth of the bank, which lowers abruptly owing to the termination of the grey silt in that direction. We seem, therefore, to have here the very spot up to which the southern margin of the Moray ice had extended. It had apparently dammed the water and formed a deep pool or lake, in which the sand and mud coming down the flank of Ben Aigan and the Spey was quietly deposited. Here, as in the Valley of the Dee, the local glacier of the Spey appears to have shrunk back lone before the thick ice coming from the West Highlands gave way. This harmonizes well with the evidence derived from the Parallel Roads of Glen Roy, which showed that the water of the glacier-lakes. of Lochaber found an open way out by the Spey Valley when all the country to the west was still heavily clad with ice. Terraces of gravel occur on both sides of the Spey at Rothes, at corresponding heights, the top of them being fully 400 feet above the sea. They also run up the Mulben hollow, where there are deep. banks of Drift, in some of which I could distinguish a middle zone of red, with grey material both above and below. When the Moray ice-stream gradually shrank, it must have dammed the river at lower and lower levels. Accordingly we find remnants of gravel-terraces here and there, at decreasing heights as we proceed down the stream, at Cairnty and elsewhere. When I first examined these terraces 47 years ago, I thought that they had been caused by the sea,’ but no marine fossils have ever been found in any of them. These terraces 1 See Quart. Journ. Geol. Soc. vol. xiv (1858) p. 527. 32 MR. T. F. JAMIESON ON THE [Feb. 1906, probably date from near the close of the Glacial Period. At all events, they must be of later age (I should think) than the mollusea, the remains of which are got in the silt at Clava, to be mentioned in the sequel (pp. 33-34). XIV. PrExiop of SUBMERGENCE—SHELL-BEDs. We find that during a certain stage of the Glacial Period there was a marked retreat of the ice, and that water occupied the coastal district eastward from Inverness. In my paper on the Red Clay of Aberdeenshire, I adduced evidence to show that, from the mode of its deposition and the altitude to which it extends, the submergence in the district where that clay occurs must have exceeded 300 feet. As much of the country stretching round the corner of Aberdeenshire to the Moray Firth lies below this level, a submergence of that amount would extend into both districts. The question now arises, What was the nature of that sub- mergence? Was it caused by a depression of the coast below the sea, or was the district covered merely by a sheet of fresh water enclosed between the ice and the land? In the Red Clay itself, although we occasionally find a few scanty remains of sea-shells, these on the whole are rare, and generally more or less broken. No bed of them has been found in place. At Annochie, however, on the coast a little to the south of Rattray Head, where a tile- work formerly existed, there is a bed of fine dark-blue silt, passing underneath the beach and extending some distance inland, in which I found Arctic shells having all the appearance of being em- bedded in their native mud. Although the position is close to the present level of the sea, the assemblage of species is decidedly a deep-water one, and indicates a very considerable amount of submergence. ‘The prevailing forms were Nucula tenuis and Leda pygmcea, both occurring in a perfect state, with the olive-coloured epidermis quite unruffied. There were also specimens of the large Arctic form of Saxicava arctica, a Cylichna which appears to be the C. alba of Brown, likewise the little Bulla turrita of Moller (other- wise known as Utriculus pertenuis of Gould, var. turritus), and one or two small shells belonging to the genus Aainus (otherwise termed Lucina or Cryptodon), One of these seemed to be A. ferruginosus. Another was thought by the late Dr. Torell to be the Awvinus Sarsii of Philippi. All these species are Arctic, and go down into very deep water, as will be seen from the following table giving the range of depth according to G. O. Sars, in his ‘ Mollusca Regionis Arctica: Norvegie’ Christiania, 1878 :— Fathoms. ENUGUIGLCCHIUS VA ee ene, 20 to 3800 LENG PUGH © 5.0. Noncteenns. 20 ,, 6d0 SQHICOUG ATCLICA ...00c00.cciaee O.4,,. 7300 COVA GIGI CUCU TF ce erec 8 bo oe cee 10> ~ 3°) 300 / EIR GEGEN EDGE ps Ae ee RO Pe 10> 5 60 Axinus ferruginosus...- ....... 40 ,, 800 ALIS SATSIO Wend nee oe 60 ,, 800 Also a large specimen of Cornuspira foliacea and other foraminifera. Vol. 62.] GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 3d There was a complete absence, not only of littoral species, but even of those characteristic of shallow water, and likewise of broken shells. Owing to the decay of the shelly matter it was difficult to extract specimens satisfactorily. At King Edward, about 6 miles south-south-east of Banff, a - considerable number of Arctic shells (some 30 species or more) have been found in the banks of a small stream which is a tributary of the Deveron. These have been collected by Dr. Milne of that locality, Dr. John Horne, F.R.S., and myself. Most of them occur in confused beds of sand, gravel, and pebbly clay. The most abundant are Tellina balthica, Bela turricula, Natica groenlandica, and JV, islandica. Many of the univalves are entire, but the others occur only as single valves or broken fragments. A similar group of shells occurs in a bed of fine sand, on the Banffshire coast at Gamrie, and at a similar height above the sea. Lists will be found at the end of my paper on the ‘ Last Geological Changes in Scot- land.’? In one spot, however, of the King Edward banks, I found Arctic shells embedded in a fine bluish silt, apparently where the organisms had lived and died. The prevailing species was the Tellina proxima of Brown (T’. calcarea of Chemnitz). The specimens were of large size, with the two valves in conjunction and shut, part of the dark-brown epidermis still remaining on them. The valves were, however, more or less cracked, the upper one sometimes quite squashed, as if caused by pressure from above; but the fragments were not shifted out of their place. Streaks, too, of carbonaceous matter were observed near the shells, as if derived from the decayed seaweed. Leda pernula and a Natica likewise occurred, both still retaining the epidermis, but they were much rarer than the TYellina. The silt also yielded foraminifera and ostracoda. This bed of bluish silt is about 160 feet above the present sea-level, and 5 miles south of the coast of the Moray Firth. It must be borne in mind that, subsequent to this period of submergence, there was a recurrence of very intense Glacial con- ditions, as I endeavoured to show in my paper on the last stage of the Glacial Period in Scotland.” This return of the great ice- sheets broke up and destroyed the shell-beds, burying some of them under heavy masses of Boulder-Clay, as we see at Clava, in Cantyre, in the south of Arran, and in the later discoveries by Mr. Smith in Ayrshire, as well as in this case at King Edward. At Clava, 6 miles east of Inverness, the shell-bed is at an altitude of no less than 500 feet. There the group of species indicates shallow water, the commonest kind being the Littorina litorea. The locality was carefully investigated by a committee of the British Association in 1892, with the result that Mr. Fraser, the original discoverer, Dr. John Horne, the late Dr. David Robertson, and myself, all inclined to the belief that the shells were really in place, and indicated a submergence exceeding 500 feet. \ > Quart. Journ. Geol. Soc. vol. xxi (1865) pp. 197 et segg. 2 Ibid. vol. xxx (1874) p. 317. Q.J.G.S. No. 245. D B84 MR. T. F, JAMIESON ON THE [Feb. 1906, - This Clava bed is covered by a great thick sheet of Boulder-Clay, showing that intense Glacial conditions must have supervened after the shells were laid down; but the occurrence of marine mollusca there shows that at the time when they lived, the mass of ice which formerly filled the basin of the Moray Firth must have melted away as far at least as Inverness, at the head of the Firth. This is further proved by a remnant of shelly silt which I observed at Ardersier, near Fort George, 10 miles north-east of Inverness. The prevailing shell in it was Astarte sulcata, but I found also Leda pernula, and a specimen of Tellana which seemed to be 7. proaimea of Brown. The shells had been entire and showed the epidermis, but were much crushed and decayed, so that it was scarcely possible to get them out in a state suitable for examination. The silt con- taining them was very hard, as if it had been heavily pressed, and was unconformably enveloped by a mass of unfossiliferous Drift of quite a-different colour and character. ; XV. THe Darx-Bive CLay oF THE BANFFSHIRE COAst AND NorTHERN ABERDEENSHIRE. The clay-beds found along the northern coast of Aberdeen and Banff are generally of a dark bluish tint, often nearly black when wet. This colour of clay extends from Fraserburgh sporadically all the way west to Cullen, near where is a bed of it at Tochineal, which has been worked for tiles for many years. A little to the east of Portsoy, I noticed remains of Arctic shells scantily dispersed in this dark clay along the high banks facing the sea. ‘These shells were in a more or less broken state. The only species that I could make out were TVellina proaima, T’. solidula, Astarte borealis, and Leda pernula. The mass of clay resting upon the rock was about 30 feet thick, mostly all of a dark indigo tint, without appearance of stratification. Near the top it was brownier 1 in colour, with more stones and even some big boulders, but no shells. This dark indigo-coloured clay often contains Jurassic fossils, which may have been derived from strata existing somewhere in the basin of the Moray Firth ; and at Moreseat, in Aberdeenshire, débris of the Greensand occurs in a dark friable clay, possibly of the same age. ‘This dark clay, however, does not seem to be the product of the last ice-sheet, for it is covered at Moreseat and elsewhere by a later Drift of a brownish or reddish tint. On the estate of Cairnfield, to the south of Buckie, Mr. T. D. Wallace (of Inverness) noticed that the dark bluish-black clay was. covered by red Boulder-Clay; and Mr. Martin (of Elgin) mentions that, on the eastern slope of Coulert Hull, near Lossiemouth, red Boulder-Clay wraps round a bed of fine blue clay containing belemnites and other Oolitic fossils. Mr. John Milne at Atherb also remarked that the ice-sheet which brought the blue fossiliferous clay must have come first, as it is always lowest, the red or yellow clay when present lying Eber it: such, indeed, is the case near Brucklay Castle. , : ait Vol. 62:] GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 35 At. Turriff, 10 miles south-east of Banff, there is a mass of dark blackish-blue clay on the south-eastern brow of an eminence about 330 feet high, known as the Market-Hill. It has been dug for the manufacture of bricks and drainpipes, is a pure solid clay devoid of stones, and shows no distinct stratification or lamination. At the time of my visit, I found it to be 13 feet thick at the deepest place, and was told by the manager of the works that it passes down into a dark bluish silt or sandier clay. It has an irregular surface, as if it had suffered denudation, and was covered by a.few feet of gravelly earth of a lighter brownish colour. It is not seen on the western slope of the hill. No shells or fossils had been found in it. ‘There is a tract of low ground running from the eastern coast of Aberdeenshire up the Ythan Valley to Turriff, and then down the Deveron Valley to the coast at Banff. The summit-level of this hellow between the Ythan and the Deveron, where it is very narrow, is only about 180 feet above the sea, or perhaps a trifle less, wherefore a submergence of 200 feet would connect the basin of the Moray Firth with the sea on the eastern coast of Aberdeenshire by a channel of water crossing the country in a curving line from south-east to north-west. The Moray- Firth ice, when at its maximum, probably sent a lobe down this way, but during the period of submergence it would be occupied by water. The shells at Gamrie, on the Banffshire coast, occur in a deep mass of fine sand containing seams of fine dark clay, forming a large mound rising to a height of nearly 300 feet above the sea, and resting upon a cliff of sandstone-rock about 80 feet high facing the bay. In the course of two visits I collected nearly thirty species of shells. They seemed to be confined pretty much to one seam in the sand, but as there is a want of sections in the upper part of the deposit, they may very likely occur there also. The group, as a whole, closely resembles that found at King Edward. The com- monest were Astarte borealis, Cyprina islandica, Tellina proxima, Tedlina balthica, Cardium grenlandicum, Bela turricula, and Natica grenlandica. The shells are very tender and much decayed, conse- quently they are apt to go to pieces, especially the larger bivalves; but many of the smaller valves and the univalves are entire. In some places there is a good deal of shelly débris in the sand. The specimens of Tellina proxima are numerous, some few of them with the valves connected by the ligament and shut. These were filled with sand. Size generally large; many of the valves measured 17 to 13 inches in length. The Astarte was of all sizes down to very young spe- cimens, and in detached valves, the larger ones in fragments. ‘The Cyprina was of various sizes, always in fragments. Bela and Natica were often entire. All the shells had lost their epidermis. Frag- ments of Balanus also occurred, some foraminifera, and the otolith of a fish. The nature of the section I found to be as follows, cou eae at the bottom :— D2 36 MR. T. F, JAMIESON ON THE [Feb. 1906, Feet. 1. Red sandstone-rock, rising up from the beach to a herphinot about. ances sabwen sche eaasee vecekerm ap sana Sane eee 80 2. Fragments of clay-slate and grit, partly waterworn, and mixed with small earthy débris of the same............... 4 3. Compact, very firm, fine blackish clay, containing a few small stones, some of which are glacially striated ; also a few crumbs of shells too small for identification. There are likewise subordinate seams of brownish Sand Wh Cts CLAY —<.¥..cucse gees anaes ad aeacecn cotanh eae eee 8 . A thin seam of waterworn pebbles of clay-slate and grit., 0% . A thick bed of fine brownish sand ................0006. about 15 . A bed of blackish fine clay, similar to No.3... ...,....055 5 . A bed of fine brownish sand, in some places full of shelly MODEIS 5: pckpvnsdanerety wor -peeil eng cometanceiess cece ence eee . A bed of fine blackish clay, thinning out into sand. CO NO Oe Above this there were no good sections, but merely small openings in places, mostly showing fine sand. No Boulder-Clay appeared. The seams of clay vary much in thickness when followed laterally. As a whole, the mass of this mound has the appearance of being deposited in water, and may have been a portion of the sea-bed during the period of submergence. Many of the shells have been pierced by boring mollusks, as at King Edward. The King-Edward locality, however, seems to me one of more especial interest, inasmuch as it appears to afford good evidence of the demolition of marine strata by the subsequent action of glacier-ice ; for I found there the fine dark silt containing complete specimens of Tellina proxima, passing up into a sandier seam containing a greater abundance and variety of shells, among which Tellina balthica was one of the most plentiful. Above this lay a heavy mass of unstratified pebbly clay with ice-scratched stones, and in the lower part shell-fragments. This Boulder-Clay of King Edward is of a dark-grey colour, varies in depth from 9 or 10 to 30 feet or more, and along the valley is covered by a mass of waterworn gravel 10 to 20 feet thick. In one piace I found the Boulder-Clay passing down at the bottom into a bed of fine silt, with the strati- fication much deranged, and containing seams of gravel, broken shells, and shelly mud. The ice-scratches on the stones as they lay in this Boulder-Clay seemed as a rule to point west 5° to 40° north, which may afford some indication of the line along which the ice moved. Here, therefore, we seem to have caught the ice in the very act of smashing up the marine strata, and converting them into a shelly Boulder-Clay like that of Caithness. At a place called Black pots, on the coast 2 miles west of Banff, I examined an interesting exposure of the clay at the tile-work there. The top of the section rises to about 80 or 90 feet above sea-level, and the depth of material lying on the rock amounts to 60 or 70 feet. The greater part of the bank at the time of my visit was seen to consist of a fine, very dark, blackish-blue clay of a soft silty nature, which graduated laterally in the space of 20 yards into fine sand at one side of the section. This sand contains some ae [re — Vol. 62.] GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 37 thin seams of gravel and small pebbles, in which I observed many small crumbs of shell. A few bits of shell occur also at times in the clay, and here and there a small pebble. The top of the section showed a bed of somewhat different nature, about 10 feet deep, consisting of a coarser earthy and pebbly brownish clay. I saw a good many stones of various kinds and sizes lying in the pit, which the workmen told me came out of this coarse clay at the top. There were three big boulders of greenstone or syenite, one of them 5 feet in length, ice-worn and scratched; several small pieces of grit or quartz-rock; and several of Jurassic shale, some of them ice-worn and containing ammonites. Bits of belemnites were likewise to be seen, and pieces of red laminated sandstone, also granite of different kinds. The base of the fine clay and sand was not exposed, and probably some Boulder-Clay may lie between it and the rock. This black clay, as I have before mentioned, occurs here and there all along the coast-district from Fraserburgh westward to Banff, Portsoy, and Cullen; but, between Cullen and Inverness, the material covering the rocks is of a more varied nature, and generally lighter in hue. What has imparted this very dark tint to the Banffshire Clay is not very evident. Perhaps it may have been the waste of the slaty rocks and serpentine of that district, or of some Jurassic beds in the basin of the Moray Firth. This clay has evidently suffered much denudation since it was originally laid down, for in some of the little valleys that run inland from the coast I observed patches of it here and there along their sides, as in the Tore Glen near Troup Head; and in a gully at Northfield, in the same neighbourhood, I noticed wasted masses of it enveloped in a coarse Drift of browner hue. Ina railway-cutting at the south side of Fraserburgh, I also observed that this fine black clay was covered by a few feet of brown clay of a coarser nature, containing ice-worn stones. The dark clay may be traced right across the northern extremity of Aberdeenshire, where it ranges on to Peterhead, becoming gradually less black and more of a leaden-grey where it meets and mingles with the red. Along the top of the high cliffs from Banff eastward to Troup Head, there is little cover of any sort on the rocks, but at Melrose Head I obtained the following section, beginning at the surface :— Feet. . Coarse pebbly clay of a brownish colour..............sscsesees 3 . Ferruginous laminated clay, without stones ...............04. 4 . Fine black clay, like that of Blackpots, with few stones, but some of-themiice-marked | 6. joe le a aschaw veers daeevabeenee . Brownish débris of an earthy character, with some ice- MORTAL OR gc etic s ani> eioias aeons seleslsin sb aa gaainssineciea teva . Débris of slate-rock, often ice-marked............ccsseeceseseees 2 . Slaty rock, forming the cliff beneath. At Mill of Rathen, 4 miles south of Fraserburgh, a deep bed of bluish clay occurs, covered by a few feet of gravel. Fragments of Balanus and shells are found in this clay. The fine dark clay of the Banffshire ore)! He (JO NO ol 38 MR. T. F. JAMIESON ON THE [ Feb. 1906, coast seems to have been formed under conditions very similar to those under which the Red Clay of Aberdeenshire was laid down. It also ranges up to a similar or even greater height, being found occasionally at fully 300 feet above the present sea-level. Like the Red Clay, it often passes into, or is interlaminated with, beds of fine sand. Seeing that this dark-blue clay minglés with the red in the district where they meet, near Peterhead, there is every reason to believe that both belong to the same stage of the Glacial Period. Both have evidently been deposited under water, at least the finer- grained stoneless varieties. Both have undergone much denudation, and are frequently covered by coarser material containing ice- seratched stones. The dark-blue clay is the repositery of the Jurassic débris, which seems to have been derived from the waste of strata in the Moray-Firth basin; while the Red Clay of Aberdeen- shire is the repository of material from the Old Red Sandstone of Kincardine and Forfar, with its associated volcanic beds. It is inthe Red Clay that we also find the curious débris of Crag-shells and yellow Secondary limestone, the source of which is difficult to conjecture. The water under which these clay-beds were deposited must have extended all round the coast from -Aberdeen by Peterhead, Fraserburgh, and Banff, to the Moray Firth, seeing that they can be traced along the whole of that distance to near Inverness. If their deposition followed close upon the decay of the preceding cover of land-ice, we may suppose that they would form first where the ice was thinnest, and where consequently it would melt away soonest. This would probably be at the angle of the coast near Fraserburgh, while latest of all would be the clay-beds near Inverness (such as that at Clava), and those to the south of Aberdeen where the Red Clay came from. The remains of the shell-beds at Clava and Ardersier show that, at the time of their deposition, the sea-water must have occupied the head of the Moray Firth; and when such was the case, deposits of marine silt must have been formed all along its shores. What has become of them all? No other instances:are known along the head of the Firth, except the two that I have mentioned ; and it was by a mere lucky accident that the Clava one was discovered, Mr. Fraser having heard from some farmers that shells had been noticed in a clay-pit there, which led him to make a search. The recurrence of intense Glacial conditions, which is proved by the heavy bed of Boulder-Clay covering the shelly silt at Clava, affords an explanation of the destruction that has overtaken these marine beds. Although no other instances of their occurrence are known between Inverness and the mouth of the Spey, yet farther eastward, between Cullen and Fraserburgh, remains of Arctic shells are not so extremely rare. Now, this harmonizes with the belief that it was the recurrence of Glacial conditions which caused the destruction, for the ice would be heaviest and most destructive at the head of the Firth whence the flow proceeded. The preservation of the mass of stratified sand and silt at Gamrie, Vol.-62. |. GLACIAL PERIOD IN ABERDEENSHIRE, ETC. 39 in which the shells are found, was probably due to its sheltered position in the lee of the great rocky cliff of Gamrie Head. - The amount of submergence at Clava must have. been decidedly over 500 feet, which is greater than any of which we have evidence along the coast farther east. If the submergence was due to a depression of the land caused by the weight of ice laid upon it, then the depression should have been greatest where the ice was heaviest, namely, at the head of the Firth just where Clava is situated, and where, as [have mentioned, the depth of ice may have probably approximated to 5000 feet, or even more. Discussion. — The Prestppnr remarked that it would be strange indeed if a cordial welcome had not been accorded to a paper by one who had been a Fellow of the Society for nearly half a century, whose papers on Glacial geology had been frequently published by this Society — papers which, although theoretical questions were by no means avoided in them, were specially characterized by the sreat number and importance a carefully-recorded facts. Prof. P. F. Kmnpatt observed that the Author, whose name was one of the most-revered in Glacial geology, had added a new division to the Glacial Series of the district with which he dealt, in the shape of the dark clay with the deep-water Arctic fauna. It was very similar to a transported mass of Drift which he had seen and described in the Isle of Man. ‘Two problems still confronted the geologist in that district; one was the mode of deposition of the Glacial deposits, the other raised the question of the land-levels in that particular area during the Glacial Period. The distribution of the erratics, the striations, and the contents of the deposits seemed to group themselves in a remarkable manner. The direction of the striations was entirely corroborative of the lithological features: thus, the grey clay containing crystalline rocks derived from the interior of the country was striated from the south-west, while the other from the Old-Red-Sandstone area was striated from nearly due south. In the Inverness district, on the other hand, there appeared to be a low-level system of deposits associated with striz from a south-westerly or north-westerly direction, while the high grounds south of Clava bore striations running from due south. These two results seemed, at first sight, to be contradictory, and it would be most interesting to ascertain the mutual relations of the two parallel sets of strie. The time had now arrived when it became necessary for glacialists to group their facts with a freer hand, and to deal with larger stretches of Drift-country. The available evidence rather pointed to the influence of the Scandinavian ice- sheet, than to the wane of the Highland Glacier and the waxing of the Tay Glacier. As to the Clava shelly deposit, the speaker at one time concurred with the late Dugald Bell in declining to regard it as a marine deposit 7 situ. He subsequently abandoned that view, but it might be necessary, with this new evidence, to return. to his former opinion. 40 MR. A, HARKER ON THE GEOLOGICAL [Feb. 1906, 4. The Guotoctcat Srrvcrvure of the Sever of Eicc. By Arrep Harker, M.A., F.R.S., F.G.8. (Read December 6th, 1905.) [Puatss III & IV.] ConrTEnNTs. Page il SEMEVOGICEION 6. fi sieaedan sets veearenamea nan oepiedereneen heme 40 IT. General Structure of the Pitchstone-Complex...... 43 III. Fragmental Accumulations and Fossil Wood ...... 52 ty) Concluding Considerations):s..0.s..c.400-0-ce-see nee 64 I. Inrropvucrion. Tue Tertiary igneous rocks of the Inner Hebrides have from time to time engaged the attention of many eminent geologists; and the results of their labours, interpreted by their skill and insight, have become part of the common stock of British geology. In certain cases the investigation has been conducted on an extended scale. ‘This is true more especially of the work of Macculloch in the earlier days, and of Sir Archibald Geikie and Prof. Judd among those who are still with us. Nevertheless, it is only recently that any consider- able part of that large and complex area has been mapped with the close attention to detail requisite in a complete geological survey. Even apart from the time demanded by such a minute examination of the ground, the large-scale Ordnance maps which make it possible have been issued only within the last few decades. Thus, in the ten years which the present writer has spent on this work, it has been his good fortune to enjoy facilities denied to his predecessors, besides benefiting by the fruits of their labours in the same field. Fortified by these advantages, he has been enabled to pass with less diffidence from observation to inference, even when the latter is at variance with the views of geologists of much wider experience. For, as might be anticipated, the results of the detailed survey, while generally confirming the conclusions of the recognized authorities, have sometimes extended, restricted, or modified them, or have even led to a decided difference of opinion. The individual questions on which such difference of opinion arises are not many, but some of them have far-reaching consequences. They are, in all cases, questions the settlement of which demands both close examination of the ground and comparison of different parts of the region; and this is doubtless the reason why they have not been raised, at least in any very definite form, until the present time. The most important modification of hitherto received opinion in consequence of the results of the survey, is the recognition of the intrusive nature of a large part of the rocks which have been generally regarded as extruded lavas. The rocks in question include fully half of those which build up the ‘ basalt-plateaux” constituting the greater part of the Inner Hebrides. As illustrating Vol. 62. ] STRUCTURE OF THE SGURR OF EIGG, 4] the kind of evidence that is relied on, and as having also a direct bearing on the main subject of this communication, a summary discussion will not be out of place. Part of the evidence in the case of the Isle of Skye has already been set forth elsewhere,* and the conclusion there arrived at has been even more firmly established in the survey of the ‘Small Isles’ to the south. The basalt-plateaux are made up of a monotonous succession of sheets of basic rocks, individually from 2 or 3 to 100 feet thick, and reaching an aggregate thickness of sometimes 2000 or even 3000 feet. The rocks composing these sheets are, broadly, of two kinds, which may be distinguished as basalts and dolerites. The basalts are usually amygdaloidal, and more or less affected by secondary changes. They are comparatively soft and crumbling under exposure to atmospheric weathering, and are very largely concealed under superficial accumulations of Drift and peat. ‘They are of fine texture, and have typically the microstructure which has been styled granulitic. The dolerites are much fresher rocks, and much more durable, forming all the strong features in the plateau-type of landscape. Their escarpments show marked vertical jointing, sometimes as- suming a véry regular columnar habit. The rocks, excepting only the thinnest sheets, are of decidedly coarser texture than the basalts, and their microstructure is typically ophitic. These dolerites are, in aggregate thickness, at least equal to the basalts with which they alternate; and their salient outerop gives them an appearance of very decided preponderance. Such is the constitution of all the north-western half of Skye, of Kigg in general (excluding the upper part of the Sgurr), of Canna, Sanday, and Muck, and of the greater part of Mull. . My observa- tions go to prove that, while the basalts represent super- ficial outpourings of lava at an early stage of igneous activity, the intercalated sheets of dolerite are intru- Sive sills belonging to a later date. The dolerite-sheets run with remarkable regularity for very long distances, but occasionally they may be observed to shift their horizon among the basalts, to come together, or to separate. When they encounter a mass of coarse volcanic agglomerate, they often behave much more irregularly, being sharply diverted or abruptly terminated. They sometimes enclose abundant fragments of the amygdaloidal basalts, and in a few instances it has been verified that a sheet of dolerite cuts through a dyke of earlier date. These are more or less conclusive proofs of intrusion for the individual sheets which afford such direct evidence; but the cases in which some of these tests can be applied are, naturally, not very frequent. Much more convincing evidence is furnished by a comparative survey of the areas in which the rocks are developed. In the first place, it is to be remarked that sheets of dolerite in all respects identical with those in question occur, not only among 1 «The Tertiary Igneous Rocks of Skye’ Mem. Geol. Surv, 1904, chap. xiv. 42 MR. A. HARKER ON THE GEOLOGICAL [Feb. 1906, the basalts, but also among the subjacent Jurassic strata along the eastern coast of Skye and in the northern part of Higg. Here their intrusive nature has long been recognized. It seems an arbitrary interpretation to make the dolerites intrusive when they occur below the base of the basalt-group, and extrusive when they occur above that line.’ More striking evidence is afforded by the volcanic agglomerates which are found in the basalt-group, mostly near its base but also at higher horizons. The fragments in these agglomerates are of basalt, sometimes with an admixture of sandstones and other rocks of extraneous origin ; but I have sought in vain for any undoubted fragment of dolerite similar to that forming the strong sheets in the neighbourhood. The same remark applies to the fluvyiatile conglomerates intercalated at various horizons in the basalt-group in Skye, Canna, Rum, and Mull. Since the dolerites play at least an equal part with the basalts in the succession as now seen, and are moreover of much more durable nature, the absence of frag- ments of the dolerites from these accumulations points. to the conclusion that they were not there when the latter were formed ; or, in other words, that they are intrusive sills of later date. ‘Equally noteworthy is the horizontal distribution of the dolerite- sheets, as brought out by the mapping of Skye. ‘The plutonic intrusions—peridotites, gabbros, and granites—which make the central mountain-tract of this island are surrounded by a belt of metamorphism, in which the basaltic lavas have a hard and tough consistency, with an absence of bedding, markedly in contrast with their usual characters. As the dolerite-sheets are traced from the plateau-country towards the mountains, they are found to die out, and they all terminate without entering the belt of hard massive basalt of the mountain-border.. Further, in the mountains themselves, the gabbro and granite enclose many patches of basalt in a more or less metamorphosed state, some as much as a mile in length; but the dolerite-sheets never form part of these enclosed patches. Having regard to all these facts, the proof seems to. be com- plete, not only that the dolerites are intrusive in the basalts, but that there is a great difference in age between the two groups, the basalts being older than the plutonic intrusions and the dolerites younger. Many other facts might be cited as confirming, directly or indirectly, these conclusions. The general sequence of the Tertiary igneous rocks established in Skye is found to be borne out in the other islands, and in particular the recognition of three suc- cessive phases of igneous activity in this region. These three phases are characterized respectively by (1) volcanic extrusions, (ii) plutonic intrusions, (111) minor intrusions of hypabyssal habit; and the first event of the third phase is represented by the great group of dolerite-sills. * Sir Archibald Geikie long ago recognized the intercalation of intrusive sills among the basalts, but he allowed them no Lupa: ESGE DT at une eS: of the series. Vol. 23). STRUCIURE OF THE SGURR OF EIGG. 43 Il. GeneRat SrRucruRE oF THE PircHstonE-CoMPLExX. The general geological structure of Higg is too well known to need any detailed account in this place. The Jurassic strata dis- played in the northern part of the island have sunk below sea-level in the southern part; and here the crumbling basalts, and the more durable dolerite-sills which alternate with them, rise inland in a terraced slope, culminating in the Sgurr. Resting on this pile, as if poised upon a pedestal, is the great mass of the pitchstone. Its crowning position and its abrupt outline make it a conspicuous object, even in a distant view. The bold ridge which it makes has a curved course with a general direction from north-west to south-east, terminating abruptly at the highest point (1289 feet). Hxcepting this most elevated portion to the south-east, the crest of the ridge runs usually at an altitude of 1000 to 1100 feet. The pitchstone presents generally precipitous walls on both sides: its base on the north- eastern side being in most places between 900 and 1000 feet above sea-level, while on the opposite side it sometimes sinks to 700 feet, showing a general inclination in this direction (south-westward or southward). ‘The hase also declines along the direction of the ridge, being at about 1060 feet at the easterly termination, and sinking to about 550 feet at the opposite end, where the ridge is truncated by sea-cliffs. There are also minor variations of level, and it is clear that the lower surface of the pitchstone has in places an undulating form. The breadth of the ridge varies up to about 500 yards, exclusive of some arm-like extensions on the north- eastern side, which make the hills named Beinn Tighe and Corn- bheinn. The maximum thickness of the pitchstone is about 400 feet. In its south-eastern portion the base, in so far as it can be regarded as an inclined plane, has a general dip of about 15° to the south. At the actual termination eastward the inclination is much steeper, the observed dip being about 40°. The fine natural section at this last-mentioned place shows clearly that the pitchstone truncates obliquely the alternations of basalt and dolerite upon which it rests (see Pl. III). The same is true in every part of the boundary, as is well brought out by the mapping of the hill. The relation of the pitchstone.to the underlying rocks is, then, of a transgressive kind. Further, the partly undulating form of the surface of contact is not due to any subsequent folding ; for the basalt-group below is undisturbed, with a gentle south- westerly dip. (See map, fig. 1, and section, fig. 2, pp. 44 & 45.) The rocks which build the Sgurr are not a single body, but constitute a number of sheets, which in most places are sharply divided from one another. The dominant rock is the well-known porphyritic pitchstone, containing clear crystals of felspar in a velvety-black resinous-looking groundmass. Intercalated in this are sheets, usually of no great thickness, composed of a dull-grey porphyritic felsite, which, as seen in the escarpment, weathers more rapidly than the pitchstone, and so forms grooves-or recesses Fig. 1.—Sketch-map of the south-western part of the Isle of Figg, on the scale of 2 inches to the mile. i{ | BASALT y MB Dowerite [The above map shows how the pitchstone truncates obliquely the basalt-and- dolerite succession; also the position of the Grulin felsite-intrusion, possibly representing the feeder of the pitchstone. ] Vol. 62.] tron ured the Sgurr of Higg, near the summit, ina d 6 inches = 1 mile.) (Scale : S. 10° W.-N. 10° E. h the south-eastern part of Fig. 2.—Section throug S 10° W N 10°E The minor undulations of the base of the ale pitchstone are omitted. | i as those on the map, f (The symbols have the same signification GEOLOGICAL STRUCTURE OF THE SGURR OF EIGG. 45 in the face of the wall. This rock differs from the pitchstone chiefly in having a compact stony groundmass instead of one largely made up of glass, and the two rocks are beyond doubt closely related tooneanother. In places there is indeed, as we shall notice, a gradual transition between them. The felsitic sheets come in at various horizons in the complex, but are more abundant in the lower parts. They are best seen in the higher eastern portion of the ridge, and especially in its southerly face. They are not continued inde- finitely, but die out in the pitchstone. The latter rock, so far as any visible evidence goes, may be a single mass from top to bottom. The pitchstone and its felsitic modification, taken together, are clearly the youngest igneous rocks in this neighbourhood. Not only are the subjacent basalts and _ dolerites obliquely cut off at the face of junction, but also, as Sir Archibald Geikie remarked, certain basalt- dykes, which intersect; these lower rocks, ter- minate abruptly at the same surface. Further, the ‘ plateau’ - faults, which displace the basalts and dolerites, do not affect the pitchstone. At the base of the 46 ‘MR. A, HARKER ON THE GEOLOGICAL ~- __[ Feb. 1906, pitchstone-complex there is seen in many places a rock which at first sight seems to be quite distinct, though evidently related to the complex above. It shows more or less abundant blocks of the black pitchstone in a soft pale-grey matrix, which is evidently formed by the decay of the pitchstone itself or of its felsitic modification. The weathering-out of this soft basal band has often given rise to a recess at the foot of the pitchstone- escarpment (see Pl. III & fig. 3, YZ; also fig. 5, p.57). Itis well seen along the southern face for a distance of 500 yards from the eastward termination of the ridge, its thickness here being usually from 3 to 7 feet, or exceptionally as much as 10 feet. The pitch- stone-blocks embedded in the soft pale-grey matrix range up to a Fig. 3. (Key to Pl. LII.)\—The Sgirr of Higg seen from the ) east-south-east, from near Galmisdale. \ a ee : ~ meet cee ‘ ee The lower slopes are made by numerous alternations of basaltic lavas (B) with intrusive sills of dolerite (D). These are truncated obliquely by the thick sheet of pitchstone (P) which forms the summit-ridge, having a columnar structure in a roughly vertical direction, or locally inclined and divergent: The pitchstone is intersected by thinner sheets of a felsite (F), the devitri- fied equivalent of the pitchstone itself, with a generally parallel disposition but deviating in places. Along the line YZ is a recess made by the weathering-out of the brec- ciated and decayed basal band of the pitchstone, and below this at Z is the breccia with foreign rock-fragments and fossil wood. foot or two in diameter, and exceptionally to 4 feet. Their varying orientation, as indicated by the flow-structure in them, proves that they are not merely relics of an unbroken sheet, but have suffered relative displacement. Sir Archibald Geikie considered this decayed band to represent ‘a kind of brecciated base or flow of the main pitchstone-mass,’ * and this is doubtless its true nature. It is very clear in places where, for a short distance, the base of the pitchstone ' Quart. Journ. Geol. Soc. vol. xxvii (1871) p. 307. Vol.62.). STRUCTURE OF THE SGURR_OF EIGG. 47 has escaped breaking up, and the pale band is seen abutting upon undisturbed and unaltered rock (see P’ in fig. 6, p. 60). In addition to the blocks of pitchstone (and felsite), the decomposed matrix encloses in places a few pieces of a blacker and more completely vitreous rock, without conspicuous felspar-crystals, which may be taken to represent a more perfectly glassy selvage to the pitchstone-sheet. There are also fragments of extraneous origin, which are sometimes rather abundant at the base of the pale band, but become rare towards its top. Hxcepting one locality, to be described below, these extraneous fragments are exclusively of basalt and dolerite, evidently picked up from the subjacent rocks. The inclusion of these is not necessarily connected with the brecciation, for they are found also in the base of the pitchstone where it is unbroken and unaltered: for example, in the spur on the north-eastern side of the ridge near Loch na Mna Moire. The general geological relations of the pitchstones to the basic rocks have at different times received very different interpretations. Macculloch? (1819) does not explicitiy discuss the point. He remarks that the rock of the Sgurr may in places be ‘seen resting on the saine trap with which it alternates, and thus forming.... the uppermost rock of this interesting island.’ The ‘trap’ here referred to is the felsite which is so intimately bound up with the pitchstone itself; but the concluding words seem to imply that pitchstone and felsite together constitute merely the uppermost member of the general sequence. This ignores the obviously transgressive junction, so clearly exhibited at the eastern end of the ridge. Von (ynhausen and von Dechen’ (1829), as appears from their brief reference and the accompanying figure, considered the pitch- stone to be intruded through the basalts, and indeed they liken it to a dyke. Hay Cunningham? took the same view, which he illus- trated by a most uncompromising diagram; and James Nicol + arrived at a like conclusion. It is difficult to understand how this dyke-hypothesis could ever be entertained, in view of the manifestly stratiform disposition of the pitchstone-mass. At numerous places aleng the base it is possible to penetrate for several yards under the great sheet, and to see its base, like an inverted pavement, overhead. Hugh Miller (1858), like Macculloch, realized that the pitchstone definitely overlies the basalts; and, though his words are more picturesque than precise, he seems to have regarded it as the result of a volcanic outpouring. He speaks of the rock of the Segurr as haying been ‘laid down ....1in one fiery layer after another’.’ It is, however, to Sir Archibald Geikie° (1865) that we owe a clear 1 ‘Description of the Western Is, of Scotland’ vol. i (1819) p. 522. > Karsten’s Archiv fir Min. vol. i (1829) pp. 105-14 & pl. iv. °> Mem. Wern. Nat. Hist. Soc. vol. viii (1839) pp. 144-63 & pl. v. + *Guide to the Geology of Scotland’ 1844, pp. 232-33. ° *Cruise of the Betsey’ 1858, p. 35. : ® «Scenery of Scotland’ Ist ed. (1865) pp. 278-82; Journ. of Travel & Nat. Hist. vol. i (1868) pp. 14-16 ; Quart. Journ. Geol. Soc. vol. xxvii (1871) pp. 279- 3810; and later works. 48 MR, A. HARKER ON THE GEOLOGICAL [Feb. 1906, exposition of the view which regards the pitchstone of the Sgurr of Higg as a lava-flow, or succession of flows, poured out over an eroded land-surface of the basalts. This may be taken as the generally accepted interpretation for the last forty years. It has been endorsed by Prof. Judd,’ Prof. Heddle,? and others; but it does not appear to have been critically tested on the ground. In 1892 the late Prof. Heddle* made an interesting contrib- ution to the subject, recognizing that a porphyritic pitchstone identical with that of the Sgurr of Eigg makes the isolated cluster of low rocky islets named Oigh-sgeir (often written Hyskeir); an observation subsequently confirmed by Sir Archibald Geikie.* Oign-sgeir, rising only 34 feet above sea-level, lies between 18 and 19 miles from Bidein Boidheach, the seaward termination of the ridge of the Sguirr, and bears about W. 15° N., agreeing with the general trend of the pitchstone-ridge of Higg. If, as Heddle supposed, Oigh-sgeir and the Sgurr of Eigg are relics of one and the same lava-stream, we must infer that the two islands formed, at the epoch of the pitchstone, parts of one extensive land-surface., Assuming that the nearly submerged pitchstone of Oigh-sgeir has a thickness comparable with that seen on Higg, Sir Archibald calculates that the base declines westward at the rate of about 35 feet per mile. When, in the course of the geological survey of the Inner Hebrides, the present writer came, in 1903, to map the southern part of Higg, the work was begun on the received hypothesis that the pitchstone of the Sgurr represents the remains of a lava-current filling an old Tertiary river-valley. The detailed examination of the ground, however, discovered circumstances which were difficult to reconcile with that hypothesis, and eventually led to a reconsid- eration of the whole question. The results of my observations, though not sufficient to decide some of the minor questions which arise, go to support the view that the pitchstone-sheet is intrusive in the basait-group. I have twice revisited Higg during the present year (1905), and on one visit had the advantage of accompanying Dr. Peach, who brought to light other facts tending to the same conclusion. J am indebted to him for useful criticism, as well as for positive additions to the evidence bearing on the question, I desire also to express my thanks to Mr. R. L. Thomson, the proprietor of the island, who has interested himself in the enquiry, and whose kindness has made the work easy and pleasant. The first point relates to the general form of the pitchstone- complex as a whole. The elongated and curving trend of the ridge, 1 Quart. Journ. Geol. Soc. vol. xxx (1874) p. 267. 2 Proc. Roy. Soc. Edin, vol. xi (1882) pp. 632-383; also (Appendix C) ‘A Vertebrate Fauna of Argyll & the Inner Hebrides’ by J. A. Harvie-Brown & T, E. Buckley, 1892, pp. 247-50. 3 Op. supra cit. p. 249. 4 Rep. Brit. Assoc. for 1894 (Oxford) pp. 652-73; and Quart. Journ. Geol. Soe. vol. lii (1896) pp. 371-78. 2 Lee — Vol. 62. | STRUCTURE OF THE SGURR OF EIGG. 49 with its projecting arms, seems at first to accord with the supposi- tion that the pitchstone occupies an old valley with smaller tributary glens ; and it is also very noticeable that, at the base of the escarp- ment, the lower surface of the pitchstone has in most places an inward slope. Even on a broad view, however, the actual courses of the supposed valley and its tributaries are not easily laid down on any plausible lines; and the main valley must be supposed to narrow to a gorge at Bidein Boidheach, although it is in this part of its channel that the assumed river-gravel has been accumulated to a depth of about 100 feet (see PI. IV & fig. 4, p. 53). Closer examination of the actual sections raises more serious diffi- culties in the way of this hypothesis. The junction of the pitchstone with the underlying rocks is sometimes a steeply-inclined surface, and in certain parts of the boundary it slopes outward. This is very noticeable near Loch Beinn Tighe, where there is a sharp dip down towards the tarn. In the south-western part of Beinn Tighe the base of the pitchstone inclines towards the main body; but in the north-eastern part of the hill the dip is in the opposite direction, implying that the supposed tributary at this place ran uphill. The same is true of the smaller projecting tongue between Beinn Tighe and Cornbheinn. Along Cornbheinn the base of the pitchstone has apparently an undulating inclination, without any general slope towards the west; nor is it easy to see how the supposed tributary glen at this place could enter the main valley. Similar difficulties are encountered at places on the opposite side of the main ridge, notably at the projecting spur north of the Grulin crofts. The detailed mapping of the ground thus leads to a different interpretation of the structure of the Sgirr. The ridge and its offshoots, as we now see them, seem to be rather the relics of a more extensive sheet which had an irregularly-undulating lower surface. The position of the escarpment in different places has been determined by lateral erosion operating in relation to the columnar jointing which is so striking a feature of the pitchstone- complex. Where the columns leaned outward, they were easily destroyed, and the escarpment was thus cut back to a line along which the columns leaned inward. This correspondsas a rule with an inward dip of the base of the sheet, since the columns tend usually (but with exceptions) to be perpendicular to the lower surface. The exceptions are seen at such places as Loch Beinn Tighe, where the columns jean inward, although the base of the pitchstone dips outward. The visible form of the lower surface of the pitchstone-complex, as seen in detail along all parts of the base of the escarpment, is a highly significant one. The pitchstone, as we have seen, truncates an alternating succession of basalts and dolerites, the latter of which are much more durable rocks than the former. The slope leading up to the Sgurr, or any other hillside in the plateau-region, shows accordingly a strongly terraced form which is very characteristic, and is well seen in Pl. III. Such should be the form of the base Q.J.G.S8. No. 245. E 50 MR. A. HARKER ON THE GEOLOGICAL [Feb. 1906, of the pitchstone, if it coincided with an old surface of erosion. The actual form, as shown in the same plate, is very different. Despite undulations, the junction is always a clean-cut one. The undulations may often be seen to correspond with the places where the pitchstone-base passes from basalt to dolerite; but they show only gentle curvature, never abrupt change of direction. In this the junction agrees with an intrusive one, and differs from the erosion-form which is a necessary part of the river-valley hypothesis. | The steep inclination of the base of the pitchstone in certain places has already been remarked. Exceptionally, as on the north side of the Bidein-Boidheach section, it seems to approach the vertical. But it is further to be noticed that a steep slope is some- times found in individual sheets of the complex, which on the extrusive hypothesis should represent separate lava-flows. The most conspicuous example is the eastern precipice of the Sgurr proper (see Pl. IIT). Here the base consists of a sheet of felsite some 30 feet thick, with parallel upper and lower surfaces, which dips southward at an angle of 40°. It is clear that a thin lava- flow could not rest in such a posture as this, and examination of the felsite leaves no doubt that it is intrusive. It truncates obliquely the columns of the overlying pitchstone; and, when followed down the southern slope, it is seen to leave the base, cutting into the pitchstone itself and running out into two tapering tongues. Other sheets of felsite, which are numerous on the southerly face of the eastern part of the ridge, show in different cases indications of intrusive relations towards the pitchstone. A small one near the base passes into a nearly vertical dyke cutting up into the pitchsone. Most of the sheets have an approximately horizontal position, but with a wavy course and tapering form,’ and bifurcation may be observed in several places. ‘The felsites do not share the regular columnar structure of the pitchstone, but often truncate the columns obliquely, while they have a less pronounced and ruder jointing perpendicular to their own contact-surfaces. These circumstances seem to afford satisfactory evidence of intrusion. We have remarked, however, that there are also gradual trans- itions between felsite and pitchstone; and it is of interest to note that this relation is sometimes to be observed in connection with sheets of felsite which still give evidence of being intrusive in the pitchstone-mass. ‘The sheet first mentioned, at the eastern base of the Sgurr, is probably an example of this, for, where it bifurcates towards the south, it seems to lose itself gradually in the pitchstone. In several other cases, which can be more closely scrutinized, such behaviour is quite manifest. A sheet which sharply cuts the pitch- stone can be followed along its length until the line of demarcation is lost, and the felsite merges into the general body of the pitchstone * See Sir Archibald Geikie’s sketch, Quart. Journ. Geol. Soc. vol. xxvii (1871) pl. xiv, fig. 3. Vol. 62.] STRUCTURE OF THE SGURR OF EIGG. 5 by insensible gradations. Such a sheet would seem to represent a portion of magma forced out from a place where solidification was not yet complete and piercing a part of the mass which had already consolidated as pitchstone. The relationship between the two rocks must thus be of a peculiarly intimate kind, since they are on this supposition products of one magma, intruded in the first place as a single body. It would not follow from the relation suggested that the felsite- sheets should agree exactly in chemical composition with the pitch- stone (which itself seems to be somewhat variable) or with one another. Indeed the crystallization of the felsite and the vitrifica- tion of the pitchstone may be determined, in part, by a certain difference in composition.. The following partial analyses are furnished by Dr. W. Pollard, the felsite being from the conspicuous sheet at the eastern base of the Sgurr :— Piichstone. Felsite. 10) ee 63°34 68:43 INO 4-56 476 SO rh. 4-50 631 Another consideration relates to the channel by which the pitch- stone ascended through the basalts. We may safely assume that the magma which gave birth to the rock possessed a considerable degree of viscosity. As a subaérial lava it could not flow to any very great distance. If it was a subterranean injection, the same remark applies, though with somewhat less force. Among our Tertiary intrusions, sheets of acid or subacid composition, although they may equal or exceed the dolerite-sills in thickness, in no case rival them in lateral extent. On either alternative, then, we might hope to find some indication of the source of the pitchstone at no great distance from the Sgurr itself. On the lava-flow hypothesis we should seek such indication in an easterly or south-easterly direction. Here we find no intrusive mass which can be supposed to mark the channel of uprise ; but, since we arrive at the sea in a distance of about 13 miles, the negative evidence is of little weight. ‘Close to the ridge of the Sgurr, however, on its south-western side, we find a remarkable intrusion of porphyritic felsite, which may possibly represent the feeder of the pitchstone-intrusion. This mass, recognized by Sir Archibald Geikie and marked on his sketch-map,’ has an elongated form, and extends for more than a mile close to the base of the pitchstone-escarpment (see map, fig. 1, p. 44). In its western or north-western part, above the deserted crofter town- ships of Grulin, the outcrop has a width of 100 to 200 yards, and it is evident that the rock cuts abruptly through the basalt-lavas and dolerite-sills with a quasi-vertical boundary (see section, fig. 2, p. 45). Hastward, near the summit of the Sgurr, the outcrop tapers away, and it is possible that the mass has in this part the form of an inclined sheet. The rock is not only later than the dolerite-sills, * Quart. Journ. Geol. Soc. vol. xxvii (1871) p. 286. oO py No; 52 MR, A. HARKER ON THE GEOLOGICAL [Feb. 1906, but probably also later than the basic dykes; for one at least of these stops abruptly on encountering the felsite, as if cut off by it. So far as the evidence goes, therefore, this Grulin felsite may well be of the same age as the pitchstone. The felsitic rocks near Laig Farm and at the extreme north of the island are earlier, being intersected by the basic dykes; and they are thoroughly acid rocks, while the Grulin felsite is essentially felspathic. Otherwise unique in the district, this rock has petrographical characters in common with the felsite-sheets of the Sgurr, although, as is to be expected, it is not of such fine texture. A partial analysis by Dr. Pollard enables us to compare its composition with that of the pitch- stone :— Pitchstone. ae SO ra oe 63:34 5813 ROPOR che Ch oe 4-56 457 TON Sa eee: 4:50 4°85 Lhe resemblance in respect of the alkalies is very close. II]. Fragmenrat ACCUMULATIONS AND Fosstz Woop. In two places accumulations of fragmental materials are seen immediately beneath the pitchstone, or beneath the pale decomposed band which has been included as a component part of the pitch- stone-complex. These fragmental accumulations are of importance, as constituting a part of the evidence upon which Sir Archibald Geikie has based his interpretation of the structure of the Setrr of Kigg. He considered them to be of the nature of river-gravels, laid down on the floor of a valley at this place, and subsequently covered by a pitchstone lava-stream. The two localities present different features, and will be described separately. We take first the section at Bidein Boidheach, where the ridge, running north-westward, is abruptly cut off by the sea-cliffs. Sir Archibald Geikie has given a figure’ of this place, and by the kindness of Mr. A. 8. Reid, F.G.8., I am enabled to illustrate it by a photograph, taken by him, under considerable difficulties, from the sea (Pl. 1V and fig. 4, p. 53). The lower part of the section shows. the usual alternations of basalt-lavas and dolerite-sills lying nearly horizontally. At the top is the pitchstone-sheet, its base at this place dipping inward on both sides of the ridge in a way which, taken alone, might seem to accord well with the supposition that the pitchstone occupies the bed of a valley. Below this, and filling a hollow in the basalt-group, is the fragmental accumulation to be noticed. As seen in the cliff-section, it has a maximum thickness. of perhaps 100 feet, rapidly thinning away in both directions. It might represent either a cross-section of a deposit extending along the direction of the ridge, or a small funnel-shaped mass which happens to be intersected by the present line of the cliff. I interpret 1 Quart. Journ. Geol. Soc. vol. xxvii (1871) p. 308. Vol. 62.] STRUCTURE OF THE SGURR OF EIGG. 53 it in the latter way, as a mass of volcanic agglomerate filling a small vent. Only a small part of it, on the south side, is accessible to direct observation. Fig. 4. (Key to Pl. IV.)\—Bidein Boidheach, showing the sea-cliff i which the ridge of the Squrr terminates north-westward. Thin = r : aT = ple pep rah f Onnaay ie ya ie Sey mn me me eee = oe : ‘ it ; ee N77 a | we re ae i te + f The outcrops of the basaltic lavas and dolerite-sills run in nearly horizontal lines. PPP is the base of the pitchstone, and A is the volcanic agglomerate which underlies it in the cliff. The dolerite-sill D, impinging upon the agglomerate and deviated by it, is not distinctly seen in the photograph. At B are basalt-dykes, cutting the basalt-lavas and dolerite-sills, and one of these is seen to terminate abruptly at the agglomerate. The first point to note is the evidence that this fragmental accumulation is older than the dolerite-sills. The large sill which makes the top of the cliff on the south side is seen to abut upon the agglomerate, but does not enter it. It is, however, not cut off, but sharply deviated. As Dr. Peach observed, it turns upward, with diminished thickness, along the boundary of the agglomerate, and then, curving away, rapidly dies out. No corresponding sill is seen on the opposite side, although the unbroken succession below shows that the steep boundary on that side is not made by a fault. Instead we find some thinner sills which, so far as can be seen, appear to terminate abruptly at the agglomerate. The sharp deviation or abrupt termination of sills where they come against a mass of coarse agglomerate is a phenomenon seen in many places in Skye, Canna, and Sanday. These and some other massive rocks appear to have the property of arresting the progress of a minor intrusion, either sill or dyke, which encounters them.’ Some basalt- dykes occur, as noted by Sir Archibald Geikie, in the cliff of Bidein Boidheach, and one of them is seen to terminate abruptly at the 1 See ‘ The Tertiary Igneous Rocks of Skye’ Mem. Geol. Surry. 1904, pp. 242, 293. o4 MR. A. HARKER ON THE GEOLOGICAL [Feb. 1906, agglomerate. Since it cuts the dolerite-sills, it cannot be older than the agglomerate, and its apparent truncation must be due to the difficulty which it experienced in piercing that mass. Both in its fofm and in its relations this fragmental accumulation of Bidein Boidheach closely resembles some of the small volcanic vents of Skye, notably one in the cliff at Sgurr-an-Duine, between Lochs Brittle and Hynort. The nature of the material is also in accord with this inter- pretation. The principal element is amygdaloidal basalt, in blocks measuring up to a foot or two in diameter, subangular or rounded ; the rounding being of the kind characteristic of our volcanic agglo- merates, that is tending to the spherical form, without the ovoid shapes so common in waterworn pebbles. Torridonian rocks, in smaller pieces, are fairly plentiful, and Sir Archibald Geikie found fragments of the white sandstone of the Oolites. Both formations are doubtless in place below sea-level at this locality, and the occurrence of their fragments in a volcanic agglomerate is very natural. Dr. Peach detected also a piece of granite and other fragments of a decayed rock, probably of the gabbro-family. These plutonic rocks occur also in the volcanic agglomerates of Skye,’ and possess a certain interest, which lies, however, outside our present subject. ' The most significant fact is the absence of any fragments of dolerite from the sills. In a river-gravel derived from the erosion of the basalts and dolerites the latter should be as abundant as the former, or more so in virtue of their superior durability. The character of the material, therefore, confirms the conclusion that we have to do here with a volcanic agglomerate of the basalt-period, much older than the dolerite-sills and very much older than the pitchstone. The conjunction of the pitchstone with the agglomerate is, on this view, merely accidental. The second locality to be examined is near the other end of the ridge, on the southern side of the higher part of the Sgurr and about a quarter of a mile from its easterly termination. It is somewhat above the former occurrence, both in altitude and as regards geological horizon. In this neighbourhood the brecciated and decayed base of the pitchstone has been cut back to make a recess or piazza, in places 10 feet or more in height. In the floor of this recess a breccia with fragments of various rocks is poorly exposed for some distance; and at one place, near some shelters for sheep, is the locality of the well-known fossil wood. Although the fossil wood of the Sgurr of Eigg was discovered 90 years ago, and has been referred to by numerous writers, the published information concerning its mode of occurrence is not only very scanty, but remarkably confused. This is due partly to misunderstanding of records obtained at second-hand, partly to the fact that wood occurs at this place in more than one situation and of more than one kind. * “The Tertiary Igneous Rocks of Skye’ Mem. Geol. Surv. 1904, p. 24. Vol. 62.] STRUCTURE OF THE SGURR OF EIGG, 55 The earliest notice is by Macculloch (1814). He states that the pitchstone les in a bed of ‘marie’ 3 or 4 feet thick, which rests upon a bed of hard reddish sandstone, and that ‘ Large masses of wood, bituminized and penetrated with carbonat of lime, are found in the marle stratum, not at all flattened. Portions also of trunks of trees, retaining their original shape, are seen in the same bed, silicified, and their rifts filled with chalcedony, approaching in aspect to semi-opal.’? Later * he made a correction, stating that what he had termed ‘marle’ was a ‘mixture of limestone, clay, and sand.’ He found reason to believe that it was not a continuous bed, but either a lenticle entangled between the pitchstone and the subjacent mass or a part of the latter. This underlying mass he described as ‘a conglomerate consisting of small and large fragments of red sandstone, trap, and silicified wood, imbedded in a basis of trap’ (loc. cit.). Macculloch’s ‘marle’ is the pale brecciated and decayed base of the pitchstone, and the ‘conglomerate’ below is the rock which will here be described as a volcanic agglomerate. ‘The large pieces of wood are correctly described as occurring in the former, and Macculloch is equally accurate in recording fragments of wood in _ the latter also. Von (iynhausen and von Dechen® apparently failed to find wood, but they unwittingly contributed to the confusion of the subject. Their paper mentions, as occurring immediately beneath the pitchstone on the south side, a conglomerate with pieces of pitchstone, hornstone, chalcedony, and calcspar; and this they identify with Macculloch’s conglomerate containing red sandstone etc., though it must certainly be the brecciated and decomposed base of the pitchstone. Fossil wood from beneath the Sgurr of Higg was described and figured by Witham* (1831), and shortly afterward by Lindley & Hutton.’ From both accounts it might be inferred that this wood, named Pinites eciggensis, occurred in place in Jurassic strata at the base of the Sgurr. This misapprehension was corrected by William Nicol,’ who had supplied the material described, and who states that it was found ‘among the débris of the prismatic columns of porphyritic pitchstone.’ The specimens described and figured, then, were not obtained 7 situ, but from loose fragments. It is fairly certain, however, that their source was the former of the two occurrences recorded by Macculloch, namely, the decaying base of the pitchstone. ! Trans. Geol. Soc. ser. 1, vol. ii (1814) p. 408. 2 ‘Description of the Western Is. of Scotland’ vol. i (1819) p. 522. 3 Karsten’s Archiv fir Min. vol. i (1829) p. 107. - 4 «Observations on Fossil Vegetables’ 1831, p. 37 & pl. v, figs. 13-14. Also ‘The Internal Structure of Fossil Vegetables’ 1833, pp. 63-65, pl. xiv, figs. 13- 14, & pl. xv, figs. 6-9. a ° «The Fossil Flora of Great Britain’ vol. i (1831-33) pp. 91-92 & pl. xxx. 6 Edin. New Phil. Journ. vol. xvi (1833) p. 154. See also Rep. Brit. Assoc. for 1834 (Edinburgh) Trans. Sections, p. 660. 56 MR. A. HARKER ON THE GEOLOGICAL [Feb. 1906, Hay Cunningham,’ upon no grounds ascertainable from his memoir, flatly contradicts Macculloch’s statements as to the mode of occurrence of the fossil wood. He asserts that the wood has never been found zm situ, and adds, inconsequently, ‘it is evident from the-.point where its fragments may be obtained, that. its matrix is the rock of the Scuir...... : Hugh Miller described quite clearly the manner of occurrence of the wood, as follows :— ‘We were successful in procuring several good specimens of the Eigg pine, at a depth, in the conglomerate, of from 8 to 18 inches. Some of the upper pieces we found in contact with the decomposing trap out of which the hollow piazza above had been scooped; but the greater number ...... lay imbedded in the original Oolitic grit in which they had been locked 72 WO eaharee It is to be remarked that at that time, notwithstanding Maccul- loch’s earlier work, the basaltic rocks were generally believed to be lavas of Oolitic age. Assuming this, Miller included the con- glomerate in the same series, and regarded the fossil wood as the relics of contemporaneous vegetation entombed in the ‘ deep-sea bottom.’ Not only did he speak of the Eigg pine as ‘an ancient tree of the Oolites,’ but he stated that it is found in this formation in other parts of Scotland :— ‘The fossil trees found in such abundance in the neighbourhood of Helmsdale that they are burnt for lime,—the fossil wood of Hathie in Cromartyshire, and that of Shandwick in Ross—all belong to the Pinites eiggensis’ (op. cit. p. 39). It does not appear, however, that this statement was based on any minute comparison of specimens; and, indeed, in a later work Miller did not insist upon the specific identity, but spoke of the examples from the eastern part of Scotland as ‘ branches and portions of the trunks of a similar pine’ to that of Higg.’ It should be noted that Miller’s Eigg specimens came partly from the decom- posing base of the pitchstone, but mostly from the underlying breccia or agglomerate. My own specimens from the latter situation include some which do not belong to Pinites; but small pieces of this also occur here, as well as in the band above. Sir Archibald Geikie * seems to have found only the small pieces of wood which occur mingled with fragments of various rocks in the breccia, as distinct from the brecciated and decaying base of the pitchstone, although his brief account does not explicitly separate the two fragmental bands which occur at this place. I have no doubt that the two are quite distinct, and that their conjunction here is accidental. ‘The brecciation of the basal part of the pitchstone (or its felsitic modification) and alteration to a soft Mem. Wern. Nat. Hist. Soc. vol. viii (1839) p. 156. ‘Cruise of the Betsey’ 1858, p. 37. ‘Sketch-Book of Popular Geology’ 1859, p. 138. Quart. Journ. Geol. Soc. vol. xxvii (1871) p. 307. Re Oo NY a a Vol. 62.] STRUCTURE OF THE SGURR OF EFIGG, 57 rock of dull-grey aspect may be seen for long distances in many parts of the escarpment of the Sgirr. The breccia below, com- posed largely of non-volcanic material, is a local accumulation, which is seen at this place for a length of perhaps 50 yards. Moreover, I shall offer reasons for believing that it is interbedded in the basalt-group, and therefore does not follow the irregular base of the pitchstone. The locality has sufficient interest to merit a more particular description. It is clearly the same as that which was visited by Macculloch, Miller, and others; and the fossil wood preserved in various collections has doubtless come from this place. Despite Fig. 5.—NSection at the southern base of the pitchstone, about a quarter of a mile west of the eastward termination of the ridge. P=Base of fresh columnar pitchstone. X=Brecciated and decayed pitchstone, enclosing blocks of the fresh rock and rare pieces of other rocks. Z=Breccia, with blocks and fragments of foreign rocks. W =Large log of fossil wood. S= Horizon of sandstone and plant-bed seen farther east. B=Basalt-series below. Hugh Miller’s allusion to a ‘ prostrate forest ’ underlying the pitch- stone, I cannot learn from the published accounts, from the inform- ation of residents, or from personal search, that wood has ever been discovered under the Sgurr, except at this place ; and the large pieces of pine which have been obtained at different times may very possibly be portions of a single tree. To exhibit the relations of ‘the rocks, as well as to procure good specimens of the wood, some work with pickaxe and spade is necessary. Excavations have been made x 58 MR. A. HARKER ON THE GEOLOGICAL [Feb. 1906, under my eye on two occasions, once in the presence of Dr. Peach ; and Mr. D. Tait, who visited the place to collect for the Geological Survey, has also furnished me with some notes to supplement myown. . The base of the fresh columnar pitchstone, forming the roof of the recess or gallery, has, as usual, an inward dip, although the general inclination of the great sheet is, in this part of the Sgurr, in the opposite direction. ‘The base of the pale decomposed band (X in fig. 5, p. 57) has a similar dip, and the thickness of the band is here about 7 feet. As elsewhere, it contains, besides the blocks of unaltered pitchstone and felsite, some pieces of basalt, which become more abundant in the lowest part. In addition we find, what are not seen elsewhere, fragments of Torridon Sandstone and little pieces of fossil wood. In the lower part of the pale decom- posed pitchstone, but distinctly enclosed in it, occurs the main mass of wood (W) which has furnished the specimens of Pinites ciggensis. There can be no doubt that. this (as well as the Torri-. donian fragments) has been taken up from the underlying breccia presently to be described. Most of the wood, both in the main mass and in the scattered pieces, is completely silicified, of a lustrous black, and often shot through with slender threads of white calcite; but there is some which is not silicified, being more or less carbonized, with a similar black lustrous aspect. Mr. Tait’s account is as follows :— ‘Here a log of fossil wood les embedded, with its length parallel with the length of the Sgirr. The piece is about 5 feet long, as it now remains. [I subsequently saw about 8 feet exposed.—A. H.] Probably the heavier end of the log lies towards the east, but this could not be definitely ascertained. The log is about 8 inches in vertical height, and somewhat more in horizontal breadth : this gives it a flattened oval outline in transverse section. It is cracked and fissured in many directions, and pieces evidently belonging to this trunk are completely separated from it. That these pieces do belong to this trunk is seen by their similarity of texture and their orientation...... Numerous fragments of smaller branches occur at this level. They are often flattened into ribbon-shaped pieces, that lie flat on the bedding-planes. At places 15 to 20 yards apart the matrix in which the wood is embedded is quite different. In this woody layer, small angular stones are embedded. These stones increase in number and size downwards, and so form the underlying breccia exposed at the outer edge of the platform.’ The last sentence would imply a passage from the pale band of decomposed pitchstone (X) to the underlying breccia or agglome- rate (Z). Dr. Peach and I satisfied ourselves that there is a clear separation between them, although the pieces of extraneous rocks in the former naturally become more numerous at its junction with the deposit which supplied them. This deposit (Z) is of a very variable character. It consists chiefly of fragments of various sizes, with a smaller proportion of matrix, which is not always of the same nature. Sandy or basaltic (perhaps ashy) material may predominate, and there may or may not be sufficient calcareous and ferruginous matter to make a binding cement. The fragments are of red sandstone and other Torridonian rocks and of basalt, Vol. 62. | STRUCIURE OF THE SGURR OF HIGe. 59 the relative proportions of these two varying, though the former element is always well represented. There are angular blocks of Torridonian sandstone up to 4 feet in diameter, besides smaller fragments. Some of the pieces of basalt measure from 1 to 2 feet in diameter, and they are mostly subangular in shape. Of more local distribution in a recognizable state is the soft Oolitic sand- stone. Pieces of this are abundant in places, generally in a crumbling condition, and the sandy element in the matrix of the deposit may be attributed to this source. Mingled with the pieces of Oolitic sandstone are fragments of brown wood, different from the black wood already mentioned. Other materials than those now enumerated seen to be very rare. I possess a piece of flint found here many years ago by Mr. C. 8S. Middlemiss. At the spot where the section (fig. 5, p. 57) is drawn it is not possible to say whether any other deposit intervenes between the breccia with Torridoniau and other extraneous débris (Z) and the underlying basalts. Some 10 yards to the east, however, a bedded sandstone is seen lying below the breccia, or constituting the lower part of it (S in fig. 6, p. 60). It encloses blocks and smaller fragments of Torridon Sandstone and basalt like those in the breccia proper, but here the matrix makes up the greater part of the mass. It is a fine-textured grey sandstone, composed of quartz-grains, with a cement mainly calcareous, though basaltic material, perhaps of the nature of volcanic ash, also enters into its composition. The thick- ness seen is about 2 feet. Im addition to rock-fragments, the sandstone encloses abundant fossil wood, and there is a band com- posed almost wholly of vegetable remains in the condition of lignite. The stratification, which was indistinguishable in the coarse breccia, is here very evident; and there is a general resemblance to the thin and inconstant stratified deposits with plant-remains intercalated at various horizons in the basalt-group in Skye and other parts of the region.' Sir Archibald Geikie considered the breccia to be a fluviatile deposit of the pitchstone-epoch, the sandstone-fragments being brought down by a river from more or less distant sources, and the pine- wood representing’a contemporaneous vegetation. This interpret- ation seems, however, to involve considerable difficulties. The high proportion of extraneous fragments as compared with the local basalt, the large size and angular form of some of the blocks, and the mingling of coarse and fine material, are all difficult to explain on this hypothesis. Nor is it easy to find a source for the Torri- donian and Jurassic rock-fragments. It must be remembered that the epoch of the pitchstone was long subsequent to any differ- ential movements of elevation and subsidence known in the region ; and therefore the relative altitudes of the several formations were ‘Substantially as they are to-day. We can scarcely suppose that the pre-Tertiary strata were more freely exposed at the surface then 1 See, for instance, Sir Archibald Geikie, Quart. Journ. Geol. Soc. vol. lii (1896) pp. 341-42. 60 MR. A. HARKER ON THE GEOLOGICAL [Feb. 1906, than now: the reverse is more probable. The white sandstones of the Oolites on Higg itself are below the level of this breccia. In the cliffs cf Raasay and the north of Skye they rise higher; but these places are 25 or 30 miles distant, and the abundance of the sandstone - fragments in the breccia, of which they compose at certain spots the chief element, remains unexplained. Of the red Torridonian rocks there are at the pre- sent day extensive - exposures at altitudes higher than the base of the Sgurr; but the nearest is 12 miles distant, on Rum, and we can scarcely sup- pose angular blocks measuring up to 3 or 4 feet in diameter to have travelled so far. Prof. Judd* has re- marked that ‘the characters of the buried conglomerate are suggestive of a mountain- ravine subject to the passage of violent floods rather than of an ordi- nary river-channel.’ Syn 30 ne) [The letter Z indicates the place where the preceding section is taken, at right angles to the present one. | chstone, as in preceding figure. ements. , with remains of contemporaneous vegetation. 40 30 Brecciated and decayed pit Te es ED Z0 Breccia, with foreign fra Sandstone x Z Ss ae a cx (Ss DLE SO —= —— To the present writer it seems to resemble more closely a vol- canic agglomerate, partly rearranged by water-action. In the _ volcanic agelomerates of the Inner Hebrides it is a common experience to find, mingled with the basaltic material, fragments of pre- 1 Quart. Journ. Geol. Soc. vol. xxx (1874) p. 267. YPINE: iy = fs AK aa crushed and unaltered. Fresh columnar pitchstone. P!= Actual base of pitchstone, where it remains locally un- EL oe Fig. 6.—Section (about 40 yards) along the base of the pitchstone-escarpment, at the same locality as fig. 5, p © aoe per P = = ° Vol o2.] STRUCIURE OF THE SGURR OF EIGG. 61 Tertiary rocks brought up from below by explosive action, and these sometimes exceed the basalt itself in amount. In the present ease, the non-volcanic fragments that occur are such as might be expected. The Oolitic sandstones underlie the basalt-group probably a little below sea-level at this place, and below that we should confidently expect a good thickness of Torridonian strata. The occasional presence of flint, suggesting the presence of concealed Cretaceous rocks, is of some interest ; for, on the coast some 2 miles to the north, Mr. Tait has detected an exposure of Cretaceous sand- stone containing flints. The larger blocks in the breccia show no. sign of arrangement by water-action, and the manner in which fragments of a given rock tend to cluster in one place is much more characteristic of a volcanic agglomerate than of a fluviatile deposit. Most significant of all is the fact that here, as at Bidein Boidheach, fragments of the strong dolerites are wholly wanting. This ean only be interpreted as showing that the accumulation, whatever its. nature, is older than the dolerite-sills, and belongs to the basalt- group, not to the epoch of the pitchstone. The visible relations at this place are scarcely sufficient in themselves to determine whether the deposit in question is interstratified in the basalt-group or behaves with the overlying pitchstone; but it is certain that no. corresponding breccia is to be seen at the base of the pitchstone on the opposite side of the ridge. If, as the above evidence would indicate, this fragmental deposit is interstratified in the basalt-. group, it should be found on the north side at some distance below the base of the pitchstone, unless it has died out in the interval. I sought it at this place without success. Dr. Peach, however, detected signs of it on the north side of the Gualainn, or shoulder, which forms the eastward prolongation of the Sgurr ridge. Here, at the proper horizon but not elsewhere, are found on the surface many small pieces of Torridonian sandstone mingled with the fragments of basalt. These probably mark the occurrence of the breccia at this place, but some excavation would be required to test this point. On the river-gravel hypothesis the deposit should occur, if at all, on the opposite (southern) side of the Gualainn ; and Sir Archibald Geikie recorded an outlier of it here, resting against the denuded edges of the basalts, the stones being much more rounded and smooth than in the former locality.’ Repeated search has not enabled me to verify this occurrence. The only fragmental accumulation to be seen here is one of Glacial origin. It consists of dolerites and basalts, with some quartz-porphyry and other foreign rocks, but without Torridon Sandstone. These rocks occur as pebbles and boulders up to a foot or more in diameter, more or less rounded, but frequently planed, and sometimes showing well-marked striation. A distinct but related question is that of the true age of the fossil wood. Sir Archibald Geikie regarded it as the remains of 1 Quart. Journ. Geol. Soc. vol. xxvii (1871) pp. 307-808. 62 MR. A. HARKER ON THE GEOLOGICAL [Feb. 1906, contemporaneous vegetation swept down by a river of the pitchstone- epoch and entombed in a river-gravel. On the interpretation put forward in the present communication the question, how far the wood is contemporaneous, how far derived in a fossil state from older formations, is of secondary importance, but has a certain interest of its own. Tertiary wood and coal, entangled even in coarse volcanic agglomerates, are not uncommon in Skye and Canna,’ and the Eigg wood may conceivably have a like mode of occurrence. On the other hand, it is quite possible that fossil wood might be derivative in the same sense as the fragments of sandstone with which it is associated. Examining the question on the ground, I inclined to this view so far as regards the greater part of the wood ; and various considerations go to support this opinion. It is to be noticed that three distinct kinds of fossil wood have been mentioned above as occurring at this place :— (a) the black, mostly silicified, Pinites eiggensis, embedded in the decomposed base of the pitchstone as well as in the breccia, but evidently derived from the latter source ; (5) the fragments of brown wood (Araucaroxylon, as noted below, p. 63) associated with pieces of Oolitic sandstone in the breccia; and (c) the wood associated with the sandstone of the breccia-horizon, and partly occurring as a regular bed in that deposit. The last is manifestly contemporaneous with its matrix, and therefore of Eocene age; but concerning the other two kinds of wood the two alternative explanations may be entertained. As regards the Pinites, Mr. Seward informs me that it is impossible to decide, from botanical evidence, whether it is of Jurassic or of Tertiary age. The state of preservation, however, is perhaps significant; for British Tertiary wood is not in general silicified, while that of Jurassic age frequently is. Wood is found in place in the white Oolitic sandstones on Higg itself. At Camas Sgiotaig, a small bay near Laig, famous for its ‘musical sand,’ it occurs in two varieties, as represented by Mr. Tait’s collection : (a) black and perfectly silicified; (6) brown and ligneous, the resemblance to the two corresponding varieties from beneath the Sgurr amounting (to the eye) to absolute identity. It might perhaps be suggested that the big log enclosed in the decayed pitchstone derived its silica from the decomposition. of that rock, and it is true that the brecciated and decayed mass in this place sometimes contains siliceous veins, like flint. Butthe character of the fracture seen on the detached branches and the places where they have been broken from the trunk is such as to suggest that the wood was silicified before it was broken up. Neither Mr. Tait nor myself could find any adherent matrix other than that in which the wood now lies. Nevertheless, the association of this kind of wood with that next to be noticed, both here and at Camas Sgiotaig, leads us 1 An interesting early record is that in the old ‘ Statistical Account,’ vol. xvii (1794) p. 293, of a log of wood embedded in the coarse agglomerate of Alman, -on the eastern coast of Canna. —"* Vol. 62.] STRUCTURE OF THE SGURR OF EIGG. 63 to group the two together as probably having the same origin and age. The fragments of brown wood mingled with sandstone-débris in the breccia lie, as a rule, quite detached in the general matrix; but in one or two cases I thought, without being certain, that they were partly embedded in the crumbling sandstone-fragments. Moreover, one piece of wood, selected for slicing, shows the unmistakable white sandstone actually embedded in the substance of the wood, as if filling a hollow made by decay. ‘This conclusion, in which Mr. Seward concurs, would imply that the fragments of wood have been derived, together with the sandstone, from Oolitic strata in place, and are of that age. Mr, A. C. Seward, F.R.S., has had the kindness to examine thin slices of this fossil wood, and I am permitted to quote here the notes which he has drawn up. Brown wood from the Great Estuarine Sandstone of Camas Sgiotaig, Higg. ‘Imperfectly preserved coniferous wood. The material is not sufficiently good to enable one to determine the nature of the plant with any certainty ; it is indubitably a Conifer, and from its arrangement of pits on the radial walls of the tracheids I think that it is probably referable to the genus Araucaroxylon. This generic type includes Araucaria and Agathis, the two surviving repre- sentatives of the Araucariez, and numerous fossil species. I think it most probable that the wood is of Jurassic age.’ + Brown wood from breccia below the pitchstone of the Sgirr of Higg (Z in fig. 5, p. 57). ‘In one specimen the wood contains an included patch of rock which was I believe, introduced during the decay of the tissue. The rock and wood are, I believe, contemporaneous. The generic identification is very difficult, not to say impossible, owing to the lack of well-preserved material. The characters do not agree with those of Pinites etggensis, as represented by the well-preserved specimens in the British Museum (Natural History). Similar characters are shown in two other sections.’ { Another specimen. | ‘Preservation very poor; but, from the occurrence of multiseriate and contiguous bordered pits on some of the tracheal walls, the wood appears to be Araucaroxylon. This type exists in Paleozoic rocks, it occurs in the Mesozoic and Tertiary formations, and is characteristic of the recent Araucariee.’ Conclusion. ‘The wood from Camas Sgiotaig may be generically identical with some of the material from the Sgirr of Higg: both show Araucarian characters. As regards age, my impression is that the evidence favours a Jurassic rather than a Tertiary horizon. The reasons in favour of tbis view are (i) the fact that Araucarian wood is very abundant in Jurassic rocks (for example, the Liassic rocks of Whitby, where this type of wood appears to have been mainly responsible for the production of jet; Araucarian plants also occur in some abundance in Inferior Oolite rocks); (ii) the nature of the material, which is more like that known to occur in Jurassic than in Tertiary stratay’ 1 [Mr. Seward was apparently not aware that these specimens had been collected (by Mr. Tait) from Jurassic strata im situ, a circumstance which adds weight to his following conclusions.—A. H.| 64 MR. A. HARKER ON THE GEOLOGICAL [Feb. 1906, LY. Conctupine CoNnsIDERATIONS. The question of the true relations of the pitchstone of the Sgurr has been discussed above in the light of such direct evidence as can be adduced. Other considerations, which may legitimately find a place here, arise when we regard the petrographical characters of the pitchstone and its probable place in the long sequence of British Tertiary igneous rocks. The pitchstone itself has been described by Prof. Judd,’ and analysed by Mr. Barker North. The analysis, reproduced in column I (below), proves that it is of subacid composition, and not essentially different from some of the Arran rocks. Puitchstones, rhyolitic, trachytic, and dacitic, referred to the Tertiary suite, occur as sills and dykes in Arran, Ardnamurchan, Eigg, Rum, and Skye, and in Counties Down and Donegal. There are also, in Arran, Kigg, and elsewhere, rocks of the same general characters but devitrified, and these are often intimately associated with the glassy varieties. Most of these occurrences are of small dimensions, but the sheet of devitrified pitchstone which forms the upper parts of Ashval and Sgurr-nan-Gillean, in the Isle of Rum, exceeds in extent and thickness the pitchstone of the Sgurr of Eigg. This Rum occurrence and one or two others belong to an earlier epoch of the hypabyssal phase; but all the rest are to be assigned to the very latest manifestations of igneous action in the British area. Wherever evidence is obtainable, these rocks are clearly intrusive. I, 1 LL: SO ear eee 65°81 63°34 66°03 PM Oe os ae 1401 ee 12:55 ORO i) gas phe batids 4°45 2°75 NE Os eae ers oe 0-89 2°38 OO tte wan 2°01 2°80 INR OT sea So sens 415 4°56 5:02 SO peresiet ees tes 6:08 4:50 4:15 Loss on ignition... 2°70 4-20 100-08 99°8 J. Porphyritic pitchstone, Sgtrr of Higg: anal. Barker North, Quart, Journ. Geol. Soc. vol. xlvi (1890) p. 379. II. The same: partial analysis by Dr. W. Pollard. TIX. Pitchstone, formore, Arran: anal. M. M. Tait, Bryce’s ‘ Geology of Arran & Clydesdale’ 3rd ed. (1865) p. 185. The only case (omitting the Sgurr of Higg) in which this state- ment is not one of common agreement is that of Beinn Hiant in Ardnamurchan. ‘The sheets of andesite and pitchstone which form that hill were regarded by Prof. Judd* as lava-flows, but Sir Archibald Geikie® has given good reasons for treating them as intrusive sills. The rock of the Sgurr of Eigg, then, if an intrusive Quart. Journ. Geol. Soe. vol. xlvi (1890) p. 380. Ibid. pp. 873-76. Trans. Roy. Soc. Edin. vol. xxxy (1888-90) pp. 118-19. Wee WS) is 2...” >. = Vol. 62. ] STRUCTURE OF THE SGURR OF EIGG. 65 sheet, falls naturally into its place as a member of this group of intrusions, with which it has much in common. It further resem- bles the rest in the absence of steam-vesicles, a negative character which is not easily reconcilable with a glassy subaérial lava con- taining 2°7 per cent. of water. The irregularly undulating lower surface of the sheet also agrees with what is seen in many other members of the group which have assumed the stratiform habit. Indeed, it is very noticeable that the remarkable regularity observable in the basic sills of the region is rarely realized in the sheet-formed intrusions of more acid rocks. The history of Tertiary igneous activity within the British area is, in its main divisions, a simple one. Setting aside the Sgurr of Higg, all the superficial volcanic manifestations, including the general outpouring of basic lavas and the more varied local out- bursts at certain centres, were comprised within the earlier. portion of the time. Later, the various groups of plutonic rocks— ultra- basic, basic, and acid——-were intruded inorder. Later still came the long succession of minor intrusions—sills and dykes—embracing numerous episodes, and doubtless representing a prolonged lapse of time. This general sequence of events in a complete cycle of igneous activity may be paralleled in various other regions. If the pitch- stone of the Sgurr of Higg, belonging to the very latest epoch, is held to represent a belated volcanic outpouring, it must stand as a unique exception. Without setting up such considerations against direct evidence, we may welcome any well-grounded interpretation which brings the apparent anomaly into harmony with what seems to be a significant law. One other point remains to be noticed in conclusion. The picture of the pitchstone-lava of the Sgurr filling an old river-valley has excited general interest, chiefly as affording a striking monument of erosion. Now, the unequivocal proof of enormous erosion which - meets the eye in every part of the Inner Hebrides is, on any reading of the phenomena, sufficiently impressive. Such sections as that of Beinn Mor in Mull prove that, in some parts of the region at least, the alternations of basalts and dolerites have had an aggregate thickness of more than 3000 feet. Assuming, as is generally supposed, that sill-intrusions can be formed only under a considerable cover, the alternating succession must have been capped originally by some thickness of basalt-lavas free from sills, adding perhaps another thousand feet to the total. Comparing this with the existing relief and geological constitution of the isles, we see reason to allow extraordinary activity to the agents of erosion in this region. In some parts, as on Rum and over about one-half of Skye, the great thickness of basic rocks has been entirely removed, and the plutonic masses, which in a general sense underlay them, have been left standing out to heights of 2500 and 3000 feet. This indicates a vertical erosion amounting, in some places and over considerable areas, to perhaps 5000 feet. The actual removal of material has, however, been vastly greater Oo)..G..S. Nosi24s: F 66 MR, A, HARKER ON THE GEOLOGICAL [ Feb. 1906, than that implied in this laying bare of the deeper rocks., In the precipitous sea-cliffs of Mull, EKigg, Canna, and the northern and western coasts of Skye, the truncated edges of the basalts and dolerites are exposed in sections 500 to 900 feet high. It is clear that the horizontal extension of the rocks has once been far wider than the present land-surface; and it is highly probable that the stratiform basic rocks of the several islands, now divided by soundings of 50 to 100 fathoms, are merely relics of one continuous tract. Faulting and differential subsidence have played an important part in producing the actual map of these western isles ; but we know of no such movements later than a somewhat early epoch in the phase of minor intrusions, and a truly surprising work must be credited to the co-operation of subaérial and marine erosion, perhaps with some assistance from ice.’ The basaltic lavas being of subaérial outpouring, their destruction must have begun as soon as they were formed; and the upper flows, without the protection of dolerite-sills, might be destroyed with comparative rapidity ; but it is none the less manifest that by far the greater part of the erosion was accomplished subsequently. It was not only later than the sills, but later than all the basic dykes, which occur up to the highest summits: that is, 1t was subsequent to the cessation of igneous activity. On the other hand, it was practically com- plete, except in the higher mountains and possibly in some of the sounds, before the advent of Glacial conditions. Between these limits we have to find scope for the agents of erosion to accomplish their work; and, even with all allowance for the land standing then at a higher elevation than now above sea-level, the results are not a little surprising. Bearing these considerations in mind, we see that the lava-flow hypothesis of the Sgurr of Higg confronts us with very great difficulties. It takes this small area quite out of relation with the rest of the region, and demands for it a chronology irreconcilable with that deduced elsewhere. The valley-bottom, now the highest ground on the island, appears thus as a relic of a vanished topography, of which we have no other trace. The carving out of such a valley, down to within 600 feet of the base of the basalts, implies a remarkable amount of erosion, which must be compressed within extremely narrow time-limits. For, by the same reasoning as before, it must have been inthe main subsequent to the intrusion of the dolerite-sills and dykes, and, ew hypothesi, it was completed prior to the pitchstone-epoch. The greater part of the work must in this case have been accomplished in a small fraction of the time covered by the phase of minor intrusions. Nor can we evade the difficulty by supposing the pitchstone of the Sgurr to represent an isolated episode, of much later date than all other igneous rocks in the region; for the evidence of immense erosion subsequent to 1 The soundings on the Admiralty-Charts indicate well-marked basins in numerous places. Good examples occur in the Sounds of Raasay and Applecross, on either side of Raasay and Rona. ‘O9Ia AO ¥HNDS AHL "07702 ‘asorutag ‘0704q ‘play ‘S ‘°K bs mace > 7 — ss .F*. (9DId AO 31S) HOWSHGIOS Nisaia "07]0.) ‘asorulag ‘070Yq. ‘plaN °S “KF ESE CLP ee OO en a oe Mol 625|- STRUCTURE OF THE SGURR OF EIGG, 67 that date is as clear here as elsewhere. It includes the removal of the greater part of the then existing mass of Eigg, leaving the valley exalted to a summit-ridge. It includes perhaps the separation of this island from Oigh-sgeir (Hyskeir), 18 miles to the west, now divided by a sea 40 to 50 fathoms deep. On the lava-flow hypothesis it certainly includes the cutting back of the coast of Eigg for miles, until the cliffs truncated the valley at an altitude of more than 500 feet above the present sea-level. If the pitchstone of the Sgurr of Higg is an intrusive sheet, this dilemma vanishes with the other difficulties which have been pointed out above; and, in its erosion as in other respects, Eigg is brought into correlation with other parts of the region to which it belongs. EXPLANATION OF PLATES III & IV. Puate IIT, The Sgurr of Higg, from a photograph by Mr. A, S. Reid. See key, fig. 3, p. 46. Prats LV. Bidein Boidheach, Isle of Kigg, from a photograph by Mr. A. S. Reid. See key, fig. 4, p. 53. Discussion. The PrustpEent said he was sure that, from the sentimental point of view, the feeling of the Fellows was one of regret that doubts should be cast upon an explanation of phenomena which had long been dear to all geologists. Sir ARcHIBALD Gerikie said that, owing to the lateness of the hour, it was not possible for him to criticize the paper in detail, but he hoped to have an opportunity of doing so after it had been published. He would have been glad to go over the ground with the Author and discuss the points in dispute on the spot, but he only heard of the Author’s views late in the summer, when it was no longer possible for him to return to the Western Isles. None of the points raised in the paper seemed to the speaker to be incapable of explanation in accordance with the interpretation of the Scuir which he had himself proposed many years ago. It appeared to him that the Author had very greatly exaggerated the proportion of sills among the plateau-basalts. It was difficult to conceive the physical possibility of these basalts being invaded by such a multitude of intrusive sheets, and yet that they should show so little visible sign of disturbance. That there were true sills in the district was well known, but the Author seemed to have somehow come to the conclusion that every columnar sheet must be a sill. * Reckoned from the sea-level at the epoch of the great erosion, the altitude would doubtless be considerably greater, F2 68 THE GEOLOGICAL STRUCTURE - [Feb. 1906, The speaker, on the contrary, believed that a large proportion of these columnar sheets were as true lava-flows as the amygdaloidal basalts with which they so regularly and abundantly alternated. The same tendency to see sills everywhere had apparently led the Author to frame his theory of the Scuir. Sills were commonly supposed to have forced their way along the planes of least resist- ance, as for instance between the bedding; but the pitchstone of the Scuir, instead of inserting itself between the nearly-level sheets of basalt, had filled up a hollow in these rocks, which, if not there _ before the intrusion, must have been ploughed out by the intrusive mass— an operation which would require a good deal of explanation, But that this hollow did exist before the advent of the pitchstone seemed to the speaker to be demonstrated by the section shown in the vertical cliff at the western end of the ridge. The pitchstone was. there distinctly seen resting in the hollow, the bottom of which was. occupied by a mass of coarse shingle or conglomerate. The hollow and the detrital material filling it had a close parallel in a remark- able section, which the speaker had described from the neighbouring: island of Canna, where a gully with vertical walls had, during the volcanic period, been eroded out of the bedded basalts and had been, buried under succeeding outflows. The Author asserted that the. mass at the western end of the Scuir was a volcanic agglomerate, but he adduced no evidence in support of that assertion. Any ordinary volcanic neck would have descended vertically through the rocks below, but the basalts underlie the Scuir conglomerate, and preserve. their unbroken regularity along the face of the chff. In lke manner, in Canna the gully referred to shows its shingle lying upon a flat unbroken pavement of basalt. It is manifestly the buried gorge of a watercourse, and the speaker held that the same expla- nation applied to the similar structure at the Scuir of Higg. The supposition that the pitchstone is a sill raises some insuper- able difficulties. The rock consists not of one continuous mass, but of a series of layers or beds, superposed one upon the other. According to the Author, the original glassy pitchstone has been invaded by a number of sheets of pale felsite. He offered, however,no explanation of how these sheets could be supposed to have split up the original solid sill by insinuating themselves, not between the columns, but across them, in a direction along whichit might have been supposed that they would meet the greatest obstacle to their progress. These felsites resemble the brecciated bottom of the main pitchstone, like which they were regarded by the speaker as parts of a succession of flows which filled up a leng and wide hollow worn by eresion out of the plateau-basalts. The Author of the paper seemed disposed. to minimize the amount of denudation during the volcanic period. The speaker, on the other hand, believed that it must have been large, and he referred in proof to the evidence of erosion and widespread accumulation of waterworn shingle to be seenin Canna and elsewhere. In con- clusion, he submitted that, while he was. ready to abandon his. published conclusions when ‘they were shown to be erroneous, much. Vol. 62.) OF THE soURR OF EIGe. 69 better evidence must be adduced against them than had been laid before the Society that evening. The AurHor, in acknowledging Sir Archibald Geikie’s criticisms, echoed his regret that a joint visit to the locality had not been found practicable. He was not to be understood as denying that the lines pro- cesses of erosion went on during the basalt-period. The fluviatile conglomerates intercalated in the succession were evidence of con- temporaneous erosion, although it is probable that these derived much of their material from the voleanic agglomerates rather than from the lavas. The intrusive nature of the dolerites had been deduced from the various direct criteria mentioned in the paper, the generalization that the vertical jointing is a characteristic of this intrusive group being a consequent conclusion. ) 70 PROF, E. H. L. SCHWARZ ON THE CoAst-LeDGES [ Feb. 1906, 5. The Coast-Lepexus in the Sovra-West of the Care Cotony. By Prof. Ernest H. L. Scuwarz, A.R.C.8., F.G.S., Rhodes University College, Grahamstown, South Africa. (Read November 8th, 1905.) THE subject of continental ledges has recently been brought into prominence in Europe and America by Prof. Hull, Prof. J. W. Spencer, and others; it is one in which attention to details has revealed many surprising results, which have modified our conception of the relation of land to sea, with the host of cognate problems. I offer the present contribution on the same subject, with the object of further emphasizing the importance of such investigations; for in South Africa we have the same set of phenomena, although we are able to study the movement in quite a different phase from that in which we find it in Europe and along the eastern coast of North America. The writings of the above-mentioned authors, together with those of Prof. J. Geikie, Prof. A. Issel, and Prof. W. C. Brogger, have made the subject so familiar of recent years, that I need not enter into a discussion of the general principles involved. Suffice it to say that, as the land rises, there are set-backs which allow the wash of the breakers to cut level plateaux of marine denudation; when a maximum elevation is reached, the land sinks and these ledges or fringing plateaux can be traced in the soundings set down on charts of the coast. In Europe, the first to draw attention to the subject were Godwin-Austen* and De la Beche*; and since then, the surveys for the submarine cables have furnished the information for a very complete statement of the occurrence in the North Atlantic. In the following pages I wili set forth what evidence we have along the southern coast of Cape Colony for the existence of these fringing plateaux, and then at the end compare them with the European and American ledges. The most striking of the coastal plateaux is that rising from 600 to 800 feet above sea-level, which extends from Caledon to Port Elizabeth. In the west it is cut in Bokkeveld (Devonian) Slates, with here and there inliers of sandstone that rise from the plain in the form of hills. To the eye of a casual traveller there is - very little evidence of a plain in this country; for there is a perfect labyrinth of steep-sided gorges, which cut the land into narrow ridges, or ruggens, as they are called in Dutch. If the observer, however, climbs on to an eminence and looks across the valleys, he notices that all the ridge-tops are cut to a level which slopes gently ‘ *The Valley of the English Channel’ Quart. Journ. Geol. Soc. vol. vi (1849) p. 69. 2 * Geological Observer’ 2nd ed. (1853) pp. 90-91. — Vol. 62.] IN THE SOUTH-WEST OF CAPE COLONY. 71 to the sea. The same thing is noticed in travelling: for the road that has to be followed, in order to get from one place to another, usually describes an immense circuit, but all on the level; whereas to go straight would necessitate climbing up and down a succession of very steep hills. The plain, or the tops of the ridges which now represent the plain, in the western part, is in places covered with surface-deposits. Near the mountains there are heaps of great, rounded, water- worn boulders, often cemented together in a siliceous matrix; farther away we find gravels with pebbles of all sizes, either loose, or cemented by silica or compounds of iron; still farther away there are sandy clays, which, according to circumstances, harden on the surface to an ironstone, or to a siliceous rock—burrstone, or freshwater quartzite. Fig. 1.—Ironstone-gravel forming beneath the sour soil of the Uplands plateau, seen in a railway-cutting south of George Town. ¢ ‘ c Le OG ‘ ‘ c c SoGet fe 1G 6 y: 6G.G ER Ue (2 SAS Cex tct Bee < cy ‘ A=Quartz-vein traversing the rotten granite, G; B=layer of subangular fragments of vein-quartz; C=sand, cemented into a granular mass by hydrated oxides of iron; D=grey sandy soil ; E=humus. [Compare the description of ironstone found beneath the sandy soil of Wolmer Forest, in Gilbert White's ‘Natural History of Selborne’ letter iv, Ist series. | \ To the east—or, to be precise, from George Town to Port Elizabeth —the’plateau-form becomes much more prominent ; the rock out of 2, PROF. E, H. L. SCHWARZ ON THE CoASt-LEDGES [ Feb. 1906, which it is cut is mostly Table-Mountain Sandstone or granite, and both of these are far more resistant to the action of weathering agencies than the clay-slates of the Devonian Bokkeveld Series. Very deep gorges do traverse the plateau here, but they run straight Lay es Ss "~ a Oo wm 4 OO eS SERS $a = iS) ~ & Cisfaa te) | = = Oo ps a KS) a S Cua. ee soe ea eee SS <<) lal S> So o oS .- & “—~ mS Sea oS '2 Oo a g fia Cae eS on ee eee 38 io St lo ee eS eS NS iS) @& ire] 5S pie omt> Si Le 2 © a & a Sg @ 40-8 4S s oy Be 20 lee ca ‘ ) a 2 sr ese Se nt 8 : > n & 2 ® an O ao fS Sa See = Seeing Soot Seas) ao a | 2's B 3 Ss 3 oe oe oowT a SOS ets = Ac) Bd e SS 5 e Se mice. ot 3 Ss Lape Sse at ee & ==) rs ££ = Mog BO 5 2 Sees ee Se oe Se NSS fr ee OS SS 2 S22, oO 4S Sy =) Ss o@Snm—26 8 fi 3 He 5S SS S 8 pss Gf, mS gay NSN ee Sa ar a = Sos oes 2 OB Sa aS SS) i we ee) if S 284338 °/ 9 i= ° Ge, (6) Fe [S| = ©o@ OA HY ff | FS A Sos . ran ea a AN Oras! 1D os yf : eo & / 20 reas om / Bs f- ~~ Hilly ih}! /, pyle it itd VMs ean ep Mibey ht Bee STORMS RIVER VILLAGE. down to the sea from the mountains, and such water as falls upon the plateau sinks in to fertilize the soil, and drains away through the cracks and joints of the rock. The character of the vegetation is very different on the plateau east ‘and west of George Town. In the ruggens-country there 1s very Vol 62.1 IN THE SOULH- WEST OF CAPE COLONY. fr) little soil on the hills, the bare rock in most cases showing through. The bushes were originally the woody Karroo-bushes, which force their roots down in the crevices formed by surface-expansion and contraction, and the necessary soil exists only in these small fissures ; a useless, imported, rhenoster-bush, however, is dis- placing the natural vegetation, to the great detriment of the country. In the east, the vegetation grows on a thick, sandy soil, and consists of grasses, sedge-like plants, proteas, heaths, and many of the beautiful Cape-iris family. The ruggens-veld is sweet veld, but the heath-country is sour. Beneath the sand on the sour veld there commonly forms an ironstone, due to the action of water dissolving out iron-compounds from the rocks, and this shuts off the roots of the plants from the substratum of rock ; where this is granite, which contains the finest mineral foods for plants, such as lime, phosphates, and potash, the surface-vegetation is as poor as that on the Table-Mountain Sandstone, which is made up of pure silica with scarcely a trace of argillaceous matter. If, by any chance, the soil is removed from off the ironstone-gravel, the latter hardens on exposure to a massive bank of ironstone; it is then what is sometimes called laterite, although there is no alumina to render the application of the term strictly correct. The deposits on this coastal pigeon have given rise to much controversy, and their nature seemed so essentially that of river- deposits, that for a long time the plain was thought to be a base- level of river-erosion. In Komgha, in the extreme east of the Colony, some of the freshwater quartzite was found to contain seeds of Chara, thus pointing to the fact that the deposit was formed in fresh water. But it does not necessarily follow in this particular case, because the deposit on the plain is evidently fresh- water, that consequently the plain itself must have been cut by river-action; for Mr. W. Anderson * has recorded Chara-seeds in a deposit on the shores of a lagoon on the Zululand coast, which lies evidently on a plain of marine denudation that has just emerged from the surface of the ocean. In Caledon and Swellendam, where the width of the plateau is some 15 or 20 miles, and the seaward end shut off by a series of outstanding hills, the problem is complicated ; but, standing on the top of a peak of the Zitzikamma Mountains, and surveying the ledge at one’s feet, which is only some 3 or 4 miles broad, it seems impossible that rivers coursing down from the mountains could, in so short a distance, acquire that sluggishness and lack of velocity which would be required to enable them to cut their channels sideways, and level the rugged country. ‘There is no possibility of invoking a different cause for the levelling in the two places, for the - same plateau is continuous, right away from the western end of Caledon to the town of Port Elizabeth ; the same level, the same gradient of fall from the mountains to the sea, the same general 1 2nd Rep. Geol. Surv. Natal & Zululand, London, 1904, p. 51. 74 PROF, E. H. L. SCHWARZ ON THE CoAsT-LuDGES [ Feb. 1906, nature of the deposits, are found on the ledge from east to west, and an explanation of its formation in one part must hold good for the whole. As I think that we have evidence for ruling out the possibility of river-erosion as a cause for the cutting of the ledge in the Zitzikamma, so must we abandon the theory of the ruggens- plateau being a peneplain. Another plateau, commencing on the east of Port Elizabeth, and beginning at the sea-front with an elevation of some 150 to 200 feet, is covered with undoubted marine shingle: it rises to 400 feet, and at the base of the next plateau it is 460 feet above tide- level. All over the plain there are these coarse boulder-beds, sometimes associated with shell-deposits characterized by a large Pectuncuius. Most of the shells are of species living on the coast at the present day, but the larger forms are now found only along the Natal coast. On the top of the plateau and in the surface-gravels, is a circular depression called the Groot Pan or Zwartkop’s Saltpan; it annually fills with rain-water, which on evaporation throws down salt of great purity, 100,000 bushels being extracted yearly. I do not think we can say that this salt- pan, by itself, is an argument in favour of the marine origin of the plateau; though the simplest explanation of the occurrence would be that the salt results from the leaching-out of the eva- porated sea-water that was held up in between the boulders as the beach rose above sea-level. The enormous amount held up in the pan is the puzzling feature; for the apparently-inexhaustible supply would seem to necessitate a soil impregnated with salt in the neighbourhood, yet the dense bush of prickly plants, aloes, euphorbias, mimosas, scutias, etc., grows up almost to the water's edge. The saltpans in Bredasdorp and at Bethelsdorp, nearly at sea-level, seem also to receive an adequate explanation on the theory that they are the result of sea-water evaporated and impregnating the neighbouring soil; but one can hardly so explain the saltpans, at Prieska, for instance, which are hundreds of miles away from the coast. In stating that there was no break from east to west in the Uplands plateau, I spoke of the land as we have reconstructed it ; for nowadays there are many gaps, caused by the inclusion of immense deposits of loose Cretaceous rocks, which have readily yielded to the impetuous torrents that coursed over the plateau. These deposits, which we term Uitenhage Beds, are made up of a basement- series of gravels known as the Enon Conglomerate, and of sands, clays, and marls, characterized by Teniopteris and Clado- phlebis denticulata. They once covered a very much greater extent than they do at present, but after their deposition some extraordinary cross-folding took place, which let them down in hollows in the older rocks upon which they had been lying; the basins in which they are now found are bounded by steep walls, sometimes a buried fault- face, sometimes a very steep fold, and are entirely closed. During Vol. 62.] IN THE SOUTH-WEST OF CAPE COLONY. 19 the levelling process all trace of the Cretaceous deposits was removed, except in these hollows, where the unbroken rim of the more consoli- dated, older rocks protected them. When the plateau was raised, these Uitenhage Beds exhibited the same levelled surface as the rest of the rocks, and recent gravels were deposited upon them ; but the enormous powers of downward erosion which the rivers now acquired, attacked these loosely-made rocks and carried them away to the sea: hence there are wide, open basins on the plateau wherever the Cretaceous deposits occur. As regards the recent deposits found on many of the ridges of the Ruggens, I think that the greater part of them are formed in situ—that is to say, before the country was cut up into these ridges, there were extensive plateaux covered with thick sandy soil, such as we now See is the case to the east in Knysna, on the same plateau, though on a harder substratum of rock. Water percolating through this soil would form the ironstone-gravel by cementing the sand-grains with hydrated iron compounds, and the surface- quartzite could be formed in the same way by the deposition of silica, In Uniondale [ have seen the process of formation of the burr- stone going on. The plateau there has been cut, very thick deposits of clayey sand have been piled upon it, and the gradual hardening upwards, till the surface is glassy quartzite, can be studied.’ Gravel does exist on the ridges, and this cannot, except within very narrow limits, be explained in a similar manner: when a hard reef projects above the general level of the plateau (for instance, a quartz-vein through granite), fragments, apparently waterworn, of the same material are found in the loose sandy soil in the neighbourhood ; but they have never been noticed actually to form gravelly deposits (see fig. 1, p. 71). Yet gravel is not formed by river-action alone: beaches are aS common as river-gravels ; and Godwin-Austen as long ago as 1849 recorded shingle on the English platform at from 80 to 100 fathoms below the surface of the water. Behind the mountains, we find the remains of plateaux cut in various rock-systems, resembling the Uplands plateau in certain features, only they are of considerably greater elevation. They are covered with surface-quartzite, ironstone-gravel, and shingle, exactly as in the coast-plateau. After studying both, I am unable to distinguish the two deposits one from the other, and yet I see as complete evidence for ruling out the action of waves from the explanation of these inland plateaux, as I see for abandoning the action of running water in explaining the formation of the coast-ledge. We are not directly concerned with these inland plateaux in the present paper; they were described in the memoir on the ‘ High-Level Gravels of the Cape, &c.’ referred to above. ' See photograph in ‘ High-Level Gravels of the Cape, &c.’ Trans. Phil. Soc. S. A. vol. xv (1904) pt. 2, pl. iii. 76 PROF. E. H. L. SCHWARZ ON THE COAST-LEDGES [Feb. 1906, In Bredasdorp and Riversdale there is an immense accumulation of blown sand on the coast. The dunes rise to 600 and 800 feet above sea-level, and rest upon shelves of varying heights. In Bredasdorp, near Cape Infanta, the base is seen some 100 feet above sea-level, and is composed of hard quartzites ; to the west, however, the base goes below sea-level.’ In Riversdale the base seems to have been generally only a few feet above sea-level, and the strata consist of exactly the same rocks as those which go to form the ruggens-country behind—namely, Bokkeveld Slates. ‘Sand on the southern coast of South Africa is shell-sand. ‘The waves throw up the valves of dead molluscs, which get rolled up and down by the ebb and flow of the tide, and finally are broken into such small fragments that the wind can carry them away. ‘This sand, then, masses up in these great dunes, sections of which show very clearly that the whole of the deposit is wind-borne; for the false bedding is so abrupt and striking, that no streams or currents could have produced such marked effects. The angular fragments of shells sometimes break down to a powdery, chalky mass; but as a rule the sand remains just as it was, piled up without any alteration except on the surface. Where rain-water can penetrate and come to rest, there is a process of solution and redeposition of lime which consolidates the sand into a rock hard enough to be used for building-purposes. The fossils found in the sand-dunes are all of recent age, the commonest being the snail, Cyclostoma. In the two districts with which we are immediately concerned, bones of the elephant, rhinoceros, eland, sable antelope, and large carnivores have been found—all species which have, since the coming of the white man, disappeared from this neighbourhood. In Riversdale and Bredasdorp the level of the top of the sand- dunes is very nearly the same as that of the plateau behind; there has been a little erosion at the junction, and a well-watered valley extends at the inner foot of the dunes; the rise, however, from the old high-level plateau to the dune-country is very slight. In George and Knysna there is a very similar accumulation of sand ; only here the original bluff of the plateau, rising steeply in preci- pices from sea-level to 600 feet, is bare, and between it and the sand-dunes there is a level tract of low country occupied by the far-famed Knysna Lakes. I should be inclined to atttribute the difference between the two tracts to the direction of the prevailing winds. In Riversdale and Bredasdorp this would have been slanting with respect to the land, whereas in Knysna the winds blew straight from the sea. In the latter case the steep cliffs would have produced a back-eddy which swept the sand away from their base, whereas in the former the sand could be piled up against them. Whatever be the explanation, the difference does exist. _ 1 See Rogers & Schwarz, ‘Notes on the Recent Limestones on Parts of the South & West Coasts of Cape Colony’ Trans. Phil. Soc. 8. As vol. x (1897-98) p. 427. Vol. 62.] IN THE SOUTH-WEST OF CAPE COLONY. we There is a further and more important difference in the Knysna dune-area—namely, the base upon which the sand rests is below sea-level. Itseems at first sight strange that so loose a material as the sand-rock should remain consolidated when it is sunk beneath the water, but we have had ample proof that it does. Near Danger Point, the bluffs of wind-deposited sand dip steeply seawards, and ' pass beneath the tide-level; and at Cape Infanta there is a great cave excavated in the sand-rock, with an old beach forming the roof. Dr. J. C. Branner mentions that, at Fernando de Noronha, wind- bedded sand occurs below tide-level.’ Off the George coast there are several ledges, separated from each other by steps of a couple of feet or so, made of consolidated sand-rock, the uppermost of which is rarely covered by the sea, so that the spray evaporates in the pools and yields salt; but the lower ledges are covered and uncovered at each ebb and flow of the tide, and the terrific surge of the breakers beats upon them incessantly. ‘The surface of the ledges is exceedingly rough, and has been formed by the rock cracking along intersecting joints; these become filled in with lime and form walls round the enclosed blocks. When exposed to the waves the softer interiors become eaten away, leaving the walls outstanding, and where two or more cracks intersect a little pinnacle is produced. Sometimes the cracks are very scanty and the included blocks correspondingly large, so that wide, shallow pools are formed, with low enclosing ridges; these are densely crowded with calcareous alge, worms, and other lime-secreting organisms, and thus the surface is further strengthened by a thick deposit of calcareous matter. The question arises: if the sand-rock can be so hardened, might these ledges not be the ordinary shore-sand hardened 2 situ, by a process similar to that which, as Dr. Branner (in the memoir just mentioned) maintains, built up the stone-reefs of Brazil? At the western end of the George sand-belt the limestone-ledges do not appear above tide-level; but they exist below it, as shown by the numerous fragments that are thrown up along the shore, yet there is no tendency for the beach-sand to form similar ledges. As certain proof exists in the caves of Cape Infanta that the sand-rock has at one time gone beneath sea-level, and as all sections along the coast where the ledges occur are favourable to the view that the sand-rock does pass under the sea, we can accept with confidence the statement that these lmestone-ledges are formed of wind- bedded sand Hardened by sea-water. There is a very simple explanation of this hardening. When the tide recedes, it leaves behind it in the pools and hollows of the ledges a certain amount of sea-water, which contains carbonate of lime in solution, in virtue of the carbonic acid which it holds; when the water is exposed in shallow sheets to the heat of the sun, the gas is driven off and the lime deposited, the process thus necessitating, in the first place, a more or less solid ground for the 1 Bull, Mus, Comp. Zool. Harvard, vol. xliy (1904) p. 229. 78 PROF, E, H, L. SCHWARZ ON THE COAST-LEDGES [ Feb. 1906, water to come to rest on, and explaining at the same time how it is that loose sand is not similarly consolidated. At Port Elizabeth there is a beach-deposit cemented into what is known in South America as a stone-reef; this may have been formed 7m situ, for the sea-water, when the tide ebbs, would be retained under the rounded stones, and carbonate of lime deposited as the water became heated. At Hermanuspetrusfontein, on the Caledon coast, there is a well- marked surf-cut ledge about 50 feet above sea-level,’ which would correspond to the ledge of rock underlying the sand-dunes at Cape Infanta; while farther out, at Danger Point, the sand-dunes rest upon a ledge that has sunk beneath the sea. At Agulhas Point the ledge is just at sea-level round the lighthouse, but it is very narrow here. Most of these places along the coast are so inaccessible, and the problems with which we are concerned in this paper have been so recently advanced, that the exact heights of any one of these ledges has not been determined with any accuracy ; but the point which I wish to bring out is that there are more or less extensive ledges, admittedly surf-cut, at various levels above and below tide-mark along this part of the coast. The Cape Peninsula extends from Cape Point to Table Bay, and is a precipitous mass of sandstone resting upon a basement of pre-Cape rocks and granite. It is separated from the mainland by a wide tract of low-lying country covered with drifting sand, known as the Cape Flats. The sand is gradually being fixed by plantations of trees, and quickgrass and other creeping plants, but much of it is still pure, loose sand. If one bores through the sand one meets first with a layer of ironstone-gravel (the so-called laterite), and then a thickness of clayey material derived from the weathering of the Malmesbury Beds, and finally the hard blue clay-slate itself. It is very difficult to say at what level the true rock-shelf exists under the sand, owing to the want of available evidence ; but I think that the ironstone may be said to rest upon the original surface. This, then, is commonly found near sea-level, taking an average, sometimes going below, sometimes occurring above that datum. Geologists have always been very chary in expressing an opinion as to the nature of the cause which cut this level fiat ; but I see no reason for propound- ing a separate explanation for this particular plain, when we have similar ones, clearly surf-cut, all along the coast to the east. The laterite-, gravel-, and lignite-deposits seem to point to these having been formed on land, but the shingle and tree-trunks may have been rolled to their places by the sea, and the laterite can and does form in situ beneath any sandy soil. Therefore I-see no hindrance, in the nature of the deposits found in the Cape Flats, to the acceptance of the agency of sea-waves in cutting the original plain. There need mot here be any question of a strict mathematical plane: for we 1 A, W. Rogers, ‘The Geology of Cape Colony’ 1905, p. 380, — Pt ea Se . . Wai. 62. IN THE SOUTH-WEST OF CAPE COLONY. 79 know that, at many places where there are shallow lakes and lagoons, there are hollows in the rock-shelf; but these are distributed so irregularly that they can offer no support to the theory that running water was the cause of the cutting of the Flats. Near Cape Point there is a raised beach about 100 feet above sea-level; and Mr. A. W. Rogers describes some very well-marked shingle-deposits on the same level on the western coast, near the Van Rhyn’s-Dorp area.’ At the base isa conglomerate strongly impreg- nated with iron-oxide ; the pebbles are of all sizes up to a foot in diameter, and consist of a considerable variety of rocks—sandstones, quartzites, schistose quartzites, granites, quartz-porphyries, vein- quartz, and slates. Above this layer come more or less ferru- ginous sandstones, which are in places silicified. In Bamboes Bay, 3 miles south of Strandfontein, the ferruginous conglomerate passes upwards into sandstones, and these again into calcareous beach- deposits containing shells of species still living on the coast. ‘The country behind the coast, at a distance of 4 miles from it, lies between 500 and 600 feet above sea-level. It would be very interest- ing to follow the surf-cut levels northwards. At Port Elizabeth and Uitenhage, the beach-deposits and shell-banks lie on a shelf about 200 feet above sea-level; but this slopes gradually up to 600 feet, and I am inclined to reckon the 100-foot level as a substage of the Uplands plateau. We have, then, a plateau 700 feet above sea-level, and one at about sea-level. JI will now discuss one at considerable depths below: this is the plateau the edge of which is known as the Agulhas Bank. — Soundings off the south-western corner of South Africa soon get into depths of 40 to 50 fathoms; and then, out to sea, there is a gradual slope of the sea-bottom to 90 or 100 fathoms. The marine charts seem to express a succession of ledges rather than one continuous slope, one of the plateaux being at 45 to 60 fathoms, another at 70 to 80 fathoms, and an outer one at 90 to 100. Close in-shore, along the Zitzikamma, is a narrow shelf at 30 fathoms, with a sharp drop off the edge to 50 fathoms. At one place on the plateau there is a sounding of 134 fathoms which may be simply an embayment, but also it may be a rock-channel, which once carried the waters of the Breede River out to sea at this point, nearly 100 miles from its present mouth. The effect of a rise of the continent of 600 feet would prolong the land some 150 miles to the south, with the apex of the point shifted somewhat to the east of where it now lies. We should then have two well- marked plateaux: one at 600 feet, more or less, and one at 1300 feet, more or less, above sea-level, besides the one at or near sea-level. As it is, the remnants of one are found to exist at about 1500 feet, * ©Geological Survey of Piquetberg, Clanwilliam, & Van Rhyn’s Dorp’ Ann. Rep. Geol. Comm. 1903 (Cape Town, 1904) pp. 161-62. 80 PROF. E. H. L. SCHWARZ ON THE COAST-LEDGHS _[ Feb. 1906, and this I will shortly describe. Thus our sub-continent seems to have been subject to lifts of 600 to 700 feet or so, with smaller intermediary halts and set-backs. Shell-sand, green sand, mud, and boulders have been dredged off the surface of this submarine plateau. For some years now the s.s. Pieter Faure, the Government trawler, has been making investigations over this area with regard to the supply of fish; and when the records are worked up we shall have, perhaps, some more definite data upon which to go. Some large quartzite-boulders, dredged 40 miles off Mossel Bay, showed that there at least land must have been at one time. The occurrence of this plateau, corresponding so closely with that which surrounds the eastern shores of America and the western shores of Europe, led me, in a recent paper, to suggest that it was caused by the bending of the edge of the continent seawards: owing to the carrying-away of material from the interior by denudation, and consequent relief of weight, and the deposition of this material as sediment off the shores, and consequent weighting of this area. If, however, we consider the coast-ledges to be plains of marine denudation, this explanation is unnecessary, aS we have evidence of block-uplift of the whole continent, and not of bending of only small portions of it. Whether there are plateaux below the Agulhas Bank, must remain problematical, until more extensive exploration and sound- ings have been made. The evidence, so far as we now can judge, goes to suggest that the ground off the Bank belongs to the absolute base-level of erosion—that level beyond and below which the action of running water was never effective, taking the continent as the unit which it now is. In former ages there may have been elevations and depressions along the coast which had an absolute base-level different from that which we assign to the movement that we see going on at the present time: an elevation, for instance, that may have connected the island of Madagascar with the mainland, or a depression that permitted the deposition of the Cretaceous marine rocks or the Karroo Beds. These periods are, however, remote, and our powers of reconstructing the physical conditions of the continent are faulty, from the want of data upon which to go; consequently, we are not likely to have more than one absolute base-level at any one place to deal with, and the confusion of speaking of several absolute base-levels will be avoided. There is a change in the fauna exhibited in the 200-foot shell-deposits at Port Elizabeth and Uitenhage; and if the 2500-foot plateau had remains of animal life upon it, they would indicate a further change, but there is nothing to lead one to suppose that even the highest undoubted surf-cut shelf was made in times as far back as the Tertiary. The absolute base-level of erosion, as I conceive it, refers simply and solely to the movements. of con. tinents as they now are, and has nothing whatever to do : : | | | : : 4 Vol. 62.] IN THE SOUTH-WEST OF CAPE COLONY. 81 with the question of the permanence of the continental outlines in past ages: that is a question which may be decided one way or another, without affecting the absolute base-level. I have intro- duced the expression because I wish to imply that continents are in a state of motion, upwards or downwards, and that the absolute base-level represents the ocean-floor, including the shelf or level of erosion cut by the surf and off-shore currents that came near the water’s edge when the depression of the present land-masses com- menced ; it is a datum from which to reckon those upward or downward movements of continents which are actually in progress. Fig. 3.—The De Vlugt plateau (about 1500 feet), looking north from Paarde Kop (2397 feet), across the valley of the Keurbooms River. Ne 4 u x ~ iN Ki \ \ | XX XN { i eae ae eR wos oe eal ea Sami Sai INS hia il i i) [In the distance are the Long Kloof Mountains, which may be the remnants of a higher plain, 5000 to 6000 feet above sea-level; the crests are seen to run in a remarkably-level line, and the highest peak of the range (Krakeel River) has an altitude of 5500 feet. The valleys are occupied by outliers of Bokkeveld Slates folded in with the Table-Mountain Sandstone. |} 1h Wp: PAL Te ne i es Uae ce sits A = In the Division of Knysna there is a complex of ridges under the mountains, formed by folded Table-Mountain Sandstone, with syn- clines filled in with long outliers of Bokkeveld Slates. Although the synclines pitch so that the slates are often cut out, and the sand- stone-strata in places stand vertical, at others with very high dips, yet the tops of the ridges all rise to a level sky-line about 1500 to 2000 feet above sea-level. Peaks rise above this ; but the evidence is satisfactory for the existence in past ages of a plateau at an altitude of about 1500 feet. Behind the mountains are the remains of a vast plateau, at about Q.J.G.8. No. 245. 3 G 82. PROF, E. H. L. SCHWARZ ON THE CoAsT-LEDGES [TF eb. 1906, 3500 to 4000 feet above sea-level,! which I have endeavoured to explain by the base-level of erosion being caused by impassable obstructions in the course of the rivers; and I still do not see my way to offer any other explanation. Had the base-level been the resuit of the depression of the land, we should obtain a most remarkable physiography for the western part of the Colony: there would have been a narrow fringing reef of mountains, with an immense central lagoon, like an atoll magnified. There are no traces of a 3500-foot terrace on the coast-sides of the mountains— a fact, however, which might be explained by the great denudation that takes place on the seaward side of mountain-chains ; even in the dry interior the remains of this plateau are exceedingly scarce, and there would have to be some exceptional cause to allow of such being protected on the coast, did they exist. The only example of which I can think, is the summit of Table Mountain at Cape Town, 3549 feet above sea-level; but I do not wish to press this into the theory, unless many other confirmatory facts can be gathered. I have often sat on a mountain-top, and, surveying the enormous landscape that opens out before one in the clear air of South Africa, have noticed how many of the chains of folded mountains cut a level sky-line. I have thought, too, that there might have been a vast peneplain at elevations of from 5000 to 6000 feet (see fig. 3, p. 81). Such a plain would have left only a few peaks of the folded mountain-region projecting as islands, and it would have included the top of the escarpment of the great inland plateau. This escarpment is now seen as the dolerite-capped hills that face the Great Karroo, the crests of which form the watershed between the streams running into the Orange River and those flowing southwards to the Indian Ocean. Turning eastwards, we find the remarkably well-preserved marine plateau near Port Elizabeth and Uitenhage, to which I have already referred in discussing the marine origin of the rock-shelves. The Uplands plateau comes eastwards to Port Elizabeth, but is, I think, not quite so high above sea-level as at George or Knysna. The plateau, with the shingle and shell-deposits, is only some 150 to. 200 feet high along the shore, but rises inland to 463 feet at Coerney ; at this level it reaches the next step or escarpment of the higher plateau of about 1000 feet. We know too little about the country beyond this point to discuss the facts further; but, as. I gazed over all this part from a peak on the Baviaan’s Kloof, there appeared to be a succession of level plains, cut off by steep cliffs on the seaward side, beginning with the 4000-foot plateau and sinking gradually to the coastal one of an altitude not exceeding 150 to 200 feet. At Kast London there is a remarkable succession of plateaux. The following levels along the railway-line will convey some idea — of the nature of the country, 1f one considers that each of the ‘High-Level Gravels of the Cape, &c.’ Trans. 8.A. Phil, Soc. vol. xy (1904), pt. 2, p. 43. ee ee Vol. 62.] _ IN THE SOUTH-WEST OF CAPE COLONY, 83 localities mentioned lies on a flat plain, or one that slopes gently towards the sea, and is terminated by an abrupt drop :— Feet ‘Baste bound onye28. 88: 3 ek: Poi Camibridge 3. -sa.--4.28 etek 467 Hort ackKSOW, 72.6 a2 .00cckadeossBes 1108 BET Yap aint a ala 1638 MS ae sees avin econ ee eee 1775 Sterkstroome » coe eee Ge. 4406 Cypherrat. 2. .cctics aden. 5450 In the Transkei and Pondoland the coast-shelves are very strongly marked. The main plateau is elevated about 2500 feet above sea- level, and still, in a very few cases, retains masses of surface-gravel and quartzite. One hill standing on the plateau, called Kentani Tree Hill, where there was a fierce fight in the Gaika-Galeka war, has an elevation of 2420 feet; the peculiar freshwater quartzite, with the many-coloured clays and sands not yet consolidated by the siliceous cement, and the conical shape which the small cap of hard rock has given to the hill, have puzzled many generations of people, both black and white. Not so very long ago it was declared a diamond-pipe, and the whole white community in the district resolved itself into a syndicate to work it. The general level of the plain is some 500 feet below the level of the hard caps; for, although the whole area is covered with grass, there is a vigorous denudation going on which is rapidly reducing the level of the country. At Kentani village there is a cliff letting the plateau down 1000 feet, and thence by other steps to a low ridge only 50 to 100 feet above sea-level. In the neighbourhood of the St. John’s River the coastal geology becomes very complicated; but in Eastern Pondoland, where there is a simple shelf of Table- Mountain Sand- stone, the subsidiary plateaux are as sharply defined now as when they rose from the water’s edge. ; The task of gathering together facts from levels carried out by eye-surveys, with very few points fixed in height, cannot lead to satisfactory results unless appeal can be made to more detailed work. The reason why I have brought in these eastern plateaux is to show, firstly, that plains of marine denudation do clearly exist all along this coast ; and, secondly, that the level of the plains observed at the various localities varies considerably. TaBLe oF Coast-SHELVES IN THE OCAprr CoLony. He Western. Midlands. Eastern. ee Feet Feet Feet Feet Cyphergat ...... 5000 to 6000? 5000 to 6000 ? 5450 ? — Sterkstroom ... 98500 to 4000? 4000 ? 4406 ? 4500 ? Beomtani .20c.c86: eee ioe 2500 2500 De Viuet......... 1500 1000 1500 . 1500 Wplands’ ......... 700 463 467 600 Bamboes Bay ... 50 to 100 200 157 50 to 200 Sea-level ......... —- -—— — — enlhas ......0.- — 600 a2 84 PROF. E. H. L. SCHWARZ ON THE COAST-LEDGES [ Feb. 1906, In the foregoing table I have cited only the main plateaux ; the figures with a query after them refer to cases where the action of marine denudation is doubtful. The highest undoubted rock- shelf is the Kentani plateau, in which the-lower road through the Native Territories is laid, along which are built the villages of Kei Road, Komgha, Butterworth, Idutywa, Umtata, Flagstaff, and Bizana. In Europe we have traces of similar terraces. The edge of the shelf supporting the British Isles is generally known as the 100-fathom line. In front of this is the great escarpment of the continent, descending steeply to 7000 or 8000 feet, sometimes precipitously ; the level at the bottom is what I have called the absolute base-level of erosion. Off Scotland, the margin of the plateau lies about 600 feet below the surface ; but in front of the English Channel, the soundings indicate depths of 1000 to 1200 feet. South-west of Spitsbergen lies a plateau depressed 2418 feet below sea-level, with depths immediately beyond of 8100 feet; and, farther off, the sea-floor sinks to 15,900 feet. Between Greenland and Iceland, the submerged plateau is depressed 1500 feet. The drowned valleys of Europe open on to the sea-floor at depths of from 7000 to 9000 feet, the greatest of them being the valley of the Adour.’ South of the Straits of Gibraltar the continental plat- form descends to 7200 feet, with a gradual slope probably, ‘ consist- ing of a succession of minor terraces breaking off in cliffs.’* In the Northern Atlantic and Arctic Ocean, littoral shells have been dredged up from depths of 8500 feet.’ I would, therefore, make the abso- _ lute base-level of erosion at an average of 9000 feet on the eastern side of the North Atlantic. Prof. Hull,in the paper just referred to, publishes a chart of the ocean-floor off the mouth of the Congo, and shows a drowned valley opening out on to an absolute base-level of erosion at 12,600 feet, so that there is apparently a depression of the sea-floor. South of this point, I find evidence of elevation of the absolute base-level of erosion, from this great depth to only some 1000 feet ; and, in passing, I should like to anticipate the conclusion of this paper by drawing attention to the nature of the river- systems north and south of the Congo: that is to say, the land- relief, in areas where the absolute base-level of erosion is very deep below tide-level, is favourable to broad valleys and permanent rivers, but where it is high the land-surface is furrowed by deep gorges and traversed by dry river-channels. On the western side of the Atlantic, there is a corresponding continental shelf, submerged some 200 or 250 feet below the ocean- surface, with an external fringe submerged another couple of hundred feet. Off this, there is a long, gradual descent to deep water, mostly unbroken by ledges; but the Blake Plateau interrupts it in places by a shelf from 2500 to 3500 feet deep, with an average * Alphonse Milne-Edwards, Bull. Soc. de Géogr. vol. iii (1882) p. 118. * HK. Hull, ‘Sub-Oceanic River-Valleys of the West-African Continent, &e. Trans. Victoria Institute, vol. xxxii (1900) p. 148. ° W.C. Brogger, Norges Geol. Underségelse, No. 31 (1900-1901) pp. 682-83. Poet Sis Vol. 62. } IN THE SOUTH-WEST OF CAPE COLONY. 85 of 2700 feet. The submarine valleys which cut through the plateaux reach to depths of 12,000 feet below sea-level. Even at these great depths there still appear to be plains, which probably were once near or above the surface of the water; for instance, south- east of New Jersey the contour-lines between 9000 and 10,500 feet are some 60 miles apart." The greater depths of 20,000 feet, such as are reached in the Bartlett Deep south of Cuba, probably indicate structural depressions below the absolute base-level of erosion. If we admit the expression absolute base-level of erosion, we have then evidence which goes to show that this surface is depressed 4000 feet more on the western shores of the North Atlantic than it is on the eastern, and that the same depression is found off the mouth of the Congo on the eastern shores of the South Atlantic, but that south of this there is a rapid elevation to within 1000 feet of ocean-level, We can, then, compare the following plateaux :— Norru ATLantic, | NortuH ATLANTIC, Soutm AFRICA. AMERICAN SIDE. EuROPEAN SIDE. Feet Kentani plateau . 2500 De Viugt plateau. 1500 Feet | Feet Uplands plateau. 700 Bea level. ccc. — | Sea-level ......... = Sea-level ......... — Coast-shelf...... 300 | Coast-shelf ...... 600 Agulhas plateau. 600 Blake plateau . 2700 | Iceland shelf ... 1200 A.B.L.E. ...... 1200 ie Gai Sage 9000 ee the. 12,000 | (A. B. L. E. = ‘ Absolute base-level of erosion.’| Whatever may be the objections to the term absolute base-level of erosion (and I myself can see many), it expresses some surface that must exist ; and once we grasp this fact, we have a most fertile medium through which to look at the unexplained differences in the surface-conditions of continental areas. By its means we determine that Europe and America are, on the Atlantic coast, on the down- grade of the oscillatory motion which all land-masses are under- going, and have very nearly reached bottom. South Africa, on the other hand, is on the upgrade, and probably near the top. Compare the topography of Europe and South Africa. In Europe, one finds flowing streams which have cut their beds so that adjoining stream- systems meet at the crest of the dividing-ridge: in South Africa, the stream-beds are separated by great distances of level country, the actual river-valley occupying only a narrow V-shaped gorge or channel. In Europe, the valleys are wide, great tracts of. alluvium spread along the borders of the streams, and deltas form at the mouths: in South Africa, the valleys are cut almost always in rock, and at the mouths the rivers run out to sea through rock- bound gates. In Europe, when a dam is to be built, a moderate ' J. W. Spencer, Bull. Geol. Soc. America, vol. xiv (1908) p. 218 ; see also zbid. vol. vi (1895) p. 103. 86 PROF, E. H. L., SCHWARZ ON THE COAST-LEDGES | Feb. 1906, height of wall will impound an enormous quantity of water; in South Africa, big dams are almost out of the question, because the grade of the rivers is so great that, in order to impound any quantity of water, the dam-wall requires to be built so high that the pressure on the base becomes greater than can be safely undertaken, unless at enormous cost. In Europe, the land having gone on sinking for many ages, the rivers have been constantly checked and the downward erosion stayed; in South Africa, the uplift has been so continuous that the rivers have been converted into agents that sawed downwards, and have often found the rate of uplift beyond their powers of erosion, so that waterfalls along their courses have resulted. This remarkable contrast of the two continental surfaces—the one, the European, very old; the other, the South African, very young— is so striking and so certain, that the relative positions of each of them in the oscillation of land-surfaces, which I have sketched above, cannot be doubted. By establishing the term absolute base-level of erosion, I have endeavoured to give some quanti- tative expression to.a fact which otherwise is difficult to grasp. There are many problems that will yield to treatment by this touchstone, such as the tilting of continents, the African north- wards, the European westwards, the American eastwards; but in South Africa, one has no opportunity of getting together literature. on world-wide questions, and I must conclude with the facts as I have observed them locally, and leave to others the task of testing them and applying them to other regions. Discussion, The Presiprenr welcomed the first communication read to this Society by one of that enthusiastic band of geologists, the value of whose work had been recently judged on the spot by the members of the British Association who had visited South Africa. He (the speaker) regarded the checking of the results obtained by study of the geology, by reference to the present conditions of non-change, as a matter of great importance. , Prof. Hurz said that he entirely agreed with the Author, that these terraces were the outcome of marine erosion, while the gorges by which they were intersected were due to river-action. The two forms of erosion were here displayed in striking contrast. The occurrence of the marine shells on the 150-to-200-foot shelf was clear evidence, so far as regarded the Bamboes-Bay shelf ; but the evidence of marine sculpturing could not be limited by this occurrence, which must be extended to the remarkably-uniform terrace of the ‘Uplands, if not to those above. The submerged Agulhas Bank was of especial interest, and (in the speaker’s opinion) was probably the representative of the great ‘continental shelf, or platform, on which the continents of Europe and Africa were planted, and through which the sub-oceanic river-valley had been cut down for several thousand feet. The speaker thought that the paper was well Wol.62. |. IN THE SOUTH-WESL OF CAPE COLONY. 87 calculated to throw increased light on those great oscillations of the coast-lands which had taken place from the Pliocene Period down to the present time. Mr. H. W. Monoxron remarked that raised rock- platforms of marine origin, apparently of the same class as those described in the paper, were found along the. Norwegian coast, and had been termed strandflade or ‘coast-plane’ by Dr. Reusch. The speaker thought the term a good one, and better than ‘coast-shelf.’ He observed that the raised beach of Gower rested upon just such a coast-plane as those found in Norway. He was glad to hear that the Author found in Africa evidence of periods of comparatively- rapid earth-movement, separated by comparatively-long periods of repose, for the speaker had noticed much evidence of similar alternate periods in both Great Britain and Norway. 88 DR. J. W. EVANS ON THE ROCKS OF THE [Feb. 1906, 6. The Rocks of the Cataracts of the River Mapertra and the Apsornine Portions of the Brent and Mamort. By Joun Wituram Evans, D.Sc., LL.B., F.G.S. (Read November 22nd, 1905.) [Prare V—Microscopz-SeEcrions. | CoNTENTS. Page I, General Configuration of the Region ................0esescee ues 88 fee revmious: Work <...)\2l2¢.cede sce ee eee eee eee ee 92 ITT, Geological and Petrological Observations :— (1) Between the Andes and the Cataracts ................4- 93 (2) The Cataract of Esperanza on the Beni ............... 95 (3) The Cataracts of the Mamiorés...2 5: ...2- e.-<0s.0 seer 98 (4) The Cataracts of the Madeira .s.............s0s0se0e-er 99 (5) The Granulites and their Genesis ...............sceeeee 119 (6) The Black Coating of the Rocks of the Cataracts ... 121 (7) Below the Cataracts. ..5 cee caccesecne 2 eee 122 V., “SUMMA oncikc does eee ate hee ae ee 123 I, GenerRaL ConFIGURATION OF THE REGION. More than eleven years have passed since the publication of a paper by this Society on ‘The Geology of Matto Grosso,’* in which I described the rocks that I met with in the area drained by the Upper Paraguay and its tributaries. They include granitoid gneiss, highly-cleaved slates, limestones, sandstones, and shales, and have a prevailing north-easterly and south-westerly strike parallel to the coast-ranges of Southern Brazil. To the northward they are covered unconformably by a younger series of sandstones and shales con- taining imperfectly-preserved Devonian fossils.” These dip gently northward, and form a broad undulating tableland extending from east to west. Still farther north, between latitudes 10° and 4° 8. on the Xingu (pronounced Shingu), Dr. Karl von den Steinen * found granite, gneiss, and other crystalline rocks. Nothing is said about the strike, but it seems that the trend of the hills was approximately south-west and north-east, as in the basin of the Upper Paraguay. East of the River Paraguay the high ground changes its trend, and stretches in a north-westerly direction to the cataracts of the Mamoré and Madeira, under the names of the Serra dos Parecis and ’ Quart. Journ, Geol. Soc. vol. 1 (1894) p. 85. * O. A. Derby, ‘Nota sobre a Geologia e Paleontologia de Matto Grosso ~* Archivos do Museu Nacional do Rio de Janeiro, vol. ix (1890) pp. 59-88. Mr. Derby believes that the two series of sandstones and shales are identical ; but, although I nowhere saw them in contact, they appeared to be in every way distinct. * *Durch Central-Brasilien’ Leipzig, 1886, pp. 330 (227), 333 (268), and maps ITI & general. Nol’ 62. CATARACTS OF THE RIVER MADEIRA, ETc. 89 Cordilheira Geral. At first it constitutes the watershed between the tributaries of the Paraguay and those of the Tapajos, while farther to the north-west it forms the north-eastern boundary of the basin of the River Guaporé or Itenes, which pursues its way in a parallel direction for nearly 400 miles on the north-eastern margin of the alluvial plain of South-Eastern Bolivia, until it unites with the Mamoré. The combined stream turns almost due north across the line of elevation, and, at a height of nearly 500 feet above the sea, its waters commence their descent in a long succession of cataracts to the Amazonian plain. After passing the falls of Guajard-merim, Guajard-guacu, Bana- neira, Pao Grande, and Lages (in Spanish ‘ Layes’), the Mamoré arrives at its confluence with the Beni, which has already passed the cataract of Esperanza. The joint stream now takes the name of the Madeira, and flows first northward and then north-eastward across the line of elevation, descending cataract after cataract, until it reaches the settlement of Santo Antonio, at a height of only about 200 feet above the sea. Thence, after a winding course of some 500 miles through swampy forest-clad alluvium, it empties itself into the Amazon. At the cataracts the rivers make their way through low ridges ot crystalline rocks, which form the subject of the present paper. Where these are foliated their strike is, as a rule, north-west and south-east, in the same direction as the ridges. The total width of this hilly region is about 170 miles, although the rivers in their windings cover a much longer distance in traversing it. The hills continue to the north-westward beyond the cataracts, but how far they extend is not known. We are told of falls on the River Ituxy, a tributary of the Purus, which flows on the north- west of, and more or less parallel to, the Madeira, so that the line of crystalline rocks probably reaches at least thus far. There is also a rapid, known as the ‘Cachoeira,’ on the Purus, still farther in the same direction, which may perhaps be referred to the same cause. It is noteworthy, too, that the Purus, the Tarahuacd-Jurua,' the Yacarana or Yavary, and the Ucayali-Marafion, all have a change of direction corresponding to that of the Mamoré-Madeira at the cataracts, and that the points at which these changes of direction occur are in continuation of the line of strike of the crystalline rocks of the cataracts of the Madeira. ' This may be explained by the fact that the general inclination of the country on the borders of Brazil, Bolivia, and Peru is slightly to the east of north, while the elevation that forms the cataracts runs obliquely across it: the result being that, when the rivers reach the ridge, they are diverted down the slope north-westward until a point of weakness or inferior elevation is reached, by which a passage is effected. It is true that there is no record of hard rocks having been observed on most of the rivers that I have mentioned, but they 1 That is, the stream formed by those rivers successively. ny SE TELS 7ST AE ATTRAC TEE es YS we fh AWN Og ‘ ay » 7 40, wo a) ay Je a Hie SSS 4 Ss \ tao Se LE ait! / oLuojzUYy OVUPSO < te fo AON OLBON! we es ee Ae oer eco tae aes snyp palflaan fijazasduiooul Pupp Oe 40 Uolzona/F £0 saulTz ~. ae Ne Sc ee Bulpjoy 40 uonwnasz fo saul7 Ne AS da Pe at ac mae ws 7 & P \ | SUIPTOF 10 woryeao]qy Jo soury Joryo CT a ws dj} JO suoljIesIp oyeuTxXo1dde oy oN NOUN XS d ot} sulsoys pre ‘\ > VOINANV HLiLNOS . DOULA) eb 2 er ery jo jaed usoyzION 9313 & wise atts jo dey-yoj99S —"y “314 Vol. 62.] THE ROCKS OF THE MADEIRA CATARACTS, ETC. 91 are still imperfectly known; and in the tropics the process of lateritization often renders slates and crystalline rocks rich in alumina almost indistinguishable from recently-formed alluvium. It must be remembered, too, that there is a wide belt east of the Andes which appears to have been depressed, and where the whole country is flooded in the rainy season. In this region there must be a considerable accumulation of alluvium, which may now cover the ancient rocks of the ridge. We have, therefore, evidence of an important axis of folding and elevation, with a south-easterly and north- westerly direction, in the centre of South America; and there is reason to believe that it extends for a distance of 1200 miles, from 144° lat. 8. and 59° long. W. to 4° lat. 8. and 73° long. W. near Iquitos. Its direction is parallel to that of the Andes in Northern Bolivia and Southern Peru, as well as to a small inter- mediate axis of elevation in Chiquitos in South-Eastern Bolivia.” These lines of folding follow the north-westerly and south- easterly strike, which is so common on the earth’s surface for some distance on either side of the Equator. I also appears to prevail in Southern Venezuela, in Guiana, and in North-Eastern Brazil, where a barrier is believed by Dr. Friedrich Katzer* to have formerly stretched across the site of the mouth of the Amazon. There appear to be four principal directions of folding or elevation in South America, namely : (1) The north-westerly and south-easterly strike already described, which is also met with in the south of the continent in Argentina and Patagonia, as well as in Tierra del Fuego and the Falkland Islands. (2) A direction roughly at right-angles to the former, approximately north- east and south-west, following the coast of Brazil south of Cape San Roque. rn se occurs on the Upper Paraguay and the Xingt, and in the Colombian naes. (3) An east-and-west strike seen in Northern Venezuela and Trinidad, and in the Amazonian basin, which consists broadly of a syncline bounded by two lines of elevation on the north and south. (4) The north-and-south line of the Andes, from Southern Bolivia to Patagonia, eas These directions are roughly indicated in the sketch-map which forms fig. L (p. 90). In the present imperfect state of our know- ledge of the geology of South America, it is impossible to give their exact positions, The movements along all these lines appear to have extended over a considerable period. The north-easterly and south-westerly folding of Matto Grosso was mainly pre-Devonian, while the north-westerly and south-easterly movements in the Andes ex- tended to at least post-Carboniferous times. The north-and-south * More exactly north 55° west. * A. @Orbigiy, ‘ Voyage dans l’Amérique Méridionale, exécuté pendant les années 1826-33’ vol. iii, pt. 3 (Géologie), Paris 1842, pp. 183-99; and J. W. Hvans, ‘ The Geology of Matto Grosso’ Quart. Journ. Geol. Soe. vol. 1 (1894) p. 96. ee der Geologie des Unteren Amazonasgebietes’ Leipzig, 1903 p. “aU. 92 DR, J. W. EVANS ON THE ROCKS OF THE [ Feb. 1906, and east-and-west lines of movement appear to be of Tertiary and Quaternary age. Along the Pacific coast, movements with lines of strike’ parallel to the shore-line have, it need scarcely be said, continued down to historical times. Il. Previous Work. With the exception of some brief notes by Joseph and Franz Keller and Col. Church,! the only description which has yet been published of the rocks of the cataracts is that given by Dr. Joao Severiano da Fonseca, who in the year 1877 descended the rivers Guaporé, Mamoré, and Madeira with the Commission for the delimitation of the frontier between Brazil and Bolivia.* As this book is comparatively inaccessible to geologists in this country, I have translated, and reproduce here, those portions which seemed to be of interest in connection with this paper. The following paragraphs refer to the geology of the cataracts as a whole :— Vol. ii, p. 280. ‘ The rocks of these cataracts are of plutonic formation, and reveal at the first glance their volcanic origin, modified perhaps by meta- morphism. Some were difficult for me to classify, on account of the obscurity of their characters; in others the mineralogical facies was satisfactorily deter- minable. The great trachytic pavements, which are nearly smooth, and either of a ferruginous hue or shining black like pitch, are formed in many places of superposed beds, which are more or less undulating and have curvilinear borders, as if they had been derived from viscid melted material poured forth in great outbursts, and forming sheets of which the later solidified before they reached the distance to which the earlier had extended. Here and there appeared large rock-masses, some prismatic in shape, others rounded: in one place were dykes of diorite and of elvan; in another loose blocks. Some were split in the middle by a mere crack, others by a gap of more than a fathom in width.’ ‘There were likewise large cauldrons, holes in the pavement, periectly rounded, the formation of which is easily explained by the attrition of stones rolled about in small depressions which, little by little, with the movement of the waters and the passage of centuries, become larger and more rounded. But it is not so easy to explain the elliptical holes in some of these pave- ments .... all are of the same dimensions, and as if arranged in uniform directions one after the other in two or three lines, so that they call to memory, though without any resemblance, human footsteps. The most notable are those of the cataracts of the Madeira, Bananeira, Ribeirao, and Paredao. Their dimensions are: 1 to 3 decimetres long, a third more or less of this broad, and nearly as much deep, always preserving an ellipsoidal form. Are they spaces formerly occupied by bodies easily disintegrated or decomposed by the waters, and that in time became vacuous? As for the pavements, in spite of their being varnished by the attrition of the water and_ brilliant with a black metallic polish, it is not difficult to classify them by their texture and system of agglutination. They are hornblende-porphyries, obsidians, syenites, petrosilex, etc.—all felspathic rocks. The canga (ferruginous con- glomerate) appears in lofty crags of a reddish-black colour, whence they have received the Tupic name tupanhona canga. At the same time, dykes of * Franz Keller-Leuzinger, ‘Vom Amazonas & Madeira’ Stuttgart, 1874; and George Earl Church, ‘The Route to Bolivia via the River Amazon (a Report to the Governments of Bolivia & Brazil)’ London, 1877, p. 186. 2 ‘Viagem ao Redor do Brasil’ Rio de Janeiro, 1880-81. —_—" Vol. 62. ] CATARACTS OF THE RIVER MADEIRA, ETC. 93 compact eurite rise to form other crags: either by breaking through the metamorphic crust, or because they are enveloped by decomposed gneiss the surface of which, now eroded by lapse of time, was formerly at the same level as that of the dykes.’ ‘In the large cauldrons, which are dry, there are not infrequently con- glomerates of small fragments of dioritic rocks, principally black diorite, which appeared to me to be cemented with hydrate of iron.’ ‘I brought back with me some specimens of the most notable of these rocks, as well as of the pebbles inserted in the cracks of the pavements, where a secondary process of agglutination with the sand of the river avd _ the clay which it carries in suspension, forms a puddingstone...... These sedimentary rocks are rare in the localities where the stream runs strongly, but very frequent in the backwaters,’ . In one of the cataracts, he believes that of Bananeira, he found a piece of petrified charcoal formed of bright shining lamelle, which is now in the Museum of the Archeological Institute of Alagoa. It will be seen from the following pages that, although I have been able to confirm some of Dr. Fonseca’s statements, I never saw any rocks that could be classed as trachytes, or that showed any evidence of having flowed at the surface. IIJ. GrorocicaL AND PErroLtoGicaAL OBSERVATIONS. (1) Between the Andes and the Cataracts. In the year 1902, when returning from Bolivia by way of the Beni and Madeira, I had an opportunity of examining the rock- exposures on those rivers. Leaving the Bala-Susi Mountains, the last outworks of the Andes, behind me at Rurenabaque, I descended the Beni until I _ reached, twenty days later, the ‘rubber-metropolis’ Riveralta, at the mouth of the Manutata (Madre de Dios). Throughout _ this distance (some 220 miles in a direct line, and 470 following the stream) the Beni flows in innumerable meanders through a wide forest-plain, between 500 and 600 feet above the sea. For the greater portion of the distance the ground is scarcely raised above the water, although the margin of the river is usually marked by a low bank or levée forming a natural breakwater, which, however, is often flooded when the river is high. Beyond are lagoons repre- - senting abandoned reaches of the river. At a greater distance from the Andes there are, however, tracts of rising ground, occasionally as much as 50 or 60 feet above the level of the stream, which some- times cuts into them, forming well-marked cliffs. These elevations are built up of argillaceous or fine sandy materials with varying amounts of iron, which is sometimes present in such quantity as to constitute a hard stone. With this exception there is no solid rock, either i situ or in fragments larger than fine sand-grains,! below the rapid of Altamarani, some 10 miles below Rurenabaque, where the river, slackening its pace after leaving the mountains, has thrown down a thick bank of more or less rounded fragments. ’ Kyen this sand is remarkably poor in heavy minerals. Araya, Santo Antonio ' Fig. 2. Sketch-Map Matese (ae of the Theotonio# Cataracts of the Rio Madeira iS & the adjoining portions of the ee Beni and Mamoré. Morrinhos | Founded on the map by Joseph & Franz Keller in Vom Amazonas und Madeira’ aes Cataracts, rapids & localities < ‘ See Nie where crystalline rocks occur, Y/Caldeirzo » alto do Gira oO; Slate and quartz ~ said to occur on this © oe <, 7 river, Y Tres Irmaos > . TParedao\ ‘2s, ‘ 2. ' ~ a. . Statute Miles ro = [For ‘ Araya’ read ‘ Aroya.’] a i i Vol. 62.| THE ROCKS OF THE MADEIRA CATARACTS, BTC. 95 After a short stay at Riveralta, literally ‘high bank,’ which stands on a cliff such as I have described, I ascended the Manutata, and thence travelled overland to the Orton. Between the two rivers is a narrow plateau, extending for many miles in the direction of its length, and raised some 40 feet above the level of the lower forest. Though it is possible that these tracts of com- paratively-high ground may in some places be formed of hard rocks decomposed by lateritization, there can be little doubt that, in the great majority of cases, the materials of which they are composed are of fluviatile origin, and were laid down by rivers under condi- tions similar to those which now exist. They are, therefore, in all probability the remains of the former surface of an ancient alluvium, the rest of which was removed by the present rivers, whose level has been lowered by the wearing-away of the barrier at the cataracts. I was unable tosee any definite evidence of local changes of relative level in these deposits, which may be provisionally referred to the ‘ Neogene’ of Dr. Friedrich Katzer.' I followed the Orton to its confluence with the Beni, and subse- quently descended the latter river. (2) The Cataract of Esperanza on the Beni. A short distance below the mouth of the Orton the Beni passes at the ‘ Correnteza, or rapid, a line of rocks, said to be thirteen in number, projecting here and there above the water, and having a general north-westerly and south-easterly direction. Although there is no perceptible fall at this point, the current runs with considerable force,’ and it was impossible to stop the boat to examine the rocks. This rapid forms the cabecera or head of the cataract of Esperanza.” A mile or two farther down stream the cataract itself is reached, where the river passes a number of ridges of crystalline rocks which. form low elevations in the forest on either side. ‘The principal fall has a height of 9 or 10 feet, and below it are rapids which con- stitute the coda or tail of the cataract, where the river passes yet another ridge of rock. Megascopically the rock (M 1)? of the Esperanza Cataract shows a granular mixture of pinkish felspar, with a little greyish quartz and specks of a black mica, the latter, however, being mostly aggregated. into parallel lenticles or irregular masses. ‘The specific gravity is 2°633. The strike of the foliation is fairly constant, in a west-- north-westerly and east-south-easterly direction. 1 *Grundziige der Geologie des Unteren Amazonasgebietes ’ 1903, p. 108. * Jn 1897 an island at this point was swept away by a flood. 3 So called from the words of the devoted Indian who accompanied Dr. Heath, the first European to descend the Beni from the Andes to its mouth: ‘ Then let us call this cataract Esperanza, for now that we have oa it we have hope of life.’ * Numbers in parentheses preceded by the letter M refer to the r eee anens and microscope-sections now in the Mineralogical Department of the British Museum (Natural History). 96 DR. J. W. EVANS ON THE ROCKS OF THE [Feb. rg06, Under the microscope, the felspar is seen to consist partly of microcline and orthoclase, both with microperthitic inclusions on a minute scale, and partly of albite, which sometimes contains more or less rectangular inclusions of microcline or orthoclase arranged along definite lines and extinguishing together. Quartz occurs in small oval or circular blebs, such as are characteristic of granulites, or in rounded lozenge-shaped individuals which appear to be sections of the double pyramid, for the extinctions are parallel to the diagonals. Other crystals are more or less hexagonal, and occasionally these are dark in all positions of the nicols. These grains or crystals of quartz are usually found as inclusions in felspars, or penetrating into their margins ; while, in some instances, the felspar appears to fill the irregular intervals between crystals of quartz, or the eroded cavities in that mineral. There can be no doubt that the crystal- lization of the quartz must, in all these cases, have preceded that of most of the felspar. Micrographic structure, however, occasionally occurs where the two must have crystallized together; and there are also cases where the quartz fills the interstices between felspars, and must have been the last mineral to separate out. There is a fair amount of biotite and brown hornblende, which is sometimes intergrown with quartz in a kind of micrographic structure. I also observed a large, rounded, greenish-yellow, highly- refracting isotropic crystal, 1-25 mm. in diameter, surrounded by hornblendes, which is altered externally to a yellow decomposition- product with low birefringence. The mineral is traversed by eracks filled partly with a mineral of high birefringence, about double that of quartz, and partly with fluor. It may, perhaps, be a silicate of the rare earths allied to pyrochlore. A few grains of apatite, zircon, and magnetite are present. The rock is rather fine-grained, the largest crystals, felspar and quartz, not exceeding 3 millimetres in diameter. The foliation was not apparent under the microscope. In some bands or veins (M 1 d) the constituents approach a centi- metre in diameter, and the dark constituents almost completely fail; ‘so that the rock may be described as a rather coarse-grained haplite, consisting of quartz, a little microcline, and albite. Inthe last are numerous minute plates, usually appearing like needles in thin section. ‘These are probably hydrous soda-mica, the result of incipient decomposition. Elsewhere (M 1 a) there are dark bands in the gneiss that reveal, under the microscope, an entirely-ditferent composition. The felspar is plagioclase, showing albite and frequently also pericline- twinning in alternately broad and narrow bands. The maximum angle of extinction with the lamelle of the albite-twinning, in sections perpendicular to the brachypinakoid, is 20°, and the refrac- tive index is well above that of the slow (extraordinary) vibrations in quartz. The felspar is, therefore, apparently an acid labradorite. It contains numerous minute inclusions of hornblende, quartz, and apatite, and occasionally lines of rectangular isotropic inclusions Vol. 62.] CATARACTS OF THE RIVER MADEIRA, ETC. 97 which are probably filled with liquid. Quartz occurs in small blebs, and shows rounded idiomorphic boundaries. There is abundance of brown prismatic hornblende, with a maximum extinction- angle between the direction of vibration of the slower wave-surface and the trace of the cleavage of about 24°, and strong pleochroism. Tt contains inclusions of quartz.’ Crystals of pyroxene are also present; and much of it is distinctly pleochroic, ranging between very pale shades of green and yellowish- brown. ‘This pyroxene appears to be partly augite and partly enstatite. Some of the augite shows signs of alteration into diallage. Irregular masses of magnetite also occur. The rock, which ‘has a specific gravity of 3-012, is apparently on the border- line between a micro- dicrite and a fine-grained dolerite. In the neighbourhood of the cataract, pebbles again appear in the Bee beaches. Some are formed of crystalline rocks ; but the majority are derived from sedimentary beds, and ae con- solidated grits or quartzites and silicified ae or chert, showing traces of stratification. The latter contain organic remains with concretions and rhombohedra of a ferruginous carbonate, perhaps ankerite, which has been oxidized to a yellow-brown tint : and a groundmass of colloid and chalcedonic silica. I submitted the microscope-sections, that I had had made, to Dr. George J. Hinde, F.R.S.. He kindly examined them, and wrote to me as ae — ‘The circular and elliptical sections in slide (M1e) are sponge-spicules. Probably they are the anchoring spicules of siliceous Hexactinellids. The inner tube is the axial canal of the spicule. They seem to be fairly common in this chert. There are traces of rods in slide (M1); these are probably also spicules, but the chert is so much altered in this slide, that one cannot be positive. ‘There are other bodies in slide (M 1 é) the nature of which is unknown to me. One is a thin shell, almond-shaped in section, which may be the carapace of some crustacean ; a smaller form may be of similar origin. ‘There is also an imperfect flask-shaped body with a reticulate structure, of which I cannot guess the character. ‘T do not see anything that can be considered radiolarian; but the chert has been very much altered and, judging trom the changes which have affected the sponge-spicules, the radiolaria, if originally present, would have been obliterated beyond recognition.’ In a subsequent letter Dr. Hinde states that he considers the chert to be marine, and that it may well be Paleozoic, although the evidence is insufficient to prove it. Rocks containing Hexactinellid sponge-spicules occur to the north of the Amazon on the rivers Trombetas and Maecurt. The accompanying fossils indicate a horizon at the base of the Silurian (using that expression so as to exclude the Ordovician).” The sedimentary rocks cannot have come from the Andes, the 1 See p. 121, footnote 1. 2 F. Katzer, ‘Grundziige der Geologie des Unteren Amazonasgebietes’ 1903, pp- 218, 222, Q.3.G.58. No. 245. H 98 DR. J. W. EVANS ON THE ROCKS OF THE [Feb. 1906, river being far too sluggish to carry stones of such size. They must be derived from strata flanking the crystalline rocks on the south-west, possibly from the reef of thirteen rocks above isperanza. The occurrence of marine beds in this region of alluvium and crystalline rocks is of great interest. I may mention that I found pebbles of silicified oolite in the Upper Paraguay, at Santa Cruz, Barra dos Bugres.* (3) The Cataracts of the Mamoré. Leaving the Esperanza Cataract behind me, I descended the river to Villa Bella, which is situated on the tongue of land between the Beni and the Mamoré. Thence I made ashort journey up the latter river,in a canoe, to the rapidsof Lages, where broad rock-pavements, which give their name to the cataract, are exposed when the river is low. Megascopically this rock (M2 &3) appears to consist of quartz and felspar speckled with flakes of black mica. The size of the constituents varies from place to place, but the grain is always rather fine. There is distinct banding, which has more the appearance of flow-structure than of the foliation of gneiss. The rock splits somewhat easily, but in a direction which is parallel rather with the surface than with the planes of banding. Its specific — gravity is 2°66. A microscope-section of the coarser (M 2) material (Pl. V, figs. 1 & 2) shows abundance of felspar, including microcline—sometimes with microperthitic inclusions, and plagioclase with lamellartwinning. The latter appears to be intermediate between albite and oligoclase; it shows in some cases inclusions the boundaries of which are parallel with those of the host; they have lower refraction and birefringence, and are probably orthoclase. Quartz occurs, both independently and as inclusions in the felspar; as in the rock of the Esperanza Cataract, it appears to have crystallized out before the latter mineral. The sections often show crystallized outlines, but are sometimes only rounded blebs. Similar rounded inclusions of felspar, especially microcline, also occur. In some cases, the felspar appears to have crystallized in the cavities of the quartz. Micrographic intergrowth is fairly common. There is much brown and green mica representing biotite in different stages of alteration into chlorite. Epidote, apatite, and magnetite are also present. The finer-grained specimens (M 3) are similar, but show crystals of yellowish-green hornblende, with an extinction-angle (between the direction of slower vibration and the cleavage) which reaches a maximum of 29°. The colours of different directions of vibration vary between yellowish-green and dark green. The largest quartz- and felspar-crystals measure about a millimetre in diameter, but numerous small quartz- and felspar (microcline)-grains occur, both 1 «The Geology of Matto Grosso’ Quart. Journ. Geol. Soe. vol. 1 (1894) p. 91. Vol. 62.] CATARACTS OF THE RIVER MADEIRA, ETC, 99 independently, and as inclusions, which in size and shape recall the minerals of a granulite. I did not ascend the Mamoré any farther. A.d’Orbigny,' who spent a few days at Beira on the right bank of the Guaporé or Itenes, some distance above its junction with the Mamoré, found, on the south- western flank of the Serra dos Parecis or Cordilheira Geral, hills formed of friable sandstones which were very ferruginous, and generally red in colour. They were of great thickness, and dipped south- eastward? at an angle of 12° or 15°. These sandstones, which appeared to extend far to the north, were covered in the neigh- bourhood of the river by ferruginous conglomerates containing much oxide of iron, and forming perfectly-horizontal beds. Dr. Fonseca states* that, at the cataract of Guajard-merim, a chain of rocks some 150 metres broad traverses the river from side to side, expanding on its margins into two enormous pave- ments of dioritic appearance. In some places there is, he says, a porous formation of a kind of varnished can ga (ferruginous quartz- conglomerate) ‘ resembling phonolite; ’ while on the bank, amid the vegetation, ‘argillotalcose schists’ without visible stratification are exposed. He also states that, above the cataract of Lages, at the mouth of the river of the same name, the rocks are covered by sands resting upon dark-grey clay with ‘ nuclei of silex’ (op. cit. p. 270). According to Col. Church (‘The Route to Bolivia via the River Amazon’ 1877, p. 187) the upper rapids are composed of ferru- ginous conglomerate, the surface of which is as black as ink, and he quotes the following from a report by the Kellers :— ‘The ferruginous conglomerate which is found on the surface of the earth, only covered with a bed of clay of from 5 to 6 metres of thickness, is a con- glomerate of gritstone, little pieces of dolerite cemented with oxide of iron, full of openings and cavities which give it the appearance of a sponge or scoria. Its beds are generally horizontal, and are from 4 to 5 metres thick. In the inferior beds the seams are smaller, at some points disappearing entirely, and forming then a more homogeneous mass of red gritstone, very argillaceous.’ They declare that this formation extends over more than 12 degrees of latitude. Col. Church (loc. cit.) further states that the rock of the Banaueira rapid is ‘distinctly granitic, but with much iron disseminated. The surface of the rocks, wherever the water has been flowing over them, is blacker thanink.... the hornblende, feldspar, and quartz are well disseminated in the rock.’ (4) The Cataracts of the Madeira. Immediately below the confluence of the Beni and Mamoré is the cataract of Madeira, where the river of the same name flows rapidly for 2 miles between rocky islands, the total fall being over 8 feet (23 metres). 1 “Voyage dans l’Amérique méridionale, exéeuté pendant les années 1826-33 , vol. iii, pt. 3 (Géologie) Paris, 1842, p. 203. 2 This is not the dip that would have been expected. Possibly it may be a verbal mistake for south-westward. ° ‘Viagem ao Redor do Brasil’ vol. ii (1881) p. 263. \ H 2 100 DR. J. W. EVANS ON THE ROCKS OF THE _—[ Feb. 1906, The rock is a pale granitoid or haplite-gneiss, in which the ferromagnesian minerals are but little developed." In some places it is fairly coarse-grained, and in pegmatoid veins the constituents measure aS much as 2 centimetres in diameter. On examination of a microscope-section (M4a) from one of these coarser portions, it appears to be mainly composed of irregular masses of quartz, microcline, and orthoclase. Sometimes the border of the quartz sends out a number of rounded prominences in crystalline continuity,, which penetrate the felspar micro- graphically. The felspar is altered in places into a colourless mineral with a higher refractive index, and a negative sign of the principal zone. ‘The relative retardation of the two directions of vibration is higher than in quartz, the birefringence reaching a maximum of about 0-029, or 29 thousandths.* This is probably a somewhat hydrous muscovite; the cleavage is, however, not well marked. Plagioclase intermediate between albite and oligoclase is also present. Elsewhere the rock (M4) shows megascopically a fine granular mixture of quartz and felspar, with small aggregates and rods of a dark-green mineral. Under the microscope it is seen to bea granu- litic rock, consisting of a mixture of rounded quartz-grains measuring up to half a millimetre in diameter, and plagioclase (albite-oligoclase) of about the same size. The general appearance has a curious resemblance to some quartzites where the grains are uniform in size. There are a certain number of felspars that show only Carlsbad-. twinning or none at all; but the refractive index appears to be identical with those showing twin-lamellation, and they have 1 Dr. Fonseca describes here the occurrence in the ‘ syenite’-rocks of the oval and elliptical holes already mentioned. ° The birefringence or difference between the indices of refraction of the two directions of vibration represents the relative retardation in a unit of length, and might be termed the rate of relative retardation. If <& be the total relative retardation between the two directions of vibration, Z the thickness of the section, and d the birefringence or rate of relative retardation, then a=", In this particular case the observed relative retardation (/) was 670 micromilli- metres (millionths of a millimetre), and the thickness was 23,000 wicromilli- metres, so that d was equal to 670+23,000=-029. It is, however, convenient to measure the thickness of the section in microns (a thousandth of a millimetre = 1000 micromillimetres), and to take as a practical unit of birefringence the number of micromillimetres of relative retardation in a micron: thus avoiding the use of small decimal fractions. If then D be the rate of relative retarda- tion in micromillimetres per micron, % the relative retardation in micromilli- k 670 ===, = 29. i 23 The unit rate of relative retardation in micromillimetres per micron is zoo0 Of the absolute unit of birefringence or rate of relative retardation, in which both the relative retardation and the distance are expressed in the same units; it is, therefore, conveniently spoken of as a ‘ thousandth.’ Perhaps a ‘millesim’ would be better; see a paper in the Mineralogical Magazine, vol. xiv (1905) p. 87. The words micron and micromillimetre are here used in accordance with the rules laid down by the British Association for the Advancement of Science. metres, and L the thickness in microns, D = Nel. 62.] CATARACTS OF THE RIVER MADEIRA, ETC. 101 probably the same composition. This view is supported by the fact that there are many felspars in which the repeated twinning is only just perceptible. Many of the plagioclases contain approximately- rectangular inclusions, arranged parallel with crystallographic directions of the host. These have a low refractive index, and are apparently isotropic ; they probably represent spaces formed by the action of solvents, and filled with liquid. No microcline is visible. The small quartz-grains are enclosed in the felspar, and are often idiomorphic. There are also a number of minute, more or less rounded inclusions that are arranged in lines passing through all the constituents of the rock. Flakes of biotite more or less con- verted into chlorite are present in considerable numbers. Veins of quartz are observed in the rocks of this cataract; and pebbles of the same material are bound together by ferruginous cement, to form a hard conglomerate that is met with here and there in the alluvium. The rocks are, as a rule, distinctly foliated on a ae sctale.; The strike is very variable. The band of erystalline rocks rises to low elevations on either side of the river; and it is said that a day’s journey into the forest to the south-east there are considerable hills. The river now runs for 10 miles through low alluvial land, until at the cataract or rapid of Misericordia, another band of erystalline rocks is passed. I landed on the left bank, a short . distance above the cataract, and found the gneiss stretching into the forest in low rounded knolls, which may have been worn by the river when it flowed at a rather higher level. To the naked eye, the rock (M 5) appears to be a gneiss, consist- ing of an aggregate of small grains of pink felspar, occasional quartz, and bands of a dark-green material. It has a specific gravity of 2:63. The strike of the foliation is north-north-westerly and | south-south-easterly, but it is apparently somewhat variable. Under the microscope, abundant microcline is seen, up to a diameter of 2 or 3 millimetres; there is also some microperthite, as well as albite-oligoclase containing minute inclusions of a colour- less material, similar to that seen in the microcline of the haplite from the Madeira Cataract: in this case they probably represent a hydrous soda-mica. There are large crystals of quartz, of about the same size as the felspars, and separated from them by a ragged irregular boundary that corresponds to the crystalline outlines of neither ; while smaller blebs or rounded crystals of quartz, such as have been already described, occur as inclusions in the felspars. In the greater portion of the rock, the ferromagnesian minerals are present only in small amount, being represented by flakes of mica and hornblendes which elsewhere form the coloured bands. a>b (namely, with the least absorption in the direction of the optic normal), while the other forms have t > >a. 116 DR. J. W. EVANS ON THE ROCKS OF THE [| Feb: 1906, A number of traverses were made through a microsection, in order to estimate the volumetric composition of the rock by the method of Rosiwal,’ who showed that the volumes of the different minerals are proportional to the sums of their intercepts on any line or lines drawn through the rock, if the number of minerals traversed be sufficient.” It was found that, out of a total length of 52°26 millimetres, the sums of the intercepts of the different minerals were as follows :— : QWaIrEZ Lip, A ace aind ee eee re 20:09 Orthioclase:! oe ott sen eee 13°37 | ° MNCrOCHHe: 0.4 oneness 9°35 } BUDIGG ne iei8 cceee ee eee eee mae 6°22 BiOtite .. un Reese eee eee 3°14 Apatite (estimated) ............... 0-04 Fluor (estimated) 2.225.220: 0:05 52°26 These figures represent, therefore, volumetric proportions of the minerals. Classing microcline with orthoclase, and allowing for anorthite in the albite, the minerals are reduced to those shown in the first column of the next table. The figures in the second column represent the volumetric proportions in which they are present. Multiplying these by the densities (taking that of ortho- clase as unity) given in the third column, we obtain the amounts in the fourth column, which represent the proportions by weight of the same minerals. In the last column these are recalculated as percentages. Volumetric Density Gravimetric Percentage proportions. (orthoclase=1). proportions. Min. Compos. Orthoclase ... 22°72 1:00 22°72 42°19 Allbite” Reaeeeert 578 1:03 5°95 11:05 Anorthite...... 0°44 1-08 0-48 0:89 Quartzeeece +, y20'09 1:04 20°89 38°79 Biotite ......... 3°14 118 371 6°89 UNIT yee seeacneane 0-04 1°24 0-05 0:09 Apatite en. eae 0:05 1:23 0:06 O11 Motals 00.5. 52°26 53°86 100-01 —————— ee The chemical composition of the rock may now be calculated from that of the minerals. The composition of the biotite is assumed to be that of the mineral from El Capitan, Yosemite Valley, the analysis 1 Verhandl. d. Geol. Reichsanst. 1898, pp. 148 e¢ segg.; ‘The Quantitative Classification of Igneous Rocks’ 1903, p. 204; J. P. Iddings, Journal of Geology, vol. xii (1904) p. 225; and Ira A. Williams, ‘American Geologist’ vol. xxxv (1905) p. 34. 2 The line must, I need scarcely say, not be so traced as to pass through one class of constituents rather than another. 3 These contain, in some cases, microperthitic inclusions of albite. — — Ss = Vol. 62.] CATARACTS OF THE RIVER MADEIRA, ETC. 117 of which is given in column (a) of Table XIV of the ‘ Quantitative Classification of Igneous Rocks’ Chicago, 1903. Mineral Percentage | Com Sate eer eeeeee seeterene Caen ewes cee eeeee sen eereee Bi O(c ombined) Less O placed re- || by F J | | | ! | Apat- | 5 | | | | | | | | Ortho- Albite, Anorth / | Quartz. Bio. | Fluor. clase. 4 tite ate. #219 11-05 080 | 3879 689 009 O11 | a7 32) 759) 038 | 3879 2-48. | Nl (ars ess ar a an ey | here i 0-39 Be zal Pe 102 ee ‘ | 0-06, peas e PHO-GBNEy i) Wsceea cil hives: | 0-18 — «004, 0:06! 0:06 | fee) ae i 0-64 | ieee bee he BT 0:03 ee | i | 0-08. ey | | TNs | 0-02 0-04 | trace Gee eeeag eta ae Hae ae —0-02 | 42-19 11:05, 0-89 | 38-79 | 6-91 0:08; O-11]] | | | | Hl i 100-02 76°56 11-52 0°39 1:02 0:06 0°68 0°34 (teh 1:54 0-08 0-05 0:06 O13 — 0:03 100-02 H} | Percentage | Chemical _ Composit. An analysis was subsequently made of this rock by Mr. G. S. Blake, with the following result :— Percentage composition. (SiL0 EO Bence ny eee 73°96 iLO Be Cesena i orien trace os UO AE eo A a eee 15°10 Le eo aa8 ee ee ape 0-74 COE eee eer ase 1:28 1 A, 0 POSER eee 0°04 gl Os SIS ee eee 0°18 EG Deel i ai ea 0-06 SAC Biren 4c aan A See Ame 0°13 SON easis eeiee asada 0-70 Og essed tn aise shane 5-05 1 El Opie se ee ae ae 3°55 1h OSE a Nene eer 0°102 Bae. peti Paktindlc shaman paki 0-041 HE eS pei Vhs ca hh 0-084 H Oabove 100°. O........+.- 0930 Leis Guy ould ie) CUR eaeeg an nie 0°230 Less O replaced by F...... —0:036 Mista). Sicaacast ees Molecular proportions. 12351 0:1284 00046 00178 0:0006 0-0045 00004 00013 0:0125 0°0537 0:0573 00007 0:0015 00044 00517 \ These figures correspond fairly well with the calculated com- The total molecular position, except in the case of the alkalies. 118 DR. J. W. EVANS ON THE ROCKS OF THE [ Feb. 1906, amount of alkalies present is nearly the same in both cases, there being ‘111 molecule inthe actual analysis and 104 in that caleu- lated from the microscopic measurements; but, whereas there are "054 molecule of potash and ‘057 of soda in the former, the calculated amounts were ‘087 and ‘014. It is evident that the orthoclase and microcline contained a considerable amount of soda —either in microperthitic inclusions, or simply replacing potash in the orthoclase and microcline. ‘The albite also appears to approach more closely to oligoclase than was supposed. This is the typical composition of a granite fairly rich in soda and lime. Granitic rocks with a closely-similar composition have been described as granitite. It is included in the subrang liparose of the American classification. Only some 29 yards or so on the up-stream side of the islands where the rock last described was exposed, is a small grey rock (M 15, Pl. V, figs. 5 & 6; specific gravity =2°63 to 2°69), visible only when the river is low. In hand-specimens this presents a saccharoidal appearance like a crystalline limestone. In colour it shows various shades of dull green arranged in lamine, so that it has a distinctly-streaky appearance. Under the microscope, it is seen to have in most places the characteristic structure of a granulite—consisting of an aggregate of small, more or less rounded grains of quartz and microcline with a considerable amount of felspar, which only occasionally shows twin-lamellation, but is probably in most cases albite. Some of the grains of quartz appear to be idiomorphic. The average diameter of the minerals is about 50 microns (0°05 millimetre). A few larger crystals of albite are met with, containing numerous needle-like inclusions, which have a birefringence of more than 11 thousandths. ‘The faster vibrations of these inclusions are at right-angles to the length, and they are perhaps a hydrous mica. There are numerous crystals of a pale-green granular augite, usually comparable in size and shape with the quartz and felspar described above, although a few may measure half a millimetre in diameter. They show little pleochroism. Some of the individuals, especiaily those sections which are near the clinopinakoid, are not quite dark in the position of extinction. This is presumably due to the dispersion of the bisectrices for the different colours. The birefringence reaches 20 thousandths. There are also occasional crystals of sphene, which are strongly pleochroic, changing from a pale grey-green (something like that of the augite) to a dull orange. The birefringence exceeds 85 thousandths. Hornblende occurs in small lath-shaped crystals or elongated plates, with an extinction-angle not exceeding 19°. Itis strongly pleochroic, changing between a pale green, identical with that of the augite, and a deep blue-green. In some places there are extensive aggregates of felspar, of much Vol. 62. CATARACTS OF THE RIVER MADEIRA, ETC. 119 ] larger size, measuring occasionally as much as a millimetre in diameter. They are separated by irregular jagged boundaries. At certain points they show the cross-lamellation characteristic of microcline, but in an ill-defined form; elsewhere the extinction often occurs in patches and streaks. The index of refraction is lower than that of quartz or albite, and there seems to be little doubt that these felspar-aggregates are to be referred to transitional forms between orthoclase and microcline. In these felspars, the smaller rounded constituents of the rock are embedded as in a glassy matrix, and at the same time recall the enclosures of blebs and rounded crystals of quartz (and occasionally microcline) in the felspars of the rocks of other cataracts. An analysis by Mr. G. 8. Blake showed the rock to have the following composition :— Percentage Molecular composition. proportion, IOs Cateewn cont andenas, pon ee Ley 11568 uN) Soe Be ohn 2 ee a 0-20 0:0025 U0 Te ace 12°83 0-1258 ERO RR eee he oa ceeb atlas tee 1-06 0-0066 SN ere eae of Soule 2 2°48 0:0344 WG a en nee ee 0:12 0:0017 J BEG Nee, 2 Spa aera 0-71 0:0177 COG). Ne a el ree i aes 4-73 0:0845 Oa ok Sets ie Soar 5°25 (0):0559 MasOWe ais ancasee uae ase 313 0:0505 HO above: 100° GC, -...4.. 0°18 0:0100 EI Orat OOO MO se. | ee cs O-11 40) 5) eas ene a 100-21 It belongs to the subrang adamellose of the American classifi- cation, and corresponds in chemical composition to the alkali- granites of ordinary nomenclature. (5) The Granulites and their Genesis. It will be seen that, in one way or another, the great majority of the rocks of the cataracts are characterized by the presence of the small rounded crystals or blebs which are typical of the granulites. But the extent to which this structure is developed varies consider- ably from point to point. In some cases it prevails to the exclusion of any other, while in others it is only characteristic of a portion, sometimes but a small portion of the rock, the rest of which is formed by larger and unrounded crystals such as are found in a normally-developed holocrystalline rock. These appear to have been formed at a subsequent date, when the crystallization of the semi-consolidated rock continued and was completed under changed 120 DR. J. W. EVANS ON THE ROCKS OF THE [Feb. 1906, conditions. Occasionally these later crystals appear to have pushed aside the granulitic element, as is sometimes the casein the rocks of - Morrinhos and Santo Antonio, but more usually they have, in the course of their formation, enveloped the pre-existent granules which are now found embedded in them. In some cases, the granulitic phase appears to have been of comparatively-short duration ; in others, the greater part, if not the whole, of the rock must have separated out before the conditions allowed of the develop- ment of the granitic structure. Sometimes the granulitic forms pass, by intermediate gradations, into those of the normal idiomorphic or hypidiomorphic granitic type; sometimes there is an abrupt change from one to the other. In ordinary granitic rocks the quartz is the last mineral to crystallize, and has no crystalline boundaries of its own, being entirely allotriomorphic ; but, throughout the granulitic element in the rocks which I am describing, a large proportion of the quartz shows idiomorphic boundaries, rounded it is true, but still recog- nizable. This is most clearly seen where the quartz-granules occur as inclusions in felspar,; but it can be recognized even where the whole rock is made up of granulitic crystals. Idiomorphic quartz is also found in many volcanic and dyke- rocks where the quartz appears to have crystallized out at least as eatly as the felspars. The most probable explanation of the difference in the order of crystallization in these and deep-seated eranulitic rocks, is that it depends on the pressure to which the magma is subjected... We may, therefore, reasonably suppose that the granulitic rocks of the cataracts crystallized at first under com- paratively low pressure, and were subjected to long-continued earth- movements, which ground together and rounded the crystals during the course of their formation, or in the period immediately succeeding it; although the superior hardness of the quartz enabled it to retain, in some cases at least, traces of its crystal-outline. The finer débris resulting from the process of rounding must have been reabsorbed by the magma. Indeed, in some cases, as in the rock of the Esperanza Cataract, we have evidence of the quartz-crystals having been attacked and partly dissolved. This resorption may be ex- plained by the increase of temperature which might be expected under the conditions that prevailed. For the prolonged earth- movements would themselves generate a considerable amount of heat, and the foldings in which they resulted would cause the accumulation of mountain-masses and a consequent rise in the isotherms. The cooling and crystallization of the still imperfectly- consolidated rock would then recommence, and proceed more slowly and under greater pressure than before, as well as under less- 1 J. A. Cunningham, ‘A Contribution to the Theory of the Order of Crystallization of Minerals in Igneous Rocks’ Sci. Proc. Roy. Dublin Soe. vol. ix (1y00-1902) p. 383. The presence of water may also have considerable influence, as has been pointed out by Prof. W. J. Sollas, Geol. Mag. 1900, pp- 295 et segq. Vol. 62.] CATARACTS OF THE RIVER MADHBIRA, ETC. 121 disturbed conditions, so that a normal granitic structure would be developed in the minerals of the second stage.* ! The haplite-dykes belong, no doubt, to the second stage of crystal- lization, although they are somewhat posterior in time; but most of the other dyke-rocks belong, I believe, to a much later period. (6) The Black Coating of the Rocks of the Cataracts. On the rocks of the island below Santo Antonio the black deposit referred to on p. 104 was well seen. A rough analysis, by Mr. G. 8. Blake, of a little of this material scraped from the surface of specimen M 14 gave the following result :— ‘ POONA cae tdi aituannaecseatneaures vetoes aes 6°6 eae tae (probably ALO.) 0... 9-3 AVIS a Oe ot ea Nise hens sista src a ecivt aise Fa 54°1 eee Bes Dens Seah emai ie an sities de pibanidclclee's ts PVE COM ere encore sens ce saiet conor state acdeecs 0:0 Undetermined components, in all probability mainly water AME ON Ma INE WALD OM ctewieteny «Soe hg goiisels weed dain dod es ere xens y Potal ..,. 100:0 The state of oxidation of the manganese was not determined, but the material resembles psilomelane (hydrated manganese-dioxide), and it is in all probability a variety of that mineral. The presence of silica and alumina is no doubt due to part of the surface of the rock having been removed with the scrapings. Similar black deposits have been described as occurring on the rocks of the Orinoco, the Upper Paraguay, the Nile, the Niger, and the Congo, especially in the neighbourhood of cataracts.” Apparently manganese and iron-oxides are more soluble in river- water in the tropics than in colder countries. It is not clear that higher temperature of the water is alone sufficient. Probably there are organic compounds dissolved in the water, and the iron and manganese either occur as salts of organic acids or as carbonates, 1 Similar inclusions in felspar of quartz with rounded idiomorphic boundaries occur in the granulites of Ceylon; see Dr. Ernst Weinschenk, ‘Zur Kenntniss der Graphitlagerstatten ’ Abhandl. der Math.-phys. Classe der k. Bayer. Akad. der Wissensch. vol. xxi (1902) p. 298 & pl. vi, figs. 1-2 ; also ‘ Grundziige der Gesteinslehre’ pt. ii (1905) fig. 20, p. 51. For inclusions of quartz in horn- blende, as in M 1a, p. 97, see pl. v, fig. 4 of the former paper, and fig. 19, p. 50 of the latter. * See a paper by MM. Lortet & Hugouneng, Comptes Rendus Acad. Sci. Paris, vol. exxxiv (1902) p. 1091 (reviewed in the ‘Geographical Journal’ vol. xx, 1902, p. 655); ‘The Geology of Matto Grosso’ Quart. Journ. Geol. Soc. vol. 1 (1894) p. 98; also Dr. W. Keert, ‘ Geologisch-agronomische Untersuchung der Umgegend yon Amani in Ostusambara,’ Berichte iiber Land- & Forst- wirthschaft in Deutsch-Ostafrika’ vol. ii (1904) pp. 152 & 163. Qo dGiis:. No. 245, K 122 DR. J. W. EVANS ON THE ROCKS OF THE [ Feb. 1906, In the latter case the réle of the organic material would be confined to preventing the further oxidation of the manganese and iron, and consequent separation of carbonic acid. Ultimately, on the evaporation of the water on the surface of the rocks, this oxidation would take place, and the black crust would be formed. Such evaporation would obviously be of most frequent occurrence near cataracts in hot countries, where the rocks are repeatedly moistened by the waves and spray, and dried by the tropical sun. It is remarkable that in none of the analyses of the water of the Lower Amazon and its tributaries, given by Dr. Katzer, is there any mention of the presence of manganese.’ Some interesting man- ganese-deposits on low ground are described (pp. 95-983), and may have an origin similar to that of the black coating on the rocks of the cataracts. Since the above was written a valuable contribution to the subject has been made by Mr. A. Lucas, Chief Chemist, Survey-Department Laboratory, Cairo,” who has examined a large number of similar deposits on the rocks of the Nile Cataracts, as well as other feebler films on desert-rovks. He inclines to the belief that such deposits are an efflorescence from the rocks immediately beneath. This is an opinion to which Iam unable to subscribe. Not only have I found large crystals of quartz and felspar coated equally with biotites or hornblendes, but I have seen in the Upper Paraguay crags of white sandstone covered externally by a jet-black coating which caused them to resemble basalt. (7) Below the Cataracts. The rocks of the island below Santo Antonio are the last crystalline rocks on the Rio Madeira. From this point in lat. 9°S. no rocks are exposed in the river-bed or banks, except argillaceous deposits usually more or less ferruginous, but sometimes nearly pure white. These may form banks 40 or 50 feet high at low water, but I saw nothing to lead me to suppose that they are the result of the alteration in situ of hard rocks. The Neogene deposits of Dr. Katzer may, however, be represented. Three or four degrees farther to the eastward on the Tapajés, diorite and other crystalline rocks are said to occur.® Their presence on the Xingu, still farther to the east, has already been mentioned (p. 88). Farther north Paleozoic rocks are met with, extending westward to the margin of the basin of the Madeira between 5° and 6° lat. 8.,* following the line of elevation on the * ‘Grundziige der Geologie des Unteren Amazonasgebietes’ 1903, p. 48. 2 «The Blackened Rocks of the Nile Cataracts & of the Egyptian Deserts ’ Cairo, 1905. 3 F. Katzer, op. supra cit. pp. 234 & 236. 4 Ibid. p. 170. Vol. 62. ] CATARACTS OF THE RIVER MADEIRA, ETC, 193 south of the Amazonian syncline. No doubt these rocks would be found below the alluvium of the Lower Madeira. IV. Summary. The crystalline rocks of the cataracts of the River Madeira and the lower waters of its tributaries are part of a ridge, with a north- westerly and south-easterly strike similar to that of the Andes in the same latitudes. This strike, which is especially prevalent in Equatorial regions, is probably due to the same causes as those that have resulted in the eastern position of the land-masses of the Southern Hemisphere as compared with the Northern. With the exception of comparatively-recent alluvial deposits, and a few pebbles of chert of marine origin but uncertain date, only erystalline rocks are met with. They all appear to be igneous, mostly massive in character, though some dyke-rocks occur. In places they are typical gneisses, and they are often banded; but in some cases they show no signs of foliation. The prevailing type is acidic, with a considerable proportion of alkalies, especially soda ; but some of the dyke-rocks are distinctly basic in character. The more acid rocks are usually fine in grain, and are often granulitic in structure. In most cases the quartz seems to have crystallized out before the felspar. ‘he causes of these characters are briefly discussed. The occurrence of andalusite of a chiastolite-type as an inclusion in a felspar is noted, as well as an unusual type of allanite. An altered basalt is described, containing minute concentric structures allied to those of a pyromeride. Above and below the region of the cataracts is a wide expanse of alluvial country, either of recent or later Tertiary date. In conclusion I wish to express my obligations to Prof. Bonney, ‘Se.D., F.R.S., who kindly examined many of the rock-sections ; also to Dr. G. T. Prior, of the British Museum (Natural History), and others, for criticism and suggestions during the preparation of this paper. J must also acknowledge my indebtedness to Mr. Thomas Crook, F.G.8., for valuable assistance in photographing the micro- scope-sections shown in Plate V. EXPLANATION OF PLATE V. Fig. 1. Section of granulitic gneiss from Lages, Rio Madeira, above the con- fluence with the Beni; magnified 21 diameters. Ordinary light; showing acid plagioclase, quartz in blebs or rounded idiomorphic crystals, biotite, and micrographic structure. (See p. 98.) 2. The same in polarized light, between crossed nicols. The plagioclase is seen to contain inclusions of orthoclase and quartz. 124 THE ROCKS OF THE MADEIRA CATARACTS, ETC. [Feb. 1906. Fig. 3. Section of granite-porphyry from the Salto do Girao, Rio Madeira ; magnified 17 diameters. Ordinary light; showing corroded quartz, decomposed albite, and groundmass. (See p. 105.) 4, Section of basalt from Theotonio, Rio Madeira; magnified 51 diameters. Polarized light without analyser; showing three spherical spaces filled with chlorite and a lime-magnesia-iron carbonate, a portion of a decomposed labradorite-crystal, and groundmass. (See . 109.) 5. es of pyroxene-granulite from the island below Santo Antonio, magnified 24 diameters. Showing pyroxene, sphene, and hornblende; the clear material is felspar and quartz. (See p. 118.) 6. The same in polarized light, between crossed nicols. Showing, at some points, only granules of the different constituents already mentioned, at others shadowy felspars in which the granulitic constituents are embedded as in a groundmass. Quart. JOURN. GEOL. Soc. VoL. LXII, PL. V. Millimetres. td Ss S > aA Millimetre 1 Millim. 4 ihe) Millims o es 2 Hundreds* N Millimetres &C. ROCKS FROM THE RIO MADEIRA, Bentrose, Collo. J.W.E,. & T.C.. Photomicro. ADMISSION AND PRIVILEGES OF FELLOWS OF THE GEOLOGICAL SOCIETY OF LONDON. Every Candidate for admission as a Fellow must be proposed by three or more Fellows, who must sign a Certificate in his favour. The Proposer whose name stands first upon the Certificate must have a personal knowledge of the Candidate. Fellows on election pay an Admission-Fee of Six Guineas. The Annual Contribu- tion paid by Fellows is Two Guineas, due on the Ist of January in every year, and payable in advance; but Fellows elected after the month of February are subject only to a proportionate part of the Contribution for the year in which they are elected, and Fellows elected in November or December pay no Contribution for the current _ year. The Annual Contribution may, at any time, be compounded for by a payment of _ Thirty-Five Pounds. The Fellows are entitled to receive gratuitously all the volumes or parts of volumes of the Quarterly Journal of the Society that may be published after their election, so long as their Annual Contributions are paid; and they may purchase any of the publications of the Society at a reduction of 25 per cent. under the selling-prices. The Library is open daily to the Fellows between the hours of 10 and 5 (except during the fortnight commencing on the first Monday in September; see also next page), and on Meeting-Days until 8 p.m. Under certain restrictions, Fellows are allowed te borrow books from the Library. Publications to be had of the Geological Society, Burlington House. Reduced Price Reduced Price TRANSACTIONS. tothe Bellows: TRANSACTIONS. to the Bellows: Bune a re ViOla th). ccd sietssccnsivesematasssesees 110 0 Ser Pee Volvgy Via baty We ccsetoene folks 010 0 Ben rilre vols. Va Part 2! sce-secccccdesseea 110 0 oe EPIL Dae Reco 010 0 Malm eeiaebarti2e c.tctnscie-cossee 0) LOO » Bambi soso. ee LelOVO Py: Part 32... Erdods sweets 010 0 Vol Vil bart lois se ere 110 0 = Part 3 (Supplement) 010 0 a AUPE At sss san sieakioas c5 6s LAO 0 “ca 1S Ue er La eee 010 0 Molva prrarti el nae 22 ene se 010 0 9 PANU AU os veto oinen elapse e's 010 0 as MALU Ot coseuocsuses sees. 010 0 QUARTERLY JOURNAL. (Vols. III to LXI, inclusive.) Price to Fellows, 13s. 6d. each (Vols. XV, XXIII, XXX, arid XXXIV to LXI, 16s. 6d.), in cloth. CLASSIFIED INDEX TO THE TRANSACTIONS, JOURNAL, &e., by G. W. Ormerop. New Edition, to the end of 1868, with First, Second, and Third Supplements to the end of 1889. Price 8s. 6d. To Fellows, 5s. 6d. [Postage 5d.|—The First, Second, and Third Supplements may be purchased separately. GENERAL INDEX TO THE FIRST FIFTY VOLUMES OF THE QUARTERLY JOURNAL (1845-1894). Part I (A-La). Part II (La-Z). Price 5s. each. To Fellows 3s. 9d. each. [Postage 3d.]| CATALOGUE OF THE LIBRARY, 1880. (620 pages 8vo.) Price 8s, _ To Fellows 5s. [Postage 6d.] GEOLOGICAL LITERATURE added to the Geological Society's Library during the years ended December 31st, 1894-1904. Price 2s. each. To Fellows ls. 6d. each. [Postage 23d. | ' LIST OF THE TYPE- AND FIGURED SPECIMENS RECOG- NIZED BY C. D. SHERBORN IN THE COLLECTION OF THE _ GEOLOGICAL SOCIETY, verified and arranged, with additions, by the | Rey. J.F. Braxe. 1902. Price 3s. 6d. (bound in cloth); 3s. Od. (in paper wrappers). - To Fellows, 2s. 6d. and 2s. Od. [Postage 3d.] HUTTON’S ‘THEORY OF THE EARTH,’ Vol. III., edited by Sir ArcuiBpaLp Gurkin, D.C.L., F.R.S. Price 3s.6d. To Fellows 2s. [Postage 4d.] | THE GEOLOGY OF NEW ZEALAND. Translated by Dr. O. F. Fiscumr, _ from the works of MM. Hocusrerrer & Perermann. With an Atlas of Six Maps. ____ Fellows may purchase One Copy of this book at 2s.; Additional Copies will be _ charged 4s. [Postage 5d.] CONTENTS, Pages Proceedings of the Geological Society, Session 1905-1906 .............cscscseceseees i-vil PAPERS READ. Page 1. Dr. A. Smith Woodward on Myriacanthus paradoxus. (Plate 1) 2. Dr. C. Davison on the Doncaster Earthquake of April 23rd, 1905. (Plate ID). 5 3. Mr. T. F. Jamieson on the Glacial Period in Aberdeenshire and the Southern Border, of the Moray Firth. ..:.. 00. ..4.8s sis evor es bskidess esses eee 18 4, Mr. A. Harker on the Geological Structure of the Sgirr of Higg. (Plates III ~ Qe LV ig fet onrae Whtcenaee Satna ae cocks dean's cmc csed fe cue thechtias Meant tiae sae 40 5. Prof. E. H. L. Schwarz on the Coast-Ledges in the South-West of the Cape OOLOMY:. 225. ses tesnscdeeg dose seove oslo. oy sebechehls dans camees tte paNgesad ss. aan 70 6. Dr. J. W. Evans on the Rocks of the Cataracts of the River Madeira, etc. CPUabegV pees ned cis ueea se teis Sein teens cites ydesemaceds atcee Momecaote etka ean se pene 88 [No. 246 of the Quarterly Journal will be published next May. | [The Editor of the Quarterly Journal is directed to make it known to the Public that the Authors alone are responsible for the facts and opinions contained in their respective Papers. | *,.* The Council request that all communications intended for publication by the Society shall be clearly and legibly written on one side of the paper only, with proper references, and in all respects in fit condition for being at once placed in the Printer’s hands. Unless this is done, it will be in the discretion of the Officers to return the communication to the Author for revision. The Library and Museum at the Apartments of the Society are open every Weekday from Ten o’clock until Five, except during the fortnight commencing on the first Monday in September, when the Library is closed for the purpose of cleaning; the Library is also closed on Saturdays at One p.m. during the months of August and September. It is open until Eight p.. on the Days of Meeting for the loan of books, and from Hight p.m. until the close of each Meeting for © conversational purposes only. é Uy Prem “Goon QL LLL DLOLODOLOFPODIOLOLI_0 O02 OIE OO Eee Vol. LXII. MAY 28rd, 1906. No. 246. Part 2. Caf Wht ; THE QUARTERLY JOURNAL OF THE GEOLOGICAL SOCIETY. EDITED BY it THE ASSISTANT-SECRETARY. , [With Thirty Plates, illustrating Papers by Mr. Jukes-Browne, Prof. Garwood, Miss Elles & Miss Slater, Messrs. Cantrill & Thomas, Mr. Rastall, Messrs. Matley & Vaughan, and Mr. Sibly. | LONDON : LONGMANS, GREEN, AND CO, PARIS:—CHARLES KLINCKSIECK, 11 RUE DE LILLE. SOLD ALSO AT THE APARTMENTS OF THE SOCIETY. Price Five Shillings. —™ a LIST OF THE OFFICERS AND COUNCIL OF THE GEOLOGICAL SOCIETY OF LONDON. nna AnO LLL Elected February 16th, 1906. aaa SS Oe Presivent. Sir Archibald Geikie, Sc.D., D.C.L., LL.D., Sec.B.S. Wice-PrestVents. Robert Stansfield Herries, M.A. John Edward Marr, Sc.D., F.R.S. Aubrey Strahan, M.A., F.R.S. J.J. H. Teall, M.A. D.Se., E.R.S. Secretaries. Prof. William Whitehead Watts, M.A., M.Sc., F.R.S. PForeiqn Secretary. Sir John Evans, K.C.B., D.C.L., LL.D., FE.R.S., F.L.S. Prof. Edmund Johnstone Garwood, M.A. Treasurer. Horace Woollaston Monckton, Treas.L.S. COUNGIL. H. H. Arnold-Bemrose, J.P., M.A. Prof. Thomas George Bonney, Sc.D., LL.D., Has. S.A. Sir John Evans, K.C.B., D.O.L., LL.D., FE.R.S. Prof. Edmund Johnstone Garwood, M.A. Sir Archibald Geikie, Sc.D., D.C.L., LL.D., Sec.R.S. Robert Stansfield Herries, M.A. Finlay Lorimer Kitchin, M.A., Ph.D. Philip Lake, M.A. George William Lamplugh, F.R.S. Richard Lydekker, B.A., F B.S. Bedford McNeill, Assoc. R. S.M. John Edward Marr, Sc.D., F.R.S. Horace Woollaston Monckton, Treas.L.8. Frederick William Rudler, I.8.0. Leonard James Spencer, M.A. Aubrey Strahan, M.A., F.R.S. Charles Fox Strangways. J.J. Harris Teall, "M.A. D.Sce., F.R.S. Richard Hill Middement M. A. Prof. William Whitehead Watts, M.A., M.Sc., F.R.S. The Rev. Henry Hoyte Winwood, M.A. Arthur Smith Woodward, LL.D., F.R.S. Horace Bolingbroke Woodward, F.R.S. Assistant-HSecretarp, Clerk, Librarian, anv Curator. L. L. Belinfante, M.Sc. Assistants in @Pflice, Library, and Museum. W. Rupert Jones. Clyde H. Black. Alec Field. STANDING PUBLICATION COMMITTEE. Sir Archibald Geikie, President. Prof. W. W. Watts. Prof. E. J. Garwood, Dr. F. A. Bather. Prof. T. G. Bonney. Mr. R. S. Herries. Prof. J. W. Judd. Mr. R. Lydekker. Dr. J. E. Marr. \ Secretaries. Prof. H. A. Miers. Mr. H. W. Monckton. Mr. H Mr. A Mr. L. Mr. H EVENING MEETINGS OF THE GEOLOGICAL SOCIETY TO BE HELD AT BURLINGTON HOUSE. Sessron 1905-1906. Wednesday, June Perce steeceeceeetesene 13*—27* [Business will commence at Hight o’ Clock precisely each Evening. | The dates marked with an asterisk are those on which the Council will meet. Vol. 62.] FOOTPRINTS FROM THE PERMIAN OF MANSFIELD. 125 7. On Foorrerints from the Permian of Mansrietp (Norrinenam- SHIRE). By Grorez Hicxiine, B.Sc. (Communicated by Prof. W. Boyp Dawkins, M.A., D.Sc., F.R.S., F.S.A., F.G.S. Read January 10th, 1906.) Ir is the object of this brief communication to call attention to a series of footprints discovered so long ago as October 1897 by Mr. Francis Holmes, of Leicester, the description of which has been delayed by the lamented death of the geologist to whom we owe their preservation, the late Mr. James Shipman, F.G.S., of Nottingham. I. Norss on THE STRATA. These impressions were obtained from the Rock-Valley Quarry, Mansfield, in the Permian rocks north-east of the town, some 500 yards from the Permo-Triassic boundary, where they are overlapped by the Lower Mottled Sandstone. The rock which is here quarried is a lenticular mass of sandstone intercalated in the Magnesian Limestone, described by Aveline in the Geological Survey-Memoir on Sheet 82 S.H. 2nd ed. (1879) pp. 10-12, as a locally-sandy type . of the limestone, passing laterally into the normal type. The stone is here yellowish-red in colour, becoming almost white on the opposite side of the town. It is by no means pure sandstone throughout, but contains irregular bands of ‘ bastard,’ a term used by the quarrymen to indicate a calcareous sandstone. Every gradation may be found between a pure sandstone and the almost pure limestone. The general succession in this quarry appears to be as follows, in descending order, though all the beds vary greatly in thickness :— Feet NPRM SUG MG a ciemacius desiieick bac tias Facto sone teak 15 2. Laminated ‘ bastard’ } 10 3, Coarse sandy ‘bastard’ { ‘rete 4. Sandstone, with bands of ‘bastard’ ......... 20 fee Wan eR COM Oa" at ste e facet dosieidk eee es doGsiestuée i Current-bedding is developed on a large scale, some of the beds of sandstone having an apparent dip of as much as 25°, although the true dip would appear to be quite trifling. Examples may be seen, too, of contemporaneous erosion. A detailed examination of the Permian round Mansfield led Mr. Shipman to conclude that these intercalated sandstones probably were formed as sandbanks off the mouth of some river flowing from the then newly-formed Pennines, and that the actual shore-line lay some 3 or 4 miles to the west of the town. , To these conclusions he was led mainly by the generally-lenticular form of the deposits, the varying coarseness of the sediments, and the nature of the current-bedding. Q.J.G.S. No. 246. L 126 MR. G, HICKLING ON FOOTPRINTS [May 1906, II. Mope or OccuRRENCE. These impressions were found on slabs from the middle of Bed No. 4 (in the foregoing section, p. 125). When first seen by Mr. Shipman, on October 21st, 1897, they were still in situ; and we learn from his notes, made on that occasion, that they formed two double rows, approximately parallel, which crossed the slab in a direction very nearly from west to east. The longer row could be traced for about 7 feet. The shorter series was by the side of the longer one at a distance of a yard, and was only to be followed satisfactorily for a little more than 2 feet.1 Of these impressions, nearly the whole of the longer series is preserved on a large slab in the Free Public Museum, University College, Not- tingham, and the prints of the right side of the shorter series are in the Manchester Museum, Owens College. ‘The left side of the latter series, and a few imperfect impressions from the former, have not been preserved. The overlying slab, with the natural casts, was used for commercial purposes before the prints were noticed. Ill. Description oF THE IMPRESSIONS. A comparison of the slabs preserved at Nottingham and Man- chester respectively at once reveals the fact that both sets of im- pressions were made by the same species of animal, although a very slight inferiority in size of the Manchester prints (which may be estimated by the stride measuring 8 inches, as against 82 inches in the Nottingham examples) would seem to point to their being due to different individuals. Practically, however, the principal measurements will apply to both series. The chief of these are as follows :— Inches. Pes: 100) 2112 Ser eRe OS sco ansed das soc: — 32 Width (tip of digit I to tip of digit V) ......... 23 ICY aL Sean De 0) ear ee eS Se Ak so 3F WRC Oe oh a eR ECs oe Nkok at 5 2% Average width between middle of right and left pes ...... z Do. do. do. do. do. do. manus... 74 Length of stride? (Nottingham slab)..............00.ccesssees 8F The general character of these impressions will be seen much more clearly from the accompanying figures (pp. 127 & 128) than from any amount of description which might be given. . They are, for the most part, very badly preserved; and only in a few cases could they be definitely recognized as footprints, had they been found separately. The individual prints are characterized by their ‘stumpy’ form, by the well-marked heel, ‘very like the heel of a boot 1 T have since learned that more of both rows was subsequently exposed, but the other portions are not preserved. 2 ‘Stride’ is here taken as the distance between successive prints of the same foot. Vol. 62. ] FROM THE PERMIAN OF MANSFIELD. 17 in form,’ and by the comparatively-slender digits—more slender, I believe, than one would at first con- Fig. 1.—Photograph of foot- clude from a casual glance at the prints on the slab preserved photographs—of which the fifth is in the Free Public Museum, set back and turns outwards and University College, Notting- forwards, As regards the presence ham. or absence of claws, it is impossible = Fas ] to speak with absolute certainty, but the pointed endings of the impres- sions of the digits, in some cases extending forwards into slender scratch - like streaks, make their presence very probable. But of much greater interest is the very smooth convexity of the surface of the stone between the impressions of the digits, seen as a rounded con- cavity in the plaster-casts. There appears to be but one conceivable explanation of this feature, namely, the presence of a membrane between the toes. Unfortunately, the state of preservation is not sufficiently good to enable one to speak with confi- dence; but it must again be regarded as probable that the animal was web- toed. ‘There is no indication of the number of phalanges in the digits. It may be added that the pes was undoubtedly pentadactylate and the manus probably so, although in the latter case I have been unable to discover a really-indubitable impres- sion of the first digit. It seems to be represented by a kind of stump on the casts photographed. Regarding the general arrange- ment of the prints in the track, Se ===} allusion may be made to their close ; eee succession, to the wide separation of {Total length=stightly over 5 feet.] Hie aotpendislotitadessand! to: the fact that the fore feet were more piaely separated than the hind feet. IV. ConcLvusions. It now remains to consider the import of these impressions. So far as I have been able to determine, they do not entirely agree with any that have been previously described, either from the Permian or from the Trias. One’s first idea is that they may, perhaps, be associated with the cheirotheroid prints from the latter formation— L2 Fig. 2.—Plaster casts of the fore and hind foot of the right side from the Manchester slab, and of the fore and hind foot of the left side from the Nottingham slab. Right fore foot. Left fore foot. Right hind foot. Left hind foot. [The photographs are on the scale of 3°, approximately. ] Vol. 62.] | FoOTPRINTS FROM THE PERMIAN OF MANSFIELD. 129 an idea which might seem to receive considerable support from the out-turned fifth digit. On more careful examination, however, the supposititious association has to be abandoned. Cheirotheroid prints have the following characteristics :—Of the five digits, 11, 111, and Iv are stout and strongly marked, the fifth is set back and curves out- wards and backwards, while the first is much less conspicuous and quite commonly leaves no impression. There is never any trace of a heel, except in the aberrant Cheirotherium Herculis, and in some cases impressions of the digits alone are seen. ‘There is no trace of any interdigital membrane. The stride was long, commonly about a yard, and all the impressions are nearly in one line, indicating an animal very well adapted for rapid progression on land. The fore feet were greatly reduced, and nearly the whole weight was supported by the hind limbs. Very different are the characters of the impressions with which we are concerned here. They were made by an animal emphatically plantigrade. The stride was very short, and the gait that of a clumsy animal ill-fitted for walking. The fore feet were only a little smaller than the hind feet, and the weight of the body was much more evenly distributed. The digits were comparatively slender, and were probably joined by a web. Even the fifth digit, though modified, was much less so than in Cheirotherium. In passing, ib may be remarked that this modification of the fifth digit seems to have been a fairly-common feature among the animals of Permian and Triassic times, and some explanation of its meaning might be interesting. The only previously-described impressions that I have been able to find, which do present any close resemblance to these Mansfield prints, are those from the Upper Permian of Thuringia, of which an account has been written byW. Pabst.’ The footprints there described under the name of Jchnium acrodactylum agree in a striking manner with those here described. Insize,in general form, in the arrange- ment of the digits, and in the general arrangement of the track no decided distinction can be found. The individual prints, however, differ markedly in one respect, namely, in the much greater stoutness of the digits in the Thuringian form, so that between the first four digits no marked interspaces appear. The digits are also somewhat longer. To some extent, though not entirely, these differences may be accounted tor by the fact that the Thuringian prints are more deeply impressed than the Mansfield prints. Two other points serve to distinguish definitely these two sets of tracks: the first, that the difference in size between manus and pes in Ichnium acrodactylum is still less marked than in our own prints; the second, that the stride of the former is much longer—approxi- mately 12 inches. The feet (at least, the hind feet) were slightly less than in the Mansfield form. Yet, despite such differences, the general resemblance is so strong that one can scarcely do other than suppose a real relationship between the Thuringian animals and those of our own ancient shores. * Zeitschr. d. Deutsch. Geol. Gesellsch. vol. xlix (1897) pp. 701 e¢ segg. 130 MR. G, HICKLING ON FOOTPRINTS [May 1906, if Mr. Shipman’s conclusions regarding the mode of formation of the rocks which yielded our impressions are sound (see p. 125), we may assume that the animal responsible for them was wandering on a sandbank some distance from land. In view of the further facts that it was probably web-footed, and that it certainly had an awkward gait, it would seem reasonable to assume that it was amphibious in habit. At least no further evidence on this point could be expected from its footprints. The two brief visits which I have as yet been able to pay to the Rock-Valley Quarry have sufficed to show that footprints are really abundant there—mostly in a very imperfect state of preservation, but including some quite good impressions. In the course of less than two hours’ search altogether, I have been able to distinguish several (at least four) distinct types of impression, which undoubtedly represent different types of animals and cannot be attributed by any chance to differences of imprint.' These other forms I hope to be able to describe in a future communication, after further search of the deposits has been possible. Meanwhile, I may note that some of them bear a marked resemblance to those found by the late Mr. G. Varty Smith in the Permian sandstones of Penrith.* The abundance and variety of the footprints in this quarry are of considerable interest, in connection with the recent discovery of a rich vertebrate fauna in the Permian of Russia.* The knowledge of the existence of quite a varied assortment of reptiles at that period—Anomodonts, Theriomorphs, Rhopalodonts, and probably Deinosaurs—widens considerably the range of beasts which we might expect to have inhabited our own shores. Formerly, the only forms to which we could attribute such footprints as those here described were the Labyrinthodont Stegocephalia, which group, indeed, satisfied very well the general characters indicated by the impressions. But, in view of Prof. Amalitzky’s discoveries, we should keep watch for indications of the higher forms. Should footprints be discovered in the bone-bearing deposits of Russia, they might, with some certainty, be referred to the animals which made them. Then the prints in our own rocks might receive their interpretation. Meanwhile, this communication may perhaps serve its most useful purpose by drawing the attention of local geologists to the existence of these sandbanks.in the Magnesian Limestone—the one at Mans- field is unlikely to be an isolated example—and to the fact of their containing traces of perhaps the most interesting fauna with which the vertebrate paleontologist has to deal. 1 [Not less than six distinct' forms have now been found.—March 15th, 1906. 2 oS Quart. Journ. Geol. Soe. vol. xl (1884) p. 479. Still more interesting is the close resemblance, if not identity, of one of these forms with that described by Huxley from the Elgin Sandstones, and figured in Geol. Sury. Monograph iu, pl. xiv (1877). 3 See V. Amalitzky, Comptes Rendus Acad. Sci. Paris, vol. exxxii (1901) p- 591. Vol. 62.] FROM THE PERMIAN OF MANSFIELD. 131 In closing this brief note I wish to express my sincere thanks to Prof. J. W. Carr, F.G.S., University College, Nottingham, for the photograph forming fig. 1 and for two of the casts shown in fig. 2 ; and to Mr. H. C. Beasley, of Liverpool, from whom personally and from whose writings I have received the greatest assistance. Discussion. Prof. Boyp Dawxrns pointed out that, as no footprints had previously been found in the undoubted Permian of this country, the finds with which the Author’s paper dealt were naturally of extreme interest. The footprints were probably made by an amphibian, analogous to the amphibian types recorded from the Permian of the Continent. 132 MR, A. J, JUKES-BROWNE ON [May 1906, 8. The Ctay-witH-Fiints; its Orierin and Distrisurion. By ALFRED JoHN JuKES-Browne, B.A.,F.G.S. (Read January 10th, 1906.) [Pirate VI—Szcrions. | ConTENTS. Page LeDetritionvand Mheoresiot Onieun cece cece ae 132% II. Composition of the Clay-with-Flints .................. 134 TI. Thickness of the Clay-with-Flints ..................0c0e0e 137 IV. wroducts of the Solution of-Challky.)..2.5 1 seseeseee 139 VY. Inferences to be drawn from the Distribution of the Clay-with- Bint. cons: niet eeeecenie a steratismele saeco 143 VWiALoSummary and (Conclusions ss-sne- esses scueseee ee renee 157 I. DerFinirion AND THEORIES OF ORIGIN. Tue peculiar deposit which has been termed ‘ Clay-with-Flints ’ well known to most English geologists, as occurring in sheets or patches of various sizes over a large area in the South of England, from Hertfordshire on the north to Sussex on the south, and from Kent on the east to Devon on the west. It almost always lies on the surface of the Upper Chalk, but in Dorset it passes onto the Middle and Lower Chalk, and in Devon it is found on the Chert- Beds of the Selbornian Group. The existence of a red clay full of flints, lying as a soil or deposit on the surface of the Chalk, was well known to geologists in the middle of last century, such as Trimmer, Lyell, and Prestwich ; and some appear'to have regarded it as a residue derived from the solution and disintegration of the Chalk, but Joshua Trimmer in 1851 maintained that it and the other ‘ soils which cover the Chalk of Kent’ were the result of aqueous transport.’ It was not until 1861 that it was described as a special accumu- lation or aggregation by Mr. Whitaker, under the name of Clay- with-Flints*; and not until 1864 ° that he ventured to suggest an explanation of its formation. His belief was and still is that ‘the Clay-with-Flints is of many ages, and may be forming even at the present day, and that it is owing in great part to the slow decomposition of the Chalk under common atmospheric action.’ He holds that, as many Chalk-districts have been exposed to the action of rain for thousands of years, much of the Chalk has been carried away in solution, leaving the flints and the insoluble earthy and ferruginous matter. He remarks that ‘To these would be added the clayey and loamy wash from the Tertiary lands, and the remains of beds of that age left in pipes and hollows of the Chalk. . 1 Quart. Journ. Geol. Soe. vol. vii (1851) p * «Geology of Parts of Oxon & Berks’ ae ae Sheet 15, Mem. Geol. Surv, p. 54. 3 ‘Geology of Parts of Middlesex, Herts, Xe.’ ees of Sheet 7, Mem. Geol. Surv. p. 64. Vol. 62. ] THE CLAY-WITH-FLINTS. 133 The clay and flints left by the dissolution of the Chalk would be present almost everywhere ; whilst the loamy materials that would be formed from the lowest Tertiary beds would most likely be more local.’ + In 1865 Mr. Thomas Codrington, describing parts of Wiltshire in which this Clay-with-Flints was prevalent, had come to a somewhat different conclusion. He accounted for the presence of the unworn flints and for the peculiar disposition of the deposit in the same way as Mr. Whitaker had done; but he thought that the original presence of an overlying deposit of clay or loam was essential to its formation. He writes :— ‘Everything seems to indicate a quiet subsidence of the overlying bed into irregularities in the dissolving Chalk. Everything here also favours the sup- position that the origin of the Clay-with-Flints is to be ascribed to the gradual dissolving away of the Chalk-with-Flints under a capping of drift brickearth. .... The Clay-with-Flints must underlie the brickearth, when the latter is present, but the whole of it (that is, the brickearth) may be absorbed into the Clay-with-Flints.’ ? Lastly, Charles Darwin, writing in 1881, does not refer to Mr. Whitaker’s explanation, but takes it for granted that the whole of the red clay (as well as the flints) was simply the insoluble residue left by dissolution of the Chalk, without the addition of any extra- neous matter.* It is clear, however, from the remarks which he makes, that he was puzzled to account for the absence of such a thick residue in many places where it might be expected, and also for its thickness in places where analyses proved that the underlying Chalk coutained a very small amount of earthy matter. Of these three explanations, I believe that Mr. Codrington’s, so far as it differs from Mr. Whitaker’s, comes nearest to the truth, for I think that he was right in asserting that the argillaceous part of the Clay-with-Flints has been derived from Tertiary material and not from the Chalk. JI dissent from Mr. Whitaker’s view, because he contends that the bulk of the clay came from the Chalk, and only admits a local admixture of Tertiary material. Lastly, I consider that Darwin was still further from the truth, because he imagined that the whole deposit had been derived from the Chalk. It is only fair, however, to say that Darwin does not appear to have studied the distribution of the Clay-with-Flints over any large area, and was only incidentally concerned with the manner of its formation. What may be termed the ‘ Chalk-residue theory’ has held the field in this country for many years, but during the last decade geologists have been losing faith in it; consequently, it seems only right and fitting that the other side of the case should be presented, and discussed more fully than Mr. Codrington had the opportunity of doing. J propose therefore to state the reasons which have induced me to abandon the faith in which I was educated, and I * “The Geology of London’ Mem. Geol. Surv. vol. i (1889) p. 282. * Wiltsh. Arch. & Nat. Hist. Mag. vol. ix (1865) pp. 180-81. 3 «The Formation of Vegetable Mould, &c.’ 1881, pp. 137-39 & pp. 298-300. 134 MR, A. J. JUKES-BROWNE ON [May 1906, trust that my old friend Mr. Whitaker will pardon me for holding a brief against him. Some of the facts which militate against the Chalk-residue theory have been already advanced by Mr. Clement Reid and by myself in the Annual Reports and Memoirs of the Geological Survey. Some of these will be quoted in the sequel, but data for a more detailed consideration of the whole subject are now available, and conse- quently it seems desirable that the whole case should be stated as clearly as possible in one paper. II. Composition oF THE CLAY-WITH-FLINTS. The material was defined by Mr. Whitaker as ‘a deposit of stiff brown and reddish clay with large unworn flints, that occurs over the higher parts of the Upper Chalk-tract.’ } He also noted that, at its base, there is often a layer of black clay, a few inches thick, which contains black-coated flints. He dis- tinguishes the true ‘ Clay-with-Flints’ from the loam or brickearth by which it is often overlain, admitting that the latter has been mainly derived from the detrition of the Reading Beds, that it often contains large unworn flints, and that it frequently passes so com- pletely into the Clay-with-Flints that it is difficult to draw a line between them.” I will, however, confine myself to the Clay-with-Flints, as defined by Mr. Whitaker, and pass on to the important point of its composition and contents. In its typical development on the west and north-west sides of the London Basin, the material is generally a stiff, unctuous, brown or reddish-brown clay, usually without any visible admixture of sand, but containing unworn flints the outer surfaces of which are generally stained brown. Where sections of such clay are seen below brickearth, the flints do not generally form more than half the bulk of the deposit ; but, where there is a surface-spread of Clay-with- Flints, the upper portion of it contains more flints than clay, and is in fact an angular flint-gravel. This is doubtless a consequence of the washing-away of the fine clay by the rain. Mr. Whitaker remarks (op. cit. p. 282): ‘ Besides the unworn flints, there are also sometimes pebbles of flint and of quartz, as well as, more rarely, pieces of old rocks.’ Again, he mentions the occurrence near Remenham, in Berkshire, of green-coated flints from the base of the Reading Beds, pebbles of quartz and quartz-rock, and a few small lumps of ironstone. Mr. Whitaker, however, does not mention the fact that broken angular flints are very common ingredients. The only passage from which it would appear that he was aware of that fact is that 1 «The Geology of London’ Mem. Geol. Surv. vol. i (1889) p. 281. 2 Ibid. p. 288. Vol. 62.] THE CLAY-WITH-FLINTS. 135 at the bottom of p. 283 (op. cit.), where he speaks of ‘the unworn and often unbroken character of the flints’ as being against the theory of transport. It is true that both kinds of flints are unworn, but there is a great difference between entire flint-nodules (as they occur in the Chalk) and angular pieces of broken-up flints. J have not had a sufficiently-wide experience of Clay-with-Flints in the South of England to be able to assert that angular flint- fragments are everywhere abundant in it. JI can only testify to their occurrence in the counties of Hertford, Bedford, and Buckingham, and again in Dorset. Mr. H. J. Osborne White, however, who has special acquaintance with the deposits overlying the Chalk in the counties of Buckingham, Oxford, and Berkshire, informs me that he considers the angular flints to be generally more numerous than the unbroken flint-nodules, and that he knows of exposures where the latter are rarely found and where all the flints usually to be seen are angular. Lastly, Mr. Clement Reid, writing of the country around Salisbury,’ describes the clay as containing ‘unworn or shattered flints,’ by which I suppose he means unbroken and broken flints. Charles Darwin, when describing the Clay-with-Flints near Down in Kent,* remarks that the flints are often broken thongh not rolled or abraded, and that the elongate flints are commonly found ‘standing nearly or quite upright in the red clay,’ a position which he ascribes to the downward movement of the mass from higher to lower levels. He also incidentally mentions that the flints in the clay ‘form almost half its bulk,’ With regard tc the unbroken flints, which have mostly been derived directly from the Chalk, Mr. Osborne White tells me that, in his experience, they seldom occur in any number, except in the bottom-layer of the clay at its junction with the Chalk, and in pipes or funnels which penetrate into the Chalk. Their greater abundance in such hollows is not surprising, since these are evidently spots where solution has taken place more extensively than else- where. Another point requiring notice is the frequent occurrence of green-coated flints, derived from the base of the Reading Beds. On this matter Mr. White writes : ‘T feel sure that green-coated flints are quite as common in the Clay-with- Flints as one would expect them to be.’ It is moreover possible that the black-coated flints, which are also common where the clay is black, may be such flints coated or stained black by oxide of manganese. Mr. Clement Reid, describing the Clay-with-Fiints in Sussex,’ says that it consists of Eocene material mixed with a certain 1 Expl. of Sheet 298, Mem. Geol. Surv. 1903, p. 64. 2 «The Formation of Vegetable Mould, &c.’ 1881, pp. 1388-39. * ‘Geology of Eastbourne’ 1898, p. 10, and ‘Geology of Chichester’ 1903, p. 38, both memoirs of the Geological Survey. le MR. A. J. JUKES-BROWNE ON [May 1906, proportion of flints, and he states that it always contains a large percentage of rounded quartz-grains, such as could not be derived from the Chalk below. Again, among its contents in the country north of Chichester he mentions the occurrence of flint-pebbles, green-coated angular flints, sarsen-stones, and quartz-sand, these ‘and probably most of the clayey matrix, being of Kocene origin.’ Mr. Reid also makes the explicit statement that ‘in this part of the Downs (near Chichester) it is confined to areas over which Eocene deposits may have spread within comparatively recent times; it is sometimes very difficult to decide whether a particular patch should be called slightly-disturbed Eocene, or mapped as Clay-with-Flints.’ (Loe. supra cit.) With the truth of this statement, as applied to other districts, I shall deal more fully in the sequel. The same observer, describing the deposit as developed near Salisbury,’ writes that ‘ parts of it are mainly composed of Kocene material,’ and ‘other parts are derived in large measure from the Chalk below.’ He adds: ‘Everywhere, however, there is a considerable admixture of material that cannot have been derived from the strata immediately below. We find in it, for instance, Chalk-flints belonging to zones which only outcrop some distance away on higher ground. It also contains pebbles derived from Tertiary deposits, which cannot have rested directly on the Chalk in that neighbourhood. On washing the matrix we obtain a sandy residue consisting of rounded grains of quartz, such as could not have been derived from the Upper Chalk.’ Again, in Dorset the proportion of material that must have been derived from the Eocene is still larger, and Mr. Reid observes : ‘At least half the deposit consists of rolled stones and rounded quartz- grains.... Even of the angular material, a close examination shows that a considerable proportion must be derived from the gravelly base of the Hocene deposits, in which angular flints usually abound,’ ? He asserts, indeed, that ‘over the area now under consideration solution of the Chalk and accumulation of the insoluble matter will not produce anything approaching in composition to the Clay-with-Flints.’ (Loc. cit.) My own observations in Wiltshire and Dorset confirm those of Mr. Reid, especially on the point that, where the deposit is most argillaceous and most essentially a Clay with flints, it is still obviously and largely composed of Hocene material. With regard to the finer portions of the Clay-with-Flints, it will be noticed that Mr. Reid has found that, both in Sussex and in Dorset, it invariably contains a large percentage of rounded quartz-grains. Mr. Osborne White tells me that he has washed many samples of the material from the neighbourhood of Reading and Wargrave, and has always found a residue consisting largely of quartz-sand. He has kindly sent me a sample of the Clay-with-Flints, overlying the Chalk in a quarry at Middle Culham, near Remenham (Berkshire) ; ' Expl. of Sheet 298, Mem. Geol. Surv. 1903, p. 64. 2 ‘Geology of the Country around Dorchester’ Mem. Geol. Surv. 1899, p. 37. Vol. 62. ] THE CLAY-WITH-FLINTs, 137 and a portion of this I have washed and examined under the micro- scope, with the following result :-— The greater part of the residue consists of rounded grains of clear white quartz, some of them stained yellowish by oxide of iron, the grains varying in size from very small to rather large. There are also many black and dark- brown grains, which seem to be rolled particles of ironstone or small concretions of oxide of iron. Both these ingredients have evidently been derived from Reading Beds. A few fragments of Jnoceramus-shell, with a few broken sponge-spicules and arenaceous foraminifera, seem to be contributions from the Chalk. Small angular chips of flint are common in the residue, but not in the finer portion mounted for microscopic examination. A sample from near Risborough, sent by the Rev. E. C. Spicer, F.G.S., yielded a very similar residue, the only noticeable difference being that there was a larger quantity of very fine quartz-sand, large grains being rare. The residue of a sample from Chaul End, near Luton, sent to me by Mr. J. Saunders, also consisted mainly of very fine red sand, which the microscope showed to be grains, presumably of quartz, coated with red clay. In all cases it was difficult to free the mineral grains from the fine sticky clay which adhered to them. III. Taickness oF THE CLAY-wITH-FLINTs. For the purpose of this enquiry, it 1s also necessary to form some idea of the average thickness or mass of the Clay-with-Flints; but this is not easy, because the clay seldom occurs as an evenly-spread deposit. On the contrary, it usually rests on an extremely-irregular surface of chalk, having sunk into hollows, funnels, and pipes which have clearly been formed by the unequal solution of the chalk. The bottoms of the hollows and depressions are often 6 or 8 feet below the summits of the intervening pinnacles of chalk, while in some places large basins occur, lined. by Clay-with-Flints and filled with masses of brickearth and gravel, from 40 to 50 feet in depth; else- where, narrower funnels and pipes filled with the same materials penetrate the Chalk for 30 or 40 feet. But although, in many places, it cannot be said that the Clay- with-Flints has any average thickness, there are limits within which it varies ; and there are some localities where its base is not too un- even to prevent a rough estimate of its average thickness from being made by eye. In other words, one can form some idea of what its average thickness would be if it were spread out on an even base. In the first place, I will quote some recorded observations and estimates of thickness :— ‘At Oakridge, west of Wycombe Abbey [Bucks], there is a thickness of between 4 and 5 feet of stiff reddish Clay-with-Flints, in places sandy and pebbly.’? (H. B. W.) ‘In a new road-cutting west of Hemel Hempstead Clay-with-Flints was seen, with a depth of up to 8 feet, resting irregularly on soft chalk w ith many flints.’ ‘Round St. Albans this irregular deposit varies from a few inches to 4 or 5 feet in thickness.’ ? W. Whitaker, ‘Geology of London’ Mem. Geol. Surv. vol.1(1889) p. 286. i A. G. Cameron, in op. cit. p. 287. 3 H. B. Woodward, ibid. Joc. cit. 1 2 138 ) MR. A. J. JUKES-BROWNE ON [May 1906, In the country round Luton its thickness is said to vary from 2to 8 feet. The following is unpublished information. Near Weston, north-east of Hitchin, I saw a fairly-even spread of dark reddish-brown Clay-with-Flints, from 14 to 3 feet in depth, underlying loamy brickearth. Mr. C. J. Gilbert, F.G.S., informs me that, on the high ground near Berkhampsted, there is much Clay-with-Flints, and that it is often from 10 to 20 feet thick when not associated with brickearth ; but that in many places brickearths, obviously derived from the Reading Beds, lie in large potholes or solution-hollows which are lined by a bed of clay and angular flints ; in such cases, the bed is thin, and is sometimes only represented by a layer of black-coated flints. Iam informed by the Rev. E. C. Spicer, F.G.S., that several of the cuttings on the Great Central Railway between Wendover and Missenden show good sections of Clay-with-Flints, that its depth varies from 1 to 6 or 7 feet, and that the average thickness might be taken at about 3 feet. At the claypits near Walter’s Ash, recently described by Mr. Spicer,’ the Clay-with-Flints forms a lining to the large basins, which are filled with brickearth, gravel, and sarsen-stones, Mr. Spicer informs me that, in other parts of the same district, where the Clay-with-Flints forms a surface-deposit, its thickness varies from 2 to 10 or 12 feet. In order to obtain some more accurate idea of the average thickness at certain spots, he has been kind enough to measure two sections for me. The first is near Loosely Row, south of Risborough, where a cutting shows a wide depression in the Chalk filled with the usual clay-deposit; the Jength of the depression is 25 yards, and its depth below the soil in the centre is 10 feet, from which it lessens gradually on each side to about 2 feet; its average depth here, therefare, is 6 feet in a length of 75 feet. | The other locality is Denner Hill, south of Great Hampden, where a slope is mantled by Clay-with-Flints, and, for a length of 440 yards with a breadth of 200 yards, the depth of the clay is from 10 to 12 feet. This is an exceptional depth to be continuous over so large an area, and especially on sloping ground. I have not been able to obtain any information as to the limits within which the thickness of Clay-with-Flints varies over Hamp- shire, for no observer seems to have paid much attention to it in that county, except to the north of Chichester (Sheet 317), where it has been mapped by Mr. C. Reid. No actual measurements of it are given in the Explanation of that sheet, but Mr. Reid observes that ‘the total thickness seldom reaches 10 feet’ (op. cit. p. 38). From the particulars now given it is clear that any theory of the formation of Clay-with-Flints must be equal to accounting for its accumulation to a thickness of 3 or 4 feet over large areas. * Quart. Journ. Geol. Soc. vol. xi (1905) p. 39. Pols-62..|\5- THE CLAY-WITH-FLINTS. 139 LV. Propucts oF THE SOLUTION oF CHALK. From the known purity of the Upper Chalk, it will readily be granted that, if the Clay-with-Flints has been mainly derived trom the dissolution in situ of Chalk-with-Flints, a very great thickness of this chalk must have been dissolved in order to furnish the material of the clay. It is well known that the soft limestones of the Upper Chalk contain a very small proportion of insoluble matter; but, until recently, very few accurate analyses of these chalks existed, and even now hardly sufficient exist to furnish data for estimating the average amount of clay to be found in each zone of the Upper Chalk. Moreover, the determinations of the amounts of insoluble residue which have been made by Dr. W. F. Hume, F.G.8., and Mr. W. Hill, F.G.S., show that in some zones these amounts vary in different parts of the country. This seems to be more especially the case with the lower zones, namely, those of Micraster cor- testudinartum and M. cor-anguinum. I have, therefore, thought it desirable to depend, as far as possible, on analyses of samples obtained from places within the area of which this paper treats. In order to increase the number of such analyses, my friend Mr. W. Hill has kindly ascertained the amount of insoluble residue in several fresh samples of chalk from the zone of Micraster cor-anguinum, and in one from that of M. cor-testudi- narvum. For two of these samples I am indebted to Mr. H. J. Osborne White, F.G.S.; another, from the zone of MW. cor-anguinum at Winnal near Winchester, was sent by Mr. Charles Griffith, F.G:S. ; and the fourth was obtained by Mr. W. Hill himself at Knebworth, south of Stevenage in Hertfordshire.! Combining the results of these analyses with a selection from those previously made and recorded,” we get the following averages for four of the zones :— (1) RestDvEs In THE ZONE OF MIcRASTER COR-TESTUDINARIUM. Grammes. Chalk from Stourpaine (Dorset) ............ 2-48 £ Hee DUNE WES) (SUSSEX)! ..20hecw sae ncne o2 0°66 ee LWover (Ment )/25.5.01 scree sanss 0-61 bh: » Remenham (Berks) ............ 1-43 ‘ .. Medmenham (Bucks) ......... 372 4 am biiie liners), 6 ones sk cis oe: 0°88 6)9°78 = average of 1°63. 1 It should be mentioned that the analysis of a sample referred to Hitchin in vol. iii of the Geological Survey-Memoir on the Cretaceous Rocks of Britain ( P., 320) was really of chalk obtained at Knebworth. 2 See ‘The Cretaceous Rocks of Britain’ Mem. Geol. Surv. vol. iii (1904) pp. 309 et segq. 14007 MR. A. J. JUKES-BROWNE ON [May 1906, (2) RestpvEs IN THE ZonE OF MICRASTER COR-ANGUINUM. Grammes. Chalk from Blandford (Dorset) ............ 119 >a 2 Durntord (Waits) 2ccce-sceees 2-00 y ee WViimmalCblants)) Beea.0c-eeee 0°81 : ) (Crondall (Elants) <.7..s--ee.eos 1:46 * oa HCullhham-(Berks)) 5; 2scc.ce soe. 1:38 z .. Knebworth (Herts) ............ 2°15 : ‘,. UDover (Kent) 2.22.5. 1:20 7)19:19 = average of 1°45. (3) RestDUEs IN THE ZONE OF MaRrsvuPizeEs. Granmes. Chalk from Blandford (Dorset) ............ 1:10 Bs » Arundel (Sussex) 2-iecss-.6- 1:24 { » . Margate (Ment) a aa-e-esces 0-99 ~ ,) Salisbury i( Wolts\)@.<-pe-eeere se 0:95 5 , (Crondalll (Hants) @es-cs-2- cone 0:87 5)5:15 = average of 1:03. (4) REsIDUES IN THE ZONE OE ACTINOCAMAX QUADRATUS. Gramimes. Chalk from Whaddon (Wilis) ............ 0-687 “ 55. Milford t(@QWallts) a) aeed 0-607 x » Culver (Isle of Wight) ...... 0°753 ms , Otterbourne (Hants)?......... 0-880 4)2°927 = average of 0°73. From the foregoing analyses it appears that there is a gradual diminution in the amount of insoluble residue as we ascend in the Upper Chalk, from the zone of Micraster cor-testudinarium to that of Actinocamawx quadratus. If we take the two lower zones together, as constituting that part of the Chalk in which flints are most numerous, we find that equal quantities of them will yield an average percentage of 1°54 of insoluble residue. The solution of equal quantities of the two middle zones will yield 1:24 per cent., and equal quantities of the two higher zones will yield only 0°88 per cent. of insoluble residue. This, however, is percentage by weight; and if we wish to calculate how many cubic inches of residue will be left by the solution of a given mass of chalk, we must allow for the difference in the weights of chalk andclay. Here, again, it has been necessary to make special experiments and estimations for the purpose of this paper, because the required information did not exist. Thus I could not find that anyone had determined the comparative weights of different kinds of chalk, I could find no record of the weight + This is taken from an analysis made by M. Duvillier, and published by Prof. Barrois in his ‘ Recherches sur le Terrain crétacé supérieur de |’Angle- terre & de l’Irlande’ 1876, p. 42, Vol. 62.] THE CLAY-WITH-FLINTS. 141 ot Clay-with-Flints, and I felt equally sure that different kinds of clay differ also in weight. In Sir C. W. Pasley’s book on ‘Limes & Cements’? it is stated that ‘one cubic foot of solid chalk, such as is obtained near Chatham, weighs when pertectly dry about 90 lbs.’ ; but in a table on p. 142 he gives the weight of a cubic foot of pure dry chalk as 94:99 Ibs., which is practically 95 lbs. The chalk to which Pasley refers was probably obtained from the zone of Micraster cor-testudinarium. Desiring, however, to have confirmation or correction of this, I asked Mr. W. Hill to ascertain the actual weight of small cubes cut from the samples of chalk sent to him. This he did, dissolving the cubes afterwards, with the results already recorded. Pieces were cut and trimmed as accurately as possible to form l-inch cubes; they were dried for 5 or 6 hours at 120° C., and then weighed, with the following results :— Grammes. Cube of Remenham chalk weighs ......... 26°05 » Knebworth _,, fhe slap hs ed 25°28 2 Culham 5 ney WERE AL ahs 32°19 » Winnal af HOM) Sp Mere te 30°25 4)113°77 = average 28-44. These results show that there is considerable variation in the weight of pure white Upper Chalk, even in the same zone of Micraster cor-anguinum, to which the last three belong. Let us first consider the Remenham chalk, for Sir C. W. Pasley’s Chatham chalk probably came from the same zone. The weight of one gramme is ‘002 of a pound, consequently 1 cubic inch of Remen- ham chalk weighs °0521 of apound. Multiplying this by 1728, we find that a cubic foot of the same chalk will weigh slightly over 90 lbs., which is the weight given for Chatham chalk—a curious coincidence. ~ The mean weight of all four samples is 28-44, and this, multiplied by -002, gives -05688 of a pound ; and again, multiplying by 1728, we get the result that a cubic foot of the average kind to be found in these two zones weighs 98°288 lbs. This is rather more than the greater weight given by Pasley, but we shall not be far wrong if we take the weight of a cubic foot of such chalk as 98 lbs. The determination of the weight of a cubic inch of Clay-with- Flints proved to be much more difficult. The plan adopted was to square roughly with a knife and then dry the cube, finally to dress it accurately with a file, and test it with calipers. Some samples, collected in dry weather, broke up so much that only half-inch cubes were obtainable; others, although moist, were sandy and friable for that reason. Out of many samples tried, Mr. Hill has only been able to make five satisfactory cuves, and three of these were half-inch cubes. From a sample obtained from the top of a quarry at Kneb- worth he cut two half-inch cubes, the mean weight of which was 1 2nd ed. (1847) pt. i, p. 4, 8vo. London. Q.J3.G.8. No. 246. Mt 142 MR. A. J, JUKES-BROWNE ON [May 1906, 4°33 grammes; this is equivalent to 34°64 grammes per cubic inch. Another half-inch cube, from the clay in a pipe at the same place, weighed 4°5 grammes, equivalent to 36 per cubic inch. From two samples obtained by Mr. J. Saunders near Luton, two good cubic inch-blocks were cut, one weighing 33 and the other 33°53 grammes. From these figures we deduce a mean value which cannot be far wrong, thus one cubic inch of Grammes. Knebworth clay.......-+.s.«- weighs 34:64 Do > dol(pipe) eee: ‘3 36:00 AGubonm (NO: 1) clay y.r-.-eee i 33°00 Dos NOM) "dor seen 55 33°53 4)137:17= average 34:04. An average cubic inch of Clay-with-Flints from these places weighs 34 grammes, which is °068 of a pound: consequently, a cubic foot of clay weighs 1728 x -068=117-5 lbs. Thus, if the Clay-with-Flints has been derived from the Chalk, its weight as compared with that of Upper Chalk is as 117-5 to 98. Further, around the London Basin the clay must have been chiefly obtained from the solution of the zones of Marsupites and Micraster cor-anguinum, for it is on one or the other of these that the Reading Beds generally rest. If, therefore, we suppose that 100 cubic feet of chalk be dissolved, half consisting of one zone and half of the other, and assuming the percentage of insoluble residue to be 1°24 by weight, the mass of the residue will be 117°5 : 98 :: 1:24:a, This works out as 1°03; consequently, the solution of a column of 200 vertical cubic feet of these two zones will yield 2:06 cubic feet of clay. In other words, in order to produce a layer of clay 2 feet thick, a thickness of 200 feet of such chalk will have to be dissolved, and the solution of 100 feet will produce only a 1-footlayer. If com- posed entirely of chalk of the Micraster cor-anguinum-zone, the layer produced by the solution of 100 feet will be 1-2 feet thick. In the area between the London and Hampshire Basins, it is mainly the chalk of the zones of Actinocamax quadratus and Mar- supites which would have suffered, and these only yield a combined percentage of -88 by weight ; the percentage by mass will be only °73. Consequently 100 cubic feet from these zones will produce a layer of clay only 9 inches thick, and it would take 200 feet of them to produce a layer of 18 inches over the whole area, supposing it to be all retained on the area and none of it lost by water-transport during the process of solution. It is very unlikely, however, that it could be so retained on the plateaux without any waste, and much more probable that half of it would be carried down into the valleys. In any case, it is clear that, if Clay-with-Flints consists half of clay and half of flints, the possibility of the derivation of the clay- portion from the Chalk must depend entirely upon the thickness of Chalk which can be proved to have been destroyed in any given district. Wol. 62.) THE CLAY-WITH-FLINTS. 143 There is, however, another way of testing the probability of such derivation, and this is to estimate the proportion of flints to clay that would result from the solution of any given thickness of chalk, for in the foregoing calculations the flints have been omitted. In the zone of Micraster cor-anguinum layers of flints are very frequent, and scattered flints sometimes occur in the chalk between them. In Kent and Sussex measured sections have shown that layers of flints frequently occur every 3 feet, and that sometimes they are only 2 feet apart; records of borings also often mention flints at intervals of 2 or 3 feet, but in the higher part of the zone they are not quite so frequent. We shall probably be under- estimating the quantity of flints, if we assume that their average distance apart in the whole zone is 34 feet; and if we also assume that the average thickness of the flints is 3 inches, then every 14 feet of chalk will contain a depth of 1 foot of flint-nodules, and 98 feet of this chalk will yield 7 feet of flints. In round numbers, therefore, the solution of 100 feet of Micraster cor-anguinum-chalk will produce a layer of flints 7 feet thick, together with enough clay to form a layer about 14 inches deep. Hence there would be about six times as much flint as clay in the mixture, and the product would not be ‘ clay-with-flints,’ but a bed of flints with about just enough clay to fill up the interstices between the nodules. In the zones of Marsupites and Actinocamax quadratus the flints are much less numerous. They do occur, however, in the latter zone at distances of from 6 to 9 feet apart, and round the London Basin flints are found in the Marsupites-zone, though sparsely. We may perhaps estimate them as occurring in these zones at 8 feet apart, but in quantity only enough to form a continuous layer 2 inches thick. At this rate there will be only twelve layers in 100 feet, with a total thickness of 2 feet of flints. The same thick- ness of chalk, if composed of equal parts of the same zones, will yield 9 inches of clay; and in this we should have something more like Clay-with-Flints, but still containing nearly three times as much flints as clay; whereas, in the real material, the proportion is only 1 to 1. If we combine 100 feet of Marsupztes-chalk and 100 feet of the Micraster cor-anguinum-zone, the 200 feet will yield a layer of flints 9 feet deep and one of clay about 2 feet deep. The mixture would be a clayey gravel, the proportion of clay to flints being nearly as 1 to 43. V. Ep aLeons TO BE DRAWN FROM THE DISTRIBUTION OF THE Cray-witH-Frints. From the conclusions arrived at in the preceding pages, it is clear that the stratigraphical relations of the tracts of Clay-with-Flints, both to the Chalk and to the Hocene Series, must be carefully considered. In this connection, it is necessary to remember that the Chalk had been raised and flexured to some extent before the M 2 144 MR. A. J. JUKES-BROWNE ON [May 1906, deposition of the Eocene beds upon it; so that, when viewed as a whole, the Eocene strata are entirely unconformable to those of the Chalk. Thus the local base of the Eocene Series rests in different districts on different portions of the Upper Chalk, from the zone of Belemnitella mucronata to that of Micraster cor-anguinum, and it 18 very probable that (in some places) it rested on the zone of Micraster cor-testudinarium, if not on that of Holaster planus. In other words, the surface which received the Hocene deposits was a surface which had been planed across a raised and broadly-curved or flexured mass of Chalk. Further, in one part of the region, namely in Surrey and Kent, elevation and erosion seem to have been renewed in early Kocene time, for Mr. Whitaker has shown! that the Oldhaven and Black- heath Beds overstep both the Woolwich and Thanet Beds, so as to rest upon the Chalk. This may have taken place over a large area in the northern part of the Weald, when the Chalk extended much farther south than it does now. Lastly, it is well known that both Chalk and Eocene were sub- jected to still greater disturbance in Miocene and Pliocene times, along lines that were more or less parallel to the great post- Oligocene flexure of the Isle of Wight. I think it almost certain that the flexuring which took place before Eocene times was of slight intensity ; and that only a broad and low geanticline was then pro- duced, the curvature of which ineluded, not only the space between the London and Hampshire Basins, but also part of the London Basin itself: the more pronounced flexures which occur within this geanticline being wholly or mainly of post-Oligocene date. These later flexures have a peculiar arrangement, occurring as discontinuous periclines, that is, as local elliptical domes and basins which are not always arranged along axial lines, but often alternate en échelon, the termination of one periclinal convexity passing below the beginning of a cymboid basin, or trough, which is flanked on each side by elongate convex periclines, all of them dying out again within a certain distance. Such an arrangement of flexures is found, for instance, in the country between Winchester and Salis- bury. The Kingsclere tract is another notable instance of a convex pericline. Now, if the Clay-with-Flints has no special relation to the Lower Kocene beds, but is a residue from the Chalk, the largest tracts of this argillaceous residue should occur over the domes and convexities which have been exposed to the greatest amount of quiet subaérial detrition since Eocene time; and the smallest and thinnest should lie in the synclines from which less chalk has been removed. On the other hand, if the Clay-with-Flints is mainly composed of residue from the basal Eocenes let down into the Chalk, it will naturally be in greatest force over those tracts from which the least amount of chalk has been removed, and on which traces or outliers of the Kocene deposits still remain; while it will be entirely absent from * Quart. Journ. Geol. Soc. vol. xxii (1866) p. 420; and Mem. Geol. Surv. vol. iv (1872) p. 240. Vol. 62.] THE CLAY-WITH-FLINTS. 145 the domes and anticlines. In this way, therefore, the distribution of the clay-tracts should clearly indicate the source from which the Clay-with-Flints has been chiefly derived. There is, however, one large area in which the relations of the Clay-with-Flints are complicated by several special circumstances, and the consideration of which I shall leave to others who have better acquaintance with it. This area is that of Kent and Kastern Surrey, where the Thanet Beds are the lowest member of the Eocene Series, and where also the Oldhaven Beds seem to have passed over them southwards onto the Chalk. This area also was submerged in Lower Pliocene times, and there seems to be some uncertainty about the relation of the Lenham Beds to the Clay- with-Flints. Mr. Reid states that, on Lenham Down,’ patches of Lenham Beds underlie extensive sheets of Clay-with-Flints ; on the other hand, Prestwich had long ago figured and described a layer of brown and black clay with flints, as underlying the Lenham Sands and resting on the Chalk, not far from the same place. Its position according to Prestwich seems natural; but its position according to Mr. Reid seems difficult to understand. I shall, therefore, confine myself in the following pages to a consideration of those areas where the Reading Beds form the lowest member of the Eocene Series, and where no marine Pliocene Beds occur as a disturbing factor. Let us turn first to the country which lies west and north-west of the London Tertiary Basin. Anyone who will study the published Drift-editions of the l-inch maps of the Geological Survey, such as Sheets 267, 268, etc., will see that there appears to be a very close connection between the Reading Beds and the areas mapped as Clay-with-Flints. It must, of course, be remembered that the latter include the overlying loam and brickearth where such materials occur; but these are admitted to be inseparable from Clay-with-Flints. The noticeable point is, that these loams, with their basal layer or fringe of Clay-with-Flints, behave as if they were merely dis- integrated portions and outliers of the Reading Beds. In the far west of the London Basin, that is, in the western parts of Wiltshire and Berkshire, the tracts of Clay-with-Flints occur on ridges and plateaux which are clearly portions of an inclined plane rising towards the outer escarpment of the Upper Chalk. The material seldom descends far down the slopes of these ridges, except in places where there is reason to believe that landslips have occurred; as a rule, the average level of the boundary-lines is little below that at which outliers of Eocene occur or might occur. In other words, the surface on which the tracts of Clay- with-Flints le is practically a prolongation of the basal Eocene plane. Moreover, the tracts do not become larger and thicker in the direction of the Chalk-escarpment, as should be 1 “The Pliocene Deposits of Britain’ Mem. Geol. Surv. 1890, p. 40. 146 MR. A. J. JUKES-BROWNE ON [May 1906, the case if both clay and flints have been derived from the solution of the Chalk; on the contrary, they become smaller and thinner, with a surface-soil that is more crowded with flints, as we should expect if these outer tracts had been exposed to detrition for a longer time than those nearer to the main mass of the Reading Beds. It will also be noticed that the positions thus occupied by the tracts of Clay-with-Flints quite preclude the idea that any great thickness of Chalk can have been dissolved from beneath them since the removal of the Reading Beds. It is true that, on the Marlborough Downs, and to the north and north-west of Lambourn, the Clay-with-Flints rests upon a lower zone of the Chalk than does the main mass of the Eocene; but there is good reason to believe that the Eocene passed transgressively over the surface of the Chalk towards the west, so that the Reading Beds stepped from higher to lower parts of the Chalk as they spread westwards to and beyond the present escarpment-ridge. North of the Thames Valley the Clay-with-Flints behaves in just the same way, rising along the Chalk-ridges with the Eocene outliers, up to the summit-ridge of the Chiltern Hills. Farther: north, round Luton, it is associated with thick masses of brickearth which seem to pass into loam and brickearth of Glacial age. I believe that much of the so-called ‘ Middle Glacial’ loam in the counties of Bedford, Hertford, and Essex has been reconstructed from the brickearths, which were themselves derived from Reading Beds; thus, near Weston, north-east of Hitchin, I have seen Clay- with-Flints beneath loam, in a position which gave strong ground for thinking that both passed under the neighbouring Boulder-Clay. Clay-with-Flints often wholly or partly surrounds outliers of Reading Beds, or connects two such outliers, without descending far below the level at which the local base of the Reading Beds is found. In fact, if the loam and brickearth have been formed from the disintegration of the Reading Beds, as is generally supposed: then it seems impossible to avoid the conclusion that the formation of the Clay-with-Flints was concomitant with that of the brickearths, and also with the detrition and diminution of the Eocene outliers. Moreover, although but one colour is assigned to. such tracts on the Geological Survey-maps, yet the index on the margin always states that this colour includes ‘ Clay-with-Flints and brickearth’ (or loam). Now, the brickearth, which is thus associated with the Clay-with-Flints, is sometimes underlain by that material, but sometimes rests directly upon the Chalk; further, this brickearth often includes beds or masses of quartz-sand. Hence, it is clear that a theory designed to explain the formation of Clay-with-Flints alone will not suffice; it must be one that will account for the disintegration, disturbance, and subsidence of masses of Eocene clays and sands part passu with the formation of the Clay-with-Flints. It must be recognized, in fact, that Clay-with-Flints is not a distinct and definite material, like Boulder-Clay, which can be separately mapped and receive an index-colour for itself; it is inseparably associated with a more or less confused assemblage of Vol. 62.] THE CLAY-WITH-FLINTS. 147 brickearths, loams, sands, and gravels. Small patches can doubtless be found which consist entirely of Clay-with-Flints; but larger tracts almost always include masses of mottled clay, sand, and gravel, and sometimes accumulations of sarsen-stones, or of unworn flints. One good instance of such associated materials has already been cited (p. 188), and another may be quoted from a recent Memoir, as occurring in the Aldworth and Yattendon area above mentioned.’ At Buttonshaw Kiln, south-west of Aldworth, Mr. Bennett saw a section showing ‘6 feet of large unworn flints and sarsens resting irregularly on rusty-brown and black-stained clay. A hole dug for clay near the kiln showed a mixture of plastic clay and coarse red sand, capped with Clay-with-Flints. The plastic clay seemed in process of conversion into the rusty-brown clay.’ Let us take an actual example of the relative positions which the Eocene and the Clay-with-Flints frequently occupy. for this purpose I choose the large outlier of these combined deposits round Yattendon and Ashampstead, west of Reading. This tract is completely isolated by the valleys of the Pang and the Thames ; it includes no fewer than seven outliers of Reading Beds, connected and more or less surrounded by Clay-with-Flints. At the extreme north-western end (north of Aldworth) the ground rises to a height of 600 feet, and there is a very small patch of Reading Beds accompanied by Clay-with-Flints at the same level. South-west of Aldworth is a larger outlier of Reading Beds, the base of which descends to about 520 feet, while that of the Clay- with-Flints near the same point is some 20 feet lower. On the south there is a continuous tract of Clay-with-Flints, the base of which falls rather rapidly to a level of about 320 feet near Yattendon; but this evidently accords with the original slope of the Kocene basal plane, for a large outlier of Reading Beds sets in at Yattendon, its base sloping south-westward from about 350 feet to a little below the contour of 300 feet. At the same time, the Clay-with-Flints thins out rapidly against the Eocene boundary. The other Eocene outliers lie on a similar sloping surface of Chalk, and the base of the connecting tracts of Clay-with-Flints. slopes in the same direction, that is, south-westward ; but its border passes below the level at which any Eocene is exposed, until it reaches the contour of 300 feet.- Again, there are cases where a kind of Clay-with-Flints seems to overlap onto the border of an Eocene tract. On this point, Mr. Osborne White informs me that he believes such overlap to be of frequent occurrence, although it is not easy to prove the fact, because it is very seldom that a continuous section from the surface of Chalk to the surface of Eocene is actually exposed. At Cadmore-End Common (Oxfordshire) he has seen a section 10 feet deep, showing Reading Beds (mottled loam) overlain by brown loam full of white angular flints, both being covered by a loamy soil full 1 «Geology of the Country around Reading’ Mem. Geol. Surv. 1903, p. 60, 148 MR. A. J. JUKES-BROWNE ON [May 1906, of pebbles. There are several exposures of similar clay on the outlier of Reading Beds at Maiden Grove (Oxon). I have the same authority for stating that most of the Reading-Bed outliers, which occur among the Clay-with-Flints of the Berkshire Downs, are overlapped at their edges by clay and loam with angular flints. The flints in such clays are almost always white and angular unbroken flint-nodules being rare; but he believes the material to be continuous with the surrounding Clay-with-Flints. Mr. White has also noticed that this transgression of the Clay- with-Flints onto Eocene tracts occurs most frequently on the sides which face up the slope of the Chalk-surface, and much less often on the lower or dipward side of the outliers ; a fact which suggests a certain amount of slow movement from the main watershed. The intimate relation between the occurrence of Clay-with-Flints and the basal plane of the Eocene Series is brought out still more strongly by a study of the area which lies between the London and Hampshire Basins. ‘There is a further advantage in dealing with this area, in the fact that the geological mapping of it has been completed, and published on the new series of 1-inch maps (Sheets 282, 283, 284, 298, 299, & 300), so that these can be referred to in confirmation of the statements made below. In considering this area, we must remember what has been said on p. 144, that the structure is complex, and that the flexuring was not all produced at one period; that the main broad geanticline is of pre-Kocene date, and was planed down before the Reading Beds were deposited, both Chalk and Eocene being subsequently bent into smaller and narrower folds. Beginning with the western part of the area (Shects 282 & 298), we find that these include a surprisingly-small amount of Clay-with- Flints. In the far west, and just outside the limits of Sheet 282, there is a tract of some length on the highest ridge of the downs which lie between Imber and Warminster. On the north, there is a curvilinear tract on the summit of the downs south of Chirton and Wilsford (600 to 700 feet), the boundary of the flinty clay being in some places not more than 20 feet above the outcrop of the Chalk-Rock. Farther east, near Upavon, there are three small patches below 600 feet; but none has been mapped on Pewsey or Milton Hills, although the latter rises to 782 feet above O.D. The only other place within the area of Sheet 282 where Clay-with- Flints has been mapped is on Sidbury Hill, where it flanks a small outlier of Reading Beds at an elevation of about 700 feet. The southern portion of the area included in Sheet 282 and the northern part of that which comes into Sheet 298 form together the central portion of Salisbury Plain. All of this area lies below 600 feet, and is traversed by a network of branching-valleys. The surface of what was originally a plain or plateau has evidently been graded and lowered by long exposure to the action of rain, and the ridges are capped by a variable thickness of flint-débris, consisting partly of entire and unworn flints, and partly of broken angular Mol, 62.] THE CLAY-WITH-FLINTS. 149 fragments, but nothing was found that could be mapped as Clay- with-Flints, except a few patches near Wilsford and Woodford. Now, if this material is mainly a Chalk-residue, how can its absence on this part of Salisbury Plain be explained? If, on the other hand, it is mainly an Eocene residue, we should not expect to find it far below the level of the basal Eocene plane. Within the area of Sheet 298, besides the few small patches above mentioned, a long tract of Clay-with-Flints caps the ridge which lies between the valleys of the Wily and the Nadder, pointing eastward to another patch on the slope of the hill north-west of Salisbury ; and there are other spreads of varying extent in the southern part of the district. The relations of these tracts of Clay-with-Flints to the tectonic structure of the country are interesting and suggestive. The ridge between the Wily and the Nadder is capped by red and brown Clay- with-Flints for a length of 9 miles, and it lies over the central axis of a syncline which intervenes between the anticlines of the Vales of Wardour and Warminster. It is, therefore, probable that the base of the Kocene lay not very far above the summit of this ridge, and that very little chalk has been removed from the watershed in post-Hocene times, either by solution or by any other process. The presence of the Clay-with-Flints here is quite comprehensible if it is mainly an Hocene residue, but is inexplicable on the Chalk-residue hypothesis. It may be mentioned that the western end of this tract of Clay- with-Flints rises to a level of 700 feet, and that Claypit Hill on the other side of the Wily valley, north-east of Codford, is capped by a patch of materials (mottled clay and sand) which, although disturbed, are recognizable as remnants cf Reading Beds.* This outlier probably owes its preservation to its having been let down into a hollow in the Chalk, or to an ancient landslip. Its present summit is only 586 feet above Ordnance-datum, but it suffices to show that Reading Beds lay over this area at some level between 600 and 650 feet. The materials derived from the Reading Beds are here overlain by a peculiar deposit, consisting of yellow sandy clay full of angular fiints and small quartz-pebbles, which is described by Prestwich as extending along the ridge of Codford Hill, but is not shown on the Geological Survey-map. Such a deposit seems to resemble the Plateau-Gravel of Berkshire more than the Clay-with-Flints, although it may be a contemporary variation of the latter. How the flexures above mentioned are continued eastward is not certain, except as regards that of the Vale of Wardour, which is cut off obliquely by a fault striking roughly from west-north-west to east-south-east. The intermediate syneline seems to be deflected southward, so as to pass below Wilton and Salisbury, and may possibly be continued into the syncline of the Eocene tract at Alderbury. The Wily anticline appears to be prolonged to Lower 1 See Prestwich, Quart. Journ. Geol. Soc. vol, xlvi (1880) p. 144, and Geol. Mag. 1898, p. 412. 150 MR. A. J. JUKES-BROWNE ON [May 1906, Woodford on the Avon, but here it seems to die out; and I am. inclined to regard the small scattered tracts of Clay-with-Flints around Woodford as being disposed round the periclinal termination of this anticline, a position which would account for the low levels at which they occur (400 to 300 feet). On the east side of the Avon Valley another set of flexnpes come in, which seem to be independent of those on the west. Here we find a synclinal trough filled with Hocene, stretching eastward from Alderbury. This is bounded on the south by a well-marked anticline of bare chalk, and on the north by what is probably a monoclinal rise, carrying the basal plane of the Eocene Series up to a level which is much higher than the present general level of Salisbury Plain. Here it is interesting and instructive to observe several tracts of Clay-with-Flints, which behave exactly as if they were outliers of Reading Beds, their southern boundaries being at about 300 feet, while on the north they reach a height of about 440 feet. On a line with them, but at a slightly-higher level (486 feet), near Laverstock is a small outlier of Reading Beds; and to the north of them, on Thorny Down, is another outlier at 533 feet. This last is about 300 feet higher than the outcrop of the Reading Beds in Clarendon Park, only 3 miles to the south, and shows the rapid rise of the basal plane or floor, upon which the Eocene rests. These small outliers, and the others previously noted as occurring farther north in the area of Sheet 282, enable us to construct a section showing the relative level at which the Eocene floor passed over this part of Salisbury Plain (see Pi. VI, fig. 1). It also indicates the manner in which the higher zones of the Chalk were planed off prior to the formation of the Reading Beds, and thus demonstrates the compound nature of the geanticline between the London and Hampshire Basins. Passing now to the area included within Sheet 283 of the Geological Survey-map, a casual view might convey the impression that the tracts of Clay-with-Flints were distributed in an irregular manner without respect to present levels, and without any special relation to the flexuring of the district; but a closer examination shows that there is a definite relation between them and the post- Eocene flexures. Broadly speaking, the structure of the area is as follows. Its northern part is traversed by an anticline, or rather by a series of periclinal domes, which are made apparent by the outcrops of Selbornian and Cenomanian strata near Grafton, Shalbourn, Wood- hay, and Burghclere. From this irregular anticline the beds of the Chalk slope southwards into a deep periclinal basin, the centre of which must le near the town of Andover, for near this place the zone of Marsupites is found at levels of from 300 to 400 feet. If this basin be the complement of the northern anticline, and was formed in post-Hocene times by a movement which carried the basal plane of the Eocene with it, and, further, if the Clay-with- Flints be mainly an Eocene residue: then we should expect to find Welo2.|° THE CLAY-WITH-FLINTS. 151 the tracts of Clay-with-Flints continually descending to lower levels as they are followed southwards. ‘This is the case: in the northern part of the area the Clay-with-Flints is only found on the summit of ridges that rise to 700 or 800 feet, but in passing southwards, and especially from north-west to south-east, we find it descending gradually to about 300 feet. Thus, on Tidcombe Down in the north- west the Clay-with-Flints sets in at a height of about 830 feet, and there are several large tracts of it to the south-east, the boundary- lines of which gradually slope from 700 to less than 400 feet, and east of Andover it occurs on the Marswpites-chalk at a level of 300 feet. In this connection it is especially interesting to find that, in all the larger tracts, Clay-with-Flints is associated with, and often covered by, thick masses of brickearth and sandy mottled clays which have evidently been derived from the destruction of the Reading Beds. Further, in Harewood Forest, east of Andover, patches of clean mottled clay and yellow sand occur, which were originally described as outliers of Reading Beds*; but, on the more recent map issued in 1898, they are included in the ‘ Clay-with- Flints and Loam.’ From the description given in the memoir above cited, it would seem that masses of little-altered Reading Beds do occur here at the low level of 300 feet: thus confirming the suggestion above made, that the Andover Chalk lies in a post-Kocene syncline, for these remnants cannot be far removed from the position that they occupied when wz situ. It is also stated that some of the adjacent material resting upon the Chalk is a reddish-brown mottled clay containing pebbles, and having a layer of large black-coated flints at its base: the whole being overlain by a gravelly deposit, consisting of rounded and angular flints in a matrix of reddish-brown clay. We seem to have here all the materials for making the mixture usually known as ‘Clay-with-Flints’, and I cannot help thinking that we are here provided with an illustration of a phase in the making of Clay-with-Flints ; with a case, in fact, in which further developments were arrested, owing to some local conditions which may or may not be discernible by a more complete study of the district. Close to the northern edge of the area in Sheet 283, and near Woodhay Clumps, is a small but instructive outlier of Reading Beds which touches the contour of 900 feet, and is only about 180 feet above the neighbouring outcrop of the Chalk-Rock. Hence, though solution of the Chalk may have caused it to subside or slip considerably below its original level, its position seems to indicate that the thickness of chalk removed from this locality before the time of the Reading Beds was greater than usual. The Eocene outlier is surrounded by Clay-with-Flints, which extends along the southward prolongation of the ridge, its basal boundary passing gradually from 900 to 650 feet. Other tracts on the south-east 1 See ‘ Geology of Parts of Berks & Hants’ Expl. of Sheet 12, Mem. Geol. Surv. 1862, p. 28. 152 MR, A. J, JUKES-BROWNE ON [May 1906, take it from 600 down to about 400 feet, and finally patches near Whitchurch carry it down to 300 feet. The natural inference from these facts is that we are following a slope which coincides approximately with the inclination given to the basal plane of the Reading Beds by post-Eocene disturbances ; and that, when allowance has been made for solution, downwash, and landslips, the tracts of Clay-with-Flints may be regarded as having been formed out of the remnants of so many ancient outliers of Reading Beds. This view receives a striking confirmation from the fact that at East Stratton, only 7 miles south-east of Whitchurch, an outlier of the Reading Beds actually occurs at a level of about 300 feet, and is surrounded by patches of Clay-with-Flints. J shall recur to this outlier in the sequel (p. 155), as it comes within the area of Sheet 300; but its existence testifies to the low level at which the Lower Eocene deposits originally lay over this part of Hampshire, and to the small vertical thickness of chalk which has been removed from the post-Eocene surface. Proceeding to the area south of Andover and Whitchurch (Sheet 299), we find a wide tract over which Clay-with-Flints either is absent, or occurs only in small scattered patches. This tract coincides with the anticline of Stockbridge, which was first discovered and described by Prof. Barrois.. He regarded it as a continuation of the Winchester uplift; but, on this point, I venture to differ from my friend and confrére, believing it to be an independent periclinal dome which may be prolonged for some miles to the westward, but dies out rapidly to the east. However this may be, it is doubtless a post-Kocene uplift, and consequently an area over which subaérial detrition has been great. If, therefore, Clay-with-Flints were a Chalk-residue, it should be in strong force; while if it be derived from the Eocene, it is easy to understand its absence, except for two small patches near Stockbridge itself. The western part of the area in Sheet 299 adjoins that of Salisbury (Sheet 298), and the flexures noticed in the latter are continued eastward to the valley of the Test. The anticline on the south passes through Dean Hill and dies out toward Lockerley. The syncline filled with Kocene deposits which runs eastward from Alderbury also seems to disappear near Mottisfont; while the Stockbridge uplift above mentioned is probably a development of the monocline traced to the north of Clarendon. A glance at the geological map will show that tracts of Clay- with-Flints occur in connection with these flexures, and are disposed 1 “Recherches sur le Terrain Crétacé de Angleterre & de l’Irlande’ 1876, Oe tie The curvature of the Stockbridge flexure must be slight and low, for the zone of Micrasten cor-testudinarium has only been found near Stockbridge ; and Mr. C. Griffith, F.G.S., informs me that to the westward, along the valley of the Wallop, all the pits seem to be in the zone of M. cor-anguinwm, while north- ward at Grately he found Utntacrinus-chalk, and southward still higher beds come in. Wol, 62. | THE CLAY-WITH-FLINTS. i5ys° exactly in accordance with the view that they are the remnants of Eocene outliers. Thus, on Dean Hill there are four patches of Clay- with-Flints, which are so arranged that their boundaries show an -eastward slope from 500 to 300 feet; and the largest one slopes south-eastward from the ridge-summit at 440 feet to about 350 feet, its termination being within half a mile of the Eocene boundary at 300 feet above Ordnance-datum. The tracts on the northern limb of the syncline are similarly disposed on a rising plane, which appears to be the natural prolongation of the basal Eocene plane. Thus, at a point near the junction of Sheets 298 & 299, the Eocene boundary-line is at a level of 222 feet; and north of this tracts of Clay-with-Flints rise from 300 to a height of 500 feet at Winterslow, in a distance of about 27 miles. Farther east, near Winchester, we find another distinct uplift coming in, which does not seem to be connected with either of the anticlines to the westward. This may be called the Winchester pericline, but it is really the periclinal termination of the Petersfield and Meon-Valley anticline. The centre of this pericline does not coincide with the centre of the Lower Chalk-exposure, but with the eastern end of that exposure near Chilcombe, where the base of the Middle Chalk is 160 feet higher than it is below the alluvium of the Itchen. The strata are, in fact, dipping westward from this centre, as well as northward and southward, and the pericline must die out between the valleys of the Itchen and the Test, This being so, it is interesting to note the complete absence of Clay-with-Flints on the Chalk-downs west of Winchester, for a distance of about 5 miles; and further, when tracts of it do set in, they are disposed on the southern, western, and northern slopes of these downs, exactly as if they were the remnants of a sheet of Eocene clays which had overlain the Chalk and had been included in the periclinal flexure. I have little doubt that they are such remnants, modified by subaérial agencies and sunken somewhat from their original position by solution of the underlying chalk. It will be noticed that, at Braishfield, there is an outlier of Reading Beds which is equidistant from the main outcrop and from a small patch of Clay-with-Flints at a higher level, leading on to the larger tract of Farley Down. This ascends to over 500 feet, and a section across the district here would appear as in Pl. VI, fig. 2. Coming now to the country round Basingstoke (see Sheet 284), I venture to assert that the positions occupied by the various tracts of Clay-with-Flints in this area can only be satisfactorily explained on the hypothesis of their having originally formed part of a continuous sheet of material, which had been flexured together with the Chalk. If all the isolated tracts are so regarded, it will be seen that their connection across the intervening spaces would mantle the surface of the Chalk, in such a manner as to accord with the known or probable flexures, and with the position of the Hocene outliers. In short, the hypothesis which I advocate furnishes a 154 MR. A. J, JUKES-BROWNE ON [ May 1906, simple and natural explana- tion of the distribution of the Clay-with-Flints in this area, which would be very difficult to account for on the supposition of its being entirely or mainly a Chalk- residue. Thus, round Han- nington there are patches which, if united, would lie on a curved surface that would partly wrap round the eastern termination of the Kingsclere pericline, and this surface, if prolonged, would meet the outcrop ot the Reading Beds near Wolverton and Ewhurst. From the high ground near Hannington the patches of Clay-with-Flints descend southward and south-east- ward to a level of 370 feet, and some of them are traversed by the rail- way-cuttings near Church Oakiey. From this low level remnants of what was once a continuous mantle rise gradually again to over 600 feet, on the high ground south of Basingstoke. I take it that this descent of the Clay-with-Flints coin- cides witha synclinal flexure of the Chalk between the Kingsclere pericline and the Bentley-Farnham up- lift, which is little more than a monocline.* As for the large spreads of Clay-with-Flints on tie high ground around Far- leigh, Ellisfield, Herriard, and lLasham, and thence eastward along the summit- N.N.E. Main Road Rye Farm near Lr. Lane t ( ff — ae 7 = Clay-with-Flints. N.£E. S.S.W. _6 = London Clay. I, el! "| 1 inch =1000 feet. | 4 = Upper Chalk (450 feet). 5 = Reading Beds, 2 inches =1 mile; vertical: Section from Froyle Down to Rye Farm, near Crondall. [Scales, horizontal : Lower Chalk (200 feet). Middle Chalk (150 feet). Selbornian. 1 2 3 1 See W. Topley, ‘Geology of the Weald’ Mem. Geol. Surv. 1875, p. 280; and A. J. Jukes- : Browne, ‘The Cretaceous Rocks NE Gn of Britain’ Mem. Geol. Surv. vol. i (1900) p. 10. Froyle Down Lane i S.2W. Vol. 62.] THE CLAY-WITH-FLINTS. 155 ridge of the Chalk-escarpment: these seem to occur at levels which would have been occupied by a continuation of the Eocene deposits westward from Farnham and Crondall through the two outliers of Reading Beds west of the latter place. A section drawn through these two outliers and through the patches of Clay-with-Flints north and south of them, shows that the latter occupy positions which must previously have been held by the Reading Beds (see the accompanying text-figure, p. 154). Of the area in Sheet 300 to the south of that described above, it is unnecessary to give much detail. Its chief features in relation to my subject are as follows :— 1. The large continuous sheet of Clay-with-Flints which covers the eastern part of the Upper Chalk-plateau, from which the ground slopes east- wards to the Wealden area. 2. The existence of a small outlier of Reading Beds, partly surrounded by Clay-with-Flints, at Stratton in the north-eastern corner of the district. 3. The entire absence of Clay-with-Flints in the south-eastern corner of the district, which seems to be traversed by a flexure connecting the Winchester pericline with the Meon-and-Petersfield anticline. The extensive tract of Clay-with-Flints on the high plateau (above 500 feet), which forms the eastern border of the Upper Chalk-area, has a length of 12 miles from south to north, and is a continuation of that mapped to the north in Sheet 284. Itis of very irregular shape, owing to the branching system of valleys by which it is deeply incised, and is thus interesting as an illustration of the manner in which similar large wide-spreading tracts have been cut up into isolated patches, and the way in which such patches come to occupy different levels on a sloping surface of Chalk. This irregular tract of Clay-with-Flints helps us to look back to a time, somewhere between the Miocene Period and the Glacial Epoch, when the greater part of the Chalk-area between the present boundaries of the London and Hampshire Basins was covered by a mantle of this material, or of the deposits from which it has been derived. The outlier of Reading Beds on the north-west takes us a step farther, and indicates one at least of the deposits out of which the Clay-with-Flints has been formed. This outlier lies just above the contour of 300 feet, and the surrounding patches of Clay-with- Flints are at about the same level, that is, from 300 to 350 feet. The remarkable fact of its occurrence has already been commented upon, but its importance can hardly be exaggerated, because it lies about midway between the London and Hampshire Basins, and because it proves that the Eocene Beds were here carried down to a com- paratively low level in the geanticline, presumably by post-Eocene re-adjustments. As the level at which Clay-with-Flints is found rises from Stratton northward, eastward, and southward, I suspect that the Eecene outlier marks the site of a periclinal basin or ‘ cymboid,’ which probably extends for several miles westward, through 156 MR. A. J. JUKES-BROWNE ON [May 1906, Micheldever and Sutton Scotney, where patches of Clay-with-Flints occur at a still lower level, that is, down to 270 feet. It is, therefore, of great importance to ascertain what zone of the Chalk underlies the Stratton outlier; for, if it were found to rest on the zone of Micraster cor-anguinum, we should have proof of a certain amount of pre-Tertiary erosion, and should infer the existence of a broad anticline of Chalk which was subsequently indented by a local post-Hocene syncline. Whereas, if the Chalk belongs to the zone of Marsupites, or to that of Actinocamax quad- ratus, there would be no proof of a pre-Tertiary anticline, and we should infer that the flexure which brings the Eocene down to this level was entirely of post-Eocene date. Mr. Charles Griffith, F.G.S. (to whom I applied for information on this point), was kind enough to visit Stratton, but could not obtain any definite evidence. The chalk-pits which he visited were much overgrown; of one exposure he writes :— ‘The appearance of the chalk and the flints suggests the zone of Marsupittes, but I saw no plates, either of that fossil, or of Utntacrinus,’ Westward, by the side of the railway west of Micheldever and south of Weston Colley, is a quarry from which Offaster lula has been obtained; while at Micheldever Station, 3 miles north of this, the zone of Micraster cor-anguinum comes in.’ Mr. Griffith has also found Offaster pilula in a quarry at Sutton Scotney, and Actino- camaxe granulatus in another about 2 miles north of that village. Thus, all the available evidence points to the conclusion that Stratton lies on one of the higher zones, either that of Marsupites or that of G. K. Gilbert, ‘ Lake Bonneville’ U.S. Geol. Surv. Monogr. 1 (1890). 8 W. M. Davis, ‘On the Classification of Lake-Basins’ Proc. Boston Soc. Nat. Hist. vol. xxi (1882-83) p. 315. 7 H. R. Mill, ‘ Bathymetrical Survey of the English Lakes’ Geogr. Journ. vol. vi (1895) pp. 46, 135. 8 Sir J. Murray, ‘ Bathymetrical Survey of the Freshwater Lochs of Scotland’ Geogr. Journ. vols. xxii-xxvii (1903-1906). ° ‘Handbuch der Seenkunde’ Stuttgart, 1901, p. 9. 166 PROF. E.. J. GARWOOD ON THE [May 1906, have us regard the majority of the large Alpine lakes as due to the direct erosive action of ice, has gradually given rise to a detailed consideration of the genesis of each individual lake. At the present day, several eminent authorities still attribute the origin of some of the larger Swiss and Italian lakes to ice-erosion. Thus Prof. Brickner, in his recent work,’ considers that the Lake of Zurich was created by glacial overdeepening, and that it represents the end of an old valley which was overdeepened by the Linth Glacier; he also attributes the formation of the Lakes of Zug and Lucerne to the same agency.” Prof. Davis again, in his paper on ‘ Glacial Erosion in the Valley of the Ticino,’? attributes the rock-basin of the Lago Maggiore to direct ice-erosion. It is evident, therefore, that this theory is in no wise extinct. On the other hand, many Swiss and Austrian geologists, notably Prof. Heim and Dr. Forel,’ attribute all these lakes to post-Miocene earth-movements in one form or another. These lakes, which are of considerable size, all occur in the outer zone of the Alps, or the ‘ Kalk-Alpen’ of Continental writers, and are often spoken of as ‘ Randseen’ or marginal lakes, Itis not, however, with these that I am at present concerned. The lakes, the origin of which I intend to discuss in the present communication, are those to which in this country the name ‘tarn’ is frequently applied; they occur invariably in mountain-districts, and in the Alps are confined to the ‘ Hochgebirge.’ They may be defined as lakes draining into the principal Alpine valleys, whereas the Randseen receive the drainage of these valleys. These tarns, although they frequently occur as ‘ corrie ’-lakes, include also others which, strictly speaking, do not come under this category. Of these latter, some are true rock-basins ; while others owe their origin to the damming of valley-drainage by loose material. Not long ago, Dr. Marr, in writing of the ‘Tarns of Lake- land,’ pointed out that many, if not all, of them were due to accumulation, and expressed a doubt as to whether any lakes of this class would eventually prove to be true rock-basins. What- ever may be the case in Cumberland, there can be no doubt of the occurrence of true rock-bound tarns among the ‘ Hochgebirgseen’ of the Alps. In the pages of the ‘Geological Magazine’ for 1898,’ Prof. Bonney instanced a group of lakes in the neighbourhood of Airolo, which he unhesitatingly described as true rock-basins, and referred provi- sionally to excavation by ice; though, in the case of Lake Ritom, at 1 A. Penck & E. Brickner, ‘ Die Alpen im Hiszeitalter’ Leipzig, Lieferung 5 (1903) p. 525. > Op. cit. pp. 537-88. ® ¢ Appalachia’ vol. ix (1900) pp 151-52. + A. Heim, ‘Die Entstehung der Alpinen Randseen’ Vierteljahrsschrift der Naturforsch. Gesellsch. in Zurich, vol. xxxix (1894) p. 66; and F. A. Forel, Bull. Soc. Vaud. Sci. Nat. ser. 3, vol. xxvi (1890-91) Proc.-verb. pp. xii, xvi & ‘Le Léman’ vol. i (1892) pp. 211 e segg. > P. 15 [not p. 45, as stated in the index of that volume]. Vol. 62.] TARNS OF THE CANTON TICINO. 167 all events, he ‘saw difficulties.’ A glance at the geological map of Switzerland shows that these lakes lie along the outcrop of certain Mesozoic limestones. It is also noticeable that, if traced westwards from the Airolo district, other lakes, such as the Engstlensee and the Trubsee, are found to be situated on the outcrop of the same rocks. Furthermore, it is obvious that many of the chief valleys and passes, such as the Val Bedretto and the Nufenen Pass, or again the Urserenthal and the Furka Pass, have been excavated along the strike of these calcareous rocks. In noting these facts, I felt convinced that the occurrence of so many tarns along the outcrop of the same beds as those in which the valleys and passes had been excavated, must point to something more than mere chance coincidence. An especial interest attaches to these lakes, inasmuch as they include the group the origin of which Prof. Bonney has hesitatingly conceded to icc-erosion. I haye, therefore, devoted portions of my summer-holidays during the past few years to their investigation. I began work on the western group of lakes, which lie to the west and north of the Titlis. A preliminary survey showed that no definite conclusions could be reached, without some knowledge of the depth and general form of the lakes; none of these lakes, however, had been sounded, and this work yet remained to be done. One great obstacle presented itself to the immediate accom- plishment of this preliminary step, in the fact that the majority of the tarns are destitute of boats. After consideration, I abandoned the idea of a portable boat, as the lakes are seldom free from wind — after sunrise, and the accuracy of the soundings obtained from a drifting coracle seemed to me to leave a good deal to be desired. Eventually, with the assistance of a grant from the Government- Grant Committee of the Royal Society, I succeeded in constructing a sounding-machine which could be used from the banks without a boat, capable of yielding very accurate results, being self-registering both as to the depth and the positions of the soundings. With this instrument I made a detailed survey of the group of lakes of the Val-Piora district, and later included other important tarns of the Canton Ticino. In the work of sounding these tarns I received great assistance from my friends, Miss Violet Waine- wright, Mr. Haworth Moberly, and Dr. Ernest Kingscote, to each and all of whom I tender my sincere thanks, not only for the results that they accomplished, but for the way in which, by their ever-ready help and interest, they converted a somewhat tedious piece of work into a summer-recreation. In the following pages I have endeavoured to bring together such facts as seem to bear on the origin of these lakes; but:I have not yet had the opportunity of visiting all the tarns in the Canton Ticino, and my remarks must therefore be taken to apply only to those lakes which are actually described. 168 PROF. E. J. GARWOOD ON THE { May 1906, ‘IL Tur Laxes or cnn Vat Piona. [Pl. VI1—topographical map, facing this page. | These lakes lie in a series of hollows in and around the mountain- glen of the Val Piora. This depression has been carved out of a series of limestones and schists, which have been squeezed in between two masses of gneiss—one of these forming the northern wall of the Val Levantina below Airolo, and the other constituting the Val-Cadlimo district on the north. That the majority of these lakes occupy rock-basins is clearly shown in Prof. Bonney’s careful description! I wiil reserve any further general remarks until I have described the lakes in detail, when we shall be in a better position to discuss the evolution of the district as a whole. Lago Ritom (see Pls. VII, VIII, & XVI) is situated exactly 6000 feet above sea-level, and occupies a true rock-basin. It has a length of about 2190 yards, and a width (at the upper end) of some 590 yards, becoming narrower towards the exit. It is fed chiefly by three streams, namely, the Murinascio torrent, draining the Val Piora, and the two streams carrying the overflow from Lago Tom and Lago Cadagno respectively. These last-named lakes receive the drainage from two unnamed tarns, to which, for conve- nience, I shall refer as Lago Taneda and Lago di Murinascio. The former (namely, Lago Taneda), which lies on the watershed, drains southwards through the Lago Cadagno, Jago di Murinascio has no true exit; any surplus-water that there may be finds its way through the screes which form the lowest point of its containing barrier. Lago Ritom must once have extended nearly 1000 yards farther to the north-east, over the area now occupied by a delta which forms a swampy alluvial flat; a delta which, from its peculiar situation, might well be used to ascertain the length of time that has elapsed since the close of the Glacial Period in this district. The southern shore of the lake is formed of coarse foliated gneiss, dipping north-north-westward at about 35°; while the northern bank runs along the foot of a steep escarpment of calcareous crystalline schists—a complicated series, much folded, contorted, and crushed. At the south-eastern corner of the delta, outcrops of rauchwacke and dolomite occur; and the same beds must form the floor of the lake throughout its length, for they reappear at its western end, where they are found cropping out close to the Hotel, while beds of a similar calcareous character occupy most of that end of the lake in places up to the 1980-metre contour. As this fact is not at all clear from Dr. K. von Fritsch’s map, where only a small patch of rauchwacke is shown, forming the saddle between Pian’ Alto and Fongia, a brief description of the rocks that occupy this end of the lake is necessary. Starting from the exit of the lake, we find the water of the Foos running out over the southern mass of gneiss. This rock can be 1 Geol. Mag. 1898, pp. 15 et segq. Quart. Journ. Geol. Soc. Vol. LXII, Pl. VII. [:42240. The Val Ticino lies just outside the map,below it; & the Val Canaria (pressed in metres. lies also just outside the map, sensibly parallel withits Western margin. ae fat Cede li nos | ae ge 310 a ae ’ Stabbio nnOVO —_Cagsina Rihvet7y 2280 Ze r = — - mers pes = z t= -Stabbio drpMezzo ~R eno di-M-ed @7 y, = aa, 2) ea : = = aE So 266: as Gnd Vier Oi: ~ P del Vomoy 7 Nya Hl etd, PRT Pp} Augie 2) 4 I. TI <27504 : 29 73/4 R —~F1ane Marina s ciO-R 2 eh Li °° . f rae S 7 ALTOS eee : tj 20". Pettano HE PiorA LAKES. e extreme left-hand corner of the map, between Pian alto and Fongio. Aiso bove the numerals of aititude 2013, north of the Murinascio stream. | : [2) Ty tical IGE) Quart, Journ. Geol. Soc, Vol. LXII, Pl. VII. 2542 Bocca di Cadlimo rome Natural Scale: 1:42240. The Val Ticino lies just outside the map,below it; & the Val Canaria D &. ; . se S " Altitudes are expressed in metres. lies also just outside the map, sensibly parallel withits Western margin. ¢ fens tery = SEO, Cassina Ag stabio dr Ry, CH 4 laces ina di Cimoehe Sve SLA =n = ST asa aren Fon; oy) Mar or tHe Prora Laxgs. [A triangular patch of rauchwacke should haye been indicated in the extreme left-hand corner of the map, between Pian alto and Fongio. Aiso the occurrence of gypsum should haye been indicated just above the numerals of aititude 2013, north of the Murinascio stream. ] Cpe anu: Le eae ee Vol. 62.) | TARNS OF THE CANTON TICINO. 169 followed for a few yards past the Hotel to a point where the path to Pian’ Alto turns up to the left. Here the gneiss dips under the end of the lake, and is not again seen; north of this no outcrop of rock in place is met with, until we reach a conspicuous gully which ends in a well-marked delta, the broad depression that intervenes being obscured by fallen talus. If, however, we turn up the path to Pian’ Alto, we find that the gneiss is immediately overlain by a bed of pale pink rauchwacke, closely resembling in texture the rock which crops out at the south-eastern end of the lake. The junction between the gneiss and the rauchwacke runs almost along the track leading to the little knoll behind the Hotel, marked 1930 m. in the Swiss Government map (see Pl. VII); in fact, the track has taken this line up the dip-slope of the gneiss, on account of the depression formed by the unequal weathering at the junction of these two rocks. From here the outcrop of the calcareous rock can be traced northwards, across the scree-covered depression, to a conspicuous bluff occupied by a small plantation of nine fir-trees. Here the rock is rather lighter in colour and much honeycombed, and closely resembles the rock surrounding the exit- end of Lake Tom. Masses of this rock are found in the gully- section farther north; and it crops out again farther on, by the side of the path running round the north-western edge of the lake (see Pl. XIII, fig. 1). ~ If we now follow the upper portion of the outcrop, south from the tree-covered bluff, we find a conspicuous band of the same rock, 40 to 50 feet thick, running for some distance round the eastern flank of Fongio, behind the 1930-metre hill, and dipping at some 6° to 10° northwards in the neighbourhood of the 1980-metre contour-line. After continuing for some distance in this direction, it ceases abruptly. It is found, however, a few yards below, on the path running round the hill into the Val Canaria. An analysis of this band is given below (p. 177). The rock is everywhere micaceous, and in places contains large tufts of actinolite, which appear to be arranged roughly along certain planes. A good section, showing nests of actinolite-crystals, is exposed in the little quarry at the back of the Hotel. It will be seen, then, that from the edge of the gneiss the western end of Lake Ritom is composed of calcareous rocks, up to a height of 1980 metres. The abrupt disappearance of this mass at either end of the outcrop is very suggestive ; and we may, I think, regard the area occupied by Lake Ritom as a large lenticle of calcareous rock, possibly including a bed of gypsum, enclosed between the gneiss on the south and the crystalline schists on the north. An interesting feature is brought out by a comparison of the northern and southern shore-lines of the lake. The former, dis- regarding the scree and delta-material, is seen to run in a straight line along the length of the lake ; the latter, on the other hand, is broken by numerous bays and rocky headlands (see Pl. VIII). The surface of the southern mass of gneiss, as shown above, can 170 PROF, E, J. GARWOOD ON THE [May 1906, be traced westwards under the calcareous rocks on the north side of the 1930-metre hill. It represents therefore almost exactly the former line of junction of this gneiss with the overlying rauchwacke, now removed by denudation over the area occupied by Lake Ritom; and the existence of the bays and headlands seems in some cases to have been determined by folds or crumplings, which must once have affected both sets of rocks, as will appear by an inspection of the bathymetrical contours. Some of the bays, however, are due to stream-erosion, as shown by the abruptly-truncated strike. The general shape of this lake-basin is indicated on the bathy- metrical chart(Pl.X VI). The greatest depth met with was 163 feet, in the centre of the lake. The greater part of the lake is over 100 feet deep; while an area, about 330 yards in length and 150 feet or more in depth, occupies what must originally have been the centre of the lake, if we allow for the portion at the upper end which is now filled with delta-material. The floor of Lago Ritom is thus seen to form a very symmetrical trough, the axis of greatest depth running east and west, practically down the middle of the lake. The contours near the southern side are influenced by the subaqueous folds of the gneiss for some distance from the shore. The contours along the northern bank bear no particular relation to the subaérial escarpment, being chiefly influenced by landslips and rock-falls from the steep cliffs above: were it not for the loose material thus accumulated, the northern shore would undoubtedly descend much more steeply than it now does. Lago Tom (see Pls. IX, X, XII, & XVII).—This lake lies at an elevation of 6637 feet above the sea. It occupies a hollow directly above, and to the north of, Lago Ritom. In shape it is a rough oval. Its greatest length from north to south is exactly 550 yards, and its greatest width from east to west about 350 yards. Prof. Bonney remarks in a footnote (op. cit. p. 19): ‘Its area is given as 1000 square metres, and it is said to be shallow. But L should think it would not be less than some 20 feet deep, and might be more.’ The greatest depth that I obtained was exactly 50 feet. The lake, again to quote Prof. Bonney : ‘occupies a kind of cirque, and lies in the strike of the same rocks [as Lago Cadagno], for the enclosing crags consist of similar amphibolitic and granatiferous gneiss, and at the lower end is rauchwacke, which can be traced from the southern side of the basin of the Lago Cadagno across the intermediate spur.’ The lake terminates at its southern end in a wall of rauchwacke, about 16 feet high. A short gorge leads into the heart of the rauchwacke, near the eastern end of this wall; and after heavy rain, when the lake is high, I have seen the water flowing through this channel and disappearing underground. It reappears as a spring farther west, on the south side of the barrier of rauchwacke, exactly at the junction of that rock with the underlying schist. After a spell of fine weather the level of the lake falls so that the Vol. 62.) — TARNS OF THE CANTON TICINO. I71 water does not reach the channel, and a broad, flat beach is laid bare at the south-western end of the lake; the spring, however, continues to issue at the same spot. This ‘terminal’ wall of rauchwacke is rotten and honeycombed, and just above the spot where the spring now issues is situated a large cave which evidently represents the old source of the spring, when it escaped higher in the rauchwacke (see P]. XII, fig. 1). Apparently it has gradually dissolved its way down to the level of the insoluble schist, against which it is now thrown out. The rock is a fine-grained meta- morphosed limestone containing some 16 per cent. of dolomite, and 8 per cent. of insoluble material, chiefly silica. The bathymetrical chart (Pl. XVII) shows the general shape of the lake-floor. The maximum depth of 50 feet is situated about one-third of the way from the northern end of the lake. The axis of greatest depth runs east and west, and passes through this point. It is parallel to the strike of the rocks, and appears to coincide with the junction of the rauchwacke and the overlying crystalline schist. The shore of the lake is entirely formed of delta-material and sand, with the exception of the southern wall of rauchwacke described above. Lut for this the upper end of the lake would, as in the case of Lago Ritom, undoubtedly show deeper soundings near the northern shore. The flat shallow bay occupying the southern end between the delta and the exit-channel has an average depth of about 3 feet, and is strewn with fragments of minerals, many of them presumably derived from crystalline schists. These coarser fragments must have been brought down by the stream descending between the rauchwacke and the overlying schists; and it is a very suggestive fact that, although fragments of mica clearly derived from the rauchwacke are represented, no pebbles of this rock are found, despite its occurrence here 2m situ round the end of the lake. Lago Cadagno (see Pls. XI & XVIII) lies due east of Lago Tom, at the somewhat lower level of 6802 feet. It is situated apparently on the strike of the same beds (namely, the rauchwacke on the south, and the overlying crystalline schists on the north), and occupies a slight synclinal depression. Itis surrounded by alluvium and scree-débris, with the exception of the north-eastern corner, where solid rock occurs, and the southern bank, where the rock is plastered with morainic material. The lake doubtless once extended to the rocky barrier over which the river now falls into Lago Ritom ; but the western end has been filled up by material washed down from the steep escarpment on the north and north-west. It is possible that it once drained from its south-eastern end. The lake is 875 yards long and about 330 yards in width, the greatest depth being 62 feet. The bathymetrical chart (Pl. XVIII) shews that it is a symmetrical spoon-shaped hollow, haying an axis of greatest depth running east and west, parallel to the length of the lake, situated nearer to the northern than to the southern shore. It is fed by a group of some ten or twelve streams which cascade down the precipitous scarp-face on the north. Most of these originate in 172 PROF, E, J. GARWOOD ON THE [May 1906, patches of snow, which lie throughout the greater part of the year in hollows near the watershed; but one, the lowest, occupying the centre of the cirque, originates in the little Lago Taneda mentioned on p. 168. This cirque is a good example of an escarpment-cirque, formed by a series of streams which derive a constant supply from lakes and melting snow. ‘The water from Cadagno winds through the alluvial and morainic material mentioned on the previous page, to the edge of the rock-escarpment of quartz-calc-garnet schist ; but the fall is very slight, probably not more than a few feet. The rock-basin of Cadagno must, therefore, be at least 50 feet deep. From the edge of the escarpment the water plunges abruptly into the Ritom basin, down a cliff 100 feet high, near the centre of the synclinal fold. Lago Scuro (see Pls, XII & XIX) is undoubtedly one of the most interesting of the lakes in this district. It lies at a height of 8048 feet, and is situated almost on the watershed between the Ticino and the Reno di Medels (or Medelser Rhein), its exact position being some 150 or 160 feet below the summit of the ridge on the northern side. It drains through the Val Cadlimo into the Medelser Rhein, and occupies a true rock-basin, which lies along the junction between the schist and the northern mass of gneiss. It is roughly oval in shape, and has a longer diameter of 404 yards, which runs north-north-east and south-south-west. The only drainage that it receives is derived from the melting snow round its sides, and it discharges its surplus water down a rocky channel into the Cadlimo River. The bathymetrical chart (Pl. XIX) shows it to have a maximum depth of 138 feet. The contours run roughly parallel to the shore, and the axis of greatest depth lies east-north- east and west-south-west, in the general direction of the junction of the gneiss and the crystalline schists. The shore is entirely formed by rock 2m situ with occasional loose blocks, while a patch of snow permanently occupies a portion of the eastern bank. The dip of this schist is about 40° northward, but it can be seen sweeping round in a basin-like curve across the southern end of the lake. Soundings show that ridges of rock continue under water for some distance out, and thin bands of harder material have weathered out so sharply as to cause considerable trouble in using the sounding- line. Lago Taneda (for contour-map, see Pl. XX) occupies a rock- basin, lying at a height of 7740 feet, and is situated along the same line of strike as Lago Scuro. According to Dr. Fritsch, who mapped the district in detail, the junction between the northern mass of gneiss and the underlying schist runs down the middle of the lake. The relation between these rock-masses at the exit is not very clear; but the gneiss is well seen, forming the watershed above the north side of the lake, where it appears to be iceworn. The southern shore is formed of a gneissose mica-schist dipping about 40° northward; but immediately below this, on the track leading Vol. 62.] . TARNS OF THE CANTON TICINO. 1738 down to Lago Cadagno, an outcrop of black hornblende-schist occurs. The rocks forming the shores of the lake are jagged and broken, and show no signs of ice-abrasion. ‘The cliff of gneiss above the northern and eastern shores is interrupted and broken down towards the north-west, where a low col leads over into the Val Cadlimo. The lowest point of this col hes some 150 feet above the present level of the lake. Another, but less well-marked depression, leads from the spring on the north-western shore to a small pool, which drains into the Lago Lisera in the Val Cadlimo. Lago Taneda is, therefore, situated as nearly along the watershed as may be, and itis probable that it drained at some former time over one or both of these gaps into the Val Cadlimo; possibly during a temporary obstruction at the present outlet, or before it had been cut down to its present depth. The greatest length of the lake from east-south-east to west- north-west, along the line of strike, is 490 yards; and the greatest depth is 140 feet, practically the same as that of Lago Scuro. The bathymetrical chart (Pl. XX) shows that the greatest depression is situated towards the western end of the lake, while the narrower south-eastern end forms a shallow bay. There is, however, a general axis of greatest depth running along the length of the lake parallel to the strike of the rocks, which probably coincides with the junction between the gneiss and the schistose rocks. Lago di Murinascio, alone among the tarns in this district, is situated wholly in one set of beds. It lies at a height of 7562 feet, and occupies a corrie in the northern wall of the Val Piora, above the Murinascio Alp. It belongs to the upper group of lakes represented by Lago Taneda and Lago Scuro. The lake is not entirely rock- bound, the water at the lower end being held up by scree-material. No soundings were made, as this lake is not a true rock-basin, and there is no exit, A ridge of rock, however, forms the shore on the south side, and it may possibly be a shallow rock-basin. Lago Lucomagno, situated in a corrie on the south-western flank of the peak of that name, is partly scree-dammed, but must be a rock-basin some 40 feet deep. It drains directly into the Ticino below Varengo, but in this case, as in that of Lago Tom, there is no overflow. A spring (which issues some 500 feet below) appears, however, to derive its supply from this lake. The lake is almost divided into two separate sheets of water by a narrow peninsula. The water at the lower end is partly held up by rock-falls and screes, and partly by rock wm situ, and is situated along a strike-ledge in the gneiss. The depression which the lake occupies can be seen continuing as a line of weakness to the east, where it is weathered out into a gully in the hillside. A little tarn, not shown on the Government map, occurs on the northern slope of the same mountain, but it is obviously due to a landslip, and does not require our further attention. Q.J.G.S8. No. 246. C 174 PROF. E. J. GARWOOD ON THE [May 1906, Lago Camogheée, lying on the western slope of the peak of that name, is another example of a scree-dammed tarn. Its formation may, however, have been influenced by the occurrence of a bed of rauchwacke, which occupies the pass between Pian’ Alto and the Cima di Camoghé. This rock, although not so shown in Dr. K. von Fritsch’s map, appears to strike round the south-western end of this lake, and reappears at the head of the gully at a point where the spring issues a little below the Alpe di Lago. The Origin of the Piora Lakes. So far as I am aware, only one writer on this district has expressed any opinion regarding the origin of these lakes, namely Prof. Bonney, in his paper in the ‘ Geological Magazine’ for 1898, already quoted.* In discussing the possibility of their formation by ice-erosion, he does so, evidently, with a desire to give all due weight to a theory which is perhaps that most generally held in regard to the formation of lakes of this class ; but he is obviously not quite satisfied, himself, with the application of the theory to these par- ticular lakes. I will, in the first place, point out the serious difficulties in the way of accepting this theory, and afterwards suggest what seems to me a more probable mode of origin. Let us first take Lago Ritom. In the case of this lake, there are only three directions in which ice can reasonably be assumed to have travelled, namely: (1) from the depression between Pian’ Alto and Fongio at the western end of the lake; (2) from the district on the north, now drained by Lago Tom; (3) from the Val Piora on the east, including possibly a tributary from the cirque now occupied by Lago Cadagno (see Pls. VII & XI). in the first case, there would be no sufficient gathering-ground inside the Ritom drainage-area, and ice advancing from this quarter must have originated outside the district and overtopped the western watershed; but we have no evidence of an accumulation of ice in the district to this depth, and had the ice in the Val Ticino reached to this height, it would have invaded the Ritom basin much more readily from the east through the present exit of the lake, which is nearly 1000 feet lower. The second case postulates a glacier descending from Lago Tom. This could not possibly have excavated the upper end of Lago Ritom, and consequently this supposition may be at once dismissed. There remains only the third alternative, namely that of a glacier descending the Val Piora, augmented possibly by ice from the Lago- Cadagno basin. ‘There can be no doubt, I think, that this district lay above the snow-line during the Pleistocene Period, and that snow and ice accumulated here and occupied the Val Piora. My difficulty lies in understanding how this ice could acquire any velocity in moving westward aloug the basin now occupied by Lago Ritom, and why it should have turned off at right angles and discharged itself to the south-east, unless these lines of drainage had been previously determined. 1 But see also A. Delebecque, Comptes Rendus Acad. Sci. Paris, vol. exxxix (1904) pp. 936 et segg. Vol. 62.| TARNS OF THE CANTON TICINO. 175 Prof. Bonney remarks (Geol. Mag. 1898, p. 20) :— ‘Supposing a glacier to be descending the Piora valley, we must assume this wall [at the north-eastern end of Lago Ritom] to be already in existence, or it could not acquire any plunging force, and even then the fall seems hardly adequate to produce the erosion of a basin like that of Lake Ritom. Possibly, however, the ice, just at this part, may have been “ jammed” ; for the main glacier was probably augmented by another ice-stream, which descended a shallow, but fairly well-marked valley, leading from a gneissic peak lying south-east down to the corresponding corner at the head of the Ritom basin ; while the narrow “sate” by which the water is now discharged towards the Val Bedretto would block the mass of ice above it, and this would produce more than usual friction on the bed of the valley now occupied by the lake and its delta. This basin then, the part which lies below the present contour-line of 6000 feet (in round numbers), is the utmost that, in my opinion, can possibly be attributed to the erosive action of ice. Of this action all the other dominant features in the surrounding scenery exhibit nothing more than superficial traces, and they appear to be due to the usual meteoric agencies.’ It is this probability of the ice becoming ‘jammed’ at the western end of Lago Ritom which would, I think, prevent it from excavating to any appreciable extent. In a former paper, read before this Society by Prof. Gregory and myself,’ we showed that where the lower layers of a glacier are embayed, their advance is practically checked, and shearing of the upper layers takes place over those below. Under these circumstances the greatest erosion, in the case that we are considering, would be on the surface of the rock-lip near the present exit of Lago Ritom, and it is precisely here where we find the most obvious indication of former glacial abrasion. Such erosion would, therefore, tend to destroy a rock- basin, not to produce one. Again, the presence of therocky headlands, which have been described as occurring along the south side of the lake, jutting out to the north-west, should have protected the rauchwacke, originally occupying some of the present low bays, from ice coming down from the east; but not a trace of the rauchwacke is to be found there. So far, therefore, as Lago Ritom is concerned, it does not seem easy to account for the excavation of the basin which it occupies on the supposition that it was ice-eroded. If we turn to the other lakes, Tom and Cadagno, the difficulty is increased. It is obvious that these lakes could not have been excavated by a glacier descending the Val Piora, and no glaciers could have formed on the nearly-vertical cliffs which hem them in to the north. If we postulate ice, other than local accumulations of snow, we must, as shown above, bring it from outside the basin. In this case it must have come over a watershed 8200 feet high, and there is no proof of the country having been swathed in ice to this enormous depth. But, supposing such ice to have poured down from the heights to the west, why, after digging out Lago Tom, did it ascend the slope of rauchwacke between Lago Tom and Lago Cadagno, only to dig Lago Cadagno again, out of the same material and along the strike of the same rocks ? Or, again, assuming that the ice poured over the watershed, down the precipices which hem in the lakes on the north, it could only have arrived at the bottom in fragments, and could at the most have contributed to a remanié glacier. + Quart. Journ. Geol. Soc. vol. liv (1898) pp. 203 et segq. 02 176 PROF, E. J. GARWOOD ON THE [May 1906, Lastly, if the ice came from the north, how was the Lago Scuro excavated, or the Lago Taneda, which is entirely shut in by pre- cipices on the north? It is, indeed, extremely difficult to account for the excavation of these last-named tarns by ice, evenif we attribute to that agent the utmost power that has ever been claimed for it. Let us take the Lago Scuro as a crucial case. Here is a rock- basin 136 feet deep, situated almost on a main watershed, within a few metres of its summit, and surrounded on three sides by steep cliffs. There is no gathering-ground above it for a glacier, and it is thoroughly protected from any ice invading the region from outside: yet we find a rock-basin of this considerable depth. I need hardly add that the steep reefs of rock running into the lake are quite inconsistent with ice-erosion. The case of Lago Taneda is still more difficult. No ice could possibly reach this lake from anywhere except a little névé from Taneda, and yet it is a deep rock-basin. Geological Structure.—Putting aside, then, the theory of ice-erosion, we must seek for some other reason for the existence of these lakes. Let us first of all consider, in rather more detail than we have yet done, the geological structure of the district. All the rock-basins of this group, as shown above, are situated along the junction of two distinct types of rock. The three lower lakes lie along the strike of two nearly-parallel outcrops of rauch- wacke and dolomite, and occur at the junction of these rocks with gneiss and crystalline schists. A glance at Dr. K. von Fritsch’s map shows, not only the lower lakes, but aiso the valley of Piora running along the outcrop of the rauchwacke and dolomite, while the same rocks occupy the two passes leading into the Val Canaria between Fongio, Pian’ Alto, and the Cima di Camoghée. We have, therefore, in this district differential erosion depending on differences in geo- logical structure, being greatest along the rocks rich in calcite and dolomite. In addition to*‘this) the soundings show that the axes of greatest depth of all the lakes lie along the junction of calcareous and erystalline rocks. The causes which produced the peculiar geological structure of the district still remain to be considered. In his well-known paper, ‘On the Crystalline Schists & their Relation to the Mesozoic Rocks in the Lepontine Alps’,’ Prof. Bonney has argued against the inclusion of the rauchwacke and dolomite in the same metamorphic series as the crystalline schists occurring above and between them. He regards them as an entirely-separate group of rocks, introduced among the crystalline series by faults or thrusts, and rejects the idea suggested in the sections accompanying Dr. K. von Fritsch’s map, that their present relation to the schists is the result of simple folds, which affected a stratigraphically-continuous series of beds. After my first visit to the district, and before I had seen Prof. Bonney’s paper, I had arrived at precisely the same conclusion; a conclusion which has been fully corroborated by subsequent ex- 1 Quart. Journ. Geol, Soc. vol. xlvi (1890) p. 187. Vol. 62.] TARNS OF THE CANTON TICINO. Wis amination. I found it no easy task to locate the exact dip and horizon of these planes of movement: in fact, the beds have been more crushed and rolled out than even Prof. Bonney himself, I think, imagined, Not only are there undoubted lines of movement between the crystalline rocks and the beds of rauchwacke, but, for some distance trom the schists, fragments of the latter rock have been torn off and incorporated in the calcareous beds. In the lower part of the schists again, between the two lakes, eyes of ‘calcite and dolomite have been rolled out and included in the crystalline schists ; and it is possible to collect from the western end of Lago Ritom a series of specimens, ranging from pure rauchwacke to a foliated rock consisting of alternate layers of rauchwacke and schist. The whole process must have heen analogous to what I have seen taking place in the case of advancing Arctic glaciers. Here not only are thrust- planes formed by the continual retardation of the lower layers of the ice, but, owing to the presence of shear-planes, fragments of the underlying material are dragged out along them and arranged in a roughly-stratified manner throughout the ice; many of these are even raised by this process for considerable distances above their source of origin. We may, I think, then conclude that, so far as the valleys in which the lakes lie are concerned, their formation has been deter- mined by the presence of thrust-planes, and that, in the case of the valleys in which the three larger lakes occur, the effect of these thrusts was emphasized by the introduction of wedges of softer rock. The overdeepened portions of the valleys in which the lakes occur still, however, remain to be accounted for. They cannot, of course, be due to ordinary mechanical erosion by the present stream ; but the special character of the rocks seems to suggest that the chemical action of water may have manifested itself in an unusual degree, owing to the relatively-greater solubility of the calcareous rocks. Analyses of three samples of the rauchwacke have kindly been made for me at University College (London) by Miss Edith Goodyear, B.Sc. These were collected from the barriers of rauchwacke rising from - the southern end of Lago Tom and the western end of Lago Ritom respectively. I have added for comparison a fourth analysis, made by Dr. Grubenmann, of the same dolomite where it crosses the Val Canaria.’ it II. III. IV. COM MeL Na 39-21 39:30 41:88 44-96 Croat et at 45‘61 45°41 41:90 40:12 MgO vesessneesneeise 4-10 3:83 8:80 11:96 On eens Sek gejit: - 1:19 ca ws TEC) Vena 1-34 2-01 1-28 0-42 FeO Insoluble residue ...... 8:10 787 6:04 1:96 Totals ...... 98:36 99-61 99:90 . 99-42 No. II is from the broken-down underground channel described above. It was especially selected, on account of the relatively-large amount of green mica which it contained ; while No. I is an average sample of the rauchwacke. No. IIT was collected from the cliff of rauchwacke cropping out on the east side of Fongio. ' See Quart. Journ. Geol. Soc. vol. xlvi (1890) p. 230. 178 PROF. E, J. GARWOOD ON THE [May 1906, The solubility of the rock is shown by its crumbly and cavernous character where it is exposed at the surface (see Pl. XII, fig. 1), and by the fact, mentioned above, that the surplus-water from Lago Tom, instead of finding an exit over the rauchwacke, dissolves its way underground and reappears exactly at the junction of the rauchwacke with the underlying insoluble schist. Experiments prove that the rock is rapidly disintegrated by water containing carbon-dioxide in solution. Miss Goodyear finds:— . ' Per cent. Soluble in cold dilute hydrochloric acid ...... 753 Soluble in hot dilute hydrochloric acid ...... 92 Solmmlemm ft /O.- CO; cee uisstamemdesntadeo stern 11 in 45 hours. Assuming the difference between the solubility in hot and cold acid to be due to dolomite, the composition of the rock would be as follows :— CaCO x shart ee eowe es 734 Dolomites; sc hee eee 16:2 Ho,O, aaitchecpe citar treet ree 3 Insoluble residue ............ 81 99:0 Microscopic examination shows how this solubility is assisted by the granular structure of the rock, which readily falls to pieces so soon as the crystals of calcite are dissolved. Being impressed, when sounding these lakes, with the possibility of their having originated from solution of the rauchwacke, I dredged several samples of the deposits from the floor of Lago Tom and Lago Cadagno. ‘These were taken from their southern sides, where the deposit overlies the outcrop of the rauchwacke. They consist, in both cases, of coarse fragments of various minerals mixed with a fine white micaceous sand. The coarser fragments are composed of quartz, garnet, actinolite, and mica, derived from the crystalline schists above, and washed into the lakes by streams from the north. The finer material, after sifting through a sieve of 80 meshes to the inch, consisted of little rounded grey grains, together with a féw flakes of mica, etc. On boiling in dilute acid, the former dissolve readily, and prove to be dolomite- grains, which have fallen on to the floor of the lake after the solution of the calcite and the disintegration of the rauchwacke. This finer material, therefore, represents the less-easily soluble residue of the rauchwacke. The collection of the dolomite-grains into one sieve is due to the wonderfully-uniform grain of the recrystallized rock. With the view of further testing the solubility of this rock, I collected samples of the water from Lago Tom and Lago Cadagno." The samples were collected where the water 1 Taken after a week of very heavy rain. Vol. 62.| TARNS OF THE CANTON TICINO. 179 flowed out of the lakes. These are found to contain ‘total solids :— ea OV VOM se ae.ca.8 ctase eee 0733 gramme pert litre. WbagorCadaeno” .ks.ci.cceeces 126077 - A qualitative analysis showed that the solid salts consisted of lime and magnesia, with a little iron and alumina. In the case of Lago Cadagno especially, a great deal of sulphuretted hydrogen was discharged when the mud was disturbed on the bottom of the lake, and the sounding-line and plummet, when drawn to the shore, always smelt strongly of this gas. The same phenomenon was noticeable in the case of Lago Ritom, but I never noticed the presence of sulphuretted hydrogen in the mud from the floor of Lago Tom. There can be little doubt, I think, that the presence of this gas in such quantities in the deposits of Lago Ritom and Lago Cadagno, and its absence from Lago T'om, must be attributed to the presence of tishes in the two former lakes and their absence from the last- named, While the two former are stocked with trout, all attempts to rear trout in Lago Tom have failed, and this failure is attributed by M. Lombardi to the presence of lime in its waters. [In this connection the investigation of MM. André Delebecque & Ernest Bourcart on the waters of certain Alpine lakes is full of interest.. M. Bourcart finds in the bottom-waters of Lago Ritom 2°365 grammes of dissolved salts per litre, including the following salts :— Granmes. SiO, =O. O10 CaO =P Cont MgO = 01962 SO, ss. 13/67 K,O = 0:0042 Na,O == 1 0:0027 Peval Oe 00012 The surface-water, on the other hand, contains only 0°140 gramme per litre of dissolved salts; and the three affluent streams from Lago Tom, Lago Cadagno, and the Val Piora contain a mean of 0-139 gramme per litre (op. cit. p. 937). The authors are probably right in concluding that the great increase in dissolved salts found in the bottom-waters is due to underground springs, which enter the lake through the soluble rauchwacke. It might be urged that the salts had accumulated at the bottom by degrees, owing to the formation of ice in winter and the consequent concentration of the salts in the surface-water, which would thus have its specific gravity increased, and, being at the same time cooled down, would be compelled to descend. An interesting point in this connection is, however, brought out by the temperature-determinations made by these authors at different * Comptes Rendus Acad. Sci. Paris, vol, exxxix (1904) pp. 936 e¢ segq. 180 PROF, E. J. GARWOOD ON THE [May 1906, depths. They find that anomalies of temperature occur, similar to those discovered by M. Venukoff in the Black Sea. With a surface-temperature of 13°2° C. a gradual decrease takes place until a depth of 33 feet is reached, where a minimum temperature of 5'1° is found. Below this, the temperature gradually increases until the bottom is reached, where, at a depth of 144 feet, the temperature is found to be 6°6°C. This rise of temperature in the bottom-waters of the lake appears to me to point to under- ground springs as the cause of concentration of salts in the lower layers, rather than to the sinking of concentrated surface-water in winter. Another interesting point is brought out by the presence of large quantity of sulphuric acid in the deepest layer of Lago Ritom. The authors attribute the presence of this acid to the solution of gypsum, but there is no bed of gypsum recorded among the rocks which crop out round the lake. It will be seen, however, that the bed of rauchwacke shown on Dr. K. von Fritsch’s map, as running from Pian’ Alto into the Val Canaria, includes large beds of gypsum where it thickens out to the west; and that, if Lago Ritom occupies the site of a similar ‘eye’ of rauchwacke, it is highly probable that this eye includes beds of gypsum as well as rauch- wacke, for it lies along the same line of strike. In this connection, the total absence of sulphuric acid from the waters of Lago Tom is a highly-suggestive tact. The analyses, then, of MM. Delebecque & Bourcart seem to confirm the idea that Lago Ritom owes its origin, in part at all events, to solution. | Although the presence of these lakes appears to depend chiefly on the structural peculiarities described above, accompanied, in some cases at all events, by solution, there is one remarkable feature in the drainage of the district on which I have only briefly touched: I refer to the way in which Lago Ritom is surrounded by hanging valleys. Not only is this the case with all the three streams flowing into it, but it is still more marked in the case of the La Foos torrent, which flows out of the lake, and carries away the whole of the drainage of the Val Piora. This torrent plunges by a series of cascades 2700 feet down into the Ticino River, at an average slope of 1 in 2:6. As I have shown elsewhere, this torrent forms one of a series which drain from hanging valleys into the Ticino as a consequence of the overdeepening of the main valley. While other valleys, such as the Val Canaria, have adjusted themselves to the Ticino, the Val Piora has been prevented from doing so by the presence of the lake, which has arrested erosion by holding up the solid particles on its floor. Let us trace the history of the Piora drainage a step farther back. A glance at a model or map of the district shows that La Foos is obviously not the natural outlet of the present Val-Piora drainage : it is evidently either a ‘stolen’ outlet, due to some disturbance in Wel. 62,)-" TARNS OF THE CANTON TICINO. 181 the original flow, or the relics of an earlier drainage before the present Val Piora came into existence, for it turns off at right angles to the general line of the valley. I think, however, that it flowed over its present exit previous to the formation of the lakes, and that their formation may be connected with the diversion of the drainage into the ‘ Foos’ channel. As the valley is entirely rock-bound, the diversion cannot be due to the blocking of an old outlet at the western end; it must, there- fore, be due to the direct encroachment of the Foos on the Piora watershed, through which it has cut. This can only have happened if the Ticino Valley were more rapidly overdeepened than the valley into which the original Piora river drained. Obviously, therefore, it did not drain directly into the Ticino; but where else could it drain? There are only two answers to this question. One is, that it drained over the col between Fongio and Pian’ Alto, at the western end of Ritom, into the Val Canaria, as suggested by Prof. Bonney ; the other is, that it flowed eastwards, and not westwards, over what is now the Passo del Uomo, and that the drainage to the west of this pass has therefore been reversed. This latter explanation would account, not only for the hanging valley of La Foos, but also for those at the eastern end of Lago Ritom and below Lago Tom. It is probable, however, that there were several stages in this process, which appear to have been as follows :— (1) Before the present depression of the Val Piora existed, the general drainage of the district would be very different from that of the present day, and a stream rising on the Camoghé range may perhaps have fiowed eastwards into the Medelser Rhein in Pliocene times. About this, however, it is difficult to be certain, as no relic of that period remains. (2) The overdeepening of the Ticino Valley, which then appears to have taken place, may have enabled a stream from the Val Canaria to cut back over what is now the Fongio col as far as the head of the present Tom Valley. This diversion, although unconnected with the formation of the lakes, appears to be indicated by the present depression between Fongio and Pian’ Alto. (3) This overdeepening by the Ticino, postulated above, would enable the Foos, which then rose on the northern slope of the Val Ticino, to cut back its head-waters until it captured the drainage of the present Tom Valley. This would naturally result from the circumstance that, whereas the Foos drained directly into the overdeepened Ticino Valley, the stream over the Fongio col drained indirectly into that valley, through the Val Canaria, which, at that time, would not have yet adjusted itself to the Ticino level. Relics of the drainage as it existed at that time are clearly visible at the present day, not only in the Foos Valley draining northwards and southwards from the col between the Punta Nera and the Taneda to the Ticino, but also in the low watershed (not yet completely destroyed) running parallel with this valley from the Taneda, and now separating Lago Tom from Lago Cadagno. (4) This watershed was next gradually cut through by an eastern tributary of La Foos at its weakest point, namely, the point occupied by the rauchwacke along the line of the present Lago Ritom, and this tributary extended farther and farther to the east, to form the present Val Piora. On this supposition, namely that the Foos Valley existed before the Val Piora, it is easy to account for the fact, otherwise difficult of explanation, that the present drainage from ni ve Piora turns at right angles where it discharges itself into the Ticino alley. (5) An uplift of the range to the west, possibly connected with the melting- 182 PROF. E. J. GARWOOD ON THE. [May 1906, off of the ice, caused a diminution in the flow of the drainage in this valley, and the solution of the rauchwacke gradually produced a hollow extending as far as the limits of this calcareous rock. ‘Thus at the present day, as shown above, Lago Ritom is bounded on the north and north-east by steep walls of crystalline schist, while its southern margin is formed by the dip-slope of the underlying gneiss. The same cause may reasonably be assumed to have operated in the case of Lago Tom and also of Lago Cadagno. In this survey of the physical history of the district, I have not included the formation of the two upper lakes situated near the watershed, namely Lago Scuro and Lago Taneda. These lakes are, indeed, difficult to account for on any hypothesis. One fact, however, appears certain, namely, that, to whatever they owe their existence, their occurrence along the junction of two dissimilar rock-masses must have been the prime determining cause. Iii. Tae Laco-Tremorer10 Grovr. This consists really of several groups of tarns scattered along both sides of the southern watershed of the Val Levantina. ‘They lie for the most part along the outcrop of the calcareous schists, in some cases exactly along the junction of these rocks with the gneiss. Lago Tremorgio.—I will take this as a type of the lakes of this group, as it is not only the largest, but, I think, the most interesting, and certainly the deepest, of those occupying the southern slopes of the Val Levantina. It is also one of the lakes enumerated by Prof. Bonney among true rock-basins. It lies at a height of 5996 feet above sea-level, and 2880 feet above the Ticino. It is nearly circular in shape, having a maximum diameter from north-east to south-west of 853 yards, and a width at right-angles to this of 787 yards. The overflow escapes through a narrow waterworn notch at the north-eastern corner, the lower few feet of which appear, as Prof. Bonney remarks, to be artificial, and falls into the Ticino by a series of cascades; in winter, however, this stream is dry. With the exception of this channel, the lake is entirely surrounded by steep hillsides (see Pl. XIV, fig. 1). All round the southern shore the lake is bounded by precipitous rocks, the upper 500 feet forming an inaccessible cliff, the average slope from the water’s edge being 55°, The slopes from the northern and western walls are gentler, but even these show an inclination of 30° to the point on the northern watershed which is marked 2047 m. in the Swiss Government map. The lake therefore lies in a funnel-shaped rock-basin, the rim of which runs about 700 feet above the lake, except where it is broken through at the exit: the average siope of the sides of the funnel being about 45°. In fact, if we consider a funnel having this slope to be tilted towards the north-west, we get an exact representation of the shape of the basin. Prof. Bonney aptly compares it to an armchair, which it certainly resembles, only the arms curve round Vol. 62.] TARNS OF THE CANTON TICINO. 183 -in front till they nearly meet, so that it could only be sat in with the legs crossed. It is obvious that, in considering the origin of the lake, we must take into account the whole of this funnel before it was breached at the exit, as it is probable that the water of the lake once stood at a much higher level before the notch had been eroded and cut down to its present depth. The basin does not lie in the path of any valley, for, with the exception of a small stream coming down a gully from Ponce. Tremorgio on the west, the only drainage that it receives pours over the southern rim of the funnel from the Campolungo Alp. This stream originates in a small lake, apparently a rock-basin, lying high up at 7436 feet above the sea, north of the Campolungo peak. Local tradition has it that this lake is unfathomable, which probably means that it is more than 20 feet deep. The stream from this lake appears originally to have flowed over the Campolungo Pass westwards into the Val Maggia, and to have been afterwards diverted into the lLago-Tremorgio drainage, as the valley was carved out of the dolomite in which the Campolungo Alp lies. The lake cannot, therefore, by any possi- bility be regarded as occupying an overdeepened pre-Glacial valley, and it resembles nothing so much as a large swallow-hole. The greatest depth of the lake is said to be 120 metres (394 feet) near the centre. I give this on the authority of Mr. Fraser, an engineer residing at Lugano, who took soundings from a raft. Unfortunately, my attempts to sound this lake were frustrated on two occasions; the first time by bad weather and fogs, and the second time, after partly completing the soundings, the line was so repeatedly cut by sharp submerged reefs that | was obliged to abandon the endeavour, the time at my disposal preventing a third attempt. The discovery of these reefs, however, is not without bearing on the origin of the lake. From the soundings which I obtained, the depth appears to increase very regularly from the western corner towards the centre; and if this general slope con- tinues, it would give a depth of 250 feet at the centre of the lake. This is not the case, however, near the northern shore, which falls steeply from the outlet. ‘The lake is excavated in the calcareous schists. The reefs running out from the south-western corner contain ‘ eyes’ of massive andalusite with a hardness of 74. The lime in the calcareous schist is chiefly present in the form of eyes of crystalline calcite, which measure frequently several inches across, and crumble at the touch; and, where portions of the rock were obtained from the edge of the lake, these eyes were frequently represented by holes, the calcite having been entirely or partly dissolved out. The lake cannot be considered as the result of glacial erosion. The production of a funnel-shaped hollow 1000 feet deep, with an upward slope of 30° on the exit-side, cannot have been produced by ice-action ; and the origin of the hollow in which the lake lies must, I think, be attributed to chemical solution. Two important facts seem to bear on this hypothesis; one is the way in which the 184 PROF, E. J. GARWOOD ON THE [May 1906, level of the lake sinks several metres below its outlet in winter, and the other is the occurrence of springs (issuing from the precipice below the lake) which continue to flow all the year round. This latter circumstance has been made use of by the engineers of the St. Gotthard Railway, who obtain the water for their locomotives, not from the overflow-torrent from the lake, but from these permanent springs. The solution of the calcareous portions of the schist might cause the funnel-shaped hollow in which the lake lies; but it is also possible that the dolomite which occupies the Campolungo Alp may have been folded in under the lake, as shown in the very suggestive section to the west of the Campolungo Pass (see Pl. XIII, fig. 2). If this fold be continued underground, it must certainly pass beneath the lake. The fact that the outcrop of the dolomite strikes across the schist, and terminates abruptly at either end against the gneiss, only to reappear as an isolated outcrop in the Val Piumogno on the east, is a very suggestive one, as it emphasizes once more the inconstancy of these eyes of dolomite and their fortuitous occurrence among the crystalline rocks. Of the remaining lakes of this group, those scattered along the southern side of the Ticino watershed are the most important. They drain either directly into the Val Magegia, or indirectly through the Val Bavona. The Lago di Naret has a wedge of erystalline limestone running across it, while the Lago Sciundrau is situated on the junction of dolomite and gneiss. The Laghetti or lakelets to the east of Lago di Naret, again, lie along the junction of the calcareous schist and the gneiss; their origin would appear, therefore, to be similar to that of the lakes of the Val Piora. But, without soundings and detailed investigation, it is best to leave their origin an open question for the present, as they will well repay further investigation. IV. Tux St. Gorraarp Laxes. (For topographical map, see p. 186.) At the northern end of the Canton le a small group of lakes near the summit of the St. Gotthard Pass. Of these, the Lago della Sella and the Hospice-tarns drain southwards by the Val Tremola to the Ticino, while the Lago di Lucendro, the Lago Orsirora, and the Lago Orsino supply the headwaters of the infant Reuss. The Hospice-lakes are small rock-pools, which owe their existence partly to artificial means, having been in some cases dammed by the old monks of the Hospice to serve as fishponds, which still yield fine trout. They are also partly due to the highroad em- bankment thrown across the swampy summit of the pass. The portion of the largest lake north of the road is only about 16 feet deep; the southern portion appears to be deeper, but I should think cannot exceed 50 feet in depth.’ Of the others, the Lago di Lucendro is by far the most important. + The deepest sounding met with. Vol. 62. | TARNS OF THE CANTON TICINO, 185 Lago di Lucendro (see Pls. XIV, XV, & XXI).—This lake lies at a height of 6832 feet above sea-level, in the valley between the Piz Lucendro and the valley of the Gotthard Reuss, above Hospenthal. In shape, it is roughly a parallelogram, with the longer diameter running north-east and south-west. The greatest length from east to west is rather more than 765 yards; and the greatest width is 306° yards, the average width being about 270 yards. The water is, to a slight extent, retained by the supports of the bridge at the exit, The greatest depth met with was 126 feet; and the axis of greatest depth, as shown in the bathy- metrical chart (Pl. X XI), lies parallel to the length of the lake, and rather nearer the western than the eastern bank. The prolongation of the contours along this line, into the bay at the western end, is an interesting feature of the chart. The valley in which the lake occurs is occupied at its upper end by the Lucendro Glacier. The drainage of this glacier is carried off by three streams, which form the source of the Reuss and constitute the chief feeders of the lake. ‘This also receives the overflow from the little lake to the north-west, situated at the eastern base of the Piz del Uomo, and several streams which drain the cirque-like wall on the south and west of the lake. The lake occupies a rock-basin, and lies along the junction of the Fibbia gneiss and the biotite-schist, the latter occupying the northern bank. A good section of the schist is seen along the north side of the lake, about 500 yards from the bridge. Here the schist is much crushed and contorted, and penetrated by veins of granite. If the veins are offshoots of the Fibbia gneiss, the lake must lie absolutely along the junction of the two rocks: as in most other cases, however, the actual shore- line is modified by scree and delta-material. With regard to the origin of this lake, there is no evidence of solution having played any part in its formation. On the other hand, the promontory of solid rock jutting out into the lake from the exit does not seem consistent with excavation by ice, as usually understood ; and the presence of this lake along the junction of a massive granitic gneiss and a mica-schist points undoubtedly to the conclusion that here again the main determining factor for the existence of the lake has been the occurrence of a line of weakness along the contact between a hard and a comparatively-soft rock, giving rise to differential weathering. The whole district, however, has been intensely glaciated, and a prolongation of the Lucendro Glacier must formerly have passed over the site of the lake. It is quite possible that this glacier removed the weathered surface of the disintegrated schist, while it made but slight impression on the harder granite. The uneven surface thus produced would cause the accumulation of water, so soon as the glacier withdréw from the lower part of the valley. The origin of Lago Orsirora and Lago Orsino, on the north of Lago di Lucendro, appears to be due to the same cause, as they lie along the junction of the gneiss on the south and the granulite on . ‘Aavpunog [eucjury ot[} so}vorput our] uayoaq siz, | ‘So1jJOUI Ul possoidxe o1e Sopnyiyy ‘ CES OUT ZT ’ S 1 anliosdtog rp edly BhLE \Q / vIqqry P| < . = 8696 o>) 0}J9YSVT Ip ‘ouog Sea ss pod yw isp . a) SSB IAqIO As xt a ier 7 Sdeoeee VuUes eTJapySeg bg od Sess YE i L VSOLT 9 Sic, ; cal YO AL ; mes a OULON [ep ZIT ,-¢ 4, = om x N = S ~ aes a . ‘ (IOIISIO S TLOTISIO Ted - Reon ane se *s, S1oquelg eg eeemen ts. FORE 9I88% *s, OuIsMC OD = aw ae U0” TPet ace ti AooyZ.1v.a\ tog *. V4 DS 7 GSLGS 7 A o 84 6696 20° Seev ssovyiny ee OUISI().ZIg ~o- ULOYIOPUL AA, 419 oy fo dozy / 00°09 7 fo aqpos oy) U0 *say) putDYpI0 Wel 62.) THE TARNS OF THE CANTON TICINO. 187 the north, while the smaller tarns are obviously dammed by loose material. Lago della Sella.—tThis lake forms an irregular sheet of water lying in the Val Torta, east of Monte Prosa, at a height of 7320 feet. It consists of a shallow lake on the north and a smaller pond on the south, united by a rocky ditch about 180 yards long,— the whole giving the idea of a flooded river-system (see Pl. XV, fig. 2). The actual exit is dammed by a landslip, although the lower sheet is certainly a rock-basin. ‘The northern and western shores of the main lake are formed of gneiss, while biotite-schist. occupies the southern and eastern banks. This schist resembles that which forms the northern bank of Lago di Lucendro, and is penetrated by similar granite-veins. The strike of the schist is north-east and south-west, parallel with the longer axis of the lake, the dip being some 60° north-westward. The schist forms a rocky ridge running into the southern end of the lake. This rises to the surface again in two islands, united under water by shallow depressions only 12 feet deep. Five rows of soundings were run across various portions of the lake, but the presence of the islands made a systematic survey difficult in the absence of aboat. The greatest depth met with was 25 feet, though it is possible that greater depths may exist near the upper end of the lake. The origin of the lake, however, seems to be due to precisely the same causes as Lago di Lucendro, namely, a line of junction between gneiss and biotite-schist along which unequal weathering has taken place, the more weathered portions having been removed by the ice, which must once have gathered in the Val Torta and flowed over the district now occupied by the lake. It is noteworthy, in this connection, that the stream now flowing into the lake from the Val Torta runs for over a mile exactly along the junction of the gneiss and the schist. V. Tue Laeo v’Etto.! Although situated just beyond the political boundary, this lake belongs geographically to the Ticino basin. It lies at a height of 3025 feet, or 2578 feet above the Lago Maggiore, and is situated on the watershed between the Val Molino and the Val Vedasco. It occupies an elongated triangular depression in the gneiss between Monte Borgna and Monte Cadrigna. The lake has a length of 930 yards from north-north-east to south-south-west, and a width at the upper end of 300 yards, which forms the base of the triangle: from here it gradually tapers to the exit, which lies in the acute angle of the triangle at its southern end. The deepest soundings were met with about one-third of the distance from the upper end, the greatest depth being 134 feet, while depths of 74 and 75 feet were ‘ A bathymetrical chart of this lake has been constructed; but it is unnecessary to introduce it here, as the lake has not been proved to be a rock- basin. 188 PROF, E. J. GARWOOD ON THE [ May 1906, determined near the northern shore. The lake is, therefore, deepest towards its upper end, and shallows gradually to the exit. ‘The northern shore is chiefly composed of loose material and boggy swamp, through which rise here and there hummocks of solid rock. The shape of the lake and the subaqueous contours, as well as the presence of loose material at the upper end, point to damming of a tributary valley by a landslip or moraine, and the reversal of the drainage over a gap in the watershed at its upper end. This view is borne out by the character of the exit, and also by the presence of a deep valley just beyond the accumulation at the northern end of the lake, into which it would apparently drain if the loose material were removed. There is, however, just a possibility that the lake may be to some extent rock-dammed. VI. GreneRat Conciusions. The results arrived at by a detailed investigation of the tarns in — the northern. portion of the Canton Ticino may, then, be summarized as follows. The lakes can, for convenience, be grouped under two heads :— (I) Those (a) entirely, or (0) partly dammed by loose material. (a) Lago Camoghé. (6) Lago di Murinascio. » Lisera. ,, Lucomagno. » Pettano, », Orsirora. d’Elio. », Orsino. St. Gotthard tarns. (11) True rock-basins due to (c) solution, (d) differential weathering. (c) Lago Ritom, (d) Lago Scuro. Pe Ronn, . Taneda. ,, Cadagno. , di Lucendro. , Lremorgio. , della Sella. Those grouped under I (a) require no comment. Those under I (6), in so far as they are rock-basins, may possibly be due to differential weathering, either by frost or by ice. It is, however, difficult to ascertain what depth, if any, we are to consider as rock-basins in each case, but it cannot be a very great depth. Lago di Murinascio and Lago Lucomagno lie upon ledges dipping steeply into the scarp-face at the foot of precipitous cliffs, above which there is no possible gathering-ground for a glacier, unless the ice poured over the cliffs in each case from the north ; but the cliffs are too steep for it to arrive, even then, in anything but a fragmentary state. Consequently, even Prof. Bonney’s general concession, that the ice, as it descended from the ranges above, “would impinge on the level floor, on which under these circumstances it might have some erosive force,’ (Geol. Mag. 1898, p. 19.) does not seem applicable here, as the corries would be filled with snow and protected from any bombardment of this nature. It is, of course, always possible that the slight irregular erosion of a few feet might have been produced by the ice collected in the cirque, as Wol.iG2) |Pn TARNS OF THE CANTON TICINO. 189 assumed by some authors to account for the formation of corrie- lakes in general; but it is difficult to realize the mechanics of the process, unless the ice found in these corries behaves rather differ- ently from that of a glacier, and moves as a solid mass and not differentially as in a glacier. The lakes under notice are, however, comparatively unimportant; but, if any part of their basin is truly rockbound, they are difficult to account for, as they occur entirely in one rock-formation. Rock-Basins. II (c).—The chief lakes of the Val Piora are shown to lie along the junction of soluble calcareous rocks with gneiss or schists. Detailed soundings prove that the axes of greatest depth of these lakes coincide very closely with these lines of junction. Although the district must have been below the snow-line in Glacial times, there is very little sign of glacial action visible at the present day, and it is difficult to understand how ice can have taken any real part in the formation of the lakes. Chemical evidence derived from analyses of the rauchwacke and from the composition of the water of the lakes, points strongly to solution having played a conspicuous part in the formation of these lakes. This is confirmed by the fact that, in the case of Lago Tom, the surplus-water does not overflow the containing barrier of rauchwacke at the exit-end, but, percolating into the heart of the limestone, reappears as a spring which rises exactly at the junction of the rauchwacke with the underlying schist. Other evidence, such as the high percentage of dolomite-grains in the fine deposit forming the floor of the lake near the exit, evidently representing the less soluble portion of the rauchwacke, points in the same direction. Inthe case of Lago Ritom, itisshown that the whole of the western end of the lake is occupied by rauch- wacke, although not so shown on the Swiss Geological Survey-map, and that the lake may be considered to occupy the site of a lenticular thickening of the calcareous beds similar to those which occur west and east of it along the same general strike. The marked increase in the amount of dissolved salts in the bottom- water of this lake, coupled with an equally-marked rise in tempera- ture, are facts which point strongly to underground solution in the form of subaqueous springs ; while the high percentage of sulphuric acid can only be explained on the supposition that beds of gypsum occur under this basin, similar to those found along the outcrop on the west and east. Lago T'remorgio appears, from its characteristic pothole shape, to be also due to solution. Jtis situated on calcareous mica-schist, andthe eyes of calcite in the schist are dissolved out from the rocks bordering the edge of the lake. It is also possible that the fold of dolomite, visible in the Campolungo Pass above the lake, may continue downwards to the lake-floor ; and, if this should be the case, it would amply account for the basin by solution. The fall of the Q.J.G.8. No. 246. P 190 PROF. E. J. GARWOOD ON THE [May 1906, surface of the lake several feet below the exit in winter, and ‘the continuous flow of springs which issue all the year round from the hillside below the lake, point also to solution as the probable origin of this lake. It does not appear to be possible to attribute this lake to glacial excavation, as it does not form part of a valley- system which could have been overdeepened by ice to produce the present basin. | II (d).—There remain four important rock-basins which are more difficult to explain. They resemble the foregoing group in the fact that they all lie along the junction of two different rock-types, so that, whatever the weathering agent may have been which finally produced the basin, the occurrence of the lakes in their present position seems certainly here again to have been determined by structural lines of weakness. Lago di Lucendro lies in the track of an extension of the Lucendro Glacier, and Lago della Sella also occupies a valley at the head of which a glacier-system still exists. It is possible, therefore, that differential weathering in pre-Glacial or inter-Glacial times, along the junction between the gneiss and the biotite-schist, may have caused an unequal removal of material by ice in Glacial times. This excavation does not appear to have been so much in the nature of digging, as of the removal of less resistant, possibly more weathered, material. This is shown by the marked convexity of the outlines of the gneiss even under water, by the presence of the rocky headland at the exit of Lago di Lucendro, and also by the occurrence of the islands of solid rock running through Lago della Sella. The two remaining lakes of this group, namely Lago Scuro and Lago Taneda, which occur within a stone’s throw of the main watershed and occupy basins 140 feet deep, are certainly very puzzling: for, although they lie along the junction of schist and gneiss, the theory of solution does not appear to be applicable here. Had these lakes lain in glaciated valleys, they would have undoubtedly been considered by many geologists as typical ice-excavated basins. Their occurrence, how- ever, a few metres below the watershed, as well: as the presence of submerged reefs running round the southern end of Lago Scuro, seems to preclude the possibility of excavation by ice. The lakes, then, of the Canton Ticino that have so far been examined, with the possible exception of the rock-pools of the St. Gotthard Hospice, do not seem to be due to ice-erosion in the generally-understood sense, if we adopt the most recent definition by Prof. James Geikie for this class of lake, as contained in the following statement : ’ ‘Rock-basins of glacial origin differ from all others in the fact that they are totally independent of geological structure and the character of the rocks themselves.’ «Structural & Field-Geology’ 1905, p. 416. | Vol. 62.)- - TARNS OF THE CANTON TICINO. 191 EXPLANATION OF PLATES VII-XXI. Prats VII (facing p. 168). Map of the Piora Lakes, on the scale of 1 : 42,240. Prats VIII. View of Lago Ritom, looking east from the exit of the lake. It shows the contrast between the dip-slope of the gneiss on the southern shore, with its bays and headlands, and the escarpment of cale-mica and black-garnet schists forming the northern shore, down which cascade the three waterfalls from Lago Tom, Lago Cadagno, and the Val Piora respectively, counting from left to right, and terminating at the edge of the lake in their respective deltas. Pizzo Taneda forms the highest point along the skyline ; while immediately to the left lies the col over to Lago Scuro, from which the photograph (reproduced in Pl. X) was taken. Below this, again, the notch in the middle distance marks a portion of the old Foos valley. On the skyline to the right of Pizzo Taneda is seen the gap occupied by Lago Taneda, the overfiow from which can be discerned falling © over the precipice of actinolite-schist into the Lago Cadagno, which lies immediately below. The Val Piora is situated on the extreme right of the photograph, behind the middle-distance escarpment, while Lago di Murinascio occupies the cirque above it. Puate IX. View of Lago Tom, looking towards the barrier of rauchwacke at the lower end, in which the water-channel leading into the rock is seen. The delta on the extreme right forms the upper end of the lake. The rauchwacke is seen occupying the centre of the picture. The skyline is composed of dip- slopes and escarpments, on the right and left respectively of each of the hills. PuaTE X. View taken from the col to the west of Pizzo Taneda (shown in PI. VIII), looking across Lago Tom and the western end of Lago Ritom to the Val Levantina. The observer is looking straight down the old valley of La Foos, sections of which can still be seen below Lago Tom and at the exit of Lago Ritom. The distant range forms the watershed between the Val Levantina and the Val Maggia, and, although not visible, ago Tremorgio lies on the flank of this range on the extreme left. The snow-filled lake in the foreground is scree-dammed, and lies on a ledge of actinolite-garnet-schist 1000 feet above Lago Tom. Puate XI. Panoramic view from the watershed between Lago Tom and Lago Cadagno, looking east. Lago Cadagno lies in the foreground on the left, and the gap immediately above it is occupied by Lago di Murinascio; while the delta at the head of Lago Ritom lies below on the right. The stream running into this down the wooded slope marks the isolated outcrop of rauchwacke at the eastern end of the lake. Up the centre of the picture runs the Val Piora, cut in the outcrop of the main mass of rauchwacke. Starting in the Pizzo Columbe (which forms the central peak in the skyline), this runs through Lago Cadagno, through the point from which the photograph was taken to Lago Tom, and away to the Val Canaria. The cattle-sheds in the foreground are built on the screes which dam the western end of the lake; while behind them, the moraine is seen forming its southern shore. The dip-slopes and scarp-faces formed by the actinolite- schists are well seen on the left in the middle distance; while the gorges, trough which the rivers from Lago Cadagno and the Val Piora fall into the Lago-Ritom basin, are conspicuous in the foreground. Prats XII. Fig. 1 is a view of the barrier of rauchwacke at the lower end of Lago Tom. It shows the spring issuing from the rauchwacke at its junction with the P2 192 PROF, E. J. GARWOOD ON THE [May 1906, underlying schist, which les just beyond the photograph to the right. The cave from which the spring formerly issued is seen in the distance; while the honeycombed appearance of the weathered rock is well shown. Fig. 2 is a photograph taken from the col looking down on to Lago Scuro. It shows the circular shape of the outcrop and the reefs of rock separating the lake into two portions, one of which is seen in shadow in the right-hand bottom eorner. The exit of the lake, where it overflows into the Val Cadlimo, lies on the far side of the lake, on the extreme left. Puate XIII. Fig. | is a view of the western end of Lago Ritom; the exit is masked by one of the promontories of gneiss on the left of the picture. The wooded slope behind the hotel is the 1930-metre hill. The lower portion, on which the hotel stands, is formed of gneiss; but the upper slopes, above and to the right of the house, are formed of the dolomite which sweeps round the end of the lake, also rising to form the col over the hill to the left and up to a line half way between the lake and the highest point—Fongio. Fig. 2 shows the fold of dolomite in the Campolungo Pass. The dolomite strikes towards the observer on the left, dipping south-eastwards at a high angle. The mass of dolomite occupying the centre of the picture has been folded and faulted down to the right of the main outcrop, and it is possible that it is continued under Lago Tremorgio, which lies immediately below, just outside the right-hand corner of the picture. The flat little valley is evidently eroded out of the dolomite; the stream flowing along it comes from the lake high up on the left, and appears to have been a somewhat recent diversion. Puate XIV. Fig. lis a photograph of the lower end of Lago Tremorgio, showing the steep high cliffs surrounding the lake and the notch cut down at the exit. These cliffs continue all round the rest of the lake, becoming nearly vertical opposite the exit ; and it is over this cliff that the stream, shown in fig. 2 of Pl. XIII, falls into the lake. The outcrop of the dolomite lies at the top of this cliff, to the right, outside the photograph. Fig. 2 shows the Lago di Lucendro, seen from the upper end, On the right is the Fibbia gneiss, on the left the biotite-schist, the junction running down the centre of the lake. ‘The bare convex surface of the glaciated gneiss at the lower end of the lake is well seen. The river in the foreground rises in the Lucendro Glacier, and forms the source of the Reuss. Prats XV. Fig. 1 is another view of the Lago di Lucendro, taken from high up on the left bank. It shows the tongue of solid rock running back into the lake from the exit, mentioned in the text (pp. 185 & 190), and also the glaciated Fibbia gneiss near the exit. The dark rock in the lower part of the foreground is the outcrop. of the contorted biotite-schist penetrated by granite-veins. In the distance is seen the St. Gotthard road, and beyond it the Hospice-lakes. Fig. 2 is a view of the Lago della Sella seen from near the exit, showing the islands and illustrating its river-like character. ‘The immediate foreground and the right bank are formed of biotite-schist, penetrated in places below the exit by veins of granite, where it borders on the Fibbia gneiss; while the greater part of the left bank is composed of gneiss. (See also map, p. 186.) Prats XVI. Contoured map of Lago Ritom, on the scale of 20,000:3. For ‘ Airola’ read ‘ Airolo.’ Puate XVII. Contoured map of Lago Tom, on the scale of 5000: 1. Quart. JouRN. GEOL. Soc. VoL. LXII, PL. VIII "ISAM AHL NOYSA NAaAS WOLIY OSV) ‘on0ud “5D ("7 Quart. JOURN. GEOL. Soc. VoL. LXII, PL. IX. SUSValSFedON alneeaSINOde Geli MeN: INOS OD Val *0J0U' Uor9) een QuaRrT. JOURN. GEOL, Soc. Voce: EX: PES SOO4S G10 AHL 070d “D*f 7 { i 2 ‘ 7 y ‘ ' ) regi i = | Hy Bi "2104 osv 7 fo D119 “PYIDILY IND NM ‘WOL OSV1 AAOEV WOYSs 1X Id ‘HX ON 008 “1045 ‘NYNOP ‘LUVAd N3aS “ONDVGVO ODV1 GNV VHOld WA AHL *aguenjoz O221q *PAOLT PA 1 ' ' \ *OLISDUIAN] YT O80 T WOE NA WE “-Y4nos ayy uo UL 10 IT “OUSDPD) OSD T Quart. JOURN. GEOL. Soc. VoL, LXII, PL. XII. Fic. 1. EJ. G., Photo. SPRING ISSUING FROM THE RAUCHWACKE, BELOW LAGO TOM. Fig, 2, Exit. LAGO SCURO, LOOKING NORTH FROM THE WATERSHED. gee eee Quart. JOURN. GEOL. Soc. VoL. LXU, PL: XIII. E.J.G. photo. THE WESTERN END OF LAGO RITOM. Fea poe wr . Sx : pip? Seay= E.J.G. photo. FOLD IN THE DOLOMITE, SOUTH OF LAGO TREMORGIO. Quart. JOURN. GEOL. Soc. VoL. LXII, PL. XIV. Fig, 1. * i. a 133 fh (Gop, SPLIDEO: THE EXIT OF LAGO TREMORGIO, SEEN FROM THE SOUTH-WEST Fia. 2. E.J.G., Photo. EAGO EVCENDRO;, SEEN FROM THE SOUTH. 7 ade» E.J.G., Photo. Quart. JouRN, GEOL. Soc. VoL. LXII, PL. XV. Fic. 1. VIEW OF LAGO LUCENDRO, SHOWING THE EXIT. E.J.G., Photo. LAGO DELLA SELLA, SEEN FROM THE SOUTH-WEST. i ae oe ae eo QUART. JOURN. GEOL. Soc. VoL. LXIl, PL. XVI, se QuarrT. JouRN. GEOL. Soc. VoL. LXII, PL. XVI. HEIGHTS IN METRES 2370 2340 e sia|sa do 80 88 897 © 21 * 1829 B27 Metres 536 Ge Go eo ee Ne GOSS c : : o—— 0 =—0 — 9 , CA ie, 5 = ‘ y. 4! / . €A.7. — 9 — o —o — °— 0 1 x o— o—o | OO ape Oe OO OR Of OS a er Oe OO Rauchwacke and Dolomite Actinolite, Garnet,and Mica-Schist Calc-mica and Black Garnet-Schist o—o—o—| josoo 0 0-0 —o—0- Gneiss Fibbia Gneiss ‘Oo Creer) 0 05° 20° one Bartholomew, Litho, Edin® PRES ast oie 6 sage ape Papier CONTOURED MAP OF LAGO RITOM (1829 M.) neh Ae) eee J Ae QuarT. JouRN. GEOL Soc, VoL.Lxj|, PL XVII. HEIGHTS IN METRES }/ 2250 2220 2190 | a 2160 2130 2100 2070 2040 Lake | Level DEPTHS IN FEET }} \- ° ji TRE. ie fy i, Ge ‘ali Uys iy: ; uf oa j ly, ay #) 1 V0 wt ‘| Rauchwacke and Dolomite Le ae i Uy, Nye. pei i: tf, pi, Actinolite, Garnet,and Mica-Schist |} /— Py Hah: ‘ 9 Vik bay ' ad A : ip Pr p's ly jj) Cale-mica and Black Garnet-Schist |/, ih /) ® é bas jy iy 7 bys hh han LS , vt ‘o202| Gneiss ae 7 Mul ty, Win ee, hy be Blea tio) a CU (ae: ee / 1 Pb cbilss vet Mitel itis le WT ALALA TI 4 Bartholomew, Litho, Edin? ° Siggite 200 Metres CONTOURED MAP OF LAGO TOM (2028 M.) el : a WRERRIE me ‘oe aS, ; r : c 4 y) ¢ ' . r . ‘ \ . 4 > « . 5 ‘ ) i fi ; ; t . 4 ' ‘ Pie! i * - . , ‘ aA i ‘ a ‘ ae * , 3 " i “ os o." f in. ¢ : ‘ta ; . - © : ’ pos 3 4 : « fe a < - ‘ - 7 t i 4 ; t a ' : - 5 : { ‘ r bi 1 ’ a” ; i ’ ‘ ‘ . 5 . ‘ ps ww r +s cat Wb: ols . * * . J ’ a . . . . ‘ . ° . . * 4 2 a ° « * : . . . » - . . . , . « ° . » . oe . . 7 . _ . - . . Bartholomew, Litho, Edin? HEIGHTS IN METRES}} Rauchwacke and Dolomite =====5| Actinolite, Garnet, and Mica-Schist Calc-mica and Black Garnet-Schist Gneiss Fibbia Gneiss Bartholomew, Litho, Edin? ° aoe 200 Metres CONTOURED MAP OF LAGO CADAGNO (1921 M.) PSS a > Quart. JourN. GEOL Soc. VoL.LXII, PL. XIX. HEIGHTS IN METRES Rauchwacke and Dolomite Actinolite, Garnet,and Mica-Schist TPR) Hh Calc-mica and Black Garnet-Schist HE Ley Cy Gneiss Fibbia Gneiss Bartholomew, Litho, Edi Seale CONTOURED MAP OF LAGO SCURO (2453 M.) t 100 200 Metres hd pm EE ee ee ated v QuarT. JourN. GEOL Soc. VoL. LXII, PL XXI. eS ee, a ee = Se ee ee CS QuarrT. JourN. GEOL Soc. VoL. LXII, PL. XX. 2310 HEIGHTS IN METRES} /5 = © oo. 2 / Oi lag \ "2 © 0 9,5 Oo FEET (ete : : - SoM ~ 0 Le) °° 100 20 U5 Pf 0 co 115 0? s Besar Rauchwacke and Dolomite % ; Actinolite, Garnet,and Mica-Schist : ij, Cale-mica and Black Garnet-Schist 2075 E 5 Wy Can . *5 Uf) Gneiss = A My “* u Fibbia Gneiss .| Rauchwacke and Dolomite Scale d : a Metres CONTOURED MAP OF LAGO TANEDA (2359 M.) Actinolite, Garnet,and Mica-Schi Calc-mica and Black Garnet-Schist ae 2 vs a) = . oa mmm or ae) 4 ih -* é yon oe ii ; i be sie: ae : se ANE Vol. 62.] TARNS OF THE CANTON TICINO. 193 Prats XVIII. Contoured map of Lago Cadagno, on the scale of 5000:1. PuatE XIX. Contoured map of Lago Scuro, on the scale of 5000: 1. Puate XX. Contoured map of Lago Taneda, on the scale of 5000:1. For ‘ Lizera’ read ‘ Lisera.’ Puate XXI. Contoured map of Lago di Lucendro, on the scale of 5000: 1. Discussion. The Presrpent observed that he would have liked to say much which, on account of the lateness of the hour, he must omit. He thought that the case of the deep lakelet Scuro and its companion, situated close below the watershed, was a hard nut for advocates of glacial erosion of such lakes to crack. He could not sit down without referring to the amount of labour devoted by the Author, not only to the study of the lakes, but also to the preparation of his paper for presentation to the Society. Prof. Bonney heartily concurred with the President’s remarks on the thoroughness with which the Author had worked out his subject and the clearness of his exposition. He should like to state, for the information of the Fellows, that, in those lakes which he had described in the ‘ Geological Magazine, he had not admitted that ice could have done more than form the rock-basin now filled with water. He had supposed, for instance, at the Lago di Tremorgio, that first a corrie was formed in the usual way by streams, and then, when the ice advanced, the descending mass might have a scooping effect, at the foot of so deep a slope. In regard to. Lake Ritom, he felt doubts as to whether the headlands on the south side could be due to flexures, for these would correspond with the north-north-east to south-south-west folding, which he thought was on a large scale and was older than the east-to-west folding ; he doubted also whether the nipped-up east-to-west strips of rauch- wacke were sufficiently large to cause, when dissolved, so complete a removal of the intervening wedges of much less soluble rock. The idea of reversal in the Piora-Valley drainage presented difficulties to him. If the water escaped eastward (supposing the floor to be nearly at its present level), it would have to go over a pass now at least 1200 feet above thelake. He thought also that the deepening of the main valleys, to which the ‘ hanging valleys’ were due, had been pre-Glacial, and that probably this part of the Val Bedretto would have been under ice, even in inter-Glacial times. But, while reserving judgment on these points untii he had read the paper, he was none the less sensible of its value. Dr. Jounston-Lavis asked the Author whether he had seen any signs of carbonic-acid springs in the sites of these curious lakes. 194 THE TARNS OF THE CANTON TICINO. [May 1906, . It struck him as curious that the soluble rocks should be dissolved only at some spots, and not at others where the active movement of streams would make one expect solution to be more rapid. Such irregularity pointed to some additional local agency, such as gas- laden springs, in determining the site of some of these lakes. Their occurrence also at the junction of different rocks would be partly explained by the crushing produced in earth-movement by differential shearing at such junctions. He quite agreed that glacier-action did not explain their origin. He would also remark that the minerals of the mica-group were much more soluble in a solution of carbon-dioxide than was usually supposed. The Avruor thanked the President and Prof. Bonney for their kind remarks. In reply to the latter, he said that he thought the division on Dr. K. von Fritsch’s map, between the rauchwacke and the calcareous schist at the western end of Lake Ritom, rather an artificial one, at all events so far as solubility was concerned; and there may have been no really-insoluble ridge, as he had traced massive calcareous rocks round that end of the lake as far as the gully shown in the photograph, while dolomite came in again still farther round at the lake-level. With this exception, he quite agreed with Prof. Bonney as to the difficulties in accounting for everything by solution. In regard to the excavation of Lago Tremorgio by ice, he had set out fully in the paper the difficulties that he found in accepting this explanation. He thought that the reversal of the Ritom drainage had begun at a much higher level. The lakes, however, did not depend on this. Vol. 62.] THE HIGHEST SILURIAN OF THE LUDLOW DISTRICT. 195 10. The Hicuest Stturtsn Rocks of the Luptow Disrricr. By Miss Gerrrupe L, Extzs, D.Sc. (Dublin), and Miss I. L. Starter, B.A. (Dublin), Newnham College, Cambridge. (Com- municated by Prof. T. McKenny Hueues, M.A., F.R.S., F.G.S. Read December 20th, 1905.) [Prats XXJI—Map.| ConTENTS. Page I. Introduction and Bibliography ..................+.. 195 TO aiine all OMe Per ieet thee esos toons ancskicos hee Boho 197 TELE ev Lye rears Nee a ea ee eg a ma 8 ZOU TENOR SOS OE 37 REL 8 RE ee ee ee a 201 V. Detailed Description of the Beds ....................- 201 Wil. *Conelustons 22. 0.22.4 .06. RA aie: pee Ly A Sack oles 218 WTI ese UD TS 6 Ses ee De ie ae 21% I. Iyrropucrion anpD BiBLioGRAPHY. Tue Ludlow district has been regarded as classic ground since the days of Murchison, and, although it has been much neglected by the geologists of more recent times, local collectors of fossils have placed in the Ludlow Museum a collection which is probably unsurpassed for its excellence in illustrating the geology of the neighbourhood. Many papers were written upon the district by Murchison and others of his time; but, with the exception of Miss Wood’s paper ‘on the Lower Ludlow Shales,’ there has been no revision of its geology since their day: and we are of opinion that the area presents many features of interest, both structurally and otherwise, which can only be brought out by more recent methods of research. The more important papers dealing with the district are the following :— Mourcuison, 1834. The earliest definite mention of the rocks of the Ludlow Proc. Geol. Soc. vol. i, district appears to be found in a paper published p. 12. by Murchison in 1834. In this he gives the name Ludlow to the beds at the top of the ‘Grauwacke Series,’ and groups together all the passage-beds into the Old Red Sandstone under the name of ‘ Tilestones,’ which he regards as the lowest division of that Series. Murcuisoy, 1839. In the ‘Silurian System’ he gives a few more details ‘Silurian System ’ respecting these passage-beds, and refers the lower pp. 196 et seqq. / members of them to the Silurian System, under the name of Downton-Castle Building-stone; he also notes the more important localities where they may be studied, and records the discovery of the Bone- Mvrcuison, 1853. Bed in the Upper Ludlow Beds by the Rev. T.T. Lewis Quart. Journ. Geol. Soc. and Dr. Lloyd, the fossil-contents of this bed being vol. 1x, p. 16. described by him in a subsequent paper. 1 Quart. Journ, Geol. Soe. vol. lvi (19U0) p. 415. 196 Mvrcuison, 1854. ‘Siluria’ Ist ed. pp. 137 et seqq. Mvxcuison, 1857. Quart. Journ. Geol. Soc. vol. xiii, p. 290. EGERTON, 1857. Quart. Journ. Geol. Soc. vol. xiii, p. 282. Mvurcuison, 1859. ‘Siluria’ 3rd ed. chapt. vii. HaRLey, 1861. Quart. Journ. Geol. Soc. vol. xvii, p. 542. CURLEY, 1863. Quart. Journ. Geol. Soc. vol. xix, p. 175. LigHTsBopY, 1863. Quart. Journ. Geol. Soc. | vol. xix, p. 368. Broptig, 1869. | Quart. Journ. Geol. Soc. | vol. xxv, p. 236. LigHTBODY, 1869. Geol. Mag. vol. vi, p. 353. Marston, 1870. Geol. Mag. vol. vii, MISS G. L. ELLES AND MISS I. L. SLATER 9N THE p. 408. Marston, 1882. ‘Guide to the Geology of Ludlow.’ [May 1906, | In 1854 his famous ‘Siluria’ was published, and in a later paper, dated 1857, in describing the section seen in the Ludlow railway-cutting worked by Lightbody, he notes the existence of what may be a second Bone- Bed, and draws attention to the excellent section exposed in the right bank of the Teme. He considers. that the beds in the railway-cutting are faulted against the Old Red Sandstone to the north-west. _ In another paper, published in the same year, Sir Philip Grey KEgerton dealt with the fish-remains of the Ludlow district. In the 3rd edition of ‘Siluria’ Murchison gives an excellent summary of all the work done in the district up to date. He limits the use of the term Tilestones to the upper members of the Transition Series, and uses Downton-Castle Stone as the equivalent of the term Downton Sandstone used by Phillips in the Malvern District. He notes Marston’s discovery at Norton of the Lower Bone-Bed, surmounted by beds. with Platyschisma helicites as the characteristic fossil, and a similar succession made out by Lightbody north of Whitcliffe Coppice. Nevertheless, he does not wish to separate the lower bed from the bed with Platy- — schisma helicites. Also, while acknowledging the apparently higher position in the series of the Bone- Bed found in the railway-cutting, Murchison does not seem to be convinced that it is not the same as the lower one. In a paper published in 1861, J. Harley gives further evidence of the existence of at least two Bone-Beds, one immediately below and the other just above the Downton Sandstone; he also describes the remains found in them. T. Curley, writing in 1863, deals chiefly with the super- ficial accumulations round Ludlow, and illustrates his paper by asmall map and asection. Robert Lightbody, writing in the same year, describes a section in the Aymestry Limestone at Mocktree; he also suggests the existence of two faults in the Whitcliffe at Ludlow, throwing up the lower beds. The Rev. P. B. Brodie describes the Ludlow-Lane section as follows :—(1) Downton Sandstone; (2) Platy- schisma-helicites Shale; (8) Bone-Bed; (4) Strata with fragments of Pterygotus, etc. ; (6) Upper Ludlow. He seems to be the first to recognize the Platyschisma- helicites Bed at Ludlow, and he regards it as distinct from the lower Bone-Bed. Lightbody, writing in the Geological Magazine of the same year, suggests the existence of a fault running from the Titterstone to Downton-Castle Bridge, and also infers that the gorge of the Teme at Ludlow is tectonic in origin. A valuable paper on the beds with which we are dealing was published in 1870 by Alfred Marston, a local Ludlow geologist. In this work tLe author, although obviously well acquainted with the details of many of the sections in the neighbourhood, contents him- self with connecting them together in a generalized account, whereby much of the value of his work is lost. In 1882 the same author published a useful local ‘ Guide to the Geology of the Ludlow District,’ in which the chief localities for fossils are described. Vol. 62.| HIGHEST SILURIAN ROCKS OF THE LUDLOW DIsTRIcT. 197 IL. CLassrFICATION. In the Ludlow-Downton district there exists an interesting series of rocks, limited by the Aymestry Limestone at their base and the Old Red Sandstone at their summit, and it is with these that the present paper deals. Lithologically, they present a varied series of sediments, ranging from limestones on the one hand, through calcareous flagstones and shales, to shallow-water sand- stones on the other; and these lithological changes are associated with certain changes in the fauna. Paleontologically, these rocks are characterized by the presence of Eurypteride, which, although rare in the lower beds, gradually increase in importance until they attain their maximum development in the beds immediately underlying the Old Red Sandstone. The rich brachiopod-fauna characteristic of the lower beds dwindles and almost dies out with the approach of shallow-water conditions, although the molluscs are somewhat more persistent. The recurrence throughout of conditions tending to the formation of ‘Bone-Beds’ is also worthy of note, such conditions having prevailed at four distinct times at least during the deposition of the rocks with which we are concerned. As the result of detailed stratigraphical and paleontological work, we believe that we are able to show that these highest Silurian rocks are capable of a certain degree of subdivision, both on lithological as well as on paleontological grounds. Future work will decide whether or not these divisions have a more than local value. The subdivisions that we would suggest are tabulated on p. 198. The Aymestry Group takes its name from the well-known locality of Aymestry, 8 miles south-west of Ludlow; its lower member, the massive Conchidium-Limestone is, as a rule, full of Conchidium ( Pentamerus) Knightw, and is thus easily recognized. At Aymestry this‘ limestone is only 75 feet thick; it thickens, however, in a northerly direction, being about 100 feet thick near Ludlow, and as much as 250 feet at Mocktree. The Mocktree or Dayia-Shales, which overlie it everywhere, vary according to the thickness of the limestone-bands that they contain; these thicken also to the north, and thus, while only 40 feet thick near Ludlow, have a thickness of 150 feet on Mocktree Hill, whence they derive theizxname. They are invariably crowded with Dayia navicula. These Mocktree Shales seem, as a rule, to have been regarded as forming part of the Upper Ludlow Group, but we are convinced that the brachiopod-fauna is more closely allied to that of the Aymestry Limestone than to that of the higher calcareous beds ; and, in addition, Monograptus leintwardinensis, which certainly occurs in the Aymestry Limestone, is also found in the highest beds of the Mocktree Shales. Therefore we place the Mocktree Shales in the Aymestry Group. The Upper Ludlow Group is also capable of a twofold subdivision into the Lower and Upper Whitcliffe Flags. These derive their name from the lofty cliff which forms the right ‘qybrusy ‘gy buy Uniprlyquoy) JO BUOT | (SniawuMzUuaT) wnrpryowog YIM “SoMOJSOUIT DAISSVT ‘pynaipu nikvg jo au07, AEE jo [uJ ‘seuojsouny poeppeq-A[uryy pues sopeyg ntiaatwumiGinla.. > -<- . “DInInU ‘pnanu v)jauoyouhyy YAM ‘Sovy entq shoarwoleg ppjauoyouhy a JO IuOo7, “puvg-uorja1oU0D “9 09992 "P]]27VULJS SAZAUOY/) WJLM ‘SoV] U2ALS-9AT[O SHOITBOTRA *D . D4 Ca Sp ' + 4 oi] . “pilgs sajouoyg 40 auoz ‘qnut ppnaa)a vuafiwidy YAM ‘sBv]E puUB soTBYS snosavoyeg *¢ ‘pog-auog Mo[pwy ‘9 ‘soyeys ApuBs snoroJI[Issojuy) +” (‘peg-euog uojuMOog =) "peg -auog v OJUL AT[RLoyV{ Surssed ‘pag sagrayjay-wusryoshjv) J *9 “punyu npnbhuyyT YA ‘souojspuRs MOTEL BATISSLIT *2 ‘SOUOJSpULS SNOBDBUOGIVD “p "DINhHUWYT Moy B pus | puvq-Ysy B YIM ‘seuo}spues snovdvoruL poppeq-A[ UTI, “purur vinhuvy Jo auo7, ‘aSeC OY 4B soMOyspuKs YIM ‘spIwut pue servys ATqqna pojesorie A ‘qII6 snosovorut Lai ‘epiioyd dingy YITM ‘sepeys aAT[o ‘pog-euog episeway, ‘epiioydAing YIM SeTVys VAT[G ‘pod -JUsUISBIT = 411d snoadvUOgavo Aa. Ysiuaere ‘snwaqdhungy pure pauioa nynhuvyT jo auo7 2 L ‘QUOJSPUBY pa] PIO oe} JO s[AvUL puv sauoyspuLs por-o[ding a (se ENS a es ke) "‘SOUOFSOUIIT-MNePLYIUOY to fsaysousy ‘VW "doug ‘soley AULSANAY ‘[ -pihvg 10 9d91ZZIOW trp) ‘SSULT-P2,2WoyoUhyy 4 IO IFJILOILY M- 19 MOT | r | AOTGO]T Badd ‘TT "INOWD ‘SOB[T-SA2Z0U0YO 10 oF Ftpoyry M-t9dd yg Tap) ‘SoTOJSPUBD MOT[IX | 10 9[4S¥O-U0OJUMOG “WY "aN0uy L WAISAWAT, TIT ‘saTVyg -snuazdhung 10 spisawe J, Z| Vol. 62.; | THE HIGHEST SILURIAN OF THE LUDLOW District. 199 bank of the Teme at Ludlow, and in which the entire thickness of the divisions is exposed. The general facies of the fauna is very similar throughout, and there is no very marked difference in lithological character; but, while Rhynchonelia nucula preponderates in the lower beds, Chonetes striatella is the predominant form of the upper member of the group: hence they may be called the Rhyncho- nella-Flags and the Chonetes-Flags. So far as we have been able to determine, they do not vary rauch in thickness throughout the district. The Temeside Group, as its name denotes, is well exposed along the banks of the Teme, both at Ludlow and in the neighbour- hood of Downton Castle; it comprises beds which are virtually passage-beds into the Old Red Sandstone, and may be subdivided into the Downton-Castle or Yellow Sandstones below, and the Temeside Shales above. The term Downton-Castle Sandstone is used practically in the same sense as Murchison’s ‘ Downton-Castle Building-Stone’*; while the Temeside Shales include the shales, marls, and grit-bands with Hurypterus and Lingula cornea, which lie between the Yellow Sandstones and the Old Red Sandstone. The whole fauna of the Downton-Castle Sandstone, meagre though it be, is more intimately related to that of the Temeside Shales than to that of the beds lying below the Ludlow Bone-Bed. It seems, therefore, advisable to place the beds of this subdivision in the Temeside Group, rather than in the Upper Ludlow Group, with which it has hitherto been classed. The sandstone varies in thick- ness from 30 to 50 feet from place to place, being thickest near Onibury, but everywhere much thinner than we had been led to believe from various papers. The Temeside Shales, so far as we have seen, exhibit but slight variation as regards their thickness, such variation as there is being readily accounted for by the thickening or thinning-out of various grit-bands which occur at different horizons in them. Minimum. Maximum. Feet. Feet. LW Wan ree Hugrypieris-Stales ©). 281 .a.cces-~s<0 120 BOE catona Downton-Castle Sandstones ...... 50 a ees heron Chometes- Wage wicasccenShcessescwses 160 11 ee oe Rhynchonella-Flags ..........00..0006 120 Laie eS WUC ILICISS| 0) (eli RE ey ORAS me nee 150 ona at Conchidiwm-Limestone .........++ 250 515 850 These six major divisions, which constitute six well-defined zones, have been mapped by us throughout the Ludlow district, over an area extending from Ludlow as far south as Overton and Mary- Knowl Dingle, eastward to Caynham Camp, and westward to Downton-on-the-Rock, while our northern boundary runs east and west through Bromfield. A small map has also been made of the same beds in the neigh- bourhood of Onibury and Norton, 6 miles north-west of Ludlow (see fig. 8, p. 215). * ‘Silurian System’ 1839, p. 198. 200 MISS G. L. ELLES AND MISS I.L.STatER oN THE [May 1906, Ill. Tecrontcs. The main structural features of the district appear to be due to the superposition of two sets of earth-movements—rocks upon which a Caledonian trend had been impressed, being affected at a later time by pressure from the south, the previously-folded beds being held by some rigid mass on the north, presumably that of the Longmynd massif. Hence, the main folds of the neighbourhood of Ludlow, while chiefly due to this (Armorican) movement from the south, run east-north-east and west-south-west, the dips on their northern limbs being everywhere steeper than those on their southern sides. Along Wenlock Edge the rocks retain their Caledonian trend, and consequently the beds between Wenlock Edge and Ludlow have been dragged round to accommodate themselves as best they may to the later influence. The faulting is intimately connected with the folding; the main dislocations are of the nature of septal faults, associated with the folds the axes of which run in an east-north-easterly direction. With these are connected a system of relief-faults running in a direction at right-angles to them. The unyielding nature of the Aymestry Limestone in the west has given rise to a series of dip-faults of accommodation ; while in a few places. as for example at Onibury, stresses seem to have found relief along older Caledonian lines. The main anticline, with an axis running east-north-east and west- south-west through Ludlow town, pitches away to east-north-east ; it is cut by an oblique fault all along its northern side, throwing down the beds to the north; with the result that all the different beds of the highest Silurian rocks are brought in turn against the Old Red Sandstone, though the abutting of the Mocktree Shales and Lower Whitcliffe Flags, north-east of Whitcliffe Cottages, is largely helped by the steepness of the ground relatively to the dip at that point. For the greater part of its course the fault has a downthrow to the north; but, near Downton Castle, where the beds begin to sweep round, the throw gradually changes over to the south side, and appears to die out altogether a little farther west. A well-marked relief- fault, nearly half a mile in length, crosses the river at Downton- Castle Bridve, running a little west of north; while another sub- sidiary relief-fault, running north 30° east, occurs close to Downton- on-the-Rock. A subsidiary earth-wave finds expression in the Downton-Castle inlier, the intervening trough being occupied by the Old Red Sandstone; while. a little farther north again, a third wave is faintly indicated by the east-north-easterly extension of the main north-to-south line of outcrop. Lightbody thought that the gorge of the Teme at Ludlow was due to faulting, since the lowest beds seen in the western cliff are not visible in the eastern. He also considered that the structure of the western cliff demanded the existence of two faults. The lower beds Vol. 62.] HIGHEST SILURIAN ROCKS OF THE LUDLOW Disrricr. 201 are certainly present in the position noticed by him; but we are of opinion that their presence may be explained by the folding rather than by faulting, and the apparent discordance between the eastern and the western cliff is due to the rapid pitching-away of this fold in an east-north-easterly direction. The Caynham inlier appears to be part of an anticline faulted on every side. The septa! fault on its northern limit is part of a long line of dislocation, extending right across the country from east- north-east to west-south-west, which, like the septal fault to the north, also changes its throw from north to south along its length ; a relief-fault truncates the eastern extremity of the inher, and it appears to be faulted all along its southern side also. Another fault which ends against the main Caynham dislocation- line runs up Mary-Knowl Dingle, and forms the southern limit of our map; it appears to approximate more closely to a north-west and south-east direction than any of the relief-faults of the Armorican System, and may be an old Caledonian relief-fault affected by the later movement. TV. Scenery. The most noticeable feature in the scenery of the district is the marked difference in the character of the‘ country occupied by the Old Red Sandstone and the Silurian rocks respectively. ‘The Old Red Sandstone constitutes, as it were, a plain of arable and pasture- jand where the only undulations consist of small isolated hills, due to the presence of cornstones. From this Old-Red-Sandstone plain the Silurian rocks rise everywhere to form wooded slopes of con- siderable elevation ; when the succession is complete, the slope down to the Old Red Sandstone is fairly uniform, but when broken by faults the fall of the ground is usually very abrupt. From such a structure it follows that there is a tendency on the part of the rivers to cut gorges in the Silurian rocks, while they meander over the conntry occupied by the Old Red Sandstone ; and in only one case have we found that the direction of the drainage is in any way connected with the faulting : this being the course of the River Onny above Onibury. V. Derartep Description oF THE BeEDs. In describing the geology of the district we have only dealt in detail with those sections which, being more or less continuous, give a real clue to the succession; these may be grouped geographically, as follows :— i. Sections near Ludlow. iv. Downton-Castle inlier. (a) River Teme. v. Mocktree. (b) Wigmore Road. | } (c) Deerhouse Bank. Vi. Sections near Onibury. ii. Caynham inlier. | (a) Craven-Arms Road. (1) Onibury-Norton Lane. (c) Norton. iii. Sections near Downton Castle. (a) River Teme. (b) North-east of the Castle. 202 MISS G. L, ELLES AND MISS I. L, SLATER ON THE __—[ May 1906, (i) Sections near Ludlow. Many of thesections in the immediate neighbourhood of the town of Ludlow, referred to by earlier authors, are now completely over- grown or built over, and among these is the famous section in the railway-cutting. Of those that remain, the most complete is that exposed on the right bank of the Teme, from Dinham Bridge to the Sewage-works; this includes the classic section of Ludford Lane and the section in our Temeside Shales referred to by Murchison.* (a) Teme Section. On the right bank of the River Teme the following beds are seen in ascending order :— Aymestry Group.—The highest beds of the massive Aymestry Limestone (A) are just seen a little south of Dinham Bridge; they contain Conchidium Knight, Atrypa reticularis, Strophomena euglypha, Str. rhombordalis, and Encrinurus punctatus as character- istic fossils. The Mocktree Shales (B), here about 40 feet thick, may be examined in a small quarry close to the bridge, where they have a general northerly dip at a low angle; and also at the bottom of the track which descends to the river from the Ludlow Arms Hotel, where they dip south-eastward, and the predominant fossil is Dayia navicula : Orthis lunata, O. canaliculata, and Atrypa reticularis being also abundant. ‘The whole group is characterized by honeycomb- weathering, due to the concretionary nature of the beds. Upper Ludlow Group.—The Mocktree Shales are at once succeeded by the lowest members of the Whitcliffe Flags (C). These consist of massive blue calcareous flagstones ; they form the base of the Whitcliffe for the greater part of its extent, and are also seen in the lower part of the quarry north of Clive Cottages. They are about 120 feet thick, and are characterized especially by the abundance of Rhynchonella nucula; Orthis lunata and Serpulites longissimus are also highly typical, but Chonetes striatella (though present) is rare. ‘These beds contain at their summit a well-marked band of concretions, some of which measure 4 x 24 feet; and as this band appears to occur everywhere at the horizon where the Rhynchonelle give place to the Chenetes as predominant forms, we have utilized it for purposes of mapping, and take it as marking the highest limit of the Rhynchonella-Flags. The greenish calcareous flags (D), which overlie them, are about 160 feet thick: they are somewhat massive at their base, but become more thinly-bedded in their upper portions. The lower beds of these flags are well seen in the great quarry in the face of the Whitcliffe, and also in the face of the cliff between it and Ludford Bridge; but, although the higher beds too are exposed in both places, they are somewhat inaccessible, and may be more easily studied in what was formerly known as ‘ Ludford Lane,’ now termed the Whitcliffe Road. 1 Quart. Journ. Geol. Soc. vol. xiii (1857) p. 290. Vol. 62.] HIGHEST SILURIAN ROCKS OF THE LUDLOW District. 203 The characteristic fossil of these Upper Whitcliffe Flags is pre- eminently Chonetes striatella, which literally swarms in them. Orbiculoidea rugata and Orthoceras bullatum are also abundant ; but Rhynchonella nucula and Orthis lunata, so common in the Lower Whitcliffe Flags, are rare. The ‘ Ludford-Lane’ section, which is about 72 yards in length, shows the passage between the Upper Whitcliffe Flags and the Downton-Castle Sandstones. These beds are exposed in a high bank, which has a maximum elevation of 12 feet. The general succession is clearest on the northern side, where the beds can be seen dipping south-eastward at 10° (cf. vertical section, fig. 1 below). The lowest beds seen are the Spirifera-elevata shales (Db); only 4 feet are, however, exposed, consisting of calcareous shales with sandy beds, which contain Sp. elevata mut., Pterinea retroflexa, and Pterygotus proble- Fig. 1.—Vertical section of the succession at maticus. They are Ludford Lane ( Whitcliffe Road), on the scale immediately suc- of 6 feet to the inch. ceeded by the fa- mous Ludlow Bone- f; =] Massive yellow micaceous sandstones Bed (D C), which is with Hurypterid-remains, Pachytheca, é and Lingula minima. too Me ell-known to e@ | He ~Beyrichia-Band. require description. a3 | ee ead It is best developed ze4 | —Beyrichia-Band. at the lower end ES | Ed =| Argillaceous bed with Platyschisma of the Section, on ES ~ Beyrichia-Band. [ helicites & Modiol- 1 : 5 opsis conplanata. the south side of Ea "| Mottled sandstones the road, where it igang shales: is 21 feet above Bi Die | Ludlow Bone-Bed. road - level, and S Saudstone. reaches a maxi- ° : ite 1 3 ale wl pirifera elevata mut.and’ MUM UtNICKNESS O P Shale with S: levat d thick f = SO = [Pterinea retroflexa. nearly 6 inches =, Db: Shales and sandstones with Ttis.h i a. Spirifera elevata, Chonetes 18, hOWever, very striatella, Orbiculoidea rugata, common! i =) il and Eurypteride. oe Separa ted into two thin bands of ‘bony’ material, divided by a few inches of soft mudstone. These bands occur in a more or less lenticular manner, and one or the other disappears almost entirely from time to time, even within the short distance occupied by the section. This feature is characteristic of all the bone-beds of these highest Silurian rocks. In addition to the numerous fish-remains and crustacean remains which the Bone- Bed contains, we have identified Chonetes striatella, Orbiculoidea rugata, and Orthis sp.: a similar fauna, with Beyrichia in addition, being found in the softer mudstone separating the ‘ bony layers.’ Temeside Group.—tThe succeeding sandstones (Ea) differ some- what in lithology from the beds described above; they are mottled, yellow, slightly-micaceous sandstones, with few traces of life. They seem to usher in new conditions; for, above the Ludlow Bone-Bed, 204 MISS G. L, ELLES AND MISS I, L. SLATER ON THE | May 1906, the articulate brachiopoda, so characteristic of the lower beds, have almost disappeared, Lingule and the molluscan fauna alone remain- ing; and we therefore consider that the dividing-line between the Upper Ludlow and Temeside Groups is best drawn at this horizon. The Platyschisma-Bed (E 6) is practically composed of Platyschisma helicites and Modiolopsis complanata. Beyrichia is also abundant at its upper and lower limits. This bed gives place upward to the typical massive, yellow, mica- Fig. 2.—Vertical section of the ceous sandstones with Lingula succession near the footbridge, minima, which form the bulk Sewage-works, Ludlow, on the of the Downton-Castle Sand- scale of 6 feet to the rnch. stones. — ee ee ee The higher members of the Massive porte isofthe Downton-Castle Sandstones are pou Red Bandstones. not well seen im) thisgiiem ‘Fragment ’-Bed. 3 a a section, for, with the exception Grey micaceous ene with of the lower beds just described, which are exposed at the lower Oe a mene end of Ludford Lane and in the aan adjoining Leominster Road, the Temeside Bone-Bed. eround occupied by them is largely built over. They appear to have been worked in an old quarry, in the copse east of TLudford Church, but this is now quite overgrown. The lowest members of the Temeside Shales (F) visible are seen in two exposures in the river-bank opposite the Gas- and marls, with Lingua works; and, from analogy with Bet other sections, we are of opinion that these are the beds which immediately succeed the highest beds of the Downton-Castle Sandstones. They consist of rubbly shales (seen at water- level) overlain by massive greenish micaceous sandstones with Lingula cornea, and with rubbly marls above again (F a). For some 250 yards below this the section is interrupted, and no exposures are visible. The section is, however, continued west of the Sewage-works footbridge, where the higher members of the group are extremely-well shown (cf. vertical section, fig. 2). The rubbly grey and red beds (F a), the lowest beds observed at this point, appear to correspond very closely in general characters with those seen opposite the Gasworks; 12 feet of these are seen in this lower exposure: they are probably, therefore, of considerable thickness, although there is no direct evidence cn this point. The } Grey micaceous grit. Impersistent Bone-Bed. Variegated rubbly shales Vol. 62.] HIGHEST SILURIAN ROCKS OF THE LUDLOW DIsTRICT. 205 overlying grit-bed (F 6), which is conspicuously ‘ bony ’ at its lower and upper limits, is here seen at its maximum development (2 feet), but, like the other Bone-Beds of this Series, it thins away rapidly to west and east. Traced westward it almost disappears in the space of a few yards, while traced eastward it diminishes more gradually in thickness, its place being taken by olive shales; so that, close to the footbridge, the grit is only 7 inches thick, the lower of the two ‘bony’ layers has disappeared, and the upper is present as a thin band in the olive shales (Fc & Fe). This Bone-Bed (Fd), which we designate the Temeside Bone-Bed so as to distinguish it from others at different horizons, is a characteristic feature of the Temeside Shales at this locality, and may be regarded as a grey micaceous grit, in which large fragments of bone and fish-spines are disseminated. ‘There is, in addition, a considerable amount of car- bonaceous matter, but whether of vegetable or animal origin is not clear. As a whole, this Temeside Bone-Bed is coarser and more diffuse than the Ludlow Bone-Bed, and very different from the latter in general appearance. The succeeding olive shales (F ¢) are 2 to 4 feet thick. Eurypterid- remains are abundant, but the majority are too imperfect for determi- nation ; these are succeeded by another grey micaceous grit (1 foot), Ff, at the top of which occurs a well-marked layer crowded with carbonaceous fragments, but in which bones are rare. Purple-red sandstones with shaly partings come on immediately above this «Fragment-Bed’: these differ in general lithology from anything that we have seen at a lower horizon; and since the ‘ Fragment-Bed’ at their base appears fairly constant over wide areas and is easily recognizable, we suggest its adoption as the upper limit of the Silurian System in this district. On the left bank of the Teme at Ludlow the only rocks now visible are the beds of the Upper Whitcliffe Flags (D), which are well exposed below the Castle. There is, however, a record that the Temeside Shales were exposed in Old Street, when drainage operations Were in progress. (0) Wigmore Road. The Aymestry and Upper Ludlow Groups are again well exposed in Whitcliffe Wood, in a series of small quarries on the east side of the main Wigmore Road. The beds dip northward at angles varying from 10° to 20°. Starting from Mary-Knowl Farm, where the highest members of the Lower Ludlow rocks are seen, the Aymestry Lime- stones (A) are exposed discontinuously along the road for about 600 yards. The beds show their characteristic honeycomb weather- ing, and yield their distinctive fauna. The outcrop of the succeeding Mocktree Shales (B) extends over a distance of about 220 yards, where they give place to the Lower Whitcliffe Flags(C). These are seen at intervals down the hill for the next 700 yards, their junction with the Upper Whitcliffe Flags (D) being marked, as usual, by the Gees Ne. 246 | a SRS Poy PlO=21 ‘a SOTRYS WIO=F 3 qlas snosovotu Loy = Z ‘La , pPeg-juewseay ,=9 ‘pW pegq-euog opisemay, } _ 14 | peg-euog juoystsedwy =,z (aynos pwyUoZ.LOY ay? SIWAA Q Sr ajvIS JooYwan oy) pun “aaf QOG fo 2ouUNISYP JDJWOZLOY D spuasaudat uorpoas ayy) “moppny ‘sy.om-ohnmag ‘abprwqnoof ay; wnaw ‘dn0w9 apisauay, ay2 U2 UOI9IS 1);U0z LOFT —"& “OL Vol. 62.]| THE HIGHEST SILURIAN OF THE LUDLOW DISTRICT. 207 Concretion Band (C 6). The remaining exposures to the edge of the wood are all in the Upper Whitcliffe Flags (D). Throughout this section the faunas of the beds are quite similar to those furnished by the rocks of the type-section. (c) Stream near Deerhouse Bank. The best series of exposures on the south-eastern side of the main anticline are seen in the neighbourhood of Deerhouse Bank. The Lower Whitcliffe Flags (C) are exposed along the course of a small stream running east through Ludford Park, the Concretion- Band (C6) at its upper limit being seen in an old quarry at the southern end of Deerhouse Bank itself. The Upper Whitcliffe Flags (D) come on immediately above it, and are seen again at the eastern end of the Lower Plantation, succeeded by the lower members of the Downton-Castle Sandstones (EK). These are ex- posed in the northern bank of the old roadway, where a small fault running east-north-east brings on Upper Whitcliffe Fiags again to the south. The Downton-Castle Sandstones are, however, better exposed in an old quarry south-east of Huck’s Barn, where they dip due eastward at 10°. Gi) The Caynham Inlier. In the Caynham inlier, 2 miles east of Ludlow, only the Aymestry and Upper Ludlow Groups are represented, the higher beds being faulted out in every direction. The Mocktree Shales (B), which are the lowest beds visible at this locality, are best seen immediately north of Caynham Camp, and in the old quarries south- west of Poughnhill Cottages. North of Caynham Camp the beds are excavated in the centre of the dome, which is abruptly truncated on the north, the strata being much disturbed close to the fault and dip- ping at 50°. The Rhynchonella-Beds (Ca) of the Lower Whitcliffe Flags occupy the greater part of the remainder of this inlier, being seen all along the southern face of the hill from the eastern boundary- fault to Saltmore, a distance of about 24 miles. They are exposed in a series of small quarries, where the dip when close to the southern fault may be 30° south-eastward. The Upper Whitcliffe Flags (D) are seen only at the Saltmore end of the inlier, and are best exposed in an old quarry east of the railway. (i111) Sections near Downton Castle. (a) Teme Section, between Bow Bridge and Downton-Castle Bridge. One of the finest sections in the district is exposed in the neigh- bourhood of Downton Castle, along the banks of the Teme between Bow Bridge and Downton-Castle Bridge, there being an almost continuous exposure of rock between these two localities. On the northern bank, due south-west of Bow Bridge, the Aymestry Limestones (A) are extensively quarried along a grand Q2 208 MISS G. L. ELLES AND MISS I. L. SLATER ON THE | May 1906, cliff-face ; they dip 20° northward, and, as usual, Conchidium ( Penta- merus) Knighiv is the characteristic fossil, associated with Stropho- mena euglypha, Wilsonia Wilsont, Atrypa reticularis, and other forms. These beds are also seen on the southern bank south-east of the bridge. In thé immediate neighbourhood of the bridge they are succeeded by the Mocktree Shales (B), crowded with Dayia navicula. These are seen to bend round quickly at the edge of the syncline, and dip eastward (downstream) at a low angle. These Mocktree Shales extend as far as the Hay Mill on the southern bank, but for a slightly-shorter distance on the north side, where a steeply-rising cliff brings on the massive Rhynchonella-Beds (C), down to water- level. Beyond the Hay Mill, these Rhynchonella-Beds form cliffs on both sides of the river for half a mile, where they are succeeded by the Chonetes-Beds (D). At the Weirs a fault runs obliquely across the stream, causing a repetition of the Rhynchonella-Beds on the north side; they extend thence to Downton-Castle Bridge at the base of the high cliff, the upper portion of which is formed of the Chonetes-Beds. Fig. 4.—Diagrammatic section across the Teme at Downton-Castle Bridge. S. N Downton-Castle Bridge {=/EISIE] P+ flelsla se ce : Buk /—|| — I ee | peal Eb. * SORBET ! === || = | — | ca Dd. = =e A set 3 wee =e on | | | il Fault C a=Calcareous blue flags, with Rhyn- chonella nucula, C b=Concretion-Band. Da=Calcareous olive-green flags, with Chonetes striatella. D =Calcareous shales and flags with Spirifera elevata mut. Dc=Ludlow Bone-Bed. Ea=Unfossiliferous sandy shales. E b=Platyschisma-helicites Bed. E c= Massive yellow sandstones, with Lingula minima. i d=Carbonaceous sandstones. Ee=Thinly-bedded micaceous sand- stones. Fa=Variegated rubbly shales and marls, with greenish sandstones at the base. On the south side of the river the Chonetes-Beds form the base of the cliff, which is capped by the Downton-Castle Sandstones (E), and close to Downton-Castle Bridge the Spirefera-elevata Beds (D 6) succeed the Lower Chonetes-Beds (D a), and are surmounted at road- level by the Ludlow Bone-Bed (Dc). The Spirifera-elevata Beds are about 12 feet thick at this locality, and contain a band swarming with Crania implicata. Vol. 62.| HIGHEST SILURIAN ROCKS OF THE LUDLOW DISTRICT. 209 Two feet above road-level the Platyschisma-Bed (K 4), which is here ‘ bony,’ may be noted (= Downton Bone-Bed); and at the com- mencement of the river-path, on the eastern side of the road, the basal members of the Temeside Shales (F) may be observed: they are obviously disturbed by the proximity of the fault, and plunge down steeply towards it. The basal Temeside Shales (F a), consisting of rubbly shales asso- ciated with greenish micaceous grits and containing Lingula cornea, are also well-exposed along the road running southward from the bridge, and again on the top of the hill to the west of it, along the track leading to Burrington Hays. Farther south-west along this same track, the Downton Bone-Bed and the Ludlow Bone-Bed too are seen once more. (6) North-east of Downton Castle. Another confirmatory section is also to be found along a road which leaves the main road to run north-west and south-east, along a gully between Downton-Castle garden and the Brakes Wood. In the main road close to the Downton-Castle lodge the Chonetes- Beds (D) are well seen, dipping at 8° 8. 10° E.; while close below the main road, where the side-road makes a big bend, the Spirifera- elevata Beds (Db) are exposed. A few yards farther east the Downton- Castle Sandstones (EZ) come on, and have been extensively quarried on both sides of the gully. The most interesting exposure is that seen on the south side, where an excellent junction between the Downton- Castle Sandstones and the Temeside Shales may be studied. At the northern end of this quarry, about 20 feet of Downton-Castle Sandstones may be seen, the upper 11 feet being the thinly- bedded micaceous sandstones (EK e), while the lowest beds seen are the carbonaceous sandstones (Ed). A deeply-weathered band 6 feet from the top may represent the Fish-Bed, but it is not clearly defined. The highest members of the thinly-bedded sandstones acquire a greenish tint and mealy texture, and pass gradually up into the rubbly shales (Fa), 15 feet of which are exposed at the southern end of the quarry. The section is then obscured for some little distance, but at the corner where the road bends more to the south-west, the olive shales with the Temeside Bone-Bed (8 inches. thick) may be observed ; while, in a small quarry east of the School- house, the Old Red Sandstones are seen. (iv) The Downton-Castle Inlier. In this inlier there is a practically-complete succession of the highest Silurian rocks, from the Chonetes-Beds of the Upper Ludlow Group into the Old Red Sandstone. Upper Ludlow Group.—The lowest members seen are the Chonetes-Beds (Da). These are exposed on the southern bank of the Teme north-east of Forge Bridge, where they dip south-eastward ata low angle, forming the lowest portion of a steep cliff some 60 feet in 210 MISS G. L, ELLES AND MISSI. L, SLATER ON THE [ May 1900, height. The succeeding Spirifera-elevata Beds (D 6) are about 12 feet thick ; but the Ludlow Bone-Bed (Dc) appears to be overgrown by vegetation wherever the cliff is accessible : its position, however, can be approximately determined, since the Spirtfera-elevata Beds are found about 40 feet above water-level, while the lowest members of the Downton-Castle Sandstones (E) are seen a few feet above. N. Ss. — f.a Downton-Castle?/ Sandstones D.b ection across the Teme at Forge Bridge |For explanation, see the vertical section, fig. 6, p. 212.] On the other side of the river the succession is clearer: the Chonetes-Beds, again form the base of the cliff both east and west of Forge Bridge, dipping north-westward; and, owing to the unsym- metrical nature of the fold, the Ludlow Bone-Bed is visible at a lower level, being seen in the road leading down to the bridge, along a track leading to Forge Rough, and again farther to the north- east, close to a cottage by the Old Millrace-weir. Temeside Group.—At all these localities the Platyschisma- Bed (E 6) is found, in its usual position, with the intervening 3 feet of mottled sandstone between it and the Ludlow Bone-Bed. In the road-section it is slightly ‘bony’ in character; but at the Cottage it has taken on all the characteristics of a bone-bed, and so closely resembles the Ludlow Bone-Bed that it can only be distinguished by the presence of Platyschisma, which still occurs abundantly and has never been found by us at the lower horizon. From the Cottage this Platyschisma Bone-Bed (=Downton Bone-Bed) strikes down towards the River Teme, and is seen at river-level about 100 yards farther east. The beds immediately above it are the massive yellow ‘sandstones with Lingula minima (Kc). The higher members of the Downton-Castle Sandstones form the remainder of the cliff on the north side of the river, but are only well exposed in the quarries on either side of the road leading to the bridge, where the highest sandstones (Ke) are seen. The entire thickness of these sandstones does not seem to exceed 35 feet at this locality, although there is a little variation from place to place. The beds gradually bend round Vol. 62.] HIGHEST SILURIAN ROCKS OF THE LUDLOW DISTRICT. 211 to the south, and 200 yards west of the bridge strike so as to cross the river. On the southern bank, the Downton-Castle Sandstones are seen just behind the Forge, and up the banks of a small stream which flows northward into the river at this locality ; but the best section is seen in a large quarry at the top of the southern cliff, where a path runs along in the Tin-Mill Wood, at a height of 45 feet above the river. Here 22 feet of sandstones are exposed, dipping southward at 7°. The lowest beds seen are the massive sandstones (Hc), but their base is not visible; they are succeeded by the carbonaceous sandstones (Ed), which are in turn overlain by the thinly-bedded micaceous sandstones (Ke). About 7 feet from their upper limit a coarse micaceous bed seems to represent the Fish-Bed, but the fish-remains are not so abundant as at some other localities. The sandstones above it gradually acquire a greenish tint, and a more mealy texture, and pass gradually up into the rubbly shales (F a) so characteristic of the highest division of these rocks. (Cf. vertical section, fig. 6, p- 212.) This is one of the best sections in the district for showing the junction between the Downton-Castle Sandstones and the Temeside Shales. Continuing north-east through the Tin-Mill Wood, and following the lower track where the path divides, one may again see the higher members of the Downton-Castle Sandstones (EK ¢) in several small quarries. The strike of the beds sweeps round gradually in a northerly direction, until beyond the old Tin-Mill the beds dip north- east at 8°. In the low cliff which forms the right bank of the Millrace an excellent section of Temeside Shales is exposed, passing up into the Old Red Sandstone (cf. vertical section, fig. 7, p. 213). The bed of the Millrace is occupied by soft grey flaggy beds (Fa), but these can only ‘be observed in dry weather, and are usually covered by water; the higher beds of the rubbly shales are seen at the side. A red shale- band is conspicuous near the base of the low cliff, and about 2 feet above it a gritty bed with broken ZLingule is found: this is easily recognizable, and seems to be a local bone-bed; it has yielded Lingula cornea, Onchus tenwistriatus, Ctenacanthus-like spines, and Leperditia cf. marginata. The hard grey sandstone (F 6) forms a well-defined ledge all along the section, and is succeeded by a grey shale-bed, from the weathered surface of which large specimens of Lingula cornea often project in great abundance. These pass up into the more typical olive shales with the Temeside Bone-Bed (Fd). In its general characters, this bone-bed is exactly comparable with that exposed on the River Teme at Ludlow. It is very fossiliferous at this locality, and has yielded the following forms :— Pterygotus ludensis. Lingula cornea. Pterygotus problematicus, Ctenacanthus sp. (?) Onchus tenuistriatus. Cephalaspis sp. (?) Onchus Murchisoni. Pachytheca spherica. Onchus sp. | Fig. 6.— Vertical section of the Downton-Castle inlier at Forge Bridge, on the scale of 6 feet to the inch. Thinly-bedde- yellow micace n Be sandstones, current-bedded.. Carbonaceous Ed sandstones, with ingula minima: current-bedded.. ' Massive yellow ‘sandstones, with Lingula OT VAY) atin LAWL LY) tavey wut Fxt oy an) tes arias KOLA Ee Ae e otf Np QD x . awtat . wu Fat ee Ee ne ale ee Variegated rubbly marls, with eon- cretions contain- ing Lingula cornea. minima. Green micaceous sandstones. Rubbly shale. : | Green ieee: Downton Bone-Bed, - eae tes, with Platyschisma. Sere a | Thinly-bedded, Sandy mudstone. yellow micaceous sandstones, with LTingula cornea: | current-bedded. Ludlow Bone-Bed. Fish-Bed. Shaies, sandstones, and flags, with Spirifera elevata. Olive-green flags, with Chonetes striatella. Vol.62.] THE HIGHEST SILURIAN OF THE LUDLOW DISTRICT. 213 The olive shales are also very fossiliferous, and contain numerous remains of Kurypteride, most of which are, however, very frag- Fig. 7.—Vertical section in the Temeside Group at Tin-Mill Race, on the scale of 6 feet to the inch. Old Red Sandstone. micaceous grit (F 6). the position of which is indicated by another waterfall. Massive purple-red sandstones. ‘ Fragment-Bed,’ Grey micaceous grit. Olive shales, with Eurypterid- remains. Grit-band. Olive shales, with Eurypterid- remains. Grit-band. Olive shales, with Eurypterid-remains and Lingula cornea. Temeside Bone-Bed. Olive shales. Rubbly olive shales, with large Lingula cornea. Grey micaceous grit. Variegated rubbly shales and marls, with Lingula cornea, Sandy bed, full of broken [Lingule and fish-spines. Red shale-band. Variegated rubbly shales. (Water-level.) mentary, as the beds are brittle and have an objectionable habit of crumbling to pieces when worked. At the top of the olive shales gritty bands begin to come in, and a well- defined one crowded with carbonaceous re- mains is recognizable as the ‘ Fragment - Bed’; while imme- diately above it come the massive purple- red sandstones, which we regard as belong- ing to the Old Red Sandstone. The Temeside Shales are also exposed at three other places in the inlier; on the north side at the Spring, where the rubbly shales alone are visible, and on the south side in the small stream near Forge Bridge. and west of the bridge below the track lead- ing to Downton - Castle Bridge. In the stream the following beds were seen :— The variegated rub- bly beds (F a) are ex- posed in the stream- bed for a short dis- tance, and terminate at a small waterfall which marks the position of the grey The lowest olive shales are not seen, but the highest beds may be detected below the massive red sandstones, beds dip a little east of south at 25°. All the 214 MISS G. L. ELLES AND MISS I. L. SLATER ON THE [May 1906, The succession of these highest beds is more clearly seen 300 yards west of the stream below the track to Downton-Castle Bridge. Here the olive shales are well exposed, with the characteristic gritty ‘Fragment-Bed’ at their upper limit. We were unable to detect this in the stream-section, but it is here immediately overlain by the same massive red sandstones as those which form the upper waterfall of the stream. The chief points wherein these beds in the neighbourhood of Downton Castle differ from those at Ludlow, are the thinning of the Platyschisma-Bed to a Bone-Bed and the thickening of the olive shales (Fc & Fe). (v) Mocktree. Owing to the faulting in the northern part of the area, a satis- factory line of section is difficult to obtain, but a fairly typical one can be made out along the Leintwardine-Ludlow road. The Aymestry Limestones (A) are still being worked ina series of quarries along that part of the road which runs through Wassell Wood; but in the large quarry 130 yards west of the Briery a small fault running from north-west to south-east drops the Mocktree Shales (B) dewn against the limestones, and the fault is well seen in the face of the quarry. The lowest limit of the Mocktree Shales is also visible and is somewhat peculiar, suggesting that at this horizon erosion of the limestones was being carried on during their deposi- tion." This section has been recently figured in the Proceedings of the Geologists’ Association, vol. xviii (1904) pl. xlii, fig. 1. The Mocktree Shales, crowded with Dayia navicula, are exposed more or less continuously for 600 yards along the road, but then the succession is interrupted by a west-north-west fault, and con- sequently thé same beds are still seen for another quarter of a mile along the road. They are more extensively developed here than anywhere else in the district, attaining a maximum thickness of 150 feet. On each side of Fiddler’s Elbow the basal beds of the Lower Whitcliffe Flags (C) are seen, and south-east of Hillpike Farm Upper Whitcliffe Flags (D) are well displayed in an old quarry on the south side of the road, and in the banks of the road itself. On the north side of the road near Lodge Lane, the highest members of the Downton-Castle Sandstones (EH) have been quarried in a small plantation. (vi) Sections near Onibury. (a) Craven-Arms Road. Another section showing the relationships of the highest Silurian rocks may be seen along the road from Craven Arms to Onibury, 6 miles north-west of Ludlow. The Aymestry Limestones (A) are exposed a little more than a mile north-west of Onibury, on the north side of the road, dipping 15° eastward; and from this point to the milestone there is a nearly-continuous exposure in the road and in the river-bank, along which the characteristic honeycomb-weathering is most clearly seen. The succeeding Mocktree Shales (B) are well 1 Lightbody, Quart. Journ. Geol. Soc. vol. xix (1868) p. 368. O_O Vol. 62.] HIGHEST SILURIAN ROCKS OF THE LUDLOW District. 215 exhibited in an old quarry off the road, and again in the river-bank to the south-east. The lowest beds of the Lower Whitcliffe Flags (C) are not expesed, but the higher members may be studied in an old Fig. 8.—Geological map of the neighbourhood of Onibury. (The map is oriented north and south.) TT n It oT rere Hy 1 = : its ora a alee I Sais Zi Cottage NU Old RedSandstenes ZZ Eurypterid Shales HEE Vellow Sandstones F=4 Chonetes-Flags Rhynchonella-Flags eed Dayia-Shales Conchidium—Limestones [=] Alluvium x quarry in the wood west of Upper Park Farm; while the Upper Whitcliffe Flags (D) are still worked farther south in another old quarry, where the Concretion-Band (Cd) is visible 6 feet from the base. (6) Onibury-Norton Lane. Crossing the railway at Onibury and taking the lane leading to Norton, a large quarry in the Downton-Castle Sandstones (E) is seen 200 yards above the post-office on the west side of the road. Some 32 feet of rock are exposed, and the beds dip E. 15° 8, at 12°, but neither the highest nor the lowest beds are visible. (Cf. vertical section, fig. 9, p. 216). The lowest beds seen are the massive sandstones (Kc), but they are only exposed for a few feet in the south side of the quarry, where working has been carried on to a lower level than elsewhere ; their massive bedding and irregular fracture are characteristic, and Fig. 9.—Vertical section in the Ee Ed Ee fee ele oF ef @ Le a ee Downton-Castle Sandstones at Onibury Quarry, on the scale of 6 feet to the inch. Thinly-bedded micaceous sandstones (current-bedded), with Lingula minima (few). Fish-band. Thinly-bedded micaceous sandstones: Lingula minima rare. Carbonaceous sandstones, with Lingula minima and Eurypterid- remains (current-bedded). Massive yellow sandstones, with] abundant Lingula minima. Fig. 10.— Vertical section in the Temeside Shales in Norton Lane, on the scale of 6 feet to the ench. ‘ Fragment-Bed.” Olive shales. Greenish-purple sandstone, with Lingula cornea. Olive shales, with sandy courses, containing Eurypteride and Lingula cornea, Fe ——— — ee ee Temeside === === Bony Band. EKd == Olive shale, with. bony fragments. Se Bone-Bed. Bone-Bed, dis- continuous. Olive shales, with Inngula cornea. | Rubbly olive shales. Hard grey micaceous sandstone. Fb Variegated rubbly shales and marls. Lingula-band. Red shale-band. Vol.62.] THE HIGHEST SILURIAN OF THE LUDLOW DIsrRict. 217 they contain numerous specimens of Lingula minima, while Leper- dita marginata and various seed-like bodies are also found. These massive sandstones pass up into others chiefly distinguished by their more regular and closer bedding, smaller number of Lingule, and abundant carbonaceous remains. These and the succeeding division are markedly current-bedded. ‘The carbonaceous remains become rarer in the higher beds, and the bedding becomes thinner, so that the highest sandstones are uniformly thinly-bedded, extremely- micaceous sandstones, with but few traces of life except in the peculiar bed that we have distinguished as the Fish-Bed. This Fish-Bed occurs about 12 feet down, and is well seen in the western face of the quarry: it is a coarsely-micaceous friable sandstone, full of fish-fragments, which show especially clearly on the weathered surface. It varies both in texture and thickness within short distances, and in this respect agrees closely with the various Bone-Beds of the Series. It does not seem to be present everywhere over the area, but wherever we have detected it, it occupied approximately this same horizon. The fish-remains collected from this bed were unfortunately very fragmentary and badly preserved. Our debt therefore is all the greater to Dr. Traquair and Dr. Smith Woodward for their kindness in examining them for us. They report that most of the fish-remains are broken-off cornua of the Cephalaspid Hukeraspis pusiulifera, Agassiz, while a few are fragmentary spines referable to the Acanthodian genus Climatius. Continuing in a north-easterly direction along Norton Lane, we find the next exposure about 80 yards farther up on the east side of the road in the ditch, where the rubbly shales (F a) are seen; they are precisely like their equivalents in the Tin-Mill Wood, and must be those immediately overlying the highest beds of the Downton- Castle Sandstones. About 300 yards farther up the lane, just before it turns abruptly to the left, is the beginning of one of Marston’s typical sections. The Temeside Shales are seen on both sides of the road, and although somewhat overgrown, the section is more or less continuous, and affords an excellent opportunity for studying the succession, which appears to be as follows (cf. vertical section, fig. 10, p. 216) :— The lowest beds seen are the variegated rubbly shales (F a), 15 feet of which are exposed: among these may be noted a con- spicuous red shale-band, 2 feet above road-level, and a Lingula- band 1 foot above the shale-band. The hard grey grit (F 6) forms a conspicuous ledge all along the section on both sides of the road ; the succeeding olive shales (F ¢) are considerably thicker than their equivalents in the Tin-Mill Wood; and the Temeside Bone-Bed is found well upin them. This bone-bed is perhaps less clearly defined than usual, consisting of two distinctly ‘ bony’ layers, with bony fragments in the intervening shale as well ; probably the three layers should be regarded as the equivalent of Fd elsewhere. Numerous fossils have been obtained at this locality (see lists, pp. 219-20). The greater thickness of the higher beds of the olive shales 218 THE HIGHEST SILURIAN OF THE LUDLOW pistRict. [May 1906. appears to be due to the presence of sandy beds, which are hardly represented in the Tin-Mill Wood. Just below the lowest of a series of purplish sandstones the ‘Fragment-Bed’ may be found. Therefore we place all the beds of the upper part of the section in the Old Red Sandstone, although, so far as we can make out, Marston included them in his ‘ Tin-Mill Shales.’ (c) Norton. Additional confirmatory sections are visible in the tracks leading from Norton-Camp Wood to Norton, known as Rotting Lane and Camp Lane respectively. The Aymestry Limestones (A) have been extensively quarried in Norton-Camp Wood; they yield their characteristic assemblage of fossils, although Conchidium (Pentamerus) Knight is perhaps nos so abundant as at some other localities, and Atrypa reticularis rather more so. The Mocktree Shales(B) are seen at various points im- mediately east of the Wood, and the Lower Whitcliffe Flags (C) come on quickly above them, and are well seen in Rotting Lane with the. Concretion-Band (C 6) which marks their upward termination. The Upper Whitcliffe Flags (D) are seen in continuous section along both tracks, dipping 6° eastward, the Sparifera-elevata Beds coming on just before the main road isreached. The Ludlow Bone-Bed (Dc) is exposed in the orchard behind the red cottage, a few yards to the north. It is better developed here than anywhere else in the district, being both thicker and richer in organic remains than in other localities known tous. The Platyschisma-Bed (Kb) occurs above it, with only 2 feet of mottled sandstone intervening. The massive sandstones (EH ¢) have been quarried behind Norton Farm, and are there seen with the carbonaceous beds (Ed) resting upon them. ‘The outcrops of the highest members of the Downton- Castle Sandstones and all the Temeside Shales are concealed by vegetation. VI. Conciusions. We hope to have shown in the foregoing pages that the highest Silurian rocks are capable of more detailed subdivision than has been hitherto believed. It remains, however, to be proved whether the classification that we have adopted is of more than local value, although from the observations of one of us in other areas we hare some reason to hope that this will prove to be the case. The mapping of our zones has also, we think, indicated in various places the presence of faults which have remained hitherto undetected. In conclusion, we offer our grateful thanks to Prof. Hughes for facilitating our work in every way ; to Dr. Marr for his advice and encouragement; and to all the other members of the Cambridge School, with whom various points have been freely discussed. We would also wish to record our gratitude to Mr. Charles Fortey, the honorary Curator of the Ludlow Museum, for his unfailing courtesy to us during our visits to that town. VII. Fossiz Lists. ae pe) | Aymusrry | Upper LupLow TEMESIDE GROUP. GROUP. | GROUP. } ; | j | | 2 ear . ' =| oO _— r = rare. 5 a = | | ide = aia ¢ = common. 3 | 3 PV roe dred | Se Sy ee es E Fa eh Slee ee a SPs fs C = very common. eS eS A el ee heen Seticen ey et cel e| 2 Si Bis el elel2 Sela 3 S ARUROEAELEELE Sere ois we ts eens Ween st ce dee | ool rae SUES! aS bes hosOd ele se | een hee ie beh Sei es soy Sle eas te ete ne lies Lek | 2 ae Ase eis ae eS [oOo PS Olea aes aS penile | O BRACHIOPODA. LD CUTEST Ee oe Pee) nar ree | Wey eee rt | PPE ee rer Eien) C LL. lata, Sow.. er verattes Ine 4 | | : ‘iL. Lewisii, Sow. | ea One lea: heat tee | L. minima, Sow. ; ced cecal tacert heel | as Ia Cr here aan é “Orbiculoidea rugata, Sow. . | eh Oca ra 2 hee | BEMESIFIGTA, SOW. ..<.....0000--25s: ives ct eller | Te ae Crania implicata, Sow. (gra ze sae eae Pentamerus ( Conchidium) | oe | Knightii, Sow. ae Va / | | Strophomena euglypha, His. ...| le Str. rhomboidalis, Wilck. CE e.'| | ae | Str. ornatella, Salter ............ Taek On on aa | | Orthis orbicularis, Sow. - CG | C | | | O. canaliculata, Lindstr. Teal ae et on ee Ce | O. lunata, Sow.. op ene ek OM iad Oy Ceres ox nee / | | O. cf. elegantula, Dalm. ........ er et een Cu emiees| | Atrypa reticularis, Linn. ....... C | C | | : | Chonetes minima, Sow. ......... et ee | | | | | Ch. striatella, Dalm. ............ heen a Gaal a Ce ae | | | Ch. lepisma, Sow.. Bee tt se et ao nee hee | | | Rhynchonella nucula, Sow. oe ee Ey Cue ie | | | Bh. (Wilsonia) Wilsoni, Sow.. C anes | | / Dayia navicula, Sow.. r C || | | | Spirifera elevata mut., ‘Dalm. | ox rll rope | Whitfieldella didyma, Dalm. TOS Se et | te oe | j LAMELEIBRANCHIATA. Avicula Danbyi, M‘Coy. ...... Pas it) eh se |} | soa ‘ ; Cucullella antiqua, Sow. ...... ay oe r | | oe an eras Ss | C. coarctica, Phill. r eet er aie | & Ss | C. ovata, Sow. r aaa | (ol eg: 3 | 3 ‘C. sp. oP Gea pelh ane t| gal ane | eS a | Ctenodonta sp. : r=} | Shall 2 | Goniophora cymbeformis, Sow. : C;CiC ae = | mt ee ores Cee eee Sow. . Soa ete Wee ae Wet | S . levis, Sow. ..... ee, liede se, tee Sh oe = } | ' | a mytilimeris, Conrad ......... Hees ere os | E Sp. . Be Aste r | S Og Sethonota amygdalina, Sow. . i fie Clelr = | (0. amygdalina var. ......... ME. te | ze een Se | 0. angulifera, M‘Coy .. c _ | | = 10. impressa, Sow.........0... 2... . c = | 4 eerigidd, SOW. ..........0:6:.000 06 Bstfne | =e = = | O. semisulcata, Sow. / Pale z | | S| O. solenoides, Sow. Sas ten | {ele | | a] ie | Pterinea hyans, M‘Coy .........) ... | .. || ¢ | 24 2 Pt. orbicularis, M‘Coy ......... r eal | ee leans ries | ‘Pt. retroflexa, Wahl Feih maak 1 Gt G.-C | | =| & | Pt. tenuistriata, M‘Coy ......... Ponte wee || | hee le eee GASTEROPODA. Bellerophon expansus, Sow. .... ¥ PS (ier wer | plas Cale B. Murchisoni, dOrb. ......... 1 es. oes) oe ta ht B. wenlockensis, Sow............. ec oy aay | Cyclonema corallii, Sow. ...... Y Ty Wie] | : | Holopella gregaria, Sow. See ed a cag | eee fae as ene H. obsoleta, Sow. .......... se a ee a | EL sp. . CEES Oe | | | | | M rehisonia articulata, Sow... awe eet | ) | Platyschisma helicites, Sow. . Nee ee Pas foe pe. bs Temeside Bone-Bed. a a ee eS Pe es. ES. Ce ee ee Laer ee a , io) fo) « Fragment- Bed.’ en These lists are compiled from our own collections, supplemented by that of the Ludlow Museum. | KS Fossit Lisrs (continued). AYMESTRY ||UPPER LuDLOW)|| | GROUP. Sap. | TEMESIDEK GROUP. ! i! | a 1 areata z oS | @ o DR ae 5 }) zi Bi 6ic r = rare. E a ro |2le\4 = = common. | 3 S| 2 Z | elelg S ae. C = very common. 4 gi | 3) 2 8 a || 2 | zi Z| el f z 3 | &|a TR BIS] si 318] 6) 61S 4 eee [eeScl oes sia|2)/9 3) 3] 2) 8) 2) 31ers 1S] 8 | S18) 8) Fi] S18! 2] el) Si ete = Ss S| 21 Sy] elle | Ss] a) O74 2) son Seen omc = 2 iS! s| Ss) a] 312) 2) 218) ee eee 3 Si Hee ee es | Ste ee tae a cr | O | 8 | RISO|a | S)2 S1S\(S/ eb lS ele CEPHALOPODA. Orthoceras bullatum, Sow. ...) ... { ... rows (G8 | Li | Wkgregarinim, SOW... ....csce8) a) | the (ae : Leal OMUUACNSCRSOW? 62. eth we | eck ile’ | c | | | | BTML CS ED CHS SOW se..5sencseace|) sec, Voree tee | Od | aa | | ACTINOZOA, etc. -: Chetetes sp. ....... ye taaee a ae Eas ore 16 Aulopora serpens, ding se be ata | | Favosites Forbesi, M.-Edw. ...| C Gi 62 -e Stenopora fibrosa, Goldts” 222. C c | € (se! |e | Spongarium Edwardsii, Murch.| C Go i | | VERMES Cornulites serpularis, Schloth.) ... {| ... {\...{/ ef e | | | | Serpulites longissimus, Murch.| ... | ... | ¢ | ¢€ | | | led | Spirorbis Lewisi, Sow. ......... AN SNA, ead: a weal Trachyderma coriaceum, Phill..| ... | ... | ¢ | ¢ | | | PLantT@ (?) Pachytheca spherica, Hooker.) ... | ie | Hee he) ee a ee ligeral oO. | af | | Pee lie 5: | 4 1 C | : CRUSTACEA. Calymene Blumenbachii,Brong.| c¢ Bay Encrinurus punctatus, Brinn.| ¢ ¥ r Homalonotus ney Omi oe eevee Wal oe PPRGECPS: SD.) 00 cnn ces Soltek eee) cock eoneay eG | | arypterus acummatus, Salt. |... |... || ...] 2 | % | ec. |] ne| ede | se | ce | oe OGM INEOTIS, SAMbs, bSe.c ceeds ceo | «ee bitin ee oot Ga a MARU US, Albin sue ye Stes 8e aes |) Nea 1 dae | AP aes dicen fees [dedaddl osbe) aa!| Ree teeae| ¢ Eur. megalops, Salt. ............ a cee alla Tae ca bea dll Pll kag eee ee c UUESPP- VALIOUS — o.. pices cde oes idee Pico ldlfices [eat te eaath Olcott Sst | Cy ae | tama ‘ Parka decipiens’ (egg spare SN ST ee ale 0h dee ae Pterygotus ieee Salt. = 5G, lier celia | orem dere ges eel ee ae lS c Pt. gigas, Salt... EEN Cele! le ees utlides | sch ail sete if cette secel ome Cine Pt. ludensis, Ae es SS Lee Bey il eae sk hh ee iis Sal os Sl eae Cele eiae Pt. problematicus, Ag. ......... “ah Phot ee teal eerie. alate ae c | eahae Beyricha Kiedent, MCoy....| ... | ... || C}C|C |...) r | C |e | © | co ses Reeemme B. sp.. ae chet ees tereenal teen nears ally el 1 Leperditia marginata, Keeysch cso |e Woe fescl snl || | (aboot 2 LL. phaseolus, var. gracilenta ie oe Wiis awe [tee Pots [lowe | std] eee wee een er rr (Jones) H, small*species —.5.5.5.c.02 nce emily See cMleseor| ete | Gs) Msn ss veil ace ee eer Physocaris vesica, Salt. ? ...... atl) See, |feegerboeinliace | ea dl.eve | se keel eed «Silurian System’ 1839, p. 365. Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 220 1 mile east of Llangynog village, but where no igneous rock is now tobe seen. Strangely enough, the Lambstone porphyry is not represented on his map. Of his three patches of ‘volcanic grit,’ one appears to be the Didymograptus-bifidus grit of Castell-y-garthen; a smaller patch to the north-east of this is probably a grit of like age, which crops out 400 yards west of Lambstone Farm; while the third is evidently the V'etragraptus-grit of Pen-y-Moelfre. He represents the sedimentary rocks associated with the igneous rocks as ‘ Llandeilo Flags’ and‘ Upper Silurian’ rocks: the ‘ Llandeilo Flags’ are Tetragraptus-Beds (Arenig); thesupposed ‘ Upper Silurian’ rocks are certain green marls, cornstones, and sandstones, which form the basement-beds of the Old Red Sandstone of the district. On his section (pl. xxxiv, fig. 11), which is drawn along a north-north-west and south-south-east line through Pen-y-Moelfre, Murchison represents the [Lower] Silurian rocks of that hill as undulating, not vertical, and as containing bands of volcanic grit ; and a boss of trap is shown as bursting up between the Lower Silurian rocks, and what (from the colour on the map) he appears to have meant for Upper Silurian, but which are now known to be Old Red Sandstone. These supposed [Upper] Silurian rocks are represented as faulted off from the Old Red Sandstone. The last- mentioned boss of trap does not appear on the map, and does not indeed occur on the ground through which the section is drawn. De la Beche’ refers to the presence of conglomerates among the Lower Silurian [Ordovician] rocks of the district. The original Geological Survey-map, Sheet 41 (published in 1845), shows the igneous rocks at Coomb and Lambstone as ash-beds, but the rhyolite of Capel Bethesda is included in the diabase north of Pen-gelli-uchaf and coloured as a greenstone-dyke. The diabase of Tre-hyrn is not shown, and much of the ground occupied by andesites near that place is thrown into the sedimentary series. The horizontal section, Sheet 2, Section 6, published by the Geological Survey in 1844, is drawn along a north-west and south- east line through Pen-y-Moelfre; it represents that hill and also Moelfre Wood as anticlines, with a syncline between them. The Lambstone igneous mass is shown as ‘trap-rock, and bleached shales are noted near the ‘trap.’ The late Thomas Roberts” noted that ‘the rock marked Psb* on the Survey-map, near Llangynog, appears to be a diabase.’ It is. probable, although not certain, that in this passage he referred to the diabase of Tre-hyrn. We have given a brief account of the igneous rocks in the ‘Summary of Progress of the Geological Survey’ for 1904.° 1 © On the Formation of the Rocks of South Wales & South-Western England ” Mem. Geol. Surv. vol. i (1846) p. 29. 2 Quart. Journ. Geol. Soe. vol. xlix (1893) p. 170. > Mem. Geol. Surv. 1905, p. 37. 226 MESSRS, T. C. CANTRILL AND H. H, THOMAS ON THE [May 1906, III. Tur Sepimentary Rocks. The sedimentary rocks associated with the various igneous masses consist of the following, in descending order :— Red marls and sandstones; with a group of green marls and cornstones, sandstones, and conglome- rates at the base. (Didymograptus-bifidus Beds. — Blue-black shales with one or more thick bands of grit (Bifidus-grits) towards the base. Graptolites of | D.-bifidus type characteristic. Tetragraptus-Beds.— Black and buff shales, often ORDOVICIAN ; iridescent and iron-stained, with interbedded thin (ARENIG). grit-bands. Thick bands of ashy grit and con- glomerate (Tetragraptus-grits) towards the base. Fossils rare; the shales yield extensiform and dendroid graptolites, Tetragraptus, and some horny brachiopods; Didymograptus bifidusis absent. The grits yield occasional specimens of Orthis. LowEr Oup Rep SANDSTONE. | (1) The Tetragraptus-Beds. The beds belonging to this subdivision crop out on some elevated ground extending north-eastward from the village of Llangynog towards the River Towy, and occupy an area some 4 miles long by 13 broad. They are faulted on the north against an outcrop of Didymograptus-bifidus Beds, and on the south are bounded by the Old Red Sandstone. Stretching across the middle of this shale- area is a series of elliptical hills and ridges made un of grits and conglomerates. The ridges, which are frequently bounded on all sides by faults, are disposed in a zigzag line running from west to east, from Pen-y-Moelfre and Moelfre Wood, through the Wind- mill Hill, to Glog, a distance of 2 miles. Inasmuch as the grits and conglomerates forming these ridges are more or less vertical, the true order of succession is not obvious; but we have every reason to think that the grits and conglomerates are the same as those which form the core of a dome-like anticline at Bola- haul and Allt-cystanog south-east of Caermarthen,' where they underlie the equivalents of the Tetragraptus-shales and constitute the lowest member of the Arenig Series. At Llangynog they are again brought up in the form of anticlines, and are the oldest part of the Tetragraptus-Beds. (a) Shales north of the Grits. On the north of the grits the best sections are to be found along the upper parts of Nant Crymlyn, a stream which rises on ‘some peaty ground north of the village of Llangynog, and flows north-westward to the Cywyn at Banc-y-felin. In this brook, due north of Lan-y-gors, greyish-black splintery shales and mudstones, * ‘Summary of Progress of the Geological Survey for 1904’ Mem. Geol. Surv. 1905, pp. 33, 34. Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 227 with small horny brachiopods and dendroid graptolites, are exposed. The same beds have been cut through by the road from Forest to College, which here crosses the stream; and this section shows the iron-staining and iridescence that frequently distinguishes these beds. Traces of dendroid graptolites and horny brachiopods were found here, together with Clonograptus sp., as identified by Miss E. M. R. Wood. There are other sections near College and about the village of Llangynog, and also at Lambstone Farm, in close contiguity to the Lambstone porphyry ; for instance, in a cart-road leading to the fields behind Banc-y-ffynnon highly-micaceous buff shales dip west 10° north at 50° or thereabouts, towards the porphyry which crops out on some rough ground 50 yards farther to the west. At Lambstone Farm, in a lane immediately west of a quarry in the porphyry, is a small opening in faulted and disturbed shales; ‘there has evidently been some amount of movement between the shales and the porphyry, but the actual junction is not well-exposed. Small outcrops and débris of buff shales may be seen in the adjacent fields. There are several outcrops of grit within the shale-area just described, but as they do not appear to lie on the horizon of the massive grits of Pen-y-Moelfre and Glog, and are relatively un- important, they will not be further described. They can be seen at Ffordd and Nant-y-gog; while another forms a low ridge running from Ffald south-westward towards Pen-y-Moelfre. (b) The Grits and Conglomerates. The main outcrop of grits and conglomerates first appears near Glog-ddu, nearly 2 miles east of Llangynog village. Here an elliptical outcrop of conglomerate forms a conspicuous ridge imme- diately west of the farm. The rock is exposed on a footpath at the eastern end of the ridge, and there are occasional outcrops along the crest; but the best section is afforded by a large and con- spicuous old quarry on the northern face of the ridge. This is doubtless the quarry described by Murchison (ante, p. 224) as showing the conglomerate passing down into ‘felspar-rock.’ The only rock now visible is a conglomerate, consisting of well-rounded pebbles of what appears to be rotten rhyolite, with others of fine white quartzite, vein-quartz, and shale, embedded in a hard felspathic and sandy matrix, olive-green and buff in colour. The rock is devoid of bedding-planes; and such small evidence of bedding as is shown by the arrangement of the different-sized pebbles suggests a dip of about 20° to the north. There is no obvious difference between the beds at the top of the quarry-face and those at the foot. From Murchison’s description we can only conclude that, at the time of his visit, some lower rock, now hidden, was exposed in the quarry, or that he was mistaken in his identification. A quarter of a mile farther north the grits and conglomerates 228 MESSRS. T.C. CANTRILL AND H. H. THOMAS ON THE — [ May 1906, reappear at Glog Farm, where they may be seen near the well a few yards north-east of the house. They consist of felsite-con— glomerates and interbedded black shales which have yielded Orthis: Traced westward from Glog, the grit-crop extends for about half a mile to the cross-roads north of Pen-gelli-uchaf. It is bounded on the south by a faulted strip of ground, which includes the rhyolites of Capel Bethesda, and a remarkable rhyolite-conglome- rate of doubtful age. At the cross-roads is a quarry in rotten diabase, which appears to have been intruded into the grits. The bare surface of the road about 60 yards north of the cross- roads shows pebbly grits, largely made up of rhyolite, within a few feet of the diabase which is exposed on the road-surface and also in the old quarry hard by. The grit-outcrop can be followed from this point as far as the site of Lan-ganol, 120 yards south of which a small quarry by the side of a field-road shows, within a few feet of the conglomerate, black, yellow-weathering shales. These yielded Dictyonema sp. and Orthis Actonie (?) Sow., and appear to be identical with the shales which immediately succeed the conglo- meratic grits in the Caermarthen district. At Lan-ganol the grit-outcrop abruptly turns back on itself, and trends in a west-south-westerly direction. At Pen-y-graig Farm it forms a prominent elliptical hill, on the north side of which a roadside-quarry, 80 yards west of a ruined windmill, exhibits massive pebbly grit apparently dipping northward at 70°. It has already been pointed ont (ante, p. 224) that Murchison’s map represents a patch of trap somewhere hereabouts, but we have not been able to find any such rock. South-west of this quarry are several exposures of the grits and conglomerates in the roads, by which the outcrop can be traced to Ffald. Here it appears suddenly to terminate against a north-east and south-west fault, which introduces an outcrop of higher non- conglomeratic grit. The conglomeratic grits reappear, however, in Pen-y- -Moelfre, This hill consists of an elliptical mass of grits and conglomerates ; it seems to be faulted on all sides, except on the west-north-west, where the grits appear to pass up into shales. It presents good sections by the roadside south of Llangynog School, and a few yards farther south at some roadstone-quarries, where ‘the beds are vertical. At the southern end of the hill, about a quarter of a mile farther to the south-east, a large quarry near Llangynog Vicarage shows coarse buff sandstone, with frequent pebbles and some shale- partings; the grits have yielded Orihis. Here it is very evident that they are cut off on the south by a fault, for they strike directly towards a good section of fine-grained olive-green shales and sandstones, much disturbed and contorted, at the Vicarage- ates. Z East of this section the grits make up the lofty hill of Moelfre Wood, and are exposed in the road at Moelfre Gate. The hill ends in a bold north-and-south scarp a little west of Moelfre Farm, and, Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 229 as sandy-shale débris appears at the foot of the slope, the grits evidently extend no farther eastward. The shales themselves, however, are immediately cut off by a north-and-south fault, which introduces the green basement-conglomerates of the Old Red Sandstone of Moelfre Farm. (c) Shales south of the Grits. Reverting now to the eastern part of the district, we find olive- green shales cropping out in a lane at the western end of the grit- ridge of Glog-ddu ; but west of this point they are concealed by boulder-clay, although they doubtless occupy the low ground north of Pen-gelli-uchaf. They have been quarried on the east side of Pen-y-coed Brook, halfway between Waun-das and Craig. Thence they extend over the low ground south of Ffald, and can be seen at Ffynnon-wen immediately north of the edge of the Old Red basement-beds. South of this Old Red tract, they are poorly exposed in Pen-y-coed Brook, 300 yards west of Waun-das, and consist of highly-micaceous soft shales, with thin bands of hard grit. Grey micaceous shales are to be seen at the junction of three roads half-a-mile east-north-east of Llangynog Church. In the road by Llangynog Church, and thence southward to Llangynog Farm, there is a development of thin grits in the shales ; and shales with thin grit-bands are exposed occasionally in the road between the Church and the Plough-&-Harrow Inn. West and south of this, the T’etragraptus-Beds are faulted off from the Didymograptus-bifidus Beds of Gardd-erfin and the andesites of the Coomb complex. (2) The Didymograptus-bifidus Beds. These occupy the ground to the north, west, and south-west of the Tetragraptus-Beds described above, and are everywhere faulted off from them. In the neighbourhood of Llangynog they exist as an inverted series, excellently exposed in the middle part of Cwm Crymlyn, where, although the Yetragraptus-Beds appear to pass normally under them, there is in reality a fault of considerable magnitude which throws the higher beds of the Didymograptus- bifidus shales against the Vetragraptus-Beds.’ In the dingle, the typical characters of the shales may be observed, and specimens of graptolites of the D.-bijidus type collected. A thick band of grit crosses the dingle, and can be followed south-westward past Uchel-ole to a large old quarry on the north of a ravine at Hithin-duon, where it is cut off by a north-westerly fault which runs down the dingle. The grit reappears one-third of a mile farther south, on the western side of the Lambstone porphyry-hill, The grit is followed on the north-west by shales which extend down to the Cywyn Valley. These are to be seen in + ‘Summary of Progress of the Geological Survey for 1904’ Mem. Geol. Surv. 1905, p. 47. 230 MESSRS. T. C. CANTRILL AND H. H. THOMAS ON THE [May 1906, several small exposures, and have yielded graptolites of the D.-bifidus type ; while, south-east of the grit, shales may be seen in the road bordering the western side of the Lambstone porphyry, against which they are faulted: these have yielded similar graptolites. South-west of Lambstone Farm these shales extend towards Llandeilo-Abercowin ; the grits form the northern slopes of the hill (crowned by an old camp named ‘Castle’ on the 6-inch map) west of Gardd-erfin; and shales in a cart-road 200 yards west of that farm have yielded Dichograptid stipes, a small Diplograptus or Climacograptus, and a form of the Didymograptus-bifidus type. South of this section they are exposed in a brook, in which, at a point 350 yards north-west of Gelli, they have yielded specimens of Didymograptus of the bifidus-type, Diplograptus dentatus (?) Brongn., and Didymograptus cf. stabilis, Elles & Wood, together with Trinucleus, sp. nov. (?), and Ampyx nudus (?) Murch. They are faulted against the igneous rocks of the Coomb complex, from the Coomb dingle on the north-east, by Gelli, to Pentre- newydd on the south-west. At Gelli they contain grits, and ina lane 150 yards south-west of the farm they have afforded Didy- mograptus bifidus, Hall, and D. artus, Elles & Wood, within a yard or two of the andesites. At Pentre-newydd, some included grits are exposed in the yard by the side of the house. These and the grits at Gelli are probably on the horizon of the chief band, which is to be found some 250 feet above the base of the subdivision in other parts of the district. South-west of Pentre-newydd the shales are well-exposed, although somewhat weathered and contorted, in a road-cutting west of the Cross Roads, and have yielded graptolites of the Didy- mograptus-bifidus type and D. Murchisoni(?), together with Aighna sp., Ampyx nudus, Murch., Ampywv sp., Barrandia Cordar, M‘Coy, and Barrandia sp. Immediately south of this section they are unconformably overlain by the basement-beds of the Old Red Sandstone, and faulted on the south-west against the Bala Limestone and Dicranogruptus-Shales of Llandeilo-A bercowin. (3) The Old Red Sandstone. The lowest beds of the Old Red Sandstone are a group of sage- green and buff, rocky, calcareous marls and cornstones, with some interbedded conglomerates and micaceous green sandstones. ‘They form a well-marked basement-group to the red marls of the Lower Old Red Sandstone of the district. (¢) The Main Outerop. At the eastern end of the area they are poorly exposed in Fern- hill Brook at Rhyd-lydan, where they strike towards the Tetra- graptus-Beds of Glog, and are presumably separated therefrom by a fault running along the brook. ‘This fault bends westward along eL. Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 231 the south side of Glog-ddu hill, where it apparently cuts out the green beds and brings the red marls against the Tetragraptus- conglomerates, although the former are buried under drift. At Pen-gelli-uchaf the green beds reappear, and are well-exposed in the fields between that place, Waun-das, and Pen-y-coed. They contain numerous beds of hard rocky marl full of pebbles (ranging up to an inch or so in length) of rhyolite derived probably from the igneous masses of the district. Instead of dipping south-eastward towards the succeeding red marls, they dip about north-north-west, and apparently overlie the red marls. They must, therefore, be either inverted or faulted-off from the red marls on the south. They are faulted on the north and west against Tetragraptus- Beds ; and, for some distance in the direction of Coomb, appear to be inverted, as they dip north-westward towards the Tetragraptus-Beds. North of Coomb the base of these green beds leaves the fault by which they are separated from the Zetragraptus-Beds on the north, and can be traced with complete precision along the eastern side of the Coomb dingle, crossing the successive outcrops of the andesites and rhyolites of the Coomb complex; while, on the western side of the dingle, the lowest bed, a rhyolite-conglomerate, 10 feet thick, can be seen dipping at 60° from the rhyolites along the southern side of Castell Cogan. At Pentre-newydd the green beds pass from the igneous rocks onto the Didymograptus-bifidus Beds, and the junction may be examined in the road west of the Cross Roads. (6) Detached Areas. Between Waun-das and Moelfre a small area of the basal green beds has been thrown-in to the north of the main outcrop; it is faulted on all sides, except on the north. The beds consist of green and buff marls and highly-micaceous sandstones, with occasional bands of rhyolite-conglomerate. The green marls and sandstones are best seen in Pen-y-coed Brook 300 yards north-north-west of Waun-das, and dip south-south-eastward at 25°; from beneath them rise lower beds, which contain conglomerates visible in the fields to the north. The conglomerates are, however, still better exposed at Ffynnon-wen, and also at Moelfre, where they strike nearly due east and west with a high dip, and contain large pebbles of a felsitic rock. One other, though somewhat doubtful, patch of these basal. beds remains to be described. On the southern edge of a low gorse- covered ridge, a small roadside-quarry 80 yards west of Capel Bethesda exposes some 8 feet of a very coarse conglomerate, made up of well-rounded pebbles and boulders (ranging up to about a foot in diameter) of white rhyolite, among which a few small pebbles of vein-quartz and fine white quartzite are present. The upper part of the section is much weathered, and there is very little matrix; in the floor of the quarry the rock is more coherent. We are uncertain whether to refer this conglomerate to the green 232 MESSRS. T. C. CANTRILL AND H. H. THOMAS ON THE [May 1906, beds of the Old Red Sandstone or to the Tetragraptus-Grits : the coarseness of the material, however, compared with that of the Letragraptus-conglomerate of Glog-ddu Hill (500 yards to the south-east) has alone determined us to group it provisionally with the Old Red Sandstone. Some 50 or 60 yards north-west of the quarry, and on the top of the ridge towards its western end, one or two oP angular masses of white-weathering | ac rhyolite project through the soil. These i) do not appear to belong to the conglo- merate; and if they are not blocks, ice-borne from the exposure of that rock at the Chapel, they must be the top of a rhyolite-mass projecting through the conglomerate. The rhyolite at the Chapel is striated west 22° south, a direction which points towards the possible boulders. S.E. Coomb Andesitic Series Stream Diabase | Rhyolites IVY. Tse Ieneous Rocks. The igneous rocks of the Llangynog district occur in three well-defined areas which, taken according to their im- portance, are (1) near Coomb; (2) at Capel Bethesda; and (3) at Lambstone. Tre-hyrn (1) The Rocks near Coomb. The igneous rocks in the neighbour- hood of Coomb occupy an area of about half a square mile, and are bounded on the south and east by the Old Red Sandstone, which unconformably over- lies them; while on the north and west they are faulted against Tetragraptus- Beds and Didymograptus-bifidus Beds. This area of igneous rocks is naturally divided into two unequal portions by a dingle through which a stream flows southward from near the Plough-&- Harrow Inn, past Llwyn-celyn, and ultimately falls into the estuary of the Taf. This dingle will hereafter be referred to as ‘ the Coomb Dingle.’ On the eastern side of the dingle, the rocks consist of two series of andesites and associated tuffs, separated by a mass of rhyolites and rhyolitic breccia. On the western side, to the north, occurs a series of andesites into which the diabase of Tre-hyrn has been. intruded, and to the south lies the large mass of rhyolitic rocks of Castell Cogan. A small patch of intrusive diabase appears at the extreme south-west of the area, near Pentre-newydd. . Fig. 1.—Section across Coomb Dingle, on the scale of 6 inches to the mile. Andesitic Lavas Ashes D.-bifidus Beds. Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 233 (a) The Eastern Side of the Dingle. (i) The Lower Andesitic Series.—In the main road, at a point about 300 yards north-west of Coomb House, these rocks may be seen to emerge from below the Old Red Sandstone and to dip northward at 52°, and, on the side of the dingle, north 20° west at 45°, and may be followed upward and northward for about 260 yards. They consist largely of banded tuffs, with occasional vesicular and fine-grained-andesitic lavas, which are well-exposed on the steep slope between the main road and the stream to the west. Good vesicular rocks and more compact lavas may be seen to the west of the bend which the road takes before it reaches the Old Red rocks. This volcanic series nowhere exhibits its lower limit, but the calculated thickness of the rocks seen is at least 350 feet. (Gi) The Rhyolites.—Succeeding the above is a series of rhyolites from 140 to 150 feet thick, consisting of pale-yellow, grey, or white-weathering rocks, often minutely but markedly spherulitic even in the hand-specimen. ‘These rocks follow the dip and strike of the andesites below and succeed them with absolute conformity. The highest bed consists of a beautiful rhyolitic breccia, made up of fragments which range up to an inch or more in length, and are similar in all respects to the rhyolites below. This breccia forms a well-marked feature, and a series of crags, running from the main road, down the slope, to the stream ; it is easily located, as it is almost due east of Tre-hyrn farmhouse. (iii) The Upper Andesitic Series.—Immediately above the breccia which forms the summit of the rhyolites comes a series of andesitic tuffs and lavas, with an estimated thickness of about 900 feet, and identical in appearance with those of the lower series. These beds are best seen in the wood running alongside and above the road, and are exposed from the rhyolites northward to beyond Llwyn-celyn. The first exposure occurs immediately above the breccia, and is seen to consist of a fine-grained, compact, green tuff. The road-bank and the wood to the east, between the breccia and the tributary stream which flows into the dingle at Llwyn-celyn, are occupied by banded tuffs and lavas, chiefly vesicular. The beds dip northward at from 40° to 50°, and the section seen along the road and in the wood, from the limiting-fault on the north to the rhyolitie breccia on the south, is as follows, in descending order :-— Thickness in feet. BHAI DUMB ire vnc. us ards sdves achive hese 90 Compact tuffs and andesite ................4. 20 3 12/20 016 (0 Mi 0110 gee a a a ea DS 65 Tuffs and obscure ground ......:.........5. 195 Dana eG Stein tic suse eG sal coca’ 105 [Tributary stream at Llwyn-celyn. | PHA OR AMS Set Wa scic Mie dy se ee) dich cok Se 5 70 234 Fig. 2.—Section across Castell Cogan, on the scale of 6 inches to the mile. Castell Cogan MESSRS. T. C. CANTRILL AND H. H. THOMAS ON THE [| May 1906, (Section continued.) Thickness in feet. ADU teh Paes ia ee ne ete Muar NR RETA 3 | 20 Compact andesite and tuffs .................. 20 Fine-grained vesicular andesite ............ 6 BETTS ae cee Selae P Sae at coe yc se Eee 35 Thin vesicular andesite ...............00006+ 3 yl Uhre hy Re en NINE cI tend J. ee 10 Thin vesicular andesite and tuffs ......... Ws Rhyolites” Andesitic Series D-bifidus Beds Rhyolitic breccia. In the above series six flows have been identified; but, owing to some of the ground being obscure, it 1s impos- sible to say whether any more are present in the section. (6) The Western Side of the Dingle. G) The Rhyolites of Castell Cogan.—The rhyolites of this area are faulted off from those on the eastern side of Coomb Dingle, and form the high ground to the south of Tre-hyrn. The country rises rapidly from the stream-level (100 feet above Ordnance- datum) to the rampart of the ancient camp of Cogan, at a height of 426 feet. It then gradually falls again towards Pentre-newydd, to an altitude of 200 feet. The rhyolites are pale-yellow and grey rocks weathering white, which locally show good spherulitic and fluxion-structures. Although the area occupied by these rocks is considerable and débris of them is plentiful, outcrops of solid rock are not numerous ; exposures may be seen to the best advantage, midway between the camp and the dingle, at a point situated at the top of the wood; and also about 200 yards west of the road near the junction with the andesites. At the top of the wood, the rocks dip south- eastward at 60° to 80°, while at the last-named exposure they are vertical and strike west 30° south. The exposures and dips of these beds, provided that there has been no strike- faulting, would indicate a thickness of rock of in all probability not less than 1150 feet. (ii) The Andesitic Series.—North of the rhyolites of Castell Cogan, and probably faulted from them, is a series of some Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 230: 900 feet of andesitic rocks, which, generally speaking, occupies a less elevated tract of country. The best exposures of these rocks are situated on the western side of the Coomb Dingle, east and north- east of Tre-hyrn T'arm, where the beds, which consist chiefly of olive-green banded tuffs, are well seen along and near the footpath leading from Tre-hyrn to Llwyn-celyn. These tuffs undulate con- siderably in strike, but have a general dip of 35° to 45° in a north- westerly direction. Surmounting the slope occupied by these tuffs, and forming con~ spicuous crags west of Llwyn-celyn, is a pale massive rock, probably of extrusive character, which microscopical examination has shown to have the composition of a hornblende-andesite. Its lower limit. is exposed along the side of the dingle for some distance, but its. summit is nowhere seen; however, there is sufficient evidence to. prove that it is followed by another series of tuffs similar to that which preceded it. There are very few exposures of the andesitic- rocks on the west towards Gelli, and such as there are evidently consist of the usual types of banded and pumiceous tufts. (ii) The Diabase of Tre-hyrn.—This rock is intruded into the andesitic tuffs below the hornblende-andesite, and in character varies from a dark-green, finely-crystalline rock to a more compact. variety, which is hard to distinguish in the hand-specimen from some of the andesitic tufts. The rock is well-exposed couth- east of Tre-hyrn farmhouse, and occupies the footpath for about 50 yards in the direction of Llwyn-celyn; it may then be traced along the bank for another hundred yards to the north-east. At this point the intrusion breaks across the andesitic tuffs, and con- tinues its course asa much thinner sheet at a higher horizon, tailing off to a band a few feet thick, and occupying a position on the side: of the dingle midway between the hornblende-andesite and Llwyn-- celyn. Its maximum thickness is exposed at Tre-hyrn. (iv) The Diabase of Pentre-newydd.—This rock is exposed only in the road east of Pentre-newydd, appearing on the imme- diate north of the Old Red Sandstone and merely a few feet away from its base. Passing northward up the hill to a gate at a bend in the road, in a distance of 30 yards, the section is seen to be as follows :—The most southerly exposure consists of a pale-blue rock with conspicuous pyrites-crystals; the rock is very much decomposed, but has a parallel structure which strikes about south 25° east. A little higher in the road the mass becomes much fresher in character, and lath- shaped felspars, small patches of chlorite, and crystals of pyrites may he detected in the hand-specimen. This is succeeded by a dark-greenish-grey rock with lath-shaped felspars, and that by a more compact marly-looking rock, which northward gives place to the blue variety first mentioned. Just at the bend of the road, the section is occupied by sedimentary rocks, consisting of grits with fairly- large grains of opalescent quartz, and a few thin beds of badly- weathered greenish shales. The section continues from the bend 236 MESSRS. T. C. CANTRILL AND H. H. THOMAS ON THE [May 1906, in the road northward, both in the road and in the hedgebanks for a distance of 17 yards. The hedgebank shows a section, in what looks like a more basic variety of the diabase, and this type continues up to the junction of the intrusive rock with the rhyolites, which may be located to within a few inches. At the junction the diabase is exceedingly rotten, and has been sheared against the harder rhyolites. The rhyolite at the junction has a flinty margin without spherulites, but is immediately succeeded by a beautifully- spherulitic rock. The mass of sediments in the section is either a patch caught up in the intrusive rock, or a bed lying between two tongues of diabase. These sediments, however, present no recog- nizable signs of metamorphism. Although no other exposures of solid rock are seen, from the loose pieces of diabase scattered through the soil of the tract between the road and Pentre-newydd Farm, it would appear that this area also is occupied by the intrusive rock. Fig. 3.—Geological map of the country around Capel Bethesda, on the scale of 6 inches to the mile. o} fo} O90 0 ovo°o (oo) 0999050 9.029 (oe) Py ce) re) ; 920900 69500°9°0 2900 09000000 oo f/f 0000 220 ©9000090g0f 209 (050000000° na | Ny Wetanaene lay KY] 0 id ‘af all J Ny Tetragraptus-Grits BEB Diabase E==] BasalGreen Beds of Old Red Sandstone TD) Tetragraptus-Shales Rhyolites [= Red Marls etc. of Old Red Sandstone (2) The Rocks at Capel Bethesda. The igneous rocks in the neighbourhood of Capel Bethesda occupy a narrow strip of country, following the road from Glog to the four cross-ways north of Pen-gelli-uchaf. The rocks consist of rbyolites and a rotten diabase. Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 237 (a) The Rhyolites.—The rhyolitic rocks are best exposed in the road immediately to the east of Capel Bethesda, where they dip a little north of west at 70°. They are tough blue-grey rocks which weather white, and contain small but conspicuous cubes of pyrites : occasionally they are markedly spherulitic. They are bounded on the north by a fault, which separates them from the Tetragraptus-grits ; and on the south, although the ground is much obscured by drift, there is every reason to think that they are again faulted off from the main mass of Arenig sediments. The estimated thickness of rock seen is certainly more than 400 feet ; and if, as seems probable, these rhyclites extend westward under the conglomerate in the roadside-quarry mentioned on p. 231, then the thickness would have to be greatly increased. (6) The Diabase.— At the cross-roads north of Pen-gelli- uchaf, in an angle between two of the roads, is an old disused quarry opened up in a very rotten, basic, igneous rock. The mass is full of joints and cracks, and veined with calcite. On the surface of the road to the east of the quarry the rock is well-exposed, but is quite as rotten as before ; it has a bluish colour, and contains fairly-large patches of calcite, which give it almost an amygdaloidal appearance. The road-section is about 60 yards in length, and is seen to include a thin mass of sedimentary material in the form of greenish indurated (? ashy) shales. These sediments may also be seen in the northern hedgebank of the road to Glog, at 20 yards from the eross-ways. Immediately north of these last-mentioned shales is a steep furze-covered bank, in which by far the freshest diabase is exposed. ‘The intrusion is bounded on its northern side by grits and shales of Tetragraptus-age, but the junction between the two series of rocks is nowhere exposed. (3) The Lambstone Porphyry. The mass of igneous rock at Lambstone occupies a position due west of the village of Llangynog [Llangunnock], and forms an elliptical hill to the north of Lambstone Farm, nearly half a mile long and one-sixth of a mile wide. The boundary between the igneous rock and the surrounding sediments is marked by a strong feature, which is easily traceable all round the hill. The mass is faulted on its western and southern sides, but on the north and east the junction seems to be a normal one. On the west it is faulted against black shales containing graptolites of the Didymograptus- bifidus type; while, on the south, these give place to yellow, splintery, and possibly-indurated shales belonging to the lower part of the Tetragraptus-series. In the neighbourhood of Lambstone Farm, the junction (which is still a fault) may be closely located, the shales striking almost at right-angles to the igneous mass. On the eastern and northern sides of the hill the junction seems to be an unfaulted one, for the shales strike parallel to the boundary Q.J.G.8. No. 246. s Tg. Shales Moelfre Wood. Gritsin 7etragraptus - Beds Pen-yMoelfre Lambstone Porphyry Fig. 4.—Section across Pen-y-Moelfre, on the scale of 6 inches to the male. De bitidus Shales N.W. of the igneous rock and appear to pass. below it. The way in which the shales curve round the northern end, and dip towards the igneous mass, would suggest. that the porphyry is a sill-like mass. resting within a syncline in the Tetra- graptus-shales. In hand-specimens, the rock is of a pale-blue colour, and weathers grey and white. There are many exposures of the weathered rock, in the form of crags, all over the hill and on the slopes; but by far the freshest rock is to be seen in a small road-metal quarry im- mediately north-west of Lambstone Farm. The mass has an estimated thickness of several hundred feet. From the microscopic characters of this rock, the absence of flow-structure, its massive jointed appearance, and the seeming induration of some of the surrounding sediments, we conclude that it is of intrusive nature. (4) Detailed Petrography. 1. The Coomb Rocks. (a) The Lower and Upper Ande- sites of the Eastern Side of the Dingle. As there are no petrographical dif- ferences to distinguish the rocks which occur above and below the rhyolites on the eastern side of the Coomb Dingle, they may be conveniently considered together. i) The Flows.—Specimens were collected and slices cut from well-marked flows—one [HE 4137-38-39'] 90 yards south of the lower limit of the rhyolites, and another [EK 4130] 25 yards south of the eastern tributary stream which flows into the dingle at Llwyn-celyn. In hand-specimens they are seen to consist of coarsely-vesicular rocks of a bluish-green colour, weathering brown. 1 The numerals in square brackets through- out this paper refer to the registered numbers of the slides in the Geological Survey-Collection. Vol.62.| IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 239 The vesicles are filled with a soft mineral, which appears darker than the surrounding rock. These highly-vesicular flows have a mean specific gravity of 2°70. Under the microscope the rocks are seen to be vesicular hyalopilitic andesites, with soda-lime felspar abundantly developed and occur- ring in two generations. The felspar of the first generation builds lath-shaped crystals, unzoned, but usually twinned according to the Carlsbad and albite-laws. Pericline-twinning is not common, but interpenetration-twins and stellate groups are of frequent occurrence. ‘The felspars are, as a rule, fairly fresh; but occasionally they show a development of secondary quartz, pale-green chlorite, and dusty material of a micaceous character. The series of low symmetrical extinctions on the twin lamelle, and on sections at right-angles to the acute bisectrix, would indicate that the majority of these felspars lie in composition between oligoclase and andesine. The second generation of felspars consists essentially of microliths with low or almost straight extinction, indicative of oligoclase, arranged in bunches and fan-like aggregates which make up the bulk of the groundmass of the rock, and cause it to present almost a variolitic appearance. No original ferromagnesian mineral in the unaltered state now exists In these rocks, but undoubted pseudomorphs after augite occur plentifully, showing the external form in many cases (see Pl. XXV, fig. 1). The interior of these pseudomorphs is composed of a mosaic of secondary quartz and chlorite. Felspar-laths may be seen to penetrate the augite, and we therefore infer that the ferromagnesian mineral was of a later consolidation than the first generation of felspars. The vesicles, which range up to *2 inch in diameter, are either spherical or irregular in shape, and are filled with fibrous chlorite- minerals in radiating masses, either alone or with a little calcite or secondary quartz. Many of the larger vesicles are lined with a thin laver of secondary quartz, ‘02 to ‘04 mm. thick. The ground- mass, although largely composed of felspar-microliths, contained a small quantity of interstitial glass, now more or less completely devitrified and clouded with iron-ores. The chlorites filling the vesicles are of two varieties : the commonest is pale-green, feebly pleochroic, building fibrous and (less frequently) scaly aggregates. The zone of elongation of the fibres has a positive sign, and the mineral has a low double refraction (y — « = :005). The other mineral, which is sometimes intergrown with the former, has a deeper colour and stronger pleochroism, but occurs in fibrous aggregates. The maximum absorption is for light vibrating parallel to the length of the fibres, of which the zone of elongation is positive. For light vibrating parallel to the long axis of,the fibres, the absorption gives a yellowish-brown, and for a direction at right-angles thereto yellowish-green. Ina slide a little more than -03 mm. thick, the mineral gives the yellow and red polarization- tints of the first order, which would indicate a birefringence of from ‘014 to -015. From the foregoing considerations, we are led to regard the mineral as one of those to which the name delessite s 2 240 MESSRS. T. C. CANTRILL AND H. H. THOMAS ON THE | May 1906, has been applied, more or less common in the amygdules of ancient rocks which carry ferromagnesian minerals. Another variety of andesitic rock [EK 4439] was observed to form a thin band, only 8 feet thick. It is of a dark-green colour and fine-grained texture, and occurs in the upper andesitic series 80 yards north of Llwyn-celyn. This rock has a specific gravity of 2°65. Under the microscope, it 1s seen to consist almost entirely of a mass of minute lath-shaped felspars and microliths, with little or no interstitial material, while porphyritic crystals of any kind are practically absent. Augite probably existed as granules, and its presence may be inferred from the small chloritic grains which occur between the felspars. The felspars are twinned usually but once, although occasionally they show albite-lamellation. The symmetrical ex- tinctions range from 0° to 5° on either side of the twin-plane, indicating oligloclase as the species to which the felspars belong, and showing that these microliths correspond more or less exactly with those occurring in the more coarsely-crystaliine andesites described above. This rock may be said to have a typical pilotaxitic structure (Pl. XXV, fig. 2). It is but feebly vesicular, for a very few small vesicles occur, which are filled in the usual manner with pale-green chlorite in fibrous aggregates. The felspar-microlithsshow an arrangement that roughly approximates to parallelism, indicating in a measure the direction of flow, but many of the crystals are curved or broken; and from this, taken in conjunction with the comparative irregularity of their arrangement, it would appear that the rock must have been in a fairly-viscous condition immediately prior to its consolidation. Only a very few crystals of a porphyritic character occur, and the biggest, an oligoclase-andesine felspar, measured *7 by ‘5mm. The felspars were partly decomposed, with the formation of epidote. A specimen taken from the surface of a flow in the lower andesites, at a point a little below the rhyolites, proved to be a fluxion-breccia. The rock is made up entirely of andesitic material, mostly pumiceous, with a flow-structure developed between the fragments. These fragments are identical, in every respect, with those occurring in the pumiceous tuffs described on p. 241. (i) The Fragmental Rocks.—The fragmental rocks of ande- sitic character are largely in excess of the flows, and usually are fine-grained, well-banded rocks of a buff to yellowish-green colour. Many samples were collected from the lower and upper andesites ; under the microscope they were seen to consist of alternating coarser and finer bands of sedimentary material, and to be composed of angular grains and fragments of quartz and felspars. The larger fragments are set in a finer matrix of felspar-microliths and broken felspar-crystals. Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 241 There is little doubt that, in these rocks, there is a considerable proportion of clastic material, and it is uncertain whether they should be classed as true ashes rather than as ashy sediments. The pumiceous tuffs are not of any great thickness, when compared with the banded rocks described above, but they occur, and present exactly similar characters, in both the lower and the upper andesitic series. They consist of lapilli of pumiceous and hyalopilitic andesite, of which the vesicles are filled with a pale- green chlorite (Pl. XXVI, fig. 1). The fragments are set in a fine-grained matrix, made up almost entirely of broken felspar-laths and microliths, with a certain amount of chlorite and a few clastic quartz-grains. The felspars in the lapilli are usually all oligoclase- microliths ; but occasionally some of the larger fragments may contain small lath-shaped crystals of a slightly more basic variety. (6) The Rhyolites on the HKastern Side of the Dingle. The rhyolites on the eastern side of the dingle are pale, highly- silicified rocks. They present both perlitic and spherulitic struc- tures. A specimen collected from a small quarry above the main road [EK 4143] proved to be a beautifully-spherulitic rhyolite (Pl. XXIV, fig. 1). The few phenocrysts that occur are felspars of the albite-anorthite series; they are twinned according to the Carlsbad and albite-laws, and give low symmetrical extinctions indicative of a fairly-acid oligoclase. Subordinate orthoclase also occurs. These phenocrysts range up to 1°3 mm. in length, and are set in a groundmass which was once glassy but is now completely devitrified. The matrix consists of cryptocrystalline quartz and felspar, but in parts of the slide traces of the original flow-lines and a perlitic structure may still be made out. The spherules, which are abundant, and are often observed to coalesce, average 0-4 mm. in diameter; they are well-bounded, and in ordinary light appear more transparent than the rest of the rock. They are composed of radiating and branching felspar-fibres, of which the zone of elongation has a negative sign, and giving practically-straight extinction. Between crossed nicols the spheru- lites show a well-defined cross at the centre, which becomes blurred and indistinct outwards towards the margin, owing to the branching and change of direction of the felspar-fibres. It appears that these spherulites are a direct result of the devitrification of the rock, as they occur as the nuclei of the perlitic structure. A very few minute patches of fibrous brightly-polarizing chlorite (p. 239) occur in the groundmass, and may perhaps represent vesicles ; while others, of a chlorite with lower bi- refringence, may possibly represent some ferromagnesian mineral. (c) The Rhyolitic Breccia. This rock seems to be a true fluxion-breccia, and consists of fragments of perlitic and spherulitic rhyolites set in a matrix of 242 MESSRS. 'T. C. CANTRILL AND H. H. THOMAS ON THE [May 1906, similar material, which occasionally shows a well-defined flow- structure. A section cut from one of the larger fragments showed it to be a devitrified patchy rhyolite [EK 4144], with a pseudospherulitic structure. The phenocrysts consist of oligoclase and oligoclase- andesine felspars, both considerably decomposed, set in a felsitic groundmass which has suffered more or less complete recrystal- lization both of the felspar and the silica. ‘The secondary felspar is usually clear and granular, but the slide also shows one fairly- large spherule. Another section [E 4146] shows a devitrified perlitic rhyolite, with included fragments of a hyalopilitic andesite caught up in it and seemingly in part absorbed. ‘The rhyolite has a well-marked flow-structure in places, and locally becomes perlitic, or less frequently spherulitic. The fragments of andesite consist of masses of telspar-microliths with some interstitial material, while the rhyolite often exhibits good flow-structure around these xenoliths. Two sections [EK 4438 & 4443] cut from the rhyolitic breccia at the point where it crosses the road, show patches of rhyolite with perlitic structure. This structure occurs chiefly in those patches which have either escaped devitrification, or in which the recon- struction has proceeded to no great extent. The still glassy part of these rhyolite-fragments is usually stained a pale-yellow colour, and may be distinguished by this character in the hand-specimen. A good example of a perlitic fragment from this breccia [HK 4443, specific gravity 2°69] is figured in Pl. XXIV, fig. 2. (d) The Rhyolites of Castell Cogan. These rhyolites are similar in many respects to those on the eastern side of the dingle, both in the hand-specimen and under the microscope. They have a mean specific gravity of 2°54. Slices cut from several localities [KE 4140, 4141, 4159] showed almost entirely-devitrified rocks, in which little or no trace of original structures could be identified. The phenocrysts, which are by no means common, consist of fairly-large felspars, measuring as much as 3 mm. in length, all but completely converted into sericitic decomposition-products. They show traces of Carlsbad twinning, and the crystals apparently have the sanidine-habit. Well-formed quartz-crystals (‘3 mm. in length) occur: these are in all probability of secondary origin, although they are surrounded by a zone of cryptocrystalline material, which seems to pass gradually into their peripheries. The groundmass is completely devitrified in all cases, and usually has a cryptocrystalline character; occasionally, however, it becomes somewhat coarse, and presents a patchy appearance, due to the secondary crystallization of both the felspar andthe quartz. Generally speaking, no other structures are present; but occasionally it is possible to make out the original flow-lines, and in places the rocks become spherulitic. Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 243 The best example of a spherulitic rock collected from this series of rhyolites was taken from the road-section close to, and a little north of, the junction of the rhyolites with the Pentre-newydd diabase. (¢) The Andesites on the Western Side of the Dingle. The andesites on the western side of Coomb Dingle may be divided into two groups, one consisting of augite-andesites, and the other of hornblende-andesites. The group of augite-andesites, both as regards the flows and the fragmental rocks, is petrographically identical with the upper and lower andesitic series on the eastern side of the dingle, already described (pp. 238 et seqq.). A rock exposed in the road, 320 yards west-south-west of Tre- hyrn [EH 4151], is seen to be a coarse tuff, made up largely of frag- ments of vesicular hyalopilitic andesites and pumice: a rock identical in character with those from the eastern side of thedingle. But an interesting and important feature of this tuff is, that it contains, in addition to the usual material, fragments of a devitrified rhyolite. These rhyolitic fragments mostly present a patchy appearance, such as 1s quite common in the rocks of Castell Cogan ; while some consist solely of cryptocrystalline felspar and quartz, without any definite arrangement. The hornblende-andesite is a pale-grey rock which weathers almost white, and is thus hard to distinguish in the hand-specimen from some of the rhyolites. It has a mean specific gravity of 2°66. Sections cut from the least-weathered portions of the flows [EK 4153, 4156], which form the crags above the stream, show that the rocks are considerably decomposed. Under the microscope, they are seen to consist of phenocrysts of felspar set in a fine-grained felspathic matrix. The phenocrysts include well-formed plagioclase-felspars, belonging to the oligoclase and oligoclase-andesine varieties, which are usually twinned according to the Carlsbad and albite-laws. In contra- distinction to the -augite-andesites described on p. 239, interpene- tration-twins and stellate groups are uncommon. The presence of hornblende is inferred from strongly doubly- refracting pseudomorphs, in an indeterminable micaceous mineral. These pseudomorphs (‘8 mm. in length) have the external form of hornblende, and show a corrosion-border of pale-green chlorite and dusky iron-ores. The groundmass consists of a felted mass of felspar-microliths, which give almost straight extinction, and are evidently of more acid composition than the phenocrysts. The matrix is, however, considerably decomposed, and has given rise to sericitic alteration-products and secondary quartz. The rock, as a whole, does not seem to have been vesicular; but 244 MESSRS. T.C. CANTRILL AND H,H. THOMAS ON THE [ May 1906, an indistinct flow-structure may be traced in the groundmass, especially around some of tie larger phenocrysts. (f) The Diabase of Tre-hyrn. Specimens [E 4131] of this rock collected from well inside the mass, show a moderately-fresh diabase, of a dark-green colour, in which the constituent minerals are visible to the unaided eye. Under the microscope, it is seen to be coarsely ophitic, and shows. large crystals of an almost colourless augite enclosing, and penetrated by, felspar-laths with low symmetrical extinction-angles. The felspars are seemingly of a composition intermediate between oliguclase and andesine, and similar to those of the first generation in the augite-andesites. The augite is in general partly decomposed, especially along the cleavages, giving rise to chlorite, serpentinous material, and calcite, which minerals also form irregular patches in other parts of the rock. The felspars are clouded, and partly decomposed with the formation of epidote, or calcite and secondary quartz. (See Pl. XXVI, fig. 2.) Beautiful skeleton-crystals of ilmenite occur irregularly distributed through the rock, and, owing to their partial decomposition into lencoxenc, they appear of a dirty pees colour when viewed by reflected light. Other specimens [E 4129] nearer the margin of the intrusion show an almost complete absence of ophitic structure. In these the augite seems to be of two generations: the first building fairly-large idiomorphic crystals, and the second occurring as granules in the groundmass. The felspars are the same as in the other part of the rock, and are clouded by decomposition-products, while the larger crystals are considerably epidotized. The diabase of Tre-hyrn has a distinct tendency, when traced towards its margin, to pass from the ophitic to the granular type, as. is so often the case with intrusions of this character; while, in hand-specimens, the rock becomes quite fine-grained, and is then hard to distinguish from the fine-grained pyroclastic rocks belonging to the andesitic series into which it is intruded. (g) The Diabase of Pentre-newydd. This rock undoubtedly consolidated as an ophitic diabase, but now the original ferromagnesian mineral has disappeared, and its. place has been taken by much calcite, dusty iron-ores, serpentine, and chlorite. The felspars appear to be oligoclase- andesine, and, so. far as their decomposed state will allow us to judge, compare in composition with those in the Tre-hyrn rock. The pseudomorphs after the ferromagnesian mineral, which consist largely of calcite, with a development of leucoxene along the traces of the cleavages, are clearly ophitic in character; but the cleavages are those of an amphibole, and not of a monoclinic Vol. 62.] | IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 245 pyroxene. Therefore, we are led to infer that at some period in its history this rock was more or less completely uralitized. In its intrusion it has included a thin lenticular mass of grits and shales; the grit consists of closely-packed subangular grains of quartz, felspar, and fine sediments, with a small quantity of interstitial chlorite and iron-ores. 2. The Rocks at Capel Bethesda. (a) The rhyolites of Capel Bethesda are almost identical in character with those of Coomb, especially those on the eastern side of Coomb Dingle (p. 241), and have a mean specific gravity of 2°64. The spherulitic type is the more common, although the section in the road opposite the Chapel has yielded a rock poor in phenocrysts, the groundmass of which presents that patchy appearance that was noticed in some of the rocks from the Castell-Cogan area (p. 242). : The small patch of conglomerate which rests upon these rocks is largely made up of partly-rounded masses of the underlying material. Many of these pebbles were collected, and several sections cut from them. They showed, in most cases, marked spherulitic structures : the spherulites in ordinary light standing out in contrast with the groundmass, on account of their being slightly stained by oxides of iron. Under crossed nicols they present exactly similar characters to the spherulites from Coomb. (G) The Diabase north of Pen-gelli-uchaf.—This rock is so rotten, even at its best exposure, that little can be said with regard to its original composition, except that it was at one time a diabase, and probably preserved an ophitic structure. Now it consists entirely of almost completely-decomposed felspar-_, laths, together with chloritic and serpentinous minerals resulting from the decomposition of the original ferromagnesian mineral, while much calcite is developed in patches and veins. A sufficient number of felspars have portions left such as to make determination possible, and these are seen to be identical in composition with those in the diabase of Tre-hyrn. The chlorites include the two varieties mentioned in connection with the andesites of Coomb (p. 239). 3. The Lambstone Porphyry. The rock at Lambstone, Llangynog, is of a light greyish-blue colour, very tough, and has a specific gravity of 2°68. ‘There is no sign of bedding, but two sets of rough joints may be noticed traversing the mass approximately at right-angles to each other. Under the microscope, the rock is seen to be considerably de- composed, but it is still fresh enough to reveal its true character. The constituent minerals are fairly-large idiomorphic plagioclase- felspar and smaller orthoclase, with minute but well-shaped cry- stals of hornblende, set in a fine felsitic groundmass without any Fig. 5.—Generalized section across the Llangynog district, from north-north-west to south-south-east. [Length of section = about 4 miles. | aS we ee bd - oe Tetragraptus-Beds D.- bifidus Beds well-defined structure. Some augite was pro- bably also present. The felspar - pheno - erysts consist of acid oligoclase, twinned ac- cording to the albite and Carlsbad laws, and = giving the characteristic low maximum extinction- angles measured onto the twin lamelle. The orthoclase is most usually untwinned. Both felspars are more or less decomposed, giving rise to secon- dary quartz and seri- citic material. The hornblende no longer exists as such, but is represented by pseudomorphs in pale- green chlorite with a low double-refraction. These pseudomorphs are beautifully shaped, having the angles and cleavages of the ori- ginal mineral well- preserved. The clea- vages are often marked by little strings of iron- ores, as also in many cases 1s the border of the crystal. VY. Srrucrurs, Forp- ING, AND FavLrine OF THE DIstTRICT. The structure of the district is one of great complexity, and there are many of the pro- blems that it has pre- sented that we do not pretend to havesolved ; some tew _ salient points, however, may be regarded as certain. The general struc- ture of the district is = ¢ lod Vol. 62.| IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 247 referable to two complicated anticlines ranging about east-north- east and west-south-west. Both anticlines are overfolded from the north, their southern limbs being inverted. ‘They are separated one from the other by a thrust of steep dip, hading northwards and cutting out the intervening syncline. These anticlines appear to owe their origin to a system of pre- Old-Red folding, but it is evident that ‘they have been further complicated by a series of folds and faults which have affected the Old Red Sandstone. The axis of the northern anticline’ lies beyond the northern limit of the map (Pl. XXIII), and ranges along the low ground at Llanllwch, traversed by the Great Western Railway west of Caer- marthen, the oldest beds exposed being Vetragraptus-shales. Its southern limb is represented by the inverted belt of Didymograptus- bifidus Beds ranging east and west of Cwm Crymlyn. The axis of the southern anticline has the same general direction, and ranges through the grit-masses of Llangynog, but has suffered much fracture and displacement by later disturbances, largely determined by the differential movement between the grits and the shales. Its northern limb dips normally towards the north, but its southern limb is generally vertical or inverted. The core of this southern anticline consists of the pebbly grits of Llangynog, which we correlate with similar grits that form the core of an anticline of like structure at Bolahaul and Cystanog, south-east of Caermarthen. A glance at the map (Pl. XXIII) shows that most of the faults have a general west-south-westerly trend. They frequently throw off branches, which enclose lenticular strips of ground and unite again with the original line of fault. Although the faults shown are numerous, it is more than likely that others remain undetected. Whether the faults are normal faults or overthrusts is frequently unknown, but in many cases there is ground for supposing them to be thrusts. That much of the folding and faulting was accomplished before the Lower Old Red Sandstone was deposited, is proved by the striking unconforniity at the base of that formation. Thus: a mile north-east of Rhyd-lydan it rests on Tremadoc Beds; at Ffynnon-wen and Moelfre it overlies the Tetragraptus-shales ; at Coomb and Castell Cogan it can be seen crossing the various igneous rocks of the Coomb. complex, and its haere beds are ae up largely of detritus from the igneous rocks around which it wraps. South of Pentre-newydd it can be seen resting on the Didymograptus-bifidus shales, while a little farther to the south-west it spreads over the Bala Beds of Llandeilo- Abercowin. Neither the fault which separates the rhyolites of Castell Cogan from the Didymograptus-bifidus Beds, nor that which a little farther * «Summary of Progress of the Geological Survey for 1904’ Mem. Geol, Surv. 1905, p. 47. 248 MESSRS. T. C. CANTRILL AND H.H. THOMAS ON THE [May 1906, west separates the latter from the Bala Beds (Dicranograptus-Shales. and Bala Limestone), affects the base of the Old Red Sandstone; these faults are probably both normal faults. Other faults, how- ever, clearly involve the Old Red Sandstone, as, for instance, those which bound the basal green beds east of Moelfre, and that which separates the same beds at Coomb from the Tetragraptus-Beds of Llangynog Church. VI. Sequence AND AGE oF THE IGNEOUS Rocks. On the eastern side of Coomb Dingle the sequence is quite clear, and may be made out with ease. The oldest rocks exposed are those forming the southern mass of augite-andesites and their associated fragmental rocks ; next in succession follows a series of perlitic and spherulitic rhyolites, of which the uppermost member is a rhyolitic fluxion-breccia of some importance. This breccia in turn was followed by another set of augite-andesites, exactly similar to those below the rhyolites. It is a misfortune that the relations of the rocks on the two sides. of Coomb Dingle are so obscure, for not only is there a discrepancy between the two sides, but the andesites on the west are themselves probably faulted off from the rhyolites of Castell Cogan. However, these andesites contain fragments of devitrified rhyolite, and it is highly probable that they were extruded later than the rhyolites. Although there are some slight differences to be observed between the rhyolites on the east and those on the west of Coomb Dingle, we think that there is sufficient similarity in composition and mode of occurrence to warrant us in considering them as marking one geological horizon. With the exception of the occurrence of a small thickness of hornblende-andesite (probably of extrusive origin) in the lower part of the western andesitic series, these rocks are identical in lithological characters and composition with those of the upper andesitic series on the eastern side of the dingle. As no hornblende- andesite has been met with on the eastern side, we suspect that the augite-andesite to the north of Castell Cogan may be the highest member of the series. We therefore suggest that the sequence was, first, augite- andesites followed by rhyolites, and then another series of augite- andesites, which included a small outpouring of a hornblendic rock. The whole series was, at a later date, intruded into by diabase. There is not much evidence to prove that the igneous rocks of Coomb and those of Capel Bethesda are on the same horizon, excepting perhaps their close proximity and the evident lithological similarity which exists between the rocks of the two areas. The rhyolites of Coomb and those of Bethesda are identical in character, and both are presumably followed by intrusions of diabase : although the latter rock in the Bethesda area is so rotten, that it has been found almost impossible to compare it with that of Tre-hyrn or Pentre-newydd. Vol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 249 One of the most difficult problems that confronts us is that which concerns the age of the extrusive and intrusive rocks of the district asawhole. ‘There is, however, as we have seen, little doubt that the whole series of extrusive rocks as developed at Coomb and Bethesda may be referred to one geological horizon. The mapping of surrounding districts in South Wales has shown us that, in the Ordovician System, we might expect to meet with volcanic rocks on any of three horizons: (a) near the top of the zone of Didymo- graptus Murchisona ; (6) near the base of the zone of D. bifidus ; or (c) low down in the Arenig Series ; for, at all these rae well- developed ash-beds have been noticed. At Llangynog the first two cases may be dismissed as impossible, for the quantity of ashy material is much greater on the north side of the main anticline, while Llangynog is on the south side, where the Didymograptus-Murchison ash is not developed, and the D.-bifidus ash is represented by sandstones and grits. We are, therefore, led to the conclusion that the igneous rocks may belong to the lower part of the Arenig Group, or, failing that, to some much older group of rocks of which no sediments exist within the district. East of Llangynog, in the neighbourhood of Caermarthen, there seems to be perfect conformity between the Tremadoc rocks with Peliura punctata (Crosfield & Skeat) and the Arenig Series, of which the lowest part consists largely of conglomeratic grits and sandstones. These conglomeratic grits are those which are again brought up by an anticlinal fold in the neighbourhood of Llangynog. It is improbable, although not impossible, that the Tremadoc rocks shonld have been overlapped, and so great an unconformity developed at the base of the Arenig Group within a few miles: which would be the case, if the igneous rocks were of pre-Arenig age. It isa significant fact that the igneous masses, on one side or the other, are always bounded by a member of the oldest Arenig sediments, and it seems that the bulk of the evidence points to the age of the extrusive rocks being that of a low horizon in the Lower Arenig Group. With regard to the intrusive masses, if is obvious, at Coomb, that they are newer than the other igneous rocks; while at Lambstone, Bethesda, and Pentre-newydd, it may be proved that they were intruded after a certain portion of the Lower Arenig beds had been deposited. VII. Summary. To sum up: we find that the older sedimentary rocks of the neighbourhood of Llangynog are referable to the Didymograptus- D fidus Beds and Tetragraptus-Beds of the Arenig Series. They take the form of two main anticlines due to pre- Old Red Sandstone folding, and are unconformably overlain by Lower Old Red Sandstone. The whole sedimentary series has been subsequently affected by folding and faulting, which took much the same direction. The igneous rocks occur in three areas, which all belong to 250 MESSRS. T. C. CANTRILL AND H. H. THOMAS ON THE [May 1906, the same petrographical province. The order of extrusion was: (1) augite-andesites ; (2) rhyolites; (3) augite-andesites, with some hornblende-andesite ; and (4) intrusions of diabase and porphyry. We adopt the view that the extrusive rocks are associated with the lower members of the Tetragraptus-Beds, and are consequently of Lower Arenig age; while the intrusive rocks have been injected into the extrusive rocks, and have also affected the Tetragraptus- - Beds. EXPLANATION OF PLATES XXITI-XXVI. Puate XXIII. Geological map of the igneous and associated rocks of the Llangynog district, on the scale of 3 inches to the mile. Pirate XXIV. Fig. 1. Devitrified spherulitic rhyolite [E 4148]; crossed nicols. Some of the spherulites show a rude black cross. X 17. See p. 241. 2. Partly-devitrified perlite [E 4443]; ordinary light. x17. See p. 242. Pratzr XXV. Fig. 1. Vesicular augitic andesite [E 4138]; ordinary light. x 17. The figure shows lath-shaped felspars, set in a groundmass that exhibits a structure approaching the variolitic. Atthe top of the fieldis a small . vesicle filled with chlorite, and at the bottom a pseudomorph after augite in chlorite and secondary quartz. The chlorite shows slightly darker than the quartz. See p. 239. 2. Pilotaxitic andesite [E4439]; ordinary light. The figure shows a mass of felspar-microliths, and a solitary felspar of an earlier generation at the top of the field. x 30. See p. 240. Puate XXVI. Fig.1. Andesitic tuff [HZ 4163]; ordinary light. The figure shows several angular fragments of hyalopilitic andesite, with minute microliths arranged with flow-structure. At the bottom of the field is a larger fragment of andesitic pumice, the vesicles of which are filled with chlorite. The larger fragments are set in a fine groundmass of broken felspar-crystals and microliths. x 17. See p. 241. 2. Ophitie diabase [E 4131]; crossed nicols. Shows ophitic augite and partly-decomposed felspars. x 17. See p. 244. Discussion. The Prestpent said that he was glad to find that the Authors had obtained definite evidence of two distinct sets of movements— the one before, the other after, the deposition of the Old Red Sand- stone. He himself, when working with the late Mr. Thomas Roberts in the Haverfordwest district, had been unable to obtain evidence bearing on this question, owing to the profound faulting which had there occurred. Prof. Groom said that geologists would heartily welcome another incursion into the obscure geology of the Welsh borderland. Knowing the ground himself, he was convinced of the accuracy i pt aux Le) SR ‘ieee cc ein fl i Bir 4.: ete ; XII, Pl. XXUI. o~ TIN Ee Quart. Journ. Geol. Soc. Vol. LXII, Pl, XXIII. Coed — /Pen=y- eho F d [Q0) 02 By 020° ° G) ) = ul a 5 a a) Sym! a0 7 ¢ po * =m ; a Sok on & He ve = us & sea = eH | of Old Red Sandstone — AI 5 = a ==) Red Marls, ete: , Zh Lian » ‘0 20 6! ° 20 Cy Te) TOW re) | SOON i Moelf, nd|Har ount*Zior T. C. Hall, Quart. JouRN. GEOL. Soc. VoL. LXII, PL. XXIV. IGNEOUS ROCKS FROM LLANGYNOG. Microphoto. Bemrose, Collo. iss a QuarT. JouRN. GEOL. Soc. VoL. LXII, PL. XXV. ; VE \ ~ SON. ‘ e Nines IGNEOUS ROCKS FROM LLANGYNOG, T. C. Hall, Microphoto. Benirose, Collo. Quart. JourN. GEOL. Soc. VoL. LXII, PL. XXVI. IGNEOUS ROCKS FROM LLANGYNOG. T. C. Hall, Photomicro. Bemrose, Collo. Yol. 62.] IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNOG. 251 of the Authors’ work. In regard, however, to the sandstone with a conglomeratic base unconformably overlying the Ordovician Series, described by the Authors as Old Red Sandstone, he was impressed, when he visited the district, with the idea that it was possibly of Downtonian age. The speaker described a set of beds much resembling Old Red Sandstone, but enclosing marine Downtonian fossils, from the neighbourhood of Llangadock. Remembering that. these beds thinned out, changed colour, and became less fossiliferous as they were traced towards the area dealt with by the Authors, it seemed quite possible that the grey sandstones which they described might represent Downtonian deposits that had overlapped the under- lying Silurian, and he would like to ask whether the Authors could give satisfactory proof that this was not the case. Mr. J. V. Etspen said that he had taken particular interest in the igneous rocks described by the Authors, who had suggested a possible resemblance between the Lambstone porphyry and the lime-bostonites and oligoclase-porphyrites occurring farther north, in Pembrokeshire. So far as he could judge, the Lambstone- rock differed chiefly in having a less felspathic groundmass; but only a chemical analysis could determine the question satisfactorily. The alternations of andesites and rhyolites, mentioned in the paper, were especially interesting, and suggested the possibility of the Lambstone intrusion representing part of a later dyke-phase of the same volcanic sequence. A similar explanation was also possible in Pembrokeshire, where the speaker had shown reasons for assign- ing the lime-bostonites to an earlier period than the diabase- intrusions. The absence of rhombic pyroxenes from the Llangynog diabases was significant, in view of their great abundance in the St. David’s-Head district. Mr. W. G. FerarysmpEs congratulated the Authors upon their additions to our knowledge of the Ordovician rocks of South Wales. He enquired whether the igneous series described would fall within the zones of Didymograptus eatensus and D. hirundo, remarking that in Southern Merionethshire Ramsay had observed very similar rocks within what would now appear to be the limits of these zones. He was glad to hear of one more district wherein later igneous intrusions are practically confined to the immediate neighbourhood of the rigid masses of the earlier volcanic outpourings. Mr. Canrritt pointed out that the mapping of the Llangynog and adjacent districts clearly establishes the fact that there have been at least two periods of movement along the same general east-north- east and west-south-west lines. That the first-antedates the Lower Old Red Sandstone is proved by the manner in which that formation crosses the outcrops of various members of the Cambrian and Ordo- vician systems and of the igneous rocks of Coomb. A subsequent movement is equally clearly later than the Old Red Sandstone, for that formation is not only much folded, but is faulted deeply between some of the Ordovician subdivisions. That the green beds at the base of the Old Red Sandstone cannot be assigned to the Downtonian seems clear from the existence of 252 IGNEOUS AND SEDIMENTARY ROCKS OF LLANGYNoG. {May 1906, the great break at their base, and from the fact that they consist of marls, sandstones, and cornstones, indistinguishable except in colour from the lower part of the Old Red Sandstone; moreover, they pass by alternation upwards into the ordinary red marls. Finally, they are unfossiliferous, and, until they yield a Silurian fauna, they can be regarded only as the basement-beds of the. Old Red Sandstone. In reply to Mr. Fearnsides, he considered that the Tetragraptus- Beds should be taken to include the zones of Didymograptus hirundo and D. extensus ; but in the Llangynog district the lower beds cease to yield graptolites, and therefore may belong to the D.-extensus Zone or to a lower horizon of the Arenig. Mr. l'nomas, after thanking the President and Fellows for their kind reception of the paper, said that he was in perfect agreement with the remarks made by Mr. Cantrill. In reply to Mr. Elsden, he said that it was quite possible that the two types of intrusive rocks (diabase and porphyry) might be the dyke-equivalents of the andesite and the rhyolite respectively ; with regard to the occurrence of rhombic pyroxenes in the diabases or andesites, careful search had failed to reveal any pseudomorphs after these minerals, Vol. 62. | THE BUITERMERE AND ENNERDALE GRANOPHYRE. 253 12. The BurrerMeRE and ENNERDALE GRANoPHYRE. By Roser Heron Rastatt, M.A., F.G.S., Fellow of Christ’s College, Cambridge. (Read January 24th, 1906.) [Prares XXVIII & XXVITI—Microscorn-Sxcrions. | ContEnts, Page Le Uinbrodulebio mp sya. so ane seas se ctge des esel sence 253 es Mield=Helationse sarc) cease saee 253 JIT. Petrographiceal Descriptions’ -........2........ 258 MV General Characters ..).-.c-cescesess-Wae sees sace: 268 Vi. Granoplaynic Structure occ... ye2s ne du eating 270 VL, Summary and Conelusiom. (0. ).3..4..: 00.0.5 272 I. Inrropuction, In the western part of the English Lake-District there occurs a large development of igneous rocks, which are conveniently described, collectively, as the Buttermere and Ennerdale Granophyre. This group extends for nearly 10 miles from north to south, and for nearly 5 miles from east to west. In this large area are to be found several different rock-types, which present many features of interest; and, at the suggestion of Dr. J. E. Marr, F.R.S., I have undertaken the investigation of the phenomena shown in this district. The first and only detailed description of this rock-mass was given by Clifton Ward * nearly 30 years ago; and the subject has apparently never been touched as a whole by modern methods, although Mr. Alfred Harker * has published a short description of some specimens from the Wastwater district. II. Frecp-ReEtarions. The map (fig. 1, p. 254) shows that the exposure of this group of rocks is very irregular j in form: to the west of Buttermere is a some- what lenticular mass about 23 miles long by nearly a mile wide, forming the lower part of High Stile, and the main mass of Red Pike and Gale Fell. The southern boundary of this intrusion cuts across the summit of Red Pike; and there is a strong contrast between the rugged crags of High Stile, formed of volcanic rocks of the Borrowdale Series, and the smooth outlines given by the intrusive rocks of Red Pike. This intrusion is connected with a much. larger mass to the south by a narrow neck on Little Dodd; and this neck is very conspicuous even from a distauce, by reason of the weathering-out of blocks of granophyre, contrasted with the peat-covered Skiddaw Slates and Borrowdale volcanic rocks on the west and east respectively. 1 Quart. Journ. Geol. Soc. vol. xxxii (1876) p. 14; also ‘Geology of the Northern Part of the English Lake-District’ Mem. Geol. Surv. 1876, p. 31. 2 “Notes on North of England Rocks’ in ‘The Naturalist’ 1889, p. 209. Q.J.G.8. No. 246. T Fig. 1.—Geological map of the district around Ennerdale Water, Buttermere, and Wastwater, on the scale of 2 miles to the inch. NEES WSS xx xx x*% x XXxXx% XXx XX XX XXKAK x xx xy 1 inch =::% 2 Miles. i. Alluvium OS Skiddaw Slates XY K% Xx X% XXKXX Yi ee = . ere ee ae eae ete “24 Volcanic Series 2 Aa AA oo SS te (iOS TLLGILE ae ees P+ li. °t-HEAD. ++ (jcc egee tele Cayo . -* . .