Yet C i oa “ ‘ y « 7 ca ¥ ve fo ot 4 ale tet J t ae Creches ra lee ae | fe ae ka dy ype ee Pew Te ee bv TR Le OS ee be aed y Te ee — Ponty MeN ~ 1S \CHiwea tart ated hal i " ete ibeobeibege w ee wte ty Welbe fhe tk tamndt battle t hn AE tae 1 eres ‘ 1 lana the ‘ a: ee ee MAL abet banter Rap how ‘ a AGN beth Dew beiesnegon Hahah heb te a vee ey i " Dicthinll boris line a . i} i 4 Sletten dw Hora tyes Het dete RD vate dob he , a ee) a tM Ke tbithokeg Morand ie ee ee oe bnitndl a) er) ayndd jon iv y Vsirarta eth , . tte We YAN a theater Ud 4 oe ‘ ; ‘ A Mette dbiotbe thsi , 1 ‘ uid es wow , oo " oh ‘ Wyo Pal thet , thet 1 Voweo oa ih a hy ee ee a ‘4 ‘ ie Vb ened ad aol . o tot htt Holt ted Wott Bett ‘ \ on tay rr Pie ee ee ewe ce ‘ . ’ Fe doodin toe Potbet lh idedmih = ’ ‘ ‘ i g 4 oO ' * , 4p Att mh ede tad tired An > Bh - Wohin ' ' : ’ ' A Nu ee a ' 4 we ‘ J he “ (ee nee | q ipa tee Vena Ab Corer ea) a ' Ka Sots A 0 , i ‘ setae bm tata te Moat Wet Pi dot « ci tose é “ “ vat Ltole “ fn bewoten ‘ ty ey ‘ ‘ i ‘ iat a ' 4 A ~~ ae . 4 tan Pe Dnt ‘ ‘ ‘ toteath 4 ‘ rn ‘ ‘ é ‘ any) " , ‘ ‘ V lvsih * \ i ‘ Pa tnw hoe ‘ Wd vedo ¥ j ‘ Lait hethowy bo . ’ j “ ‘ ” " i q ‘ : “i ‘ : ‘ * " _ : eb oe ae Te ie) Pome Bie eee 2 Se ; ae a a “= , + 4 : RK a Te om - 7 [en - - a ¥ ; 7 =. 7 : 7 - , Aa mt I 738 ae aes bea = aa is 4 SereGanters, (att bei: B 842 ts 2 5, Ne ee ote ee wAey At fap Ahi hieenss be Tes N. ) THE QUARTERLY JOURNAL GEOLOGICAL SOCIETY OF LONDON. EDITED BY THE ASSISTANT-SECRETARY OF THE GEOLOGICAL SOCIETY. 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 belle et probabiliter opinari, sed certo et ostensive scire; tales, tanquam veri scientiarum filii, nobis (si videbitur) se adjungant. —Novum Organum, Prefatio. VOLUME THE THIRTY-SEVENTH. CULL "7 Visas ’ 1881. KS LONDON: LURE Ta LONGMANS, GREEN, READER, AND DYER. PARIS: FRIED, KLINCKSIECK, 11 RUE DE LILLE; F. SAVY, 24 RUE HAUTEFEUILLE, LEIPZIG; T.O. WEIGEL. SOLD ALSO AT THE APARTMENTS OF THE SOCIETY; MDCCCLXXXI. List OF THE OFFICERS OF THE GEOLOGICAL SOCIETY OF LONDON. Present, R. Etheridge, Esq., F.R.S. Wice-Prestents. John Evans, Esq., D.C.L., LL.D., F.R.S. Prof. J. Morris, M.A. J. W. Hulke, Esq., F.R.S. H. C. Sorby, Esq., LL.D., F.R.S. Secretaries. Prof. T. G. Bonney, M.A., F.R.S. | Prof. J. W. Judd, F.R.S. PForeiqn Secretary. Warington W. Smyth, Esq., M.A., F.R.S. Creasurer, J. Gwyn Jeffreys, Esq., LL.D., F.R.S. COUNGEIL, H. Bauerman, Esq. J Gwyn Jeffreys, Esq., LL.D., F.R.S. Prof. J. F. Blake, M.A. Prof. J. W. Judd, F.R.S. Prof. T. G. Bonney, M.A., F.R.S. Sir John Lubbock, Bart., M.P., F.R.S. W. Carruthers, Esq., F.R.S. Prof. N. S. Maskelyne, M.P., M.A., F.R.S. Prof. P. M. Duncan, M.B., F.R.S. Prof. J. Morris, M.A. R. Etheridge, Esq., F.R.S. J. A. Phillips, Esq., F.R.S. John Evans, Esq., D.C.L., LL.D., F.R.S. F. W. Rudler, Esq. Lieut.-Col. H. H. Godwin-Austen, F.R.S. | Prof. H. G. Seeley, F.R.S., F.L.S. J. Clarke Hawkshaw, Esq., M.A. Warington W. Smyth, Esq., M.A., F.R.S. Rev. Edwin Hill, M.A. H.C. Sorby, Esq., LL.D., F.R.S. W. H. Hudleston, Esq., M.A. H. Woodward, Esq., LL.D., F.R.S. J. W. Hulke, Esq., F.R.S. Assistant-Secretary, Librarian, anv Curator. W.S. Dailas, Esq., F.L.S. Clerk. Mr. W. W. Leighton. Librarp and fMuseum Assistant. Mr. W. Rupert Jones. TABLE OF CONTENTS. ANDREWS, Rey. W. R. Note on the Purbeck beds at Teffont .... 251 BuakE, Rey. J. F. On the Correlation of the Upper Jurassic Rocks of England with those of the Continent.— Part I. The Paris erasOTPMn Obs NONOVIE hore cs sie oie eile sarees sino) o sie lel ei s'= sige e oie 497 Bonney, Rev. Prof. T. G. On the Serpentine and associated Rocks of Anglesey ; with a Note on the so-called Ser pene of Porth- TLE De 6 sin agibroidit: cid OxeRDON IG I ena EDs nee 40 On a Boulder of Hornblende Picrite near Pen-y-Carnisiog, PMNS SVM Ee Petra Pela cio ah nei Stai's, hs = oie ois! at slausiia's: oe) sileyeie ele 5 « 137 Notes on the Microscopic Structure of some Anglesey Rocks 232 Bropis, Rev. P. B. On certain Quartzite and Sandstone Fossili- ferous Pebbles in the Drift in Warwickshire, and their probable Identity, lithologically and zoologically, with the true Lower Silurian Pebbles with similar Fossils in the Trias at Budleigh ACHE ORAM CM OTIGMILO 21) a ti c/s ta «aid eet olor Sv cack me wo elaieleres ee ene 430 Buckman, JAmzEs, Esq. On the Terminations of some Ammonites from the Inferior Oolite of Dorset and Somerset ............ 57 Buckman, 8. 8., Esq. A descriptive Catalogue of some of the Spe- cies of Ammonites from the Inferior Oolite of Dorset ........ 588 CatLaway, Dr.CHarites. The Archean Geology of Anglesey; with an Appendix on the Microscopic Structure of some Anglesey fvecuompy Eton |. (G. Bonney. (Plate VE)... 3... 210 The Limestone of Durness and Assynt.................. 239 CarPENTER, P. H., Esq. On two new Crinoids from the Upper Chalk of Southern Sweden, GElater Valle arses crates seas oe 128 CaRRALL, JAMES W., Esq. Notes on the Locality of some Fossils found in the Carboniferous Rocks at T’ ang Shan, China ...... 83 @orewerr, Dr. R. W. On Soil-cap Motion..............0.006: 348 Davis, JAMES W., Esq. Notes on the Fish-remains of the Bone-bed at Aust, near Bristol, with the Description of some new Genera 207, Syoweresy (Celera 9.0.4 De i ee an nen nen 414 On Anodontacanthus, a new Genus of fossil Fishes from the Coal-measures, with Descriptions of three new Species. TElghe OCLL) Seis RRR See inns aioe a aeeientain oi ee 427 iV TABLE OF CONTENTS. Page Dawson, Dr. G. M. Additional Observations on the Superficial Geology of British Columbia and adjacent regions............ 272 Dawson, Dr. J. W. Notes on new Erian (Devonian) Plants. (Plates GHC a>. C U0) NA aim SAMmin norma oh aboDe do Gs dccooso on cor 259 Duncan, Prof. P. Martin. On the Coralliferous Series of Sind, and its Connexion with the last Upheaval of the Himalayas.... 190 Dunn, E. J., Esq. Notes on the Diamond-fields, South Africa, 1880 609 Ecciss, James, Esq. On the Mode of Occurrence of some of the Volcanic Rocks of Montana, U.S. AS. ee acer ee 399 ETHERIDGE, R., Esq. On a new Species of Zrigoma from the Pur- beck beds of the Vale of Wardour; with a Note on the Strata by the Rey. W. ER. ANDREWS 9 .0).)02 66 «elec ieee eee 246 . On Plant-remains from the Base of the Denbighshire Grits. (Plate XOXW ies ot a ae oeacteio ravens one tiene te BPRS cc i50 6 490 Hicks, Dr. Henry. On the Discovery of some Remains of Plants at the Base of the Denbighshire Grits, near Corwen, N. Wales; with an Appendix by R. EruEripeGs, Esq. (Plate XXV.).... 482 Hotiinewortu, G. H., Esq. Description of a Peat-bed interstrati- fied with the Boulder-diift at Oldham .....--.. 72 eeeeee 713 Hormss, T. V., Esq. The Permian, Triassic, and Liassic Rocks of the Carlisle Basin. (Plate X05)... 3. oc). 9. 286 Jupp, Prof. J. W. On the Occurrence of the Remains of a Cetacean in the Lower Oligocene Strata ofthe Hampshire Basin; with an Appendix by Prof. BG. Simtiny Pues ts. ee eee 708 Krrpine, H., Esq., and E. B. Tawnry, Esq. On the Beds at Headon Hill and Colwell Bay in the Isle of Wight. (Plate V.) 85 KEEPING, WALTER, Esq. The Geology of Central Wales; with an Appendix on some new Species of Cladophora, by CHARLES LAPWORTH, Hsq. (Plate VIL) 9.02.0. ee 141 KENDALL, J. D., Esq. Interglacial Deposits of West Cumberland and North Lancashire. (Plate TMI)... co. a oe eee 29 LaPworTH, CHAR Es, Esq. On the Cladophora (Hopk.) or Dendroid Graptolites collected by Prof. Keeping in the Llandovery Rocks of Mid Wales. (Plate VIL)... ......522 08 oe 171 MackinTosu, D., Esq. On the precise Mode of Accumulation and Derivation of the Moel-Tryfan Shelly Deposits; on the Discovery of similar High-level Deposits along the Eastern Slopes of the Welsh Mountains ; and on the Existence of Drift-zones, showing probable Variations in the Rate of Submergence.............. 301 Moor®, Cuaruzs, Esq. On Abnormal Geological Deposits in the Bristol District. 0 5.0.5. 620+ aes oe on ene 67 Owen, Prof. R. On the Order Theriodontia, with a Description of a new Genus and Species (4lurosaurus felinus, Ow.). (Plate IX.) 261 Description of Parts of the Skeleton of an Anomodont Reptile (Platypodosaurus robustus, Ow.). Part II. The Pelvis. (Plate) ee oa eee oes 266 TABLE OF CONTENTS. Vv Page PaRKINSON, C., Esq. p Upper cope) and Chloritic Marl, Isle of Ne Nee hea. ores wi sialcs PATS Ss Stowell aGocuarl Siw Beat 370 Puitiies, Joun ARTHUR, Esq. Note on the Occurrence of Remains gigcecent Plantsiim row? Irom-0re. .32)) ines ce aes Soe wie ee if . On the Constitution and History of Grits and Sandstones. perma SPU Orme: Meoes uateape aldlai beets cal sky os altho sya) Gai io:4 vise ale «2 36s 6 REavDeE, T. MELLARD, Esq. The Date of the last Change of Level in 1 IDOE SIDTRG eo SiS es APO Come SEO ER cP ene BOR ran 456 RuTLeEY, FRanK, Esq. The Microscopic Characters of the Vitreous Rocks of Montana, U.S. A.; with an Appendix by JAMES EccLEs, IPT opm OL ALCP NO NG clare teccrs esol, oid agetd Sa cls aie ealocenepeianc ie ptt w ofiere, Sele 591 On the Microscopic Structure of Devitrified Rocks from Beddgelert and Snowdon ; with an Appendix on the ae Rocks of Skomer Island. ’ (Plate O05) ete Peet pn see En NO 403 SrEeey, Prof. H.G. On Remains ofa small Lizard from the Neo- comian Rocks of Comén, near Trieste, preserved in the Geological Museum of the University of Vienna. (Plate IV.) .......... 52 The Reptile Fauna of the Gosau Formation preserved in the Geological Museum of the University of Vienna; with a Note on the Geological Horizon of the Fossils at Neue Welt, west of Wiener Neustadt, by Prof. Epwarp SvusEss. (Plates ETT OO. Ae ee eee cheno PRS ee Pa aa art 620 Note on the Caudal Vertebra of a Cetacean discovered by Prof. Judd in the Brockenhurst Beds, indicative of a new Type allied to Balenoptera (Balenoptera Juddi.)... 1... cece ee eae 709 SHRUBSOLE, G. W., Esq. Further Notes on the Carboniferous “PS ES YSIIEGISS J 5G hte Gasser Me NEy CRANE AEE 178 Sounas, Prof. W. J. On Astroconia Granti, a new Lyssakine Hex- actinellid from the Silurian Formation of Canada ............ 254 On a new Species of Plesiosaurus (P. Conybeart) from the Lower Lias of Charmouth; with Observations on P. megace- phalus, Stutchbury, and P. brachycephalus, Owen ; accompanied by a Supplement on the Geographical Distribution of the Genus Plesiosaurus, by G. F. Wurpporne, Esq. (Plates XXIII. & URN Pen cia ee ey ire ai Sia nah a tiene Rane! iz teieie dase the 440 Susss, Prof. Epwarp. Note on the Gosau Beds of the Neue Welt, UE SDROMAY TENCI NGUSCAGCG. «206s syn ceunredials's a Se ds be eernae we 702 Tawney, EK. B., Esq., and H. Kreprne, Esq. On the Beds at Headon Hill and Colwell Bay in the Isle of Wight. (Plate V.) 85 Ving, G. R., Esq. Further Notes on the Family Diastoporide, Busk, § pecies from the Lias and Oolite. (Plate XIX.)...... 381 Silurian Uniserial Stomatapore and Ascodictya.......... 613 Waters, A. W., Esq. On fossil Chilostomatous Bryozoa from South- west, Victoria, Australia. (Plates XIV—XVII.)............ 309 vi TABLE OF CONTENTS. Page Wutpporne, G.F., Esq. Tabular Synopsis of the Geographical Distrbucion onthieyieleslosaursrar | easels eee is ciel aes tele ioe renner 480 WILLETT, Epear W., Esq., and Henry WILLETT, sq. Notes on a Mammalian Jaw from the Purbeck Beds at Swanage, Dorset. 376 PROCEEDINGS. Page anual’ Reports. cecteeaela t reece) ite oe 8 duist of Horeien Miembersi). caer eerie eee 16 List of Foreign Correspondents .........-. +... eee eee eernneeee 17 List of Wollaston’ Medallists\2 72); oni... 0 0 -isie ere eeee 18 List of Murchison Medallists. 066.522 «joes aie eee 19 Taist of lyellt Medallists ie. se ieee ae eeeeeie 20 Mist ot Biesby Medallists) 2 ee retteeee 21 Mimancial Report. ce. cre crete peters ellie rie Rents sie “ag eae 22 Award of the Medals &. 0.05. 6 oie 1. sie she oes ile aie ketene ee 28 Anniversary Address... 2). ciel eee cine Cle elo eee 37 Donations to the Library (with Bibliography) .................. 244 Presentation of Portrait of William Smith? 2.05. oon 2 Haun, Dr. Otto. On Microscopic Sections of Meteorites ........ 7 Lones, F. D., Esq. On some Specimens of Diastopora and Stomato- pora from the Wenlock Limestone. (Abstract.) ............ 2.39 Bequest of Drawings of Fossil Fishes by Sir Philip Egerton ... ni 241 LIST OF THE FOSSILS FIGURED AND DESCRIBED IN THIS VOLUME. [In this list, those fossils the names of which are printed in Roman type have been previously described. ] Name of Species. Formation. Locality. Page. Incerteé sedis. Tubutella ambig“ua .......csceceeceecees |Jurassic &c..,.... [pe Britain’ esse Pvt PLANT. Aitheotesta devonica. PI. xii. f. 14. \ | (Scotland ...... 306 Archeopteris, sp. Pl. xiii. f.19 ... | | New Brunswick 305 Asteropteris noveboracensis. PI. xii. | | RNS e so inincio'rsiesivn's «see ce sciees New York ... 299 Cardiopteris eriana. P\. xiii. f. 18.. New Brunswick 305 Celluloxylon prim@vum ....0..ee0.00e- | | New York ... 302 CIOPLCTISN SD! gio cejsicess aise cetee eecie ss | . | New Brunswick 305 Cyciostigma affine. Pl. xii. f. 11, 12 p PSOE Bac sccu: 4 New York ... 301 Dicranophylium australicum. Pl. | sai Te, Aa ee Australia ...... 306 Equisetides Wrightiana. Pl. xii. f.10, [P| STG 15 7A Dee eee ee New York) .2: 301 Lepidodendron primevum. PI. xii. RBSRPRRC Rae ons Sieacincencinieesedseonss ) \ New York ... 302 ue ematop ee, I STITHEIN Fagabeose WaleSiteces.c.8 494 a fier oeamnosa: 4 El xi. } Devonian......... New Brunswick 305 Pachytheca. P|. xxv. f.7-9 ........ Silurian sceece-e: Wealesm.css.cce: 483 SPONGIA. Astroconia Granti ......... Gonocsdcdden UTTER ear apee | Camnada......... | 254 Vill FOSSILS FIGURED AND DESCRIBED. Name of Species. Formation. Locality. Page. C@LENTERATA. (Hydrozoa.) Acanthograptus ramosus. PI. vii. \ | CBO ga cies sanceisiveisislncis sisters gaa femaees ( 174 Calyptograptus’? digitatus. PI. vii. | : dees deaticoccingediansecucnanaeancenee 173 plumosus. PI. vii. f. 4.......... | 174 Die tyonema genome TL aL ( Llandovery ...... Central Wales { Se Caneee weelseceiacuislemece aaeacieise es | 172 delicatulum. Pl. vii. f.2 ...... | | 172 DONUSLUNE aly neville gl gece eectye: el a alee Odontocaulis Keepingii. PI. vii. f. 7) L| 176 ECHINODERMATA. Antedon impressa. Pl.vi.f.8,9... |Upper Chalk ...| Sweden ...... 135 Mesocrinus Fischeri. PI. vi. f. 1, 2: |Planerkalk ...... SaMONYV sess 130 suedicuss Pl vie fade vasneeeoss Upper Chalk Sweden meee 132 PoLyzoa. Ascodictyon radians «............+-+0. USERS A Sosecoe: Britain 618 Stellatum: cecjaddesc-cssrseneeene Caberea rudis. PI. xviii. f. 86 ...... \ (| 322 Canda fossitis. Pl. xvi. f. 51, 52... 322 Catenicella alata. Pl. xvi. f.47, 49, OOo Gos Hei aeons mee cece ee er 317 ampla. Pl. xvi. f. 46, 50...... | 317 cribriformis. Pl. xvi. f. 39... 317 elegans, var. Buskii. Pl. xvi. | fA 2 AG ee one rae ee e 317 flexuosa. Pl. xvi. f. 40, 41 . 317 —— internodia. Pl. xvi. f. 78, 79.. 318 marginata. Pl. xvi. f. 44, 45.. 317 soliday selexvie isso Or re ; 318 Cellaria fistulosa. Pl. xiv. f. 1, 2, MMICBANS Goscesoe: se Australia DOV cir eine Biaeataseiy Satara ea eee 319 qlouusosay abaleexivenie lon lezmee 321 — malvinensis. Pl. xiv. f.3...... 321 ovicellosa. PI. xiv. f. 4-6, Pl. | NVI O Qiiciae ecseeaae tee ee coe 321 Cellepora fossa. Pl. xviii. f. 89 ... 343 UT ARTIS atic benconecebuebassoe | 343 SAS DE sSecece tat scrote ce eee eee 344 Cribrillina dentipora. P\. XY. 1 33. 326 suggerens. Pl. xvii. f. 75.2... 327 terminata. PI. xvii. f. és ey (| 326 Diastopora cricopora. P|. xix. f. 18- 2 ince dale nase echt Mone tenet ea eee Ooliter2 see Britaines cee 387 — oolitica. PI. xix. f. 11-14 386 es Ee a Britain &e. ... | 384 ventricosa. Pl. xix. f.15-17.. |Oolite ............ Britain’. 2.2.4.9: 385 FOSSILS FIGURED AND DESCRIBED. 1x Name of Species. - Formation. Locality. Page. Potyzoa (continued). HEMESUCIIAS CLASS onc c.0 eevee. ca cseeee \ (| 186 DUMENSIS 55.25 sas ale sccco deen ig EST —— membranacea..................005 Carbonif Britain , 181 == COS ee ee r |Carboniferous -.. race heey 183 === GE 179 MOLY HOLAGA! 6.0 5os50cedeodeossecace ) 185 Lepralia cleidostoma, var. rotunda. \ Pls SOU, 1 Ee een ee are 336 corrugata. Pl. xvii. f. 60...... 339 monilifera, var. armata. Pl. xv. | BPM ooo Sclch oaiacis vig ciencicanbawe's 335 spumulata. Pl. xvii. f. 87...... 335 Membranipora argus. PI. xiv. f. 20, é BUN en oo Sahin tins chew eeaicerinn 24 GAME MOANA seo sicfoicie's saree <'s/ejeie as 323 cylindriformis. Pl. xvii. f. 74 323 concamerata. PI. xiv. f. 22, : PME ie icticinc nts aceeiseddesuees 24 gemmara: Pl. xvi. f..95 ...... | 325 MCE AMER ees S oSaaiso idee gies 323 lusoria. Pl. xiv. f. 14, Pl. xviii. if, eos soclgonoSO OER CHOEe Ronee ee eeenCe 324 — macrostoma. Pl. xiv. f. 18, 4095 tuaonicar | Ploxiv. £9) ..:...... 329 WVIGCTONONOPAUMUEA oo. ceccceccssaacess 326 Microporella enigmatica. Pl. xv. i, 2D AD) cocc ene g eee ae eee 331 clavata. Pl. xviii. f. 84 ...... | coscinopora, var. armata. PI. | RNS MS OMENS oe a cis vacea cases ceovelen PlVitgeemer wane .ne: S.W. Australia 331 elevata. Pl. xvii. f. 63, 64, TAL, SAT TE) Ree eee a 3 ceo Pl xvi, f. 72 ooik oc. .0s symmetrica. Pl. xviii. f. 83... violacea. Pl. xv. f. 26, Pl. xvii. yarraensis. Pl. xv. f. 27, 28... Mucronella duplicata. P\. xvi. f. 54 Clemanseme llevar. f. 91s oo 2... mucronata. Pl. xvii. f. 66 ... Porella denticulata. Pl. xvii. f. 70.. emendata. Pl. xvii. f. 69...... Porina clypeata. Pl. xvii. f. 67 ... columnata. PI. xviii. f. 88 ... | cGononatas “RI xv.t. 97 ..L... | Retepora marsupiata. Pl. xv. f. 34— 36, Pl. xvii. f. 59, 61, 76, 77 | | | | | 332 | | | \ rimata. P\. xvi. f. 48, 53...... 343 Schizoporella amphora ............4+ 341 australis. Pl. xiv. f.15. ...... | 341 conservata. Pl. xviii. f. 81 ... 340 — excubans. Pl. xvi. f. 56, Pl. PMO! Eos) ow eisicecy iocave chess 341 PMCSEGRLOR Seth S. es oh xtostanwee. ) 339 Name of Species. | FOSSILS FIGURED AND DESCRIBED. Formation. Locality. PoLyzoa (continued). Schizoporella phymatopora. Pl. xv. \ [ERO IMBY-) Sham be a remeninsarsr re cen Hors | spiroporina submersa. ventricosa vigilans. , Sp. Selenaria alata marginata. Smittia anceps. centralis Peeters eeesetesecsese CeCe ere ereeesereese esses secrcceee Pees ercereoesesssesesscoesesserecr Ceo eoeee sere eres secees Pixie! Pl. xviii. f. 94 Cece eereSoescrsenes seessseae BUGS) Cech eereceesc ese sseesceseegsorr oe Pl. xvii. f. 65 Stomatopora dissimilis inflata eoeoeecee eertesroecesoccseeesoeecs cece ‘Miocene S.W. Australia oe I Silurian Silurian eoceecocees| BILIUGAILILI w.cccvcces Mo.uuusca. (Lamellibranchiata.) Trigonia densinoda seoneseescseesessccae |Purbeck Dorsetshire ... ( Cephalopoda.) Amaltheus subspinatus ......0....004. Ammonites Braikenridgii —— Brongniarti cadomensis concavus Gervillii —- Humphriesianus linguiferus — Mansel eceveeserecsessesozsece eoceec® 08 Cesseoe eoceesseessscessss ene see eeceocesetoseocc cee ess cs eee Sececc SO Oeeeseeses seeceosecce @eeeeersccescereecosesossssecsece Subraditatusim--eencoscceneceteere Cosmoceras Caumontii Garantianum ParkimsOMlsnn.ccuscee reece Cees SOURED So5coabaccasoscocaoce Haploceras oolithicum Harpoceras adicrum concayum CONUS Cee s Pena — cycloides dispalisuimienennsesccsease cece: — Edouardianum —— fissilobatum ...............00-00- evesquellyasr sj. Seki eaccine dese Moorei Murchisone ......... ane Ne aeane ) @eececsccsccses e@oeeaeccesssece eons ectcoceesecsrce Coeee eo ee OOP eeoevesseses | > ball Inferior Oolite...| Dorsetshire ... + | \ | Page. 338 340 340 338 338 339 345 344 337 337 337 337 615 615 | 246 FOSSILS FIGURED AND DESCRIBED. Xl Name of Species. Formation. Locality. Page. Motuvsca (continued). ( Cephalopoda, continued). Harpoceras Sowerbyi...............-+. Nei (; 602 EYLOCCTUS CONFUSUIM .....00005..00000: | | | 601 ——— Budesianum ...........:.-.ses00e 601 REPRE SUIS aw ores caicie-is\sie se oer «a | || 601 Oppelia subcostata ......... Re ercniades '! 607 SPUD ACUI AAS etter o/s wie oh tive sae bos 606 PRR U ET ea eres = ctclaic cia ciciaicis'o eer ee | 606 Perisphimctes Martinsii ............... 602 BV SIESUIN ose se cca caceee ese | 602 Spheroceras Brongniarti ............ 097 ———— GEMOEPNUM « ...--.2.22-20+eseeee | 598 Mansel 02022070, F imferior Ootte....- Dorsetshire ...4 | 597 REMI CMS 80 ce ccioosciciinij> inch in diameter. Some of the quartz fragments are so traversed by the moss-like greenish mineral, often forming a constituent of the cement of sandstones, as almost to suggest the idea of their being pseudo- morphs after a mineral which has disappeared. If seen in polarized light, however, they will be observed to be each made up of several distinct grains, in the fissures between which the substance referred to has obviously been deposited. The quartz of this sandstone occa- sionally encloses a few needles of schorl or hair-like crystals of rutile ; cavities containing bubbles are rare. A rock belonging to the Coniston-Grit series from Green-quarter Feli, Westmoreland, consists of angular grains of quartz and felspar, united by a siliceous cement traversed in all directions by numerous greenish BEerouths. The average size of its constituent particles does not exceed 5,45 inch, allow there are a few larger ones, measur.ng about 1, inch in came cen! The quartz contains on fluid- - cavities with bubbles ; but when these occur they are extremely minute. This rock contains a little iron pyrites, and the cement is sometimes stained by hydrated ferric oxide; a few flakes of colour- less and dark-brown mica are occasionally seen between the grains of quartz and felspar. Devonian.—The majority of the siliceous grits of Cornwall are usually regarded as being of Devonian age; but it is probable that some of them may be of older date. -Two distinct beds of such rock, of a greenish-grey colour, which are worked for road-metal, are quarried on the farms of Tregian and Dairy, in the parish of St. Ewe, near St. Austell, and were noticed in a previous paper under the name of slaty agglomerates*. In both these localities the grit contains angular fragments of a soft clay-slate of a greenish-blue colour, and is exceedingly hard and tough. The rock quarried at Tregian is composed of a mixture * Quart. Journ. Geol. Soc. 1878, vol. xxxiv. p. 476. 10 J. A. PHILLIPS ON THE CONSTITUTION AND of angular pieces of quartz and felspar, of which some of the larger fragments have a diameter of =, inch; they are cemented by a siliceous concrete enclosing particles of granular quartz, through which minute greenish microliths are plentifully disseminated. Comparatively few fluid-cavities, either with or without bubbles, are present. In addition this rock contains a few crystals of schorl enclosed in the quartz, some water-worn flakes of silvery-white mica, a few crystals of pyrites, and perhaps a little altered mag- netite. The felspar chiefly belongs to a triclinic species, but ortho- clase is also present. The rock at Dairy differs from that at Tregian only inasmuch as it contains a few water-worn crystals of horn- blende and a little magnetite. In addition to the foregoing, through the kindness of Mr. J. H. Collins, I have been enabled to examine four other specimens of Cornish grit, namely, one from St. Allen, four miles north of Truro, two from Ladock, five miles further east, and one from Perranzabuloe, on the Bristol Channel. Hand specimens of all these rocks closely resemble one another, excepting that those from Ladock enclose numerous angular frag- ments of a greenish slate, which the others do not, and that one of them contains a number of rounded quartzose and other grains 7 inch in diameter. When examined under the microscope, the St. Allen grit differs little from those at St. Ewe; the quartz is angular and transparent, the largest fragments having a diameter of about = inch, and they sometimes, though rarely, enclose minute crystals of tourmaline. A little hornblende, with white mica and epidote, are also present. The felspar is, to a large extent, triclinic, but there is also some altered orthoclase ; the quartz contains but few fluid-cavities. In the rock from Ladock, which contains small rounded grains of quartz, felspar, and other material, these bodies are sparsely dissemi- nated throughout the mass of the normal grit; and a microscopical examination shows that some of them are fragments of volcanic rocks closely akin to the “‘ greenstones” and “ dunstones” of many parts of Cornwall, but which have often become so altered as to be recognizable only by their felspars and general structure. Fig. 2, Pl. I., represents, in black and white, a fragment of volcanic rock which occurs in this grit, as seen in polarized light, magnified 18 diameters. The second specimen from this district is made up chiefly of angular fragments of quartz and felspar, united by the usual cementing con- crete. It contains a considerable amount of felspar, a large propor- tion of which is triclinic. The largest pieces of quartz are about ;'z inch in diameter ; and fluid-cavities, although by no means abun- dant, are more plentiful than they are in the Cornish grits before described. Some of the quartz contains a notable quantity of dis- seminated epidote (?), and flakes of white mica are frequently jammed between the fragments of which the rock iscomposed. A few minute garnets are present, as well as some fragments of a volcanic rock. The microscopical structure of the gritty rock from Perranzabuloe so closely resembles that of those from the other localities that a special HISTORY OF GRITS AND SANDSTONES. J1 description of it is unnecessary. The felspar, of which a large pro- portion is triclinic, is present in large quantities; and, as in the case of the other specimens examined, the quartz contains but few fluid-cavities. It will be observed that in this respect the quartz of these grits materially differs from that of the Cornish granites, in which fluid-cavities with bubbles are abundant. Carboniferous.—A. fine-grained yellowish-white sandstone from Shalk Beck, Cumberland, belonging to the Yoredale series, much spotted by stains of hydrated ferric oxide, and rendered somewhat mealy a) the presence of kaolin, contains but few fragments of more than ;4, ich in diameter. The quartz is in angular pieces, colour- less and transparent, and, to a large extent, free from fluid-cavities, which, when present, are, for the most part, full, and consequently without bubbles. In addition to quartz, with a little felspar and kaolin, the only recognizable mineral is white mica. For the analysis, see p. 21. A sandstone from Brigham, Cumberland, belonging to the Mill- stone-Grit series, was examined both microscopically and chemically, and is essentially composed of a of quartz with a little fel- spar, the grains being usually about +45 inch in diameter, united by a siliceous cement, which j is sometimes a little cloudy. Between the - constituent fraements of this rock there are sometimes minute erystals of a mineral which may perhaps be epidote; and the quartz, which is colourless and transparent, encloses a few needles of tour- maline, besides containing occasional fluid-cavities, but few of which contain bubbles. For the chemical composition of this sandstone, see page 21. AtSpinkwell quarry, near Bradford, a foliated siliceous sandstone, which can be raised in the form of very large slabs, is worked in the Lower Coal-measures, and is much used in the construction of chemical apparatus on a manufacturing scale. This sandstone is mainly com- posed of fragmentary quartz and felspar, of the latter of which a portion is triclinic, united by the usual siliceous cement. The quartz is colourless and transparent, and contains but few fluid- cavities, although it sometimes encloses needles of schorl. In addition to quartz and felspar, this rock contains kaolin, with a few minute garnets, and flakes of dark-brown and colourless mica, which are more abundant along certain lines of foliation than else- where. A few minute crystalline scales of micaceous oxide of iron were observed in the eas of this sandstone. The component fragments rarely exceed 51, inch in diameter. For analysis of this rock, see page 21. The fine-grained yellowish-grey Coal-measure sandstone of Stony- hough, Workington, Cumberland, is, to a large extent, oe of minute crystals. or crystalline ageregations oi quartz of about 51, inch in diameter, somewhat loosely united by a siliceous cement, often much stained by hydrated ferric oxide. A considerable number of sandstones belonging to the Carboni- ferous period are chiefly composed of quartz crystals, which have evidently crystallized in situ, since they exhibit the freshness of out- line peculiar to crystals which have not been subjected to the slightest 1s J. A. PHILLIPS ON THE CONSTITUTION AND amount of abrasion subsequently to their formation, not a point being broken or an angle removed. Sometimes the original grains of quartz have, by the subsequent deposit of silica upon their surfaces, become converted into complete double-terminated crystals; but the forms are frequently less simple, and the faces bounding the ex- terior cannot all be referred tothe same crystal. That this crystal- lization is produced by a deposit of silica around the original grains of quartz, subsequently to their having become members of an accu- mulation of sand, was first noticed in British rocks by Professor Bonney*, and has been subsequently demonstrated by Dr. Sorby 7. It is almost equally certain, as will be subsequently shown, that a portion, at least, of the silica so deposited has been derived from the decomposition of felspar. Among highly crystalline Carboniferous sandstones may be men- tioned one belonging to the Mountain-Limestone series, which occurs at Yeathouse, in Cumberland, that of Parkhead, in the same county, and that of Augill, near Brough, Westmoreland; the two last of Yoredale age. Another sandstone in the Lower Coal-measures, worked at Barngill quarry, in the county of Cumberland, is also crystalline. Permian.—The St. Bees Sandstone, at Rheda, Cumberland, is a fine-grained reddish-brown rock, composed of a mixture of angular - fragments and minute crystals of quartz with a little felspar, the whole being united by a cement rendered, to some extent, opaque by ferric hydrate. The grains and crystals of quartz have usually a diameter of about 54, inch, and contain but few fluid-cavities. The colour of this sandstone, like that of the majority of similar rocks, is caused by a coating of hydrated oxide of iron over the surfaces of the grains and crystals of which it is composed, but which is readily removed by digestion in acids. Felspar is present in notable quan- tity, and is often considerably altered; no triclinic species was observed. A small amount of colourless mica is present in the form of water-worn flakes, together with a few imperfect crystals of schorl. Numerous other crystalline sandstones of Permian age might be cited; but those of Penrith, which have been described by Dr. Sorby, are probably the most interesting and remarkable examples. Triassic.—Among the Bunter sandstones of Lancashire and Che- shire are certain reddish-brown friable beds, possessing but little cohesion, and of which the constituent grains are all so completely rounded, that the disintegrated sand flows between the fingers as readily as shot. Deposits of such sandstones, which are distin- guished by the name of “ Millet-seed beds,” occur plentifully in the Lower Mottled series, and occasionally among the Upper Mottled sandstones, as well as in the Frodsham beds of the Keuper. A specimen of millet-seed sandstone from the Lower Bunter, obtained in the form of a core, at a depth of 1039 feet, from the * Quart. Journ. Geol. Soc. vol. xxxy. p. 666. + Address delivered at Anniversary Meeting of the Geological Society of London, 20th February, 1880, p. 36. HISTORY OF GRITS AND SANDSTONES. 13 Bootle bore-hole of the Liverpool water-works, was, with a large number of others from that district, kindly furnished to me by Mr. Charles HE. De Rance, of the Geological Survey, but was found to be too friable to admit of the preparation from it of thin sections. On examination by reflected ght, however, it was found to be so entirely made up of rounded grains, varying in diameter from +, to st, of an inch, that I did not hesitate to suggest to Mr. De Rance the probability of its origin being due to blown sands united by a ferruginous cement. These grains, of which the majority are quartz, are so rounded as to represent miniature pebbles, while a few, consisting of partially decomposed felspar, are often corroded into deep cavities on one or more of their sides. The granules of quartz, as well as those of felspar, have been covered by a thin coating of hydrated ferric oxide; while on the surfaces of the former a beautiful growth of crystals of trans- parent quartz has frequently taken place. These crystals do not often exceed +4, inch in length; but they are sometimes very perfect, with sharply defined angles, and frequently exhibit both plagihedral modifications and horizontal striation. A few crystals of pyrites and of calcite have also been formed on the surfaces of the rounded quartz-grains. Figs. 1, 2, 3, and 5, Pl. Il., drawn by Mr. F. Rutley from specimens which I selected for that purpose, and magnified one hundred diameters, represent crystals of quartz attached to rounded grains of the same mineral forming the basis of this sandstone. Fig. 4, magnified to the same extent as the others, is a grain of quartz which exhibits a depression at its point of contact with another similar body. Specimens presenting this appearance, which are not very numerous, may sometimes be the result of one grain of quartz having been forced or ground into the substance of another. A careful examination, however, of such depressions leads to the conclusion that in some cases when this pitting of a grain is ob- served a deposit of silica may have taken place upon all parts of its _ surface, excepting where it has been protected by contact with adjoining grains. Angular cores of a siliceous material which have been deposited in cavities at the point of junction of several grains are sometimes detached when the rock is carefully disintegrated by friction with a hard brush. When these adhere to one only of the adjacent grains, haying separated from the others, they obviously might give rise to depressions of the kind referred to. If, after treating this sandstone by hydrochloric acid, the residue be examined under the microscope, the presence of these siliceous bodies becomes at once apparent. They are sometimes slightly coloured by ferric oxide, and do not always exhibit colours when seen in polarized light; in other cases they afford evidence of im- perfect crystalline structure, and are occasionally colourless and transparent. Fig. 6, Pl. II., is a grain of felspar which has become so corroded 14 J. A. PHILLIPS ON THE CONSTITUTION AND that two distinct cavities have been produced in its substance. A erystal of iron pyrites attached to a rounded grain of quartz is seen in fig. 7, which, as well as the preceding figure, is represented as magnified to the same extent as the other illustrations contained in the Plate. Prolonged digestion in hydrochloric acid removes the oxide of iron, leaving the surfaces of the rounded quartz granules clean and colourless. The minute crystals of quartz which have been formed upon them, however, adhere firmly to the rounded grains, after this treatment, and no stain of ferruginous matter can be observed between their point of attachment and the grain of sand on which they have been formed. It would therefore appear that, although the sand had been covered generally by ferric hy- drates previous to the growth of quartz crystals, these have never- theless originated at those points only where a chemically clean surface of the quartzose nucleus was exposed. An analysis of this sandstone is given, p. 21. Sandstones chiefly composed of rounded weather-worn grains occur in the Lower Mottled series, at a depth of 80 feet from the surface at Stock’s Well, belonging to the Widnes water-works, and at Scott’s bore-hole near St. Helens, as well as at a depth of 260 feet in the Winwick boring of the Warrington water-works. Beds of loose, rounded sand of the age of the Lower Mottled Sandstone are known to occur at Chapel Bridge, Prescot, and in a boring a little east of Newton Bridge, near Warrington. Sandstones of this character are met with north of Eccleston hill, and asimilar bed belonging to the Upper Mottled group comes to the surface in the yard of the Bridgewater Foundry at Runcorn. By no means, however, do all the sandstones of Lancashire and Cheshire which belong to this geological age exhibit characteristics suggestive of their formation from eolian sands. At Wirral, in Cheshire, as well as sometimes in the neighbourhood of Liverpool, the Pebble- beds of the Bunter are represented by a brownish-yellow sandstone containing numerous pebbles, which is much employed for building- purposes. In many of these beds the quartz is almost entirely in the form of minute crystals, or crystalline aggregations, often united by a ferru- ginous cement, which has manifestly been introduced subsequently to the covering of the original grains with crystalline quartz. An excelient example of a non-ferruginous crystallized sandstone belonging to the Upper Mottled group occurs at Town Green, near Ormskirk. This rock is mainly composed of crystals of transparent quartz, of which the edges and angles are beautifully perfect. It is of a far too friable nature to allow of the preparation of thin sections ; but it appears to have little or no cementing material, and to be, to a large extent, merely felted together by the intergrowth of its constituent crystals. It will be needless to remark that the grains of sandstones of Bunter age are not always oleae rounded or enveloped i in crystals of quartz. Among the Keuper division of the Triassic candauinen which are HISTORY OF GRITS AND SANDSTONES. 15 mainly composed of well-rounded siliceous grains, may be cited a grey, friable, fine-grained rock from the base of this formation, which is exposed in the railway-cutting at the Runcorn station, and a dark-red sandstone of still finer texture, belonging to the Frodsham beds, at no great distance from the same locality. The Lower Keuper cupreous sandstones of Alderley Edge, Cheshire, are frequently made up of quartz crystals on which still more minute crystals of vanadinite may sometimes be distinguished by the aid of a lens. In many localities the quartz pebbles occurring in crystalline arenaceous rocks have their surfaces, and more particularly their upper surfaces, covered by minute crystals of that mineral. ‘This may be observed in the case of pebbles found in Lower Keuper sandstone in a quarry near Litherland. At Dymoke, Worcestershire, there is a Lower Keuper sandstone which is sufficiently coherent to admit of the preparation of thin sections. This is a fine-grained, quartzose, distinctly DALES rock, of which the component grains sr in diameter. These to a large extent consist of quartz, Jeena: enclosing hair-like crystals of rutile, and occasionally fluid-cavities, in some of which bubbles were observed. A certain amount of felspar, a portion of which is triclinic, is present in this rock. A few flakes of colourless mica, and a little of the- fibrous mineral which has been referred to as often occurring in the cement of certain sandstones, were also observed. The cementing material, which contains a little kaolin, encloses a few minute garnets, and is frequently stained by hydrated ferric oxide. The Waterstone beds belonging to this series enclose numerous angular fragments of dark-coloured slaty rock, some of which are as much as 3 inch in diameter. The quartz grains, many of which are ; inch in diameter, are usually much ‘younded, and not unfre- quently enclose fluid-cavities. In addition to quartz and the cementing material felspar is present, as is also, in small quantities, another mineral of a light yellowish-green colour, which I have been unable to identify, but which occasionally forms part of the cement. A fine-grained sandstone of Upper Keuper age, which occurs at High House, Warwickshire, is to a large extent composed of quartz erystals, while a bed of loose sand, found 25 fect below the surface at Frodsham, above the Keuper Marl, is, on the contrary, entirely made up of much-rounded grains. The rounded quartz grains of Triassic sandstones, when examined in a suitable medium, after the removal of their external ferruginous coating, are found to be colourless and often transparent. Grains con- teens ‘fluid-cavities are comparatively rare, but they are apparently more numerous in the Keuper sandstones than in rocks of Bunter age. 4y)5 mch in length. Many of the smaller grains in this rock exhibit, when examined in polarized light, that complex structure so frequently observed in the quartz of clay-slates and other somewhat similar rocks. A sand resulting from a disintegrated Portland Stone at Fonthill Giffard, Wiltshire, is largely composed of ovoid grains of calcite. After being attacked by hydrochloric acid, a rounded quartzose sand, amounting to about one quarter of the total bulk of the mixture, remains behind. This sand, of which the grains vary from =), to 545 inch in diameter, contains but few fluid-cavities, and these, as a rule, are without bubbles. The fragments of quartz are associated with, and not enclosed by, the ovoid grains of calcic carbonate. Cretaceous.—The Tilgate Sandstone, Ashdown Sands, from Mewaee Park, Sussex, 1s composed of slightly rounded grains of colourless transparent quartz, united by a cement consisting partly of calcic carbonate and partly of flint. The quartz is almost entirely free from fluid-cavities, but encloses a few hair-like crystals of a mineral which is probably rutile. If felspar be present it has become too extensively altered to admit of identification. i Sections have been examined of the chert known as Sevenoaks Stone, as well as of several others of Lower Greensand age. They all contain numerous fossils, particularly sponge-spicules, and in HISTORY OF GRITS AND SANDSTONES. 17; some cases consist, to a large extent, of amorphous silica; in others the rock becomes crystalline and distinctly chalcedonic. The flinty varieties often contain crystals of calcite, which cluster around an included fossil as a nucleus. All the specimens examined contain glauconite, and occasionally grains of ordinary quartz, some of which are much rounded, while others are angular. The “‘ Carstone” of Hunstanton, near King’s Lynn, Norfolk, and of various other localities, is a friable ferruginous sandstone belonging to the upper portion of the Lower Greensand formation, and occurring in beds of which the relations have not, as yet, been accurately determined. A series of specimens from this locality was kindly furnished me by Mr. 8. B. J. Skertchly, of the Geological Survey, who is at present occupied in working out the geology of the district. With the exception of certain variations in colour, these beds so closely resemble one another in their general characteristics that a description of one of them will suffice for the present purpose. A specimen of Carstone from immediately below the Red Chalk exposed in the cliff at Hunstanton was found to be mainly composed of a mixture of somewhat rounded grains of quartz, with small pebble-like granules of dark-brown iron-ore. The individual grains of these minerals vary in diameter from +4, inch to the most minute sand, although small pebbles of larger size than the highest limit quoted are not of unfrequent occurrence. The quartz contains schorl and rutile, together with a few fluid-cavities, of which the majority are without bubbles. In addition to ordinary quartz grains, this rock, when carefully disintegrated, exhibits numerous examples of the angular bodies resulting from the breaking-up of a siliceous deposit formed between the grains of the original sand, which have been noticed (p. 13) in connexion with millet-seed sandstones of Bunter age. By the prolonged action of hydrochloric acid the quartz of this rock is rendered colourless, while the globules of ferric hydrate are dissolved, with the exception of a siliceous skeleton which preserves the exact form of the original grains. These bodies do not usually exhibit colours when mounted in balsam and examined by polarized light; but in a few cases the presence of a dark cross indicates a pisolitic structure in the siliceous residue*. In addition to the foregoing, this rock contains a few minute scales of mica and a very small quantity of felspar. The majority of the grains of quartz have their angles distinctly abraded; in some instances they have been completely removed, and a pebble-like form has been the result. A specimen of Carstone obtained from a bed directly beneath the Red Chalk, afforded on analysis the following results :— * Both Dr. Percy and Professor Judd have described siliceous skeletons which occur in the pisolitic grains of Northamptonshire iron-ore:—Metallurgy, Iron and Steel, pp. 225, 226; Memoirs of the Geological Survey, Geology of Rutland &e. p. 119. @.5.G.S. No. 145. C 18 J. A. PHILLIPS ON THE CONSTITUTION AND Ao [fab PeTRONINE TLIO 54 ooo 55.0 v 3°85 Water (eee CEG dire a ok 6:56 SiGe hy he SeGec Peat tice eee engin A981 Phosphoric anhydride yee eee 0-42 PAWL UUAAATN Gute we loy nee Rete es eRe ee ene eC Salen *Reric Oxide y tea ena eee 29°17 errous OxIde « Yds sie: 0-35 APART) A eee) Seat a ee aaa i ocean a 2°43 Macon esiae Ac eeciee cea ee Reet ee 0°95 POEASSA EL. Sei ie Reale 0°48 SOdariicls eae Re Ray erp acters 0-84 100-03 Another variety of this rock from the same locality, but darker in colour, was found to contain 37 per cent. of ferric oxide and 4o per cent. of silica; the amount of phosphoric anhydride was nearly the same as in the first specimen analyzed. An examination of the spherules of various pisolitic iron-ores shows that they exhibit all the characteristics of the globular ferruginous grains found in these sandstones; and it may therefore be inferred that they have had a similar origin. A pisolitic iron- ore of Middle Neocomian age, which occurs at Market Rasen in Lincolnshire, consists to a large extent of spherules very closely resembling the ferruginous grains in the sandstones at Hunstanton. Tertiary.—Hertfordshire Puddingstone, Lower Kocene, is a conglo- merate of flint pebbles united by a concrete consisting of fragments of transparent quartz and greyish flint held together by a flinty cement. In this concrete the quartz is considerably in excess of the flint, and sometimes contains fiuid-cavities. Its fragments are all angular, and vary in diameter from =, to 54, inch. A specimen of sand from Hordwell, Hampshire, equivalent in age to the Headon beds, contains no recognizable felspar. All the quartz down to a diameter of +, inch is completely rounded; and even the smallest particles have had their angles entirely removed. Fluid- cavities with bubbles are abundant in some of the quartz consti- tuting this sand. Sand from the Marine beds, near the top of the Hempstead series, Isle of Wight, was, after treatment with hydrochloric acid, found to be composed chiefly of grains of quartz, of which about three fourths had a diameter of less than 4, inch. These, down to the finest particles, are much rounded, although still roughly retaining the form of the original fragments. The fine-grained brilliantly coloured sands at Alum Bay, Isle of Wight, of Upper Eocene age, usually classified as Lower Bagshot, have not, as yet, been definitely identified with the beds of the London Basin. By digestion in hydrochloric acid the quartz becomes colour- less; and, although not completely rounded, the angles even of the smallest fragments have generally been modified by attrition. Fluid- cavities are not plentiful, and when present seldom enclose bubbles. HISTORY OF GRITS AND SANDSTONES. 19 Needle-like crystals of schorl are sometimes enciosed in this quartz. The most considerable bed of sand at Bovey Heathfield, Devon- shire, no. 27 of Mr. Pengelly* (Miocene?), consists largely of quartzose fragments, nearly all of which are sharply angular, trans- parent, and colourless. ‘They contain fluid-cavities with bubbles ; but the latter appear to be less numerous than in the quartz of some Cornish granites. Schorl is present in considerable quantity, both as detached crystals and as portions of crystals, also as needles penetrating quartz. Post-Tertiary.—Sand washed from the Lower Boulder-clay at Holy- well, Flintshire, is largely composed of small quartz pebbles, rounded grains of various felspathic and other rocks, and numerous fragments of millet-seed sandstone. A few unworn quartz crystals resulting from the disintegration of crystalline sandstones, and some angular grains of quartz, were also observed. Even the smallest particles of this sand are often rounded. The larger grains of a sand of Middle Glacial age which occurs in this locality are either rounded grains of quartz or of some other rock, or small pebbles of millet-seed sandstone. Those of medium size are millet-seed quartz grains, mixed with a few unworn crystals and angular pieces. of the same mineral. A specimen of Middle Glacial sand from Bagilt in the same county differs in no respect from the foregoing, excepting that erystals of quartz derived from crystalline sandstones are rather more numerous, angular fragments are less rare, and broken millet- seed grains are of more frequent occurrence. The Middle Glacial drift at Colwyn Bay is mainly composed of small pebbles of various rocks, principally of quartz, with a few unworn crystals of the same mineral, resulting from the disintegra- tion of sandstones. In this drift the smallest fragments, although generally rounded, have not been converted into minute pebbles. At different times I have examined numerous specimens of recent water-borne sands. Among these, that on the sea-shore at Pentewan in Cornwall is, as described, p. 24, perfectly sharp and angular, as is the sand on the beach at Par, about six miles further east. Accord- ing to Dr. Sorby such is also the case with regard to the sands of the modern beach at Scarborough, and those of the river-terraces at Dunkeld. A large proportion of the quartz in the sands of the Thames valley is sharply angular, although mixed with rounded grains of the same size. The grains of the auriferous sands collected on the coast of Northern California are hkewise for the most part angular, although perfectly rounded ones are at the same time present. Among the blown or eolian sands which have been examined is one from the Great African Desert, and another from Arabia Petrea. The grains of these are, without exception, much worn; and there is no admixture of the angular fragments found in we * “The Lignites and Clays of Bovey Tracey,” Phil. Trans. 1862, vol. clii. p. 1019. c2 20 J. A. PHILLIPS ON THE CONSTITUTION AND all subaqueous deposits; the majority are in the state of minute well-rounded pebbles. As in the case of water-borne sands, the effects of attrition are more conspicuous in the larger fragments than in the smaller ones; but even the most minute particles are in these sands much rounded. Exceptionally the quartz encloses fragments of a felspathic material; and fluid-cavities with bubbles are not entirely absent in African specimens. A bed of sandstone, said to occur in a salt-producing district sixty miles south-east of Tebessa, of which some years ago I brought a specimen from Tunis to London, is entirely made up of rounded orains. In that respect this rock resembles the millet-seed sand- stones of Lancashire and Cheshire; but it is unlike them, inasmuch as no crystals of quartz, or of any other mineral, have been deposited upon the surfaces of the rounded granules. Modern blown sands, of which we have numerous examples in this country, differ from desert sands and from those of certain sandstones only in being usually somewhat less completely rounded. Among the sands of this description which have been examined are specimens from the dunes at Rhyl, Flintshire, Colwyn Bay, Den- bighshire, Lytham, Lancashire, and from Perranzabuloe and Lelant, in Cornwall. Speaking generally, the sands from the northern loca- lities have been more completely rounded than those from Cornwall, and consist of a mixture of worn quartz and various slaty and other rocks, with a little felspar and a few fragments of shells. Fluid-cavities with bubbles are rare in the quartz of these sands. In addition to rounded grains of various slaty and other rocks, quartz, felspar, and fragments of shells are present; among these quartz largely predominates. The Cornish sands contain a few partially rounded prisms of tourmaline. Chemical Composition of Sandstones §c.—A microscopic examina- tion of a large number of sections of grits and sandstones haying led to the conclusion that many of the published analyses of such rocks must be of a very imperfect character, five different specimens were selected for analysis. In making these analyses I have received the valuable assistance of Mr. E. W. Voelcker, A.R.S.M.; and in each case fusion with alkaline carbonates was adopted. The estimation of alkalies was made after an attack by hydrofluoric acid, and was checked by a fusion with carbonate of calcium and chloride of ammonium. HISTORY OF GRITS AND SANDSTONES. PAA Analyses*. I. 1 III. TV. V. Winter { hygrometric ...... 125 150 050 “400 150 Combined) J/.05.5-.+- ‘935 “700 | 1:290 "850 "300 STLE ES) coo Ogee See 80-600 | 75°750 | 87-400 | 85:550 | 87-150 ALOT eee ee eee 9200 | 8227 | 3997 | 7570| 3-948 Carbonic anhydride ......... IE OZON ers cce IES7Oa | iecsaccc: 1-200 Phosphoric anhydride ...... ‘076 0-15 | trace. 070 | trace. Sulphuric anhydride ...... trace. 171 060 | trace. ‘094 IEEEIGIOXICO) -o 20... 20. -c25e0e- trace te LO ile rsa leases ses 1-352 IERCOUSIORIGE ........5..002- 2370 1352 1366 1-915 Herre persulphide (ReS,) .|....2..5. | ..-..---- 800 ‘73 °203 Manganous oxide ............ 232 | trace. PALS) A paceoagee IDEID® 48-0 /0dsseeeee ee 1-330 532 1-932 588 | 2°681 WEEN GEIC)) ieee en oe 1-285 "360 684 612 1-080 I°QUSINEC) SASe nae ROneeee pe eeeeeee 1-647 1-059 Weal "915 1:273 SG)” cocccseeCsn senate USB | es} "332 | 1-113 "840 100-197 | 100-120 | 100-307 | 100:336 | 100-271 Specific gravity ......... 2-689 | 2-464 2710 | 2-531 2-660 J. Grit, Cambrian: Barmouth, North Wales. II. Sandstone, Carboniferoust: Yoredale Series, Shalk Beck, Cumberland. Iii. Sandstone, Carboniferous: Millstone-Grit Series, Brigham, Cumberland. IV. Sandstone, Carboniferous: Lower Coal-measures, Spinkwell quarry, Bradford. VY. Sandstone, Triassic: Bunter, Bootle Well, Liverpool. Examination of Water- borne Sands. With the view of to some extent studying the action of running water upon the mineral fragments which it transports, a microsco- pical examination was made of the sands of the St. Austell river, in Cornwall. This stream, which during the summer months is a mere rivulet of moderate size, sometimes in winter becomes a considerable torrent. Its eastern arm arises at a distance of two miles and three quarters from the town, and at a height of 470 feet above the foot of the weir at the “Old Bridge.” Its western arm, which meets, the other a little north-west of the town, takes its rise in a small valley only a mile from the point of origin of the more easterly branch. From the bridge the distance to the sea at Pentewan is four miles, while the total fall is only 114 feet. This stream formerly carried with it into the bay vast quantities of the granitic sand which is separated by washing from china-clay at the different clay-works in the district. As, however, catch-pits have of late years been employed for the purpose of retaining it at * Since the above were completed, my attention has been directed to some analyses published in U.S. Geol. Survey, XLth Parallel, vol. ii. pp. 35 & 246, which agree very closely with those here given. + A white fine-grained sandstone, much spotted with brown. - 22 J. A. PHILLIPS ON THE CONSTITUTION AND the several works, the quantities thus transported are now very small. On the eastern branch a certain amount of granitic sand escapes into the stream almost immediately at its source; this is repeated at short intervals for a distance of a mile and a half, and finally ceases a mile and a quarter above the Old Bridge, at a height of about 200 feet above the weir-foot. The first introduction of sand into the western branch takes place nearly a mile below its source, and is discontinued half a mile further down its course, but at a somewhat lower level than in the case of the eastern fork. It follows that, before arriving at the bridge, each grain of sand must have travelled over a distance of at least a mile and a quarter, with a fall of above 150 feet, while a portion of it has been trans- ported two miles and three quarters through a channel thickly strewn with granite boulders, and having a fall in that distance of 470 feet. Before arriving at the sea, therefore, the whole of the sand must have travelled at least five miles and a quarter, with a minimum fall of about 270 feet, while a portion of it will have been transported a distance of six miles and three quarters over a total declivity of 584 feet. The first samples of material were collected from the bed of the river a little below the bridge, and a mile and a half from the point at which the last granitic sands are discharged into the stream. For the purpose of facilitating a microscopical examination of these sands, they were divided into four different parcels by a series of sieves, the first sieve allowing to pass through it all fragments less tesa =|, inch in diameter, the second those haying a less diameter than 4 inch, and the third all particles having a smaller diameter fo} Wi than —~ inch. The lar rgest fragments, retained upon the coarsest sieve were about 7 inch in inmeten eraduating to a diameter of =, inch. This sand consists of a mixture of quartz, felspar, schorl, and mica, in which the last-named mineral is present in smaller pro- portion than any of the other minerals. When examined by re- flected light, and magnified 20 diameters, the edges and points of the different fragments of quartz and schorl are found to be sharp and unrounded ; the only exception being in the case of certain grains of quartz, which Dr. Sorby suggests may have been corroded by the action of alkaline waters, but which may have perhaps never possessed other than rounded outlines. The angles and edges of the felspar and mica are, on the contrary, distinctly ‘Tounded ; and although this might have been anticipated as regards the mica, it is at first’ sight not so easily understood in the case of felspar, whose density and hardness differ but slightly from those of quartz. It must, however, be remembered that the grains of felspar had become externally kaolinized while still form- ing an integral portion of the decomposed granite, and that, on the HISTORY OF GRITS AND SANDSTONES. 23 removal of this coating of china-clay by washing, a rounded central nucleus will remain. Sand which had passed through the ;5-inch apertures of the first sieve, but which was retained by the ,-inch openings of the second, has a similar composition to that of coarser grain. The quartz and schorl are both angular; the felspar is more rounded than the quartz; and the mica, which is more plentiful than in the coarser sand, is much worn on the edges. The sand, which after passing through the j,-inch sieve was retained on the ,1,-inch sieve, consists of a mixture of angular quartz, unworn crystals of tourmaline, grains of felspar (some of which are rounded) and flakes ‘of mica (which are much worn on the edges). An examination of the material which passed the sieve having apertures ;}, inch in diameter, shows that its grains are entirely unwaterworn. The proportion of mica is much larger than in the coarser sands, and there is less felspar; but most of the grains have had their angles removed. ‘The quartz is generally in the form of tabular flake-like fragments, while the schorl often oceurs as small acicular crystals. The quartz of these sands is frequently penetrated by needles of schorl; and when mounted in balsam it is seen to be full of fiuid- cavities containing bubbles—in this respect differing entirely from the quartz of the Cornish grits, as well as from that of the majority of sandstones. Specimens of the sandy deposit were taken from down the course of the river, at intervals of a mile apart, the last having been obtained at a point slightly above the sea-level at high water at Pentewan. In every case, however, they so exactly resembled those first taken from below the Old Bridge at St. Austell as not to require detailed description. The quartz and schorl are angular, the felspar is more or less rounded, and the larger flakes of mica are worn at the edges. With regard to the distribution of sand along the river-bed, it is needless to remark that the coarser fragments are found towards the centre of the stream, while the finer silt, with minute flakes of mica and quartz, accumulates in less rapidly moving currents near the banks. Having found that quartz grains below —4, inch diameter are not in the slightest degree rounded by a minimum transit of five miles and a quarter down the course of the stream, it was thought desirable to ascertain the effect of a prolonged action of the waves upon the sand lying on the sea-shore. It must be here remarked that since the first opening of china- elay works in this district, now about sixty years since, millions of tons of granitic sand have been carried into the sea by the streams into which it was discharged. The effect of this at Pentewan has been to silt up the harbour to a very serious extent ; while the whole of the sands upon the sea-beach bear evidence of having been derived from the same source. The point from which the specimens were taken for examination 94. J. A. PHILLIPS ON THE CONSTITUTION AND is situated at a distance of half a mile west from the present mouth of the stream by which the sands were brought down. They were collected from the water’s edge at half-tide; and as the discharge of granitic sands into the river has been very small during the last ten or twelve years, and this point is considerably removed from its mouth, it is evident that a large proportion of the grains taken must have been subjected during many years to the wearing action of the - waves. An pao of this sand shows that quartz having a diameter between =} go and zy 5 inch is usually angular, although some of the larger pieces are “distinctly (but not considerably). rounded. The schorl, like the quartz, generally” presents sharp angles, although somewhat abraded grains ‘of this mineral are occasionally met with. Nearly all the felspar is rounded to a considerable extent, as is also the small Graal of mica which is present. Below =; inch in diameter the angularity of the fragments of quartz and schorl is perfect, with the exception of ‘occasional ‘ corroded ” grains ; the felspar has, for the most part, rounded outlines; and mica is almost entirely absent. At page 32 of his Address to the Geological Society of London (1880) Dr. Sorby remarks :—“ Unfortunately I am not acquainted with sufficient facts to prove how long it would require to thoroughly wear down and round a grain 4, inch in diameter. It is evident it is avery different thing from the wearing of a pebble, and may require a longer period of wear than we might suspect, if we did not bear in mind that when buoyed up by water the friction of such small particles on the bottom must be always small.” Again, at page 34, he says :—‘‘ It appears to me sufficiently proved that a great amount of drifting and mechanical action must be segue to wear down angular fragments of quartz into rounded grains 7 )q inch i in diameter, ith i have taken as the standard for comparison.” Professor Daubrée states that the diameter ot erains capable of floating in slightly agitated water is about 4, millimetre, or, say, sti inch, and remarks that all smaller grains must of necessiby remain angular *. He subsequently says ‘That a current or wave capable of carrying off in suspension particles of that diameter, with- out in any way affecting their form, would cause larger fragments of the same mineral to so rub one against another as gradually to produce rounded sand. According to an experiment quoted by this STN, a sand of which the grains have a diameter of ;°, millimetre, say =, inch, to which a movement of one metre pe Tagan 1s Im- ee becomes rounded, with a loss equal to 754 pp of its weight per kilometre traversed. This experiment appears to indicate that a grain of quartz zy inch in diameter requires, before becoming completely rounded and assuming the form of a miniature pebble, an amount of abrasion equal to ‘that which would result from having travelled a distance of three thousand miles. In arriving at this conclusion the fact must not be lost sight of that, after the first rounding of the * Géologie Expérimentale, p. 256. HISTORY OF GRITS AND SANDSTONES. 25 angles and edges, the operation will gradually become slower as the surfaces become more worn and the weight of the grain decreases. That angular fragments of quartz having a diameter of less than =i; inch remain unrounded by the continuous action of breakers after many years’ exposure, is evident from an examination of the sands at Pentewan. It has been shown by other evidence, as well as by the recent experiments of Professor Daubrée, that the rounding-down of such sands by the action of running water must be an exceed- ingly slow operation, and one requiring a somewhat active current with an amount of friction equivalent to transport over enormous distances. Grains of quartz of similar dimensions are, in blown sands, completely rounded. Summary and General Conclusions. The Cambrian grits of Barmouth contain quartz and felspar, both in the form of angular fragments and also as rounded pebbles. The materials presenting these different forms have probably been de- rived from two distinct sources; while the large size and complete sharpness of the angles of mauy of the irregular grains appear to indicate that they cannot have been transported from any consider- able distance, and that the felspar cannot have been derived from kaolinized granite. All the arenaceous rocks of Silurian age which have been ex- amined contain a small proportion of felspar, the grains of the various constituent minerals being in some cases angular, and in others rounded. Many of the rocks belonging to this period are composed of a mixture of grains of both forms. Among rocks mainly composed of rounded grains are the Stiper Stones of Shropshire and the Lower Lickey Quartzites of Westmoreland. Some of the grits from the neighbourhood of Aberystwith enclose fragments of a volcanic rock of doleritic character. The grits of Cornwall, which are of at least Devonian age, include flakes of soft slaty rocks, the edges of which are perfectly sharp, together with angular fragments of the well-known “ greenstones ”’ and ‘‘ dunstones” of that county. A large number of the Carboniferous, Permian, and Triassic sandstones are composed almost entirely of quartz crystals, which have undoubtedly been produced in sztu, as they not only penetrate and interpenetrate one another, but also exhibit the most perfect sharpness and freshness of outline. As confirmatory of this hypo- thesis it may be mentioned that in a quarry at Foggen Tor, on Dartmoor, the felspar has in places become decomposed into soft kaolin, in which the liberated silica is imbedded in the form of ageregations of well-formed and transparent quartz crystals *. Unworn double-pointed crystals of quartz, likewise resulting from the decomposition of felspar, have. recently been found near St. Austell, Cornwall, in soft china-clay; one of these, more than three * Quart. Journ. Geol. Soc. vol. xxxvi. p. 9. 26 J. A. PHILLIPS ON THE CONSTITUTION AND inches in length, is now in the collection of the Museum of Practical Geology. Sandstones of this description are not unfrequently with- out any kind of cementing matrix, being merely felted together by a matted intergrowth of their constituent crystals. Professor Daubrée entertains the opinion that crystalline sand- stones frequently owe their origin to chemical agencies resulting from an outpouring of igneous rock; but, although this may some- times haye been the case, many of the most completely crystalline British sandstones are situated at distances of many miles from any known rock belonging to this class. The same author maintains that the presence of anhydrous ferric oxide in sandstones affords evidence of their having been subjected to a high temperature*. It must, however, be remembered that the carnallite of Stassfurt, which has evidently never been highly heated, contains crystals of specular iron-ore. Numerous fine-grained sandstones, particularly among those of Triassic Age, are composed of quartz grains so completely rounded as, under the microscope, to resemble well-worn pebbles. These ‘* millet-seed”’ sandstones are often coloured either red or brown by variously hydrated oxides of iron; and in some cases minute, perfectly formed, and beautifully transparent crystals of quartz have been developed upon their surfaces. On attacking the sand of such sandstones with hydrochloric acid, the oxide of iron is easily removed, put the crystals of quartz still remain firmly attached to the surface of the grains upon which they have grown. It would also appear that crystals have been formed upon those parts only of the grains which, having been free from a coating of oxide of iron and from every other extraneous material, have admitted of direct chemical contact between the silica of the rounded quartz and that of the subsequently formed crystals. In addition to silica in the form of perfect crystals of quartz, that substance has often been deposited in such a way as to fill cavities existing between the original grains of sand. As this silica has frequently been thrown down upon a thin deposit of ferric hydrate, it is generally detached by prolonged digestion in hydrochloric acid, by which the intervening ferruginous coating of the grains is ultimately removed. When such a deposit of silica adheres to only one of the adjoining grains it may give rise to a de- pression upon its surface of the kind represented in fig. 4, Pl. II. On examining a considerable number of modern sands, none of them, excepting such as had long been subjected to the wearing effects of wind action, were found to resemble those of the millet- seed sandstones in having all their grains reduced to a pebble-like form. Among these the grains of blown desert-sands most com- pletely resemble those of millct-seed sandstones. * Géologie Expérimentale, pp. 226-230. + With regard to the Torridon Sandstones of the Central Highlands it has been observed by Professor Bonney that wherever “dirt” has been deposited upon the siliceous grains their agglutination has been prevented (Quart. Journ. Geol. Soe. vol. xxxvi. p. 106). HISTORY OF GRITS AND SANDSTONES. 27 The above facts would appear to render it probable that the rounded grains of these sandstones may be of eolian origin, and that, during certain periods of Triassic time, desert areas with blown sands extensively prevailed in this country. Mr. De Rance has observed that the millet-seed beds are usually free from pebbles, shale-beds, pseudomorphs after common salt, and from all traces of life*—conditions which are characteristic of de- posits produced by wind-currents. The granules of brown iron-ore which are so plentiful in the **Carstones”’ of Hunstanton are pisolitic grains, and not fragments of that mineral rounded by attrition. An instructive example of the occurrence at the same time of rounded and angular grains is met with in the Interglacial sands of Flintshire, where seme of the pebbles are fragments of a millet- seed sandstone, while many of the smaller particles are grains detached from the same rock. . HXPLANATION OF PLATES L. & II. Puate I. Magnified 18 diameters. Fig, 1. Group of felspar crystals in Cambrian Grit. Polarized light: p. 7. 2. Grit from Ladock, Cornwall, enclosing a fragment of a volcanic rock. Polarized light: p. 10. Prats IT. Magnified 100 diameters. Figs. 1, 2,3, & 5. Crystals of quartz deposited upon rounded grains of the same mineral in Bunter Sandstone: p. 18. 4. Depression in a grain of quartz from the same sandstone. 6. Corroded grain of felspar from the same. 7. Rounded grain of quartz with attached crystal of iron pyrites. Discussion. The Presrpent expressed his sense of the value of Mr. Phillips’s communication. Dr. Sorpy expressed his agreement with the paper, to which he had listened with great interest, especially as the author had ap- proached the subject from a point of view somewhat different from his own. He was especially glad to find that his opinions were confirmed by the author, especially as to the crystals of quartz in certain sandstones. The observations as to the time required to wear down a grain of sand were especially valuable. He had found the drift sands of the Yorkshire coast almost all angular; but then those examined by Mr. Phillips were from another locality, which might explain the difference in their observations. He should only regard sand as eolian when a very large proportion of grains were rounded. * “Further Notes of Triassic Borings near Warrington,” read before the Manchester Geol. Soc. June 29th, 1880. 28 ON THE CONSTITUTION AND HISTORY OF GRITS AND SANDSTONES. So far as his observation had gone, the sands of the dunes on our coasts were not much more rounded than other sands. Dr. Hicks said that the condition of the Barmouth Grit led him to think that the materials had not drifted from far. He had also observed that there was a large proportion of rounded grains in the Stiper Stones; were we to suppose these to be blown sands? Rounded grains were still more common in the quartzites of the N.W. of Scotland ; we must believe therefore that there was, still remaining at the time they were deposited, an adjoining great land- area or the materials from one not far away. Mr. Dr Rance said that the Keuper beds under the Marls were now divided into the Water-stones, soft current-bedded sandstones called Frodsham beds (which denoted entirely different physical conditions and contained the millet-seed grains), and then the Lower Keuper building-stone (Labyrimthodon-beds). Then came a line of erosion. In the Bunter series were the Upper Mottled Sandstones (with the millet-seed grains), then the Pebble Beds (which had a different kind of current-bedding from that of the Frodsham beds), then another line of erosion and the Lower Mottled Sandstone with millet-seed grains again. The bedding of the sandhills of Lanca- shire much resembled that of the Frodsham and other millet-seed beds in their high angle and rapid change. Mr. Ruttey said he had examined eruptive rocks rather than sedimentary, but could not but express his gratification at the agreement between two such observers as Dr. Sorby and Mr. Phillips. He called attention to the presence of felspar in many of the sandstones described, and suggested that it was possible for such sandstones to be changed into felstone. There was often much difficulty in distinguishing between the finer-grained igneous and sedimentary rocks. He also called attention to the develop- ment of microcrystalline structure in felspar crystals. Mr. Branrorp said that some years ago he had examined the Indian desert, and found the grains of sand well rounded. They were mostly of quartz, with a few of felspar and occasionally of hornblende. The strongest wind there blows from the west; the sands had come from the coast and the river Indus; and the sand in the bed of the river was also rounded. The blown sand ap- peared unstratified. Mr. Puitiips said that the grains of millet-seed sandstone were much more rounded than was usual in sea-sand. He thought the rounding of felspar was often due to disintegration by decomposition. With regard to Mr. Blanford’s remarks, he could only say that he believed the grains of all the desert-sands which had been yet examined had been found to be much rounded. INTERGLACIAL DEPOSITS OF W. CUMBERLAND AND N. LANCASHIRE, 29 3. LNTERGLACIAL Deposits of West CumBertAnp and Norra Lanca- sHirE. By J. D. Kenpatt, Esq., C.H., F.G.8. (Read Novem- ber 17, 1880.) [Puate IIT.] CoNTENTS. 1. Introduction. 2. Observed Facts. 3. Deductions. 1. IyrrRopuction. Tue glacial deposits of these districts are capable of a threefold division, as below :— 1. Boulder-clay (Upper). 2. Sand, Gravel, and Clay. 3. Boulder-ciay (Lower). Seldom do we find the whole three members present in one section. Sometimes the Lower Boulder-clay alone is found; at others this is overlain by sand or gravel or clay, or by some or all of these rocks. In other cases we find the series complete. They occur almost continuously, in more or less completeness, from the sea-shore to an altitude of 500 feet; and from that level they appear in patches up to 1000 feet above the sea. The two Boulder-clays have the ordinary character, and are very much alike, except that the lower is tougher than the upper and contains larger boulders. ‘The character of the included stones is the same in both clays, as well as in the middle sands and gravel. Some most remarkable facts are presented by the distribution of these boulders*. Associated with these glacial beds, and occurring at various places, sometimes inland, sometimes along the sea-coast, between high- and low-water marks, there are a number of deposits of vege- table matter, which hitherto have been almost entirely neglected by geologists. When occurring on the sea-shore, these deposits usually pass by the name of ‘‘submerged forests.” I am, however, inclined to doubt the accuracy of this appellation. After a long and careful investigation into the nature of these deposits, I have come to the conclusion that they are not forests at all, nor the sites of forests, as will appear further on. * “On the Distribution of Boulders in West Cumberland, by iD) Kendall, C.E., F.G.S.,” vol. v. Trans. Cumberland Assoc. for the Advancement of Lit. and Science. 30 J. D. KENDALL ON THE INTERGLACIAL DEPOSITS OF 2. OBSERVED F Acts. Lindal deposit.—The first of these deposits to which I shall refer has already been dealt with by two other writers :— First, im vol. xvii. p. 274 of the Quarterly Journal, in a short paper by the late Mr. John Bolton, “On a Deposit with Insects, Leaves, &e.” ; secondly, in vol. xix. p. 19 of the same Journal, in a paper by the late Miss Hodgson. A comparison of these papers will show that the writers differed seriously about the facts with which they dealt. It is therefore not surprising that they arrived at very different conclusions. Owing to this discrepancy, I have been at some pains to ascertain the facts accurately ; and I give them below. The position of the deposit referred to by the above writers is shown on the map (Plate III. fig. 1). It occurs near Lindal in Furness. In figs. 2-4 a plan and two sections of the deposit are given. The data from which these sections were prepared are given below; they were obtained by a number of shafts and boreholes put down by the Ulverston Mining Company in search of hematite. The posi- tions of the boreholes are shown in the plan and sections. Sections of Boreholes. (Explanation of local terms used :—Pinel = Boulder-clay ; Black muck* = Vegetable deposit.) Borehole No. 41. Thickness Depth of each from stratum. surface. ft. 1. Tt, iA. Surface=soilk i. 25. cab onteenic an eee eee: 2 0 2 0 Gikey spinel. seas O 58 0 £2, DIG TLS Te EA ae eT 5 0 63°60 Meme CLAY, |... 2. -e- wee Laseedesanecesazserees 7; 0 65 0 Wellow SO8Sdn 2......i5<.csszszcccescsccctes 1 0 66. O [h.20, (UNDO) UG. Ae eee ee nen Asean erate 20 ua TCT OIHS sa cae Ae SBE aera ee eee 0 8 th ts Drarke pinelee sco< ccc) for seed cbt eee seed eons (0) TAS | TTT SST ROTI Ha 5, ae ee eee a 10226 88 2 Berehole No. 52 TL GGRCTOTL | 4 Bae eee ree tay See as YY) Zi LES, TELS) eee a ie ee ee 30 0 32 0 S15) LNG) Se nee ee ae ee 36 0 68 0 TE0) (UDnelh A On reer eee ener 9 0 TD 2 SRE ETS E Gere ee ie Aa ee ee 10 O 87 O Borehole No. 58 Brriietee SONNE och rat ee) eos oe te oo ck je car “Sse TOG) deg ee pace eee ee SiO 82 0 22 UE TSG A Bane nee ae 20 86. GO eilack Muck (WOOdY).......2.c-s.0s00es0s-+ 24. 0 ITO. © “ELISE ROTI | RA a Oe Sat Oe ee ee 20 112 @ 25G7, TID MIP Nee hee cote aera ek eee ee 3 O 11d O Red pinel and limestone .................. Dd 0 120; “0 ison ore and’ StOneS ....2...c..6+0ssececeso- 2 0 176 1 L POTTS) OTIS) gai oe eae le ae 0 10 122 10 J. D. KENDALL ON THE INTERGLACIAL DEPOSITS OF Borehole No. 59. Thickness Depth of each from stratum. surface. ft. in. vita) 00 Gin hacessOll wh esceesoceoe en Sas ee eee 2 0 20) (Civeyy jOMMVall, $4 Age ceoononoccdseosaso0q0esc0acuce 70 O 72 O Pack Much (WOOdYy))..0-.2-+ ease eee ene 10 eG Mplueusamd:) eee. va. acd. eats eee eee 2 0 75 O BLACIEGICULCK (WOOGTY, ics): 2-- cee eeecee ene il ©) 86 oO (Ble San yess vas she cciececeha ries mse eeeeeeee 4 0 90 O Regine), ese qucicec ce ss one eee eee 4 0 94 0 Weimre stones. asactien seca ee cre nce eeee gy 103 0 Borehole No. 60. Surface=soil’"... ote ccseeece scence ee eee eeee i) 0 Red Spinel. 3c pki alo dare eee 30 0 3l 0 Greyipimel ponent edocs ce coe earner ee 43 0 74 0 THAD GOOG (CiOOENS)) anncoodescosescsesnsa%a: 12 © 36 0 Bluewsand. “pe sves. cect scans cmesee cee enecene 2 0 838 0 Wellowashiclaw ero suse eee ee ce eeeeeeee 2 0 90 O Broken ground (dragged by workings).. 22 0O ZO Limestone \ian.cecenes eee eee cee eee 4 0 116 @ Borehole No. 61. Surface-soil 7 ces cece eta 0 ib Grey pinell .2) a frase aes ancien eee 69 O 70 O Vellowapinelieereseenceceres Re, eee 20 ( Y Grey spinel 284 4e seen se ee eee 6 0 fier BIGche: MUch Ne die Oe eee eee eee Go) 3 S7aas iBlnevclaya(Sandiy)) eens eee ee eeeeee eee 3 0 90 3 Limestone). ccc saanscce wer teden eae nee Oo 90 11 Redspinell #224 4: gests cere eee ee netec ene 3 0 93 11 Tumestonel. ears eee ee een eee i 94 11 Borehole No. 62. Surface-soil’ 72. canoe eon nO ei) Grey pinel. 2.0). eaeareeerten ae meeenaee 88 0 89 0 Black nick (woodsp\peeercceecsneeo ances If oO I0Oo oO Blue*clay -s.i)...ccda cree eee Caen 2A) 102 O ‘Yellow sand’ \c)2) ahi ae eee eee 6 0 108 O IRedspinelli(withtoxe) my-eetee rer eee ene 4 0 112 O Noft stone Gwithiore)wee-eeeeeeeeeee eae 3 0 115 0 Limestone with clay joints ..............- AL MD) 156 0 Borehole No. 63. Surtace sot: os cucacc.carel ce heeeene eee 1 A) he @ Greyopimel co. 2. ...025. ah voce eRe eee eee ay O 88 0 Black wniwek) (woody). ..2./s¢.cesanaa ete 10 Oo 98 oO BF uSielay i athe cncea dees Ahan nea ee 3 0 101 9 ved apimell(wathyore) eee eee eee 4 0 105 O ivedipine li (wathistone) pase -oeeree eee 8 0 113 0 iin CSLOME ransenecesscan eet ae eee eee On) ipa WEST CUMBERLAND AND NORTH LANCASHIRE. 33 Borehole No. 64. Thickness Depth of each from stratum. surface. re aya foe ue Surface-soil ..... ME Can Gah seen reaes eae 1 O 10) BERG DUONG Bee dae siccncecsarciconiek asia stnies nessa 92°, 0 93 O RICH MUCH (WOOTY) .. icccccsecesclecedenes 20 0 13 0 Bineclay ....,<.- SOLOS ROREE CNeH OAC Beare Re 2 0 Milsy PROMEEC Ne iced wainacine Sia te oresielee eter 9 0 124 0 LEDS OT AS) GRR rey a ep Re ees f O ll 124 Li MNOHEERIGHOMERE cas cc ccs si canaccaan ce wacscuncuces 0 8 WEY 7 / [TRONMEGINE: SaRs AP SOG RE a ee ean LS 126 10 PROC ISLON Omen ss one ccc. tec saan tieaein 0 8 Ie 6 PEICORE ee ech escscaesewcac sec ccseeccnusl 2 6 130 O Be ecg MM ESTOBG secs is nc wowace aaesecee' 20 0 HOO TO SMRRESCE SOMME eonece socicceaeanieswesa oaccnaeet LG 1 O RPE CHPNIN CY es sate ecieh Se tentisas sesaqas sins 99 0 100 O NEGIIUGH (WOOGY))...0.2.06..00000cess00e5 AO @ 1W6) ©) PEPER G) AY gaifoe nics chicane ce cislecicacieiviisiswesieidsis.si my AY) 122 0 SLELLORP EEOC YE ee RO 125 O HRPM WOM Ome tes secoetecattc clas felosccetionwatloalsae 3 O 128 0 Borehole No. 66 SUREKCC-SOL Uecasccecacemerss Seacoast tak ag) 1 O BCR CMM EG Wa aise is Sine ueueid se seeleatiina be tewans 96 0 ot 0 2r BULONTEP SENTING | So et Ts A) 98 O EDOELIS CIES MEE RE ee 1) 116 Oo PENITIEN SM Clere een eH Ch ecais eaten laccns wea es 2 0 118 0O BWrellFous POSsaM” Mo. c sects tenes eat venules 6 O 124 0 TL BTIDOSTGTING Sec eie ae IS Eee ae L6G 125 6 The pinel or Boulder-clay overlying the Black muck or vegetable matter in the above sections is the Upper Boulder-clay, having a greyish and ochrey matrix, and containing numerous boulders, some as much as 2’ 0" x1' 3” x1 3”; but the majority are below 3” in diameter. Coniston grits and flags, from the adjacent highlands, are the principal rocks represented. Besides these, however, there are St.-Bees Sandstone, Eskdale granite, Carboniferous Limestone, &c. The extent of the deposit has not yet been proved in cither a N.E. or §.W. direction, as shown on the section AB (fig. 3). So far as is at present known, it covers an area of about 34 acres. For a list of the plants &c. found in this deposit, the two papers above referred to should be consulted, as I had only an opportunity of seeing some of the vegetable deposit as 1t came out of the bore- holes. It was then too much broken up to enable me to identify any of the plants. Crossgates Deposits —Several deposits similar to that at Lindal have been met with at Crossgates, in working the hematite-mines. In some cases they were covered up by 9 or 10 fathoms of Boulder- clay; and they invariably rested on clay. JI have seen a large quantity of the woody matter which came from some of these de- GedaG.s. No. 145. D b+ J. D. KENDALL ON THE INTERGLACIAL DEPOSITS OF posits ; but no correct information has been preserved as to their ex- tent and thickness. Accumulations of similar material have been met with in the solid rock, as shown in the following section :— Section obtained by Boring at Crossgates. Thickness Depth of each from stratum. surface. fies a0 fie. nas Soil Danis Seer tec ate sie oars Uostasietdocioeeee aerate Het) 20 Grayeland layer. cece. s-coen cee eee 24 0 26 0 Decomposed limestone ............1--++-+-+ ge O) 43 0 Yellow clay mixed with iron-ore......... 4 0 47 O TEUIE TUG a Me RE BORCSOE PEDO DOE ORCL EO C8.00% A (0) 5r To Tron-ore (dark-coloured) ...........,..0++ 2,10 53 0 Black muck mixed with iron-ore......... Q, © 59 Oo PPOM=OVC, oe cerca sh oteael cle anen acenee eee 8 0 67 20 Decomposed limestone m-.cse eeese eee os) 74 0 LEUCOE THODUY WAVOSE see cnondccodscadnacecen: 2 86 0 Decomposed limestone ........5....-....06 (ie 92 0 Black mould and wood .......2...+.0.00+00¢ oy) 94 0 Yellow clay mixed with iron-ore......... 16 0 110 O Black mould mixed with tron-ore......... LOE 120 <0 ABU GCI I NOWLE SEN aot ee me ae ne Ly © nye) Black mould mixed with vron-ore and limestone Core cesses esesaseersessesseossossseons 250 127 116 Watney Deposit—On the western shore of the island of Walney, and about a mile south of the village of Biggar (fig. 1), there is a vegetable deposit, in many respects like that at Lindal; but not much of it is exposed. A plan and two sections of it are given in figs. 5-7. The deposit rests on Boulder-clay certainly ; but I have not been able to prove that it is overlain by the same formation. Still I think there is very little doubt about it when we look at the sec- tions, and w.cen we know that the shore is travelling rapidly land- wards. On the occasion of my visit I had not time to make any observations on the inner nature of the deposit with the view of determining the different kinds of plants enclosed in it; but I may state that, externally, it is very much like other deposits that I shall describe more in detail further on. Drigq Deposit.—On the shore opposite Drigg (Plate III. fig. 1). there is another woody deposit. A plan and several sections of it are given i figs. 8-11. In one part of it I sunk a shaft, as shown on the plan. A section of this shaft is given below. A. Vegetable matter (brown) and grey sand in alternating layers of various thicknesses, the vegetable matter predominating. I found in it the elytra of beetles, acorns, oak-leaves, hazel-wood, alder-wood and leaves, stems of common bracken, pieces of Sphagnum, seeds of various kinds and sizes, and rush-like stems and leaves, the stems standing on end and crumpled endwise, asif by downward pressure, the leaves lying on their side. The fol- lowing diatoms were found in this bed:—Cyclotella minutula, Pinnularia viridis, Pinnularia oblonga, Gomphonema acuminatum, Himantidium bidens, Himantidium pectinale. B. Yellow and brown sandy clay in irregular layers, C. Blue sandy clay. D. Red and blue sandy clay, the last 12 inches redder and more clayey. WEST CUMBERLAND AND NORTH LANCASHIRE. 30 Some of the pieces of wood in A. were as much as 12 inches in diameter, but all lying on their side and partly flattened as if by pressure. Here, as at Walney, we find Boulder-clay under the deposit, but we have no direct proof of its having been overlain by that forma- tion. St.-Bees Deposit.—A few miles further north than Driggs, along the coast, opposite St. Bees (Plate III. fig. 1), there is another deposit, which I have examined more minutely than any of the others. A plan and two sections of it are given in figs. 12-14. In this deposit I have sunk several shafts and boreholes, sections of two of which (Nos. 1 & 8) I give below. Shaft No. 1. A. Vegetable matter (brown) and grey sand in layers of different thicknesses. The vegetable matter contaims seeds of various sizes, leaves and stems of rush-like plants, hazel-nuts, leaves and wood of the oak, alder, and hazel. The rush-like stems are numerous, and vary in length from 1 to 3 mehes; they are standing erect, and crumpled endwise, like those found at Drigg. The following diatoms have been found in this layer :— Epithemia turgida, Epithemia granulata, Epithemia proboscidia, Pinnu- laria acuta. A vertebral column about the size of that of a rat was also found in this bed. B. Grey sandy clay containing a few rush-like leaves and the elytra of eetles. C. Hazel-wood, nuts, and leaves, also leaves of the oak, beech, and alder. D. Similar to A, but containing more leaves and hazel-nuts. (Boulder-clay.) Shaft No. &. A. Vegetable matter (brown), containing wood and leaves of the same kinds as those found in the same bed in Shaft No. 1. B. Fine oe clay with rush-like stems standing upright, and the elytra of beetles. C. Vegetable matter consisting almost entirely of leaves of the oak, alder, and willow, and hazel and alder wood. D. Brown sand and vegetable matter with some rush-like stems standing up- right and crumpled. On the surface of the deposit there are a considerable number of stems of oak and alder, the former turned black and the latter ee ; some of them are as much as 12 and 18 inches in diameter. T also found a large piece of the stem of a coniferous plant about 18 inches in diameter. Under the microscope it was like the yew. These stems and branches all lie on their sides. In several parts of the deposit there are root-stocks of oak and alder with about 12 inches of stem standing, as though they had grown where we see them: the rootlets, however, are imbedded only in vegetable matter ; they do not extend into the underlying earth. Near the upper edge of the shingle, on the beach, two boreholes were put down for the purpose of proving whether or not the vege- table deposit extended so far inland. One of these boreholes is shown in the section A B, fig. 13. The pees of it were as follows :— D2 36 J. D. KENDALL ON THE INTERGLACIAL DEPOSITS OF Borehole No. 2. Thickness Depth of each from stratum. surface. fos yaa fits yum. Ths SIMMS Sorccsessogossscosdoopon0d0000000 1 6 LG 2, (Senne eyarsl mph yell ccooansoancanséosnc0000 Ik 3.0 By WGI QiOUG TUCIRIEP gboodbsedooonceHs50ce 3 6 11 6 4, Boulder-clay (grey) ............se0ees ROO NEG 12 0 The vegetable matter (No. 3) in this section was very soft and spongy, easily pierced by the boring-tool, and quite unlike that found in any of the pits or bores made on the shore, which was particularly difficult to bore through; in fact No. 3 is more like the soft peaty deposit which occupies the flat ground in St.-Bees valley. Two years ago a great length of drains was cut in this peaty matter ; and I had then a good opportunity of ascertaining the nature of it. It, however, was nothing like the vegetable matter which occurs on the shore, being much more spongy, and containing a far larger quantity of water, and altogether having more the appearance of peat, which in fact itis. The vegetable matter on the beach contains very little, or almost no water, notwithstanding that it is covered twice every day by the tide. Another difference between the two deposits is that the one on the shore is quite laminated, whilst that in the valley is totally devoid of lamination. About eighteen months ago Pow Beck, during a freshet, diverted its course just where it passes on to the shingle of the beach (see Section A B, fig. 13). In the new course a large patch of peaty matter, similar to that in the valley, was exposed. After seeing that, 1 had no doubt whatever that the vegetable matter found in borehole No. 2 was the same, and not at all like that found on the shore. Maryport Deposit.About halfway between Allonby and Mary- port, and about one third of the range of the tide from low-water mark (Pl. III. fig. 1), there is another vegetable deposit; but it is not very well exposed. I have not dug through this deposit, and therefore cannot say what is below it; and the soft silt, which covers the shore in the neighbourhood, prevents any information on that head being obtained. It has the same external appearance as all the other deposits, besides being laminated and very much more compact than peat. It contains a large number of seeds about the size of gun- shot, and a quantity of rush-like leaves, as well as pieces of hazel and alder wood. The wood is in every case flattened as if by pressure. | There is another and similar deposit about a quarter of a mile further on the shore towards Allonby; but only the upper surface of it can beseen. Near Beckfoot, I am told, there is another; but that I have not yet visited ; nor have I seen one which occurs at Cardunock, on the Solway. All the deposits described, whether on the sea-shore or inland, WIEST CUMBERLAND AND NORTH LANCASHIRE. oy have the same compact nature, and, so far as I have been able to ascertain, contain the same kind of plants. 3. DEDUCTIONS. So little was known of the extent of the Lindal deposit at the time Mr. Bolton and Miss Hodgson wrote, that its real geological import- ance was missed. We now see that it is overlain by an immense mass of Boulder-clay, in some places nearly 100 feet thick; it is also underlain by Boulder-clay; so that I think we may fairly say it is interglacial. The impossibility of its being a recent introduction, carried down by means of swallow-holes, as suggested by Miss Hodgson, must be apparent to every one now that we know more of its extent and real nature. We may ascribe a like (that is, interglacial) age to the deposits at Crossgates. The deposits in solid rock were, I believe, thrown down in preexisting cavities in the Limestone at the same time as the deposits immediately below the Boulder-clay. The shore-deposits, so-called submerged forests, I think are also of the same age. We are not able to prove it so directly as in the case of the other deposits ; but I think that a careful consideration of all the facts must lead any one to the conclusion that they are interglacial. The fact upon which I most rely is their compact nature, which, to me, speaks most forcibly of their having been subjected to great pressure, such as would be the case if they had been overlain by the upper glacial beds. They are altogether unlike the spongy peat which occurs in St.-Bees valley, and quite as unlike any of the vegetable deposits which usually go by the name of peat, although many of the species of plants found in the shore-deposits are common to peaty accumulations. These shore-deposits all rest on the Lower Boulder-clay, a fact which is somewhat remarkable if they are of postglacial age. Why should we not find some of them on the Upper Boulder-clay ? Besides, how can it be for a moment doubted that the Walney deposit extends below the Boulder-clay, although this has not been directly proved. Referring to Plate III. figs. 6 & 7, we see that the deposit dips towards the land, and that it is only about 50 feet from high-water mark, where we have the Upper Boulder-clay. When to these facts we add that the shore is travelling rapidly inland, it seems to me we are bound to admit the Interglacial age of the deposit. Usually these deposits on the sea-shore pass by the name of “‘ submerged forests ;” but the conclusion to which I am led is, as already stated, that they are not forests at all, nor the sites of forests, but that the vegetable matter has been accumulated under water. All the facts but one point in this direction: the Diatoms, the rush-like stems and leaves (probably of a species of Sparganium), and the interbedded layers of sand, all speak of watery conditions ; the only fact which seems to me to indicate in any way that these deposits are the remains of ancient forests that grew on the spot, is the occurrence in them of a few root-stocks in their normal position. 38 J. D. KENDALL ON THE INTERGLACIAL DEPOSITS OF But, as I have already pointed out, the rootlets from these stocks do not pass down into the underlying Boulder-clay, but are simply imbedded in the neighbouring woody matter. Some of these root- stocks belong to trees which must have been at least 18 imches in diameter; so that if such trees could have derived their necessary nourishment from a woody soil of this kind, it is perfectly certain they would not have been able to stand in it, because there would be nothing of any weight or tenacity for the roots to lay hold of. The presence of such root-stocks, it seems to me, may be better explained by supposing them to have been floated to the positions in which we see them. The position in which they now stand is that of flotation; that is to say, a root-stock would be floated and dropped in water with the same side up as when it was growing. The inner nature of the shore-deposits being precisely the same as that of the cavernous deposits found at Crossgates is also sug- gestive of drifting ; for clearly the latter are not on the site of an ancient ferest, but have been carried to their present resting-place by water. The facts presented by these deposits seem to me to have a most important bearing on the question of the formation of coal. We have here similar underclays to those which accompany cval-seams, and the same kind of intercalated clay-bands, both of which, in the case of coal, suggest watery conditions just as much as they do in the deposits under consideration. I think it is just as impossible that trees can have grown in the underclays of coal, which at the time would be soft and inccherent, as that they can have stood in the woody matter of the so-called ‘‘ submerged forests.” ‘The Stig- ~ maria rootlets of the coal may be accounted for in the same manner as the root-stocks which are found in vegetable deposits like that at St. Bees. The trees which have been found rising out of coal-seams and passing into the overlying strata, may be explained by supposing them to have been so loaded at the root when they were deposited that their position of flotation was erect. In this way the whole Coal-measures may have been deposited during a gradual subsidence without any of the periods of cessation demanded by those who hold that coal-seams are the remains of forests which grew in situ. On the assumption that the shore-deposits are “‘submerged forests ” of recent age, it has been often held that they indicate a subsidence of the coast; and writers, in consequence, have had to make the land rise and fall in a very remarkable and erratic manner, when dealing with the phenomena of “raised, beaches” on the one hand and ‘‘ submerged forests” on the other. The view just enunciated sim- plifies matters very considerably, as, according to that, the vegetable deposits on the shore do not necessarily indicate either a rise or a fall of the land in recent times. } j | / | j } | | } | | | | >= a SwncremavEN etek = 4 = = ae = = \ as => = = S % so = seice = . ms iromeere Bash £ Low Water Fig . 3. Section on Line AB Fig 2. Fig. 2 ae Fug. 4, , Sectton on Line © D Map of Lindal District ¢ wo" | 2 c Fs ny *R , == [ew Te ; \ > = AX | — 9 ~ @LINDAL STATION | = - | th — | : to Se £ i Se ae me , = 2 ( = eee (EEN x Scale of Se 320 fe R cy Pinna are Mineslae pur ot cale of Selon Oe 7 uh. i : ° SUL SHAT. ‘ 4% FBedent of deposit. yt proved colored. nenlnal tint. } MEI SHAFT \ “ 7 [|| a Ss A Grey Pinel ( Boulder Gay) Zz x B Vegetable Deposit C Blue Sand (Cavey) D Red Pinel (Boulder Clav) E F . > Scale of Map 3 inches tw a Mile _ Journ. Geol. Soe. Vol. XATVIL Pi iit jae Ome Plan of Deposit om Shore oppose Prigg : Reference to Plan & Sections. ™ Figs 8-11. A Vegetable Depostt , B Blue Any (Sandy, ) “4 C Red Cay in some places containing stunes(Boulde Tay) c De 3.2 5¢ | : \ Dip of Varrane of Woody atter- Carboniferous Limestone Gravel Borvholes evel liere Fig 1. Section on Line €E Fig-4. 1 em A Fig. 10 Section on Line © D. Fig.4 Agt A Scale of Plan und Sections AB and CD 72 feet to an Inch. : Fig. 9. Secon on Line AB Fig. 4. Zz cS f os = = > =z oc > Zs: = Zz aa 7 rats LINDAL esssorr Fig. 6. Section on Line AB Fug. 5. | High Water __ s Eyer eon C Boulder Clay A Fig. Ih. Scale of Plan & Section AB 100 feet to an Inch SM Plan of Deposit a y 2 ~ at S! Bees. velerence 4 A Vegetable Deposu 1 ———— 7 Glow 6 | fron ee Es SS SS \ Benet iiey . / Whitehaven; FURNESS RATEWAY Sa 5 = Po, ! ‘sts ‘ oy = eon is / red by Sand and 5°" er 0.47 é i , : : SHINCLE Fig. 5. a H rid Plan of the Walney Depostt : { Fig 7 Section on Line © D Fig. 5 Chaves 8 Seale of Section CD90 feet to an Inch = Norixontal Scale of Reference to Plan & Sections Fig. $5 1214. A Vegetable Deposit compact aa D® soft and spongy ke peat B Blue Sundv Clay C Roulder ve Sonatas Fig. 14. D Sand & Stones covering veg le osth = No Dep of lamina of Woody Matter eee rae Paes CEN Eee # iP Fig. 13. Section on Line AB Fig. R. = ee |) ; ot Scale of Section © DH) fet to an mck. Hugh Water Ww cl Y === SS Plan & Section AB Iti hes 10 a Mile Vertical Scale cf Section AB IO feet to un uich. MAPS , PLANS AND SECTIONS OF INTERCLACIAL DEPOSITS OF WEST CUMBERLAND AND NORTH LANCASHIRE. - + Frad® Dengerlieké Leth Luratim, =r # a Final WEST CUMBERLAND AND NORTH LANCASHIRE, 39 EXPLANATION OF PLATE III. Plans and Sections illustrating the Interglacial Deposits of West Cumberland and North Lancashire. Rig. 1. Map of the district, showing the positions in which deposits of woody matter have been found, Figs. 2-4. Map and sections in the Lindal District. 5-7. Plan and sections of the Walney Deposit. 8-11. Plan and sections of Deposit on shore opposite Brigg. 12-14. Plan and sections of Deposit at St. Bees. Discusston. Mr. ©. Rem said that in the Cromer Forest bed he had dug up many stumps of trees asserted to have been im situ, and had found them not to be so. He thought that great caution was needed in asserting trees to be wm situ. They would usually sink in a vertical position with their roots downwards. Sometimes portions of the soil in which the trees had grown were retained among the roots, and differed entirely from the matrix. Rey. H. H. Wrywoop pointed out. that trees often grew in a soil, but the roots did not pierce the underlying clay. Might not this be the case with the Cromer trees mentioned by Mr. Reid ? Rey. J. F. Braxz spoke of a boring at York with peaty matter in the midst of the Boulder-clays. Mr. Tippeman asked whether it was quite certain that there was Boulder-clay above and below these peaty deposits. The Sxcrerary replied that the author stated so distinctly in the aper. : The PRESIDENT spoke of the importance of comparing together the results of investigations in different areas. i 40 PROF. T. G. BONNEY ON THE SERPENTINE 4. On the Serpentine and assocrateD Rocks of ANGLESEY ; with a Nore on the so-called Srrprntine of PoRtHDINLLEYN (CAERNAR- VONSHIRE). By Prof. T. G. Bonnzy, M.A., F.R.S., Sec. G.s. (Read November 3, 1880.) 1. ANGLESEY witH HonyHEAD IsLAND. TE serpentine from the vicinity of Rhoscolyn, though known within a limited area as an ornamental stone, does not appear to have received much attention from geologists. It is dismissed in Pro- fessor Ramsay’s memoir on North Wales in a few brief sentences, and has not, so far as I know, met with fuller notice in any other quarter. On the Geological-Survey map it is delineated as forming an elongated lozenge-shaped patch about a mile and three quarters in length from EK. to W., extending inland north of Rhoscolyn from the west shore of Holyhead Island, and occurring in several smaller patches on or near the opposite coast of Anglesey. In each of these localities I have studied the rock in the field, and have examined microscopically the specimens there collected. The results, I hope, may be of some use, though, after three separate visits, I have not been able to visit every one of the outcrops or examine minutely every part of the district. Owing to the intricacy of the coast in certain places, a map on a considerably larger scale than 1 inch to the mile would be required before an elaborate study could be undertaken. The questions which I have attempted to solve (as in all former studies of serpentine), were (1) its relations to the other associated rocks, and (2) its nature and origin. The ordinary rock in the vicinity of these masses of serpentine is a dull bluish or greenish schist, composed chiefly of minutely crystallized micaceous or chloritic minerals with some quartz, the foliation being parallel to the bedding. The latter is generally very distinct and sometimes exhibits very remarkable crumplings. The serpentine occurs in low rugged knolls, cropping out here and. there from the fields over which in the map the colour denoting its presence is extended—the general aspect of the rock masses reminding us of the serpentines of Cornwall, Scotland, and Italy. It differs markedly in all its characters from the ordinary schists of the district, and, where associated with the latter, occurs in such a manner as to suggest either intrusion or sporadic metamorphism of a very singular character. This is also suggested by the mapping; but it is even — more conspicuous in the field. My examination, however, showed me that under one name two distinet rocks had in some cases been confounded, viz. a true serpen- tine and a gabbro, especially in the Anglesey group. Jor instance, if we alight at the Valley Station on the Bangor-Holyhead railway, and follow the road to Four-mile bridge, we observe in the fields on our left, near Ty Newydd, a long craggy mass of serpentinous AND ASSOCIATED ROCKS OF ANGLESEY. 41 aspect, the northern edge of the largest patch in Anglesey, according to the map. But on reaching this we find it a mass of gabbro, generally rather coarse—considerably altered, no doubt, but Fig. 1.—Junction of Schist and Serpentine. ! ; \y { Ly, = t= ~ La AP ZF- : ——— Ss AEF, SSS GLY Le il ——s So : _——DS Sq TELL ee Se , —— SS Se a Se eee a A. Schist. B. Serpentine. C. Sand. unmistakable. There is, however, a small inconspicuous outcrop of true serpentine, separated from it by some grass, about seventy yards to the north of its eastern end. From the west end of this gabbro massif we made* for the shore. Here we found an irregular rocky coast-line fringed with skerries and islets, mostly dry at low tide. Some are schist, marvellously contorted, others gabbro, a few ser- pentine. Without a large-scale map, it would be impossible to make the details intelligible; but it may suffice to say that we worked carefully along the shore to the inlet by Tyddyn Gob, examining the different rocks. In one place the evidence, though not perhaps absolutely conclusive, seemed strongly in favour of the serpen- tine being intrusive in the schist; for if the junction were due to a fault, this would be a very strange one (fig. 1)7. Further on (beneath a wall) gabbro is seen intrusive in serpentine, which has assumed, as is not uncommon in such cases, a locally schistose aspect. , ‘sdagg Axren% LORS *‘q1g0dep uel ado “OnAdTIOM ‘umog moypung fo abpy ay, yo worn ay —"T “OL “£urend DEPOSITS IN THE BRISTOL DISTRICT. Tt fringe the outcrop of the clder rocks, but high up on their table- lands, as in the case of the Charterhouse lead-mine, near Cheddar. Tue BristoLt AREA. Before proceeding with a description of the Bristol area and comparing it with the above, it is desirable I should advert to a new palzontological feature which hitherto has not attracted any atten- tion, in the presence of a multitude of minute Serpula-like calcareous tubes found in various deposits under examination. Years ago I noticed them in the freshwater brick-earth of Salisbury, which is of Postpliocene age, and put them aside as minute Serpule; but after- wards learning that all species of this family are marine, I thought they might possibly be analogous to caddis-cases, and belong to some insect. When they were afterwards found in enormous numbers, and under many diverse circumstances, I saw the desirability of learning more about them, and sent them to friends who were authorities in special departments of natural history for their opinions, at the same time describing the circumstances under which they were found. Without mentioning names, it proved a “ pursuit of knowledge under difficulties.” First, they were doubtfully re- ferred to the cases of some Dipterous insect ; and if not such, it would be worth considering if they were Serpule. An entomologist decided that they were not insect-tubes; a good microscopist was of opinion they were the calcareous cases left by rootlets of vegetation passing down from the surface; next, an eminent zoologist intimated, with some uncertainty, that they somewhat resembled the genus of Ser- pulide, Filograna; then a good botanical friend and microscopist pronounced them to be vegetable; upon which I sent them to a first-class botanical expert, who was of opinion they had nothing to do with vegetables, and that he had consulted a zoological friend, who also believed they were worm-tubes allied to Annelids, and that again they presented some resemblance to /ilograna. They are to be found very abundantly in some of the vein- fissures of the Carboniferous Limestone (to be referred to hereafter) in the Bristol district, and also, though not in such numbers, in some of the later stratified marls. My impression is that they are due to freshwater conditions, and that, though they may not be rootlets which have passed down from the surface (some of them being found under conditions apparently precluding this idea), yet they may, notwithstanding, be due to freshwater vegetation. They are sometimes free, but often in clusters united by the ma- trix of the deposit. It will be desirable to determine their syste- matic position, as they will have a bearing on the age and other conditions of the deposits in which they are found. I propose that they should be recognized under the designation of Tubutella ambigua. Durdham and Clifton Downs.—I now propose to show that the same physical conditions prevail at Durdham Down and at various outlying spots near Bristol, as in the Frome district. The area to 72 C. MOORE ON ABNORMAL GEOLOGICAL _be first noticed will be that embraced in the southern escarpment of the Carboniferous Limestone, extending from Redland, its eastern end at Durdham Down, to the suspension-bridge at Clifton, over- looking the river Avon on the west. Near the Old Black Boy inn, just under the edge of Durdham Down, there are places known as The Quarry and The Quarry Steps (see fig. 1). comming on the latter, we look down into a large excavation, which musi have been formerly extensively worked for stone, but is now occupied by small houses and gardens, most of which are probably a century old. One of these houses is close under the flight of steps leading to The Quarry, and has been built against a natural vertical wall of mottled red or yellow unstratified limestone, differing altogether in colour and texture from the grey Carboniferous Limestone of the district, by which, on either side, itis bounded. It is about 8 feet in thickness, and may be traced on the east side of the excavation by its brighter colouring; and there is no doubt that it continues into the limestone of Durdham Down immediately above. As was the case with the workings on the Mendips, the old quarry- - men here extracted the purer limestone, leaving the impurer some- what conglomeratic infillings standing out, and terminated their quarry on the east by this large dyke. As it has not been opened up, little examination could be given to it for organic remains, which, from its close proximity to the Thecodont deposit, would have been desirable; but that some are present therein is sufficiently indicated by the fact that, after examining a few pounds weight of the softer material taken from the interstices or sides of the vein, I obtained numerous minute fragments of bone or teeth, one very small fish- tooth, an Hchinus-spine, a few joints of Carboniferous-limestone Encrinites, and some of the tubes previously referred to. The platform of the Quarry Steps rests upon the surface of the above dyke. Looking from it, along the Down escarpment to the west, the eye takes in Bellevue Terrace, on the edge of the Down ; and it was between these houses and the quarry, a distance probably of 200 yards, along the same face of limestone and on the same horizon, that the deposit containing the Thecodontosaurian remains was found. Unfortunately the precise spot is unknown; and, from its being built over, there is not much hope of its being again identified. My late friend, Mr. W. Sanders, F.R.S., of Clifton, gave me, some years ago, a sketch showing the conglomerate on the edge of the lime- stone, with what was then considered to be New Red Sandstone at its base (fig. 2); and it is significant that he does not so much represent it as a basin-shaped depression in the limestone as indicate a deposit following the slope of the escarpment, similar to the case of a vein the top of which was opened up, but from which the limestone still resting below against its side had not been removed. Mr. Sanders also marks the spot where the reptilian bones were supposed to be found. . At the time when these reptilia were discovered, the peculiar conditions of deposition I have indicated were unknown. In the DEPOSITS IN THE BRISTOL DISTRICT. 73 Fig. 2.—Sketch Section of the Thecodontosaurus-bed of Durdham Down. (Drawn by W. Sanders, Esq., F.R.S.) Cis Pretest BORE SEAC < ZING ‘ Z NY iN ENG ZG RS yy 4, Y, es OX COw: aN Oe & G “nN NEE N UAW <4 SO OX Sf ON, a. Conglomerate of New Red Sandstone. b. Locality of Thecodontosaurus. c. New Red Sandstone. d. Carboniferous Limestone. section described below, which embraces a line of working from the Quarry Steps to a large quarry which I call the ‘‘ Avenue Quarry ” (from its being immediately north of the Avenue Road), a distance all together of about 680 yards, I shall show the presence, as at Hol- well and elsewhere, of a series of veins with infillings derived from different geological ages; and my interpretation of the reptilian deposit therefore is that it 1s one of such a series in Carboniferous Limestone, by which, on both sides, it is surrounded. Durdham Down Section.—From the Quarry Steps to the end of Bellevue Terrace, within which area the reptilia were found, is about 300 yards. Until lately these were the last houses fringing the Down to the west; and there then stood up by itself, like a wall, a large vein-infilling (the limestone having been worked up to it on both sides) separating the garden of the last house from an adjoining quarry. Other buildings have now occupied the line of the quarry, covering up four smaller veins; but the one above referred to, from 14 to 16 feet in thickness, is still utilized as a boundary between the gardens. It has the same general hthological character as that at Quarry Steps, but has occasionally small pockets of iron-ore, and is more mineralized. I purposed to examine these veins closely for organisms, but have been prevented. Passing along the roadway from this quarry, the next vein seen was one containing hematite iron-ore, at its thickest about 16 feet. It had been worked up to the roadway on the edge of the Down. At my last visit this also was being filled up, and a house built across it. Alluvial Veins with Inassic and Rhetic Remains.—For about 200 yards from the iron-ore vein the limestone has been unworked. Following the road to a spot near the ventilating-shaft of the Avonmouth Railway, an archway facing the road leads into the Avenue Quarry, which presents some interesting features. For many years a lump of seemingly stratified yellow marl had been left in the bottom of the quarry, the limestone having been worked 74 C. MOORE ON ABNORMAL GEOLOGICAL around it. Its presence often puzzled me; but there is little doubt that it is due to an alluvial infillmg into a small limestone cavern, and is connected with one of three thin veins of alluvial material which are present in this quarry. A cavern was many years ago discovered on the Down, containing, amongst other things, the remains of Hippopotamus, which are now in the Bristol] Museum. These three veins are filled with an ochreous or brown clay, which, when critically examined in its natural condition, appears to differ slightly in character in each, as though they might have received their infilling at somewhat different times ; still, to save labour in the examination of their contents, they were mixed together. Their organic remains are very varied, and not less so their mineral con- stituents. Their Postpliocene or still later age is indicated by the presence in them of frequent portions of the incisor teeth of Arvicola, and a single shell of Hew. Although all the remains are rare, those from the Lower Lias are most abundant. These consist of young forms of Ammonites of two species, Myacites and Astarte, Cylindrites and portions of three other univalves, a Pentacrinite-stem, and a single valve of an Entomostracon. ‘There are teeth and scales of fishes, and a fragment of bone which may be either Liassic or Rheetic. The interesting presence of Rhetic remains, however, is shown by teeth of Saurichthys and Lophodus and a small palate identical with species from Holwell. Some of the teeth are bleached and worn by water-erosion. The little calcareous tubes accompany them. The residue of these veins, after washing, shows a great abund- ance of black pisolitic granules. The following analysis of its mineral contents has been obligingly made by Mr. Gatehouse, City Analyst of Bath :— Soluble in acids: Watery Mos) ioc ts Sisieoe Gade a Bee a 7°80 Calciumycarbonmate tame ar ae eee 20°50 Magnesium carbonate...... ee eB os 0°73 Oxide ‘Of Tromso.) eee os sil eens 6:50 Zime carbonate (calamine) 2-5. 02) 24.45 0°54 Manganese, dioxaideds saucy. a)-Nelcie cna 0-50 ead=sulphide\(eallena)) ye. eee 0°90 ho) Det Mra AUR eM NAME ie Bestia i ny'ss 6 8°16 PeMADNACUNGl: PREM ML HARE OIRO 6 oa.3 2°60 Pobassivamn'y 6,552.05 eng sieeucie ds oe ieee ne 0:07 . 48°30 Insoluble in acids: Calcium sulphate............ Pale ee 0-39 SUHICALC Of, ZINC Cerrar. cele n anand eae 0-25 Oxadeyor ron Gnsoluble) ie; ariel 6°55 Alumina Pe ROMA NHRC MRO Wind suis ics 10°45 Silica RS UIM Rn NENTS Mae fee ka hes 33°70 WOPDCE MN ce cas walneal ele emete trace: DEPOSITS IN THE BRISTOL DISTRICT, 75 Inassic Vein.—The working of the limestone in this quarry has ter- minated on the western side against a vein from 10 to 12 feet thick, which has received an infilling of yellow clay of an entirely different lithological character from those last mentioned. So far as the evi- “dence goes, it is apparently a vein of true Lower Lias, with occa- sional weathered blocks of Carboniferous Limestone. All the remains found are Liassic, and consist of :— Teeth and scales of fishes. Gryphea. Ammonites (young). Terebratula. Univalves (several species). Cidaris (teeth, plates, and spines). Astarte. Pentacrinites. Modiola. Cristellaria. Lima. Webbina. Cardium All the phenomena I have mentioned, from the Quarry Steps to the Avenue Quarry, are embraced in less than half a mile. Without including the Reptilian deposit, eleven veins are shown to be pre- ‘sent; and but for a considerable part of the edge of the Down being covered, no doubt many others would have been seen. The thetic Bone-bed—There are no workings between the Avenue Quarry and the neighbourhood of the Clifton suspension- bridge. ‘The road to the latter is cut through Carboniferous Lime- stone containing many veins, which are usually filled with cale-spar and other mineralized matter. Looking towards the Observatory, there are two veins. One, filled with limestone, has been left boldly standing up by the quarrymen; but nothing can be said of its age. Close to the toll-house, however, on the Clifton side, there is a deposit of considerable interest, having a face of about forty feet, in which the Rhetic bone-bed and its associated remains are present. It is partly composed of irony and yellow sandy-looking marl, with many free crystals of carbonate of lime, as in the Holwell fossiliferous infilling ; and there are also patches of finely laminated rock, similar to the Rhetic “ White Lias.” Some of the associated blocks of stone appear to be fossiliferous ; but as the deposit forms the boundary-wall of the toll-house, they cannot well be broken down for examination. The bone-bed is two inches thick, with teeth of Saurichthys apicalis, Lophodus minimus, and many fish-scales; and the clay on either side contains fish-remains of the same age. In Professor Ramsay’s ‘ Physical Geology and Geography of Great Britain, he has given a very pretty sketch of the Gorge of the Avon, with the Suspension-bridge and its beautiful scenery, which, through the kindness of Prof. Ramsay and of Mr. Stanford, the pub- lisher, I am permitted to use (fig. 3). Curiously enough, it includes the Rheetic bone-bed, and, on the shoulder of limestone looking down the river on the east, the conglomerate mentioned below. The Car- boniferous Limestone dips rapidly to the south under the Bedminster coal-beds, and disappears under beds of New Red Sandstone. A lighter line in the sketch, between the eastern abutment of the bridge and the houses on the level of the roadway, with the Observatory seen beyond it in the distance, shows the exact position of the Rheetic deposit. 76 C. MOORE ON ABNORMAL GEOLOGICAL Fig. 3.—Gorge of the Avon at Clifton. Before referring to other outlying sections near Bristol, it may be noticed that although the true New Red Marls are present around, the red and variegated marls which lie on the east of the basin between the Durdham-Down and Clifton escarpment, and extend towards Bristol, are alluvial. They have been opened up in brick-pits repeatedly, as well as the conglomerates below. From the former, after long search, I obtained several small bits of Inferior Oolite and fragments of shells of Ostrea. The conglomerates are bedded, and are usually extracted on the site of each building erected; and although hundreds of thousands of tons have been so used, I have never detected any contemporaneous organism in them, nor have I heard of such having been found. They get coarser and more irony as they approach the limestone below; and in several sections its rugged surface has received pockets of hematite iron-ore. The pebbles are comparatively small and angular, and give the idea that they may be due to a subaerial denudation of the Carboniferous tableland above. Conglomerate of the Avon Gorge.—Passing down the river, not far below the Suspension-bridge, the ‘ Dolomitic Conglomerate,” of which Mr. Etheridge has given a section, is reached. It rests on the edge of the Carboniferous Limestone at the side of the road leading up to the Down, and is, as a friend has remarked, ‘a gigantic heap of conglomerate” with pebbles of great size at the base, getting gradually smaller at the top. It is very irony, and appears to occupy but a small area. Some red marls gathered from DEPOSITS IN THE BRISTOL DISTRICT. 77 the sides and interstices of the deposit yielded me part of a fish- palate, which is probably of Carboniferous-limestone age, and great numbers of Encrinital stems, a comparatively recent freshwater operculum, and numerous examples of Tubutella. Whatever may be the age of this deposit, groups of the latter are usually surrounded by an irony matrix which seems part thereof. Vein at Ashton.—On the west side of the Avon, about a mile beyond Ashton, considerable deposits of iron-ore occur ; but I have been unable to examine them. They lie at the foot of the lime- stone escarpment. On the tableland above them, at Longwood Farm there are disturbed Carboniferous Limestones, in which a vein occurs containing occasional lumps of galena and some ¢a- lamine. Tubutella is present in great abundance. The upper part of this vein is somewhat honeycombed; and the tubes are found adhering to thin flakes of calc-spar. No other organisms were found at this spot. Westbury-on-Trym— Carboniferous Limestone with Minerals and Oolitic Remains.—On passing from the Durdham-Down quarries (previously mentioned) along the eastern edge of the Carboniferous Limestone, at a distance of two miles, some large quarries are to be found. On the eastern side of one at Southmead, worked by Mr. Kennedy, there may be seen what appears to be an ordinary mineral vein, about a foot thick, passing down through the section. It con- tains good hematite iron-ore, ochre, galena, and calamine. Although it appeared nothing but a mass of mineral matter, I still hoped a sample might yield some evidence of its age. In this I succeeded beyond my expectation; for on washing it I at once found many angular pieces of Inferior Oolite, which, from their being stained with iron, were not before visible. On a still closer examination I obtained the oolitic organisms given in the list below, asso- ciated with TZwubutella; the specimens show very little sign of attrition. No Oolitic deposit from which they could be derived, however, is in this district nearer than Dundry, six miles to the south, or the Cotteswolds, many miles to the east, with, in the latter case, the area of the Bristol coal-field intervening. In the same quarry there is at another spot a pocket of light- green clay. Although very intractable, I was able to disintegrate it sufficiently to find that it contained Alge &c.; and although there remains a little doubt as to their age, it is hkely they are of comparatively recent date. Westbury vem, Oolitic Organisms. Fish-tooth. Thecidium. Nerina. Bryozoa, several species. Solarium. Echini, teeth, plates, and spines. Univalve, sp. Entomostraca. Turbo. Serpule. Dentalium. Foraminifera ? Astarte, Tubutella, Cardium. Pentacrinite joints. Ostrea. Encrinite stems, Carboniferous. Algeze (in clay vein), several genera 78 C. MOORE ON ABNORMAL GEOLOGICAL The sections remaining to be noticed are along the eastern edge and to the north of the Bristol coal-field. On the eastern side the Carboniferous Limestone does not come to the surface between the Mendips and a very small outcrop at Grammar Rock, under Lans- downe, near Bath (which is not recorded on the Ordnance Map), and others at Wick Rocks and Codrington. From Chipping-Sodbury a narrow belt continues to Yate snd Wickwar, and entirely sur- rounds the north of the coal-basin. At Wick, where the quarries are extensively worked, mineral veins are to be seen having their usual vertical bands of barytes, galena, &c. passing down them. In their softer pockets Tubutella is abundant; in one which has a marly infilling it is present in great numbers. Occasionally small patches or pockets of grey marl lie near the surface; but I have not yet found any other organisms in them. The Yate Rock Sections—These are between Chipping Sodbury and Wickwar. On the ragged surfaces of the limestone there are here also pockets containing more mineralogical materials with Tubutella. A vein passes down from the surface of one of the quarries a foot in thickness, containing soft mineralized material in which are myriads of these little tubes; they are often attached to the broken-up pieces of barytes and other minerals. Netilebury Quarry, Yate, and Clevedon.—The Nettlebury quarry is a large one, nearest Wickwar, and is the last I shall refer to on this side the Bristol coal-basin. A section of it has been given by Mr. Etheridge *, in which are shown to the east horizontal step-like beds, overlying highly inclined Carboniferous Limestone, whilst on the west side equivalent beds dip at the same angle as the limestone towards the coal-basin. At the present time it would be difficult to recognize the locality from the above section, from the almost entire absence of the beds on the eastern side, though they are present from 12 to 14 feet thick on the west. Mr. Etheridge considers these beds to be of the age of the Dolomitic Conglomerates, and the representatives in lithological condition and age of the supposed Magnesian Limestones of Clevedon. The latter are thick, irregularly bedded, yellow limestones, used in the district for building-purposes. It is apparently a local deposit, resting in the quarry near the hotel on Old Red Sandstone, which crops out on the beach and abuts against Carboniferous Limestone to the east, the Old Red being fringed on a level with the Severn by a continuous belt of Dolomitic Conglomerate, continuing north for some miles. . Almost every parting or crack between the blocks of Clevedon stone shows the presence of galena, carbonate of copper, or other minerals. In only one instance have I found a block of stone with organic remains. This contained afew Encrinital stems, an im- perfect Rhynchonella, and also several imperfect Strophomene. Were it not that the specimens put on the lithological appearance of the enclosing matrix, I should be disposed to think them redeposited * Quart. Journ. Geol. Soe. vol. xxvi. (1870), p. 179. DEPOSITS IN THE BRISTOL DISTRICT, 79 from the Carboniferous Limestone *.” Looking at the Yate deposit as it lies on the limestone in the section, it has much resemblance to the Clevedon beds, but on closer examination this ceases. The latter are almost pure Magnesian Limestone, whilst the upper 9 feet of the former are of sandy yellow marls, which on being washed float away and leave a sandy residue. Next follow eighteen inches of yellow limestone, in two irregular beds, which in structure and colour somewhat resemble the Clevedon stone. On treatment with acid these also leave a sandy residue. Between the above and the Carboniferous beds there is a thin deposit in pockets, almost com- posed of fine grey sandstone. There are no organic remains special to these beds beyond fucoids and Tubutella ambiqua, which is rare. For these reasons, and knowing how varied are the deposits on the outcrops of the older limestones, I am disposed to think that the yellow beds at Yate are of comparatively recent age. On the surface of the Carboniferous Limestone there are occasional thin patches of what appears to be a comminuted shelly breccia with a grey sandy matrix, which I do not think the equivalent of the bouldered Dolomitic Conglomerates found in other parts of the district. Thornbury Railway and Secondary Veins—The branch line which leads from the Yate station to Thornbury exhibits some interesting geology. At Tytherington fine sections of Carboniferous Limestone are seen, and thick deposits follow of what are, no doubt, true Dolomitic Conglomerates. On emerging from the tunnel towards Thornbury these have some marly divisions, in which, in a flat over the tunnel, some galena has been found. Not far beyond, the Lower Limestone shales pass into the Upper Devonian beds, the conglo- merates also resting upon them. Just before reaching this point there are several thin veins in the limestone containing sulphate of barytes and galena, in a matrix of gossany iron-ore. A sample from the soft ochreous part yielded me three Conodonts of Carboni- ferous-limestone age—and of Secondary remains, Pentacrinite joints and a single specimen of a Foraminiferous shell, Planularia pau- perata. On a second visit I discovered deposits towards the surface which, lithologically, I cannot distinguish from a ferruginous marl of the Middle Lias, which contains hollow casts of shells and crushed specimens of what appear to be Rhynchonella tetrahedra. Tubuiella is present in great numbers, surrounded, as at Yate Rocks, by a re- deposited ferruginous matrix. Age of the Deposits—From the foregoing examples of abnormal deposition, all of which have been accidentally revealed by quarrying- * Since writing the above, I have sent the specimens to my friend Mr. Davidson, who says:—‘‘ The specimen you send for my examination you say is from the Magnesian limestone. In colour it looks like a rock of that for- mation ; but I have never hitherto seen from our British Permian rocks a Sfro- phomena or Streptorhynchus shell similar to the one I observe on both sides of the specimen, and which looks like Strophomena crenistria. 'The Rhynchonella is not sufficiently complete for specific determination. If not Carboniferous, at least one of the species would be new to our Permian rocks or Magnesian Lime- stone. I almost fear your enclosed specimen is Carboniferous.” 80 C. MOORE ON ABNORMAL GEOLOGICAL operations in the Carboniferous Limestone, it must be manifest that there are a multitude of other examples not yet opened up that would yield an interesting study to the geologist. One of the most difficult problems regarding some of them is to arrive at satisfactory conclusions as to their exact age; but there can be little doubt that the physical conditions to which the deposits are due were the same both in the Bristol and the Mendip areas. Supposing the fissures in any district had all been caused by the same shrinking or change of level, they would have been subject to the same refilling influences, and would contemporaneously have received a mixture of materials derived from the denudation of that time; but although the alluvial infillmgs in the Avenue Quarry have a mixture of organisms, it is a singular fact that in a series of parallel veins coming to the surface on the same horizon, not far removed from one another, and some of them but a few inches in thickness, each appears to have an individuality of its own, and to represent in geological time intervals clearly distinct from one another. As at Holwell, so at Durdham Down, the worked face of the escarpments reveals infillings of allnvium, Oolite, Lias, and Rheetic and Keuper beds, whilst mineralized or iron-ore veins show conditions specially their own. Reference has repeatedly been made to the Tubutella ambiqua, which I have found in almost every deposit that could be examined under favourable conditions, from Maidenhead to those of Gloucester- shire and Somersetshire. When they occur, as in the brick-earths of Salisbury, in association with freshwater shells, and also with Postpliocene mammalia, there seems little reason to doubt that they belong to freshwater deposits. If they can be traced in older formations, they may be a guide in determining the conditions under which those formations have been laid down. I have reason to suppose this may be the case between the Upper Devonian and the Carboniferous series. All veins, mineral or otherwise, come to the surface; and if the Tubutelle be found therein, they will probably indicate the presence of freshwater conditions. ‘The upper portions of those I have mentioned in this district represent the gossans of the lodes in more ancient rocks ; and if, as at Yate and Tytherington, the Tubutella is caught up or surrounded by the mineral matter of the vein, there has either been a remodification of that portion of the vein, or it must have been contemporaneous with the organisms enclosed. 7 I have already shown that most of the mineral veins of the Mendips and South Wales are at least of Liassic age; and on this point I have much confidence in the belief that a careful exami- nation of the gossans and other mineral constituents of the veins in our more ancient rocks will repay the labour, by giving either more precise indications of their age or of the physical conditions under which they were deposited. Age of the Bristol Reptilia.—The varied points mentioned in this paper have drawn me away from the chief object which led to their consideration, viz. the age of the Bristol reptilia. As before DEPOSITS IN THE BRISTOL DISTRICT, ont remarked, the peculiar circumstances attending their discovery has always left this in some obscurity. They are now assigned by Mr. Etheridge to the Dolomitic Conglomerates at the base of the Keuper. ‘These conglomerates are in great part composed of rounded boulders of the Carboniferous Limestone, some of them of great size, and seemingly requiring glacial conditions for their removal and transportation. Not only do they fringe the outcrops of the limestone from which they are derived, but they form an almost continuous deposit of considerable thickness, extending for many square miles in the Somersetshire coal-basin, the inclined ragged limestone edges of which are, in places, rendered by their denudation quite horizontal. The removal and redeposition of the conglomerates indicate very troublous times, during which it seems impossible for any organic life to have existed; for nothing could have withstood the grinding-processes to which it would have been subjected ; and it is a significant fact that no organic remains have ever been referred to this period except the reptilians under notice. For these reasons, and from my having found the teeth of The- codontosawrus and Palcosaurus in the Rheetic deposit at Holwell, and also from my subsequent discovery of the Rhetic bone-bed and remains of that age almost alongside the Clifton reptilia, I had come to the conclusion that the latter belonged to this period—a view which further investigation respecting both Keuper and Rheetic reptilia requires me to modify. Seventeen Thecodont teeth, more or less perfect, are in my Holwell series. On comparison with those from Bristol, they are more robust, have a more wrinkled or striated surface, with the serrations on the edges smaller, less oblique. and more numerous. In my paper ‘‘On Abnormal Con- ditions” &c.*, I gave a section of variegated Keuper marls at Ruishton, near Taunton, one bed of which I described as a “ Gritty conglomerate, with occasional sandy bands and intermediate layers of marl, with fish, reptile, and batrachian remains, fourteen inches thick.” In this bed I have lately found some teeth of T'hecodonto- saurus, which appear in all respects identical, in form, structure, and the character of the serrations on their edges, with those from Bristol. It contains also Acrodus keuperinus, Hybodus, Diplodus, &e. Itisrather a coarse sandy bed than a conglomerate; and, owing to its being rather unconsolidated, its remains are very fragile. There seems little doubt that this bed is on the horizon of that in Warwickshire which has yielded identical vertebrata; and if so, the Bristol reptilia will have to be removed one stage later in time, from the Dolomitic Conglomerate to the middle of the Upper Keuper. Té is an interesting paleontological fact that, although most of the generic forms of the Keuper recur in the Rheetic beds, so far as I have ascertained, the species differ, and are special to the two formations. * Quart. Journ. Geol. Soc. vol. xxiii. (1867), p. 468. Orie G.c. No, 145. G 82 ABNORMAL GEOLOGICAL DEPOSITS IN THE BRISTOL DISTRICT. Discusston. The PresipEent spoke of the great industry and skill of the author in collecting the evidence on which this paper was based. Mr. Tawney stated that Mr. Sanders held that the Dolomitic Conglomerate is of different ages in different parts of the district, and that the Vhecodontosawrus is high up in the Keuper series. He thought that the fact of these remains being imbedded in solid con- glomerate was scarcely reconcilable with the notion that they came from vein-infilling. Dr. Duncan, with reference to the origin of the so-called Tubu- tella, stated that similar tubes might be seen in course of formation by the escape of air-bubbles from the surface of oysters and other shells covered with Algze in turbid water containing carbonate of lime. Prof. Sretny spoke of the great interest attaching to the question of the age of the Thecodontosawrus. He considered the specimens exhibited to belong to at least two genera. He stated that the ilium of Thecodontosaurus is Crocodilian, with Dinosaurian affinities. Rev. H. H. Winwoop supported the views of the author concerning the position of the veins of Durdham Down. He remarked upon the difficulty of understanding the mingling of different faunas in the same vein. The Presipent supported the views of Mr. W. Sanders, as ex- pounded by Mr. Tawney, and bore testimony to the great value of Mr. Sanders’s map of the Bristol area. The AutHor agreed that the Magnesian Conglomerate is of diffe- rent ages. He thought the Thecodontosaurian remains were ob- tained from the top of one of the veins. He stated that, while the veins occasionally contained the remains of fossils belonging to more than one geological period, others contained organisms which appeared special to a single period only, implying a denudation and refilling in veins, at the present time difficult to account for. J. W. CARRALL ON CHINESE CARBONIFEROUS PLANTS. 83 8. Norzs on the Locaritry of some Fossits found in the CARBONIFEROUS Rocks at T’ane Suan, Cuina. By Jamus W. Carratt, Esq., F.G.S. &c., of H. 1. M. Customs Service, China. (Read Novem- ber 3, 1850.) Tue fossils which form the subject of this paper were found at T’ang Shan, Chiao Chia T’un, in Lan Chow of the Province of Chih Li, and about 120 miles from the treaty port of Tientsin, in a N.N.E. direction from that port. A Chinese company has been formed, known as the “ Chinese Engineering and Mining Company,” to work the bituminous coal deposits there found on the European system. Mining operations were commenced in 1878 by ascertaining the locality of the coal-seams by boring with the diamond boring- machine. The bore-holes, three in number, averaged about 40U feet apart; the third and deepest of the three driven reached a depth of 536 feet. The seams dip at an angle of 45° to the north, calculated by the angle of the strata found in the bores. The thickness of the coal-seams, not taken at the slant but parallel to the beds, is as follows :— ft. in iNowi-seam, N. of King’ seam: o..4.... 6 O odd. TSO? GOUT USS nO Ire el cee era Ibs Wosileseam, o: of King seam ...... 5... Oued No. 2 seam cy Mn a ete ec ie Ra Ia nd No. 3 sear ey abla Meeps ONS ae oe Oe Id No. 4 seam Boy) 1 aI ae ae ere ALD The seams are in curves or folds; one seam that comes to the surface at the back of the colliery, again appears above ground halfway from the colliery to Kai Ping (say three miles from the first outcrop), and again five miles further on, trending in a north- erly direction. Some very good magnetite, containing between 45 and 8U per cent. of metallic iron, has been found at Pai Mah Shan, about seventeen miles from T’ang Shan. It is intended to erect roiling-mills near the colliery, and place them also under foreign superintendence, trams being laid down (after permission has been obtained) between Pai Mah Shan and the hills for the transit of the ore. The most striking feature of the geological formation of the country round ‘l’ang Shan is, that above the Carboniferous system is first loose sand and then loam, the loam being uppermost, and extending but deepening all the way to Tientsin. Decomposed red sandstone was seen in the distance to the north. A gradual ascent commences four miles before coming to Lu T’ai by land from Han Ku, and continues all the way to T’ang Shan. The colliery is situated in this incline, and is about a mile 84 J. W. CARRALL ON CHINESE CARBONIFEROUS PLANTS. from the nearest hill, T’ang Shan, which I do not think is higher than 300 feet or thereabouts. From the roundness of the hills and the way the chain is detached in places, there is not the least doubt in my mind that this part of the country was under water, and 1s now gradually rising, this also being proved by the fact that, according to Chinese history, the city of Tientsin was at one time situated on the seaside, and itis now some twenty-five miles inland. Note on the Specimens, by W. Carrurumrs, Esq., F.R.S., F.G.S. The specimens from China belong to a single species of Annularia ; and I have no doubt that it is A. longifolia, Brongn., which is so abundant in our coal-measures in Britain, and is found on the con- tinent and in North America. It occurs throughout the whole of the coal-bearing beds, from the lowest to the highest. Newberry and others have described fossil plants from China asso- ciated with beds of coal; but these belong to Trias or still later formations. The great interest of this communication is that it records the existence of the true coal-measures in China, and gives a range to a well-known Carboniferous plant, in harmony with what we already know of other western contemporaneous Paleozoic plants. DISCUSSION. THe PrEstipent remarked on the great interest attaching to the discovery of these Coal-measure fossils. Mr. Carrurners pointed out the interest attaching to the com- mencement of coal-mining in China. He remarked upon the world- wide distribution of certain Paleozoic species of plants. Pror. Jupp spoke of the importance of this discovery in a country the geology of which was so little known. Mr. Branrorp suggested that possibly future observation might show that these plants were associated with Mesozoic forms. The Palseozoic flora had not a worldwide distribution. Mesozoic types of plants had been observed in Australia in Paleeozoic beds. Pror. Seetey demurred to the conclusions of Mr. Carruthers as to the distribution of Paleozoic plants. Mr. Buaxs supported Mr. Carruthers’s views. Mr. Carrutuers, in reply to Mr. Blanford, said he believed that the Glossopteris-beds of Australia are really Mesozoic and not Palee- ozoic in age. . Mr. Buanrorp said the late Rev. W. B. Clarke had found Glos- sopteris-beds intercalated amongst Paleozoic marine fossiliferous beds in coal-pits in New South Wales. THe PRESIDENT supported the views of Mr. Carruthers and Mr. Blake, in opposition to those of Prof. Seeley. Pror. Jupp supported Prof. Seeley’s views as to the existence of life-provinces in Palzeozoic times. tor Ho fl Geol. Soc. Vol. XXXVII. PLV. WIGS _ 400 te. yf / Bagshot “Sands ‘s tation. How-ledge Lime: stone RL Lyne Ue Dangertield | : 1 Lith 20 DPB d. S* Coven S* Covent Garden 786 : e ¥ bat é . \\ ‘ i peels , ‘ i ; 2 2 \ Wiss. H Headon Hill OT ft. if 0 t 1 a nr d B a y = . 2 i Gravel | i | 200 # ee ' Thick Lymna@a Teclend Bey Firting Ree | = : Bote a Z Vera bed: ; GCrio-bed gs C12 i How Ledge limestone Neritina -bed Varden sands Wardein sands G 0 Zz n e l L I @ v : } = N 10° E<—> S95 W VIGE SBY were ji 4 NIGE<—> S16°W i Warden. Battery (Eoerbeppenth : pe Bramble Post- Tertiary Gravel cap eer Petiory Linstone Colwell Limestone —— ee U. Head: Np eee U. Heador Chine = : ____ Me Headon ee f Ue Sands & hazelnut bed » Hea ae z tipper ee ae — } Grass slopes - EN = Sp 7 A > How Ledge Warder Chawa.-limestones pero = TO ry : with Lymnea Limestone = Cventricosaor bed. gime re = 5 0 prcpratan zone i Oyster Ledg: bank > ot 7 ani Scale for Vertical Heights 2500. Scale for Horizontal Distances 5280, or 12 inches w the Mile (one inch 0 46 yards.) aa ~ | Fral® Dangerfield Zit: 22 Bedtord. S* Covent Garden =e get geile had THE BEDS AT HEADON HILL AND COLWELL BAY, ISLE OF WIGHT. 895 9. On the Beds at Heapon Hitt and Cotwett Bay in the Isle of Wight. By H. Kunrrne, Esq., and E. B. Tawney, Esq., M.A., F.G.S., of the Woodwardian Museum, Cambridge. (Read De- cember i, 1880.) [Puate V.] I. InrRopuctIon. In a recent communication laid before the Society* the opinion was expressed that a serious error had been made by almost all previous writers in regarding the marine beds at Colwell Bay and Headon Hill as on the same geological horizon ; we read :—‘* We shall now demonstrate that the Colwell-Bay marine beds are not, as has been hitherto supposed, the equivalent of those of Headon Hill and Hord- well Cliff, but that they occupy a distinct and much higher horizon.” Upon the correction of this supposed error a new classification and nomenclature for the Upper Eocene formation of Britain was pro- posed. . The author further, after a review of the paleontclogical evidence, arrived at the conclusion that, on the one hand, the fossils in the Headon Hill and Hordwell Cliffs were identical, while, in the second place, those of Colwell Bay, White Cliff, and Brockenhurst presented the closest agreement among themselves. Then, comparing the former two localities, taken together, with the latter three, taken together, he considered (1) that the fauna of the first group was largely estuarine, and that of the second group marine; (2) that less than half the forms found in the former occur in the latter; (8) that the fauna of the former approximated more to that of the Barton beds, having about one third in common with them, while not more than one fifth of those from the latter three localities occur at Barton ; (4) that the fauna of the former two agreed with that of a series of beds on the Continent which underlay and were older than beds containing the fauna of the last three. In the following communication the authors attempt to traverse these points in the paper above referred to, in succession. By reference to detail-sections they argue that the stratigraphical evidence is plainly demonstrative of the identity of the Colwell-Bay and Headon- Hill marine series, the beds being continuous through the cliffs and easily traceable. Referring to the paleeontological evidence, it is shown from collec- tions, made with their own hands this year, (1) that the fauna of the Colwell-Bay and Headon-Hill beds are identical; (2) that this fauna differs considerably from that of the Brockenhurst bed, which occupies a lower horizon; (3) that the Colwell-Bay bed has less than one third of its species common to Barton beds, while the Brockenhurst fauna has nearly one half in common with Barton beds. * «On the Oligocene Strata of the Hampshire Basin,” by Prof. J. W. Judd, E.R.S., Sec. G.S., Quart. Journ. Geol. Soe. vol. xxxvi. p. 1387. Q.J.G. 8. No. 146. H 86 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT We conclude therefore that no reason has been shown for upsetting the classification of the strata adopted by the Geological Survey, and which, for nearly a quarter of a century, has been received among geologists. Certainly we wish to uphold in its integrity the work of the late E. Forbes *, and the classification of beds adopted by him when Palxontologist to the Geological Survey, and subsequently confirmed by Mx. H. W. Bristow, F.R.S., in the second Survey Memoir? on the Isle of Wight. One of the authors, from his long residence in the district and his constant occupation with these beds, has been long satisfied that E. Forbes’s account of the beds is true to nature, and his classification fully borne out by lithological identity of beds, as well as by distri- bution of the fossils. The present notes, then, are based on his part upon an aquaintance with the district, and the work of the Geological Survey there during its progress, supplemented by subsequent visits, and specially this summer by a joint examination by both, including measurement of beds and collection of fossils, which, however in- complete, was made bed by bed, and represents the prevailing fauna of each—a point on which we lay great stress. We do not wish to delay over the history of previous opinion, which has been sufficiently treated in Forbes’s and Prof. Judd’s memoirs ; but the latest criticism of Forbes’s work (op. cit. p. 141) may be alluded to. II. Tornanp anp Cotwett Bays. The Survey Horizontal Section east of Headon Hill_—The first point at issue between Forbes supported by “ nearly all observers, ” on the one hand, and Prof. Judd on the other, is whether certain marine beds known as the Middle Headon (including the “* Venus-bed” of local collectors) in Colwell Bay are rightly associated with similar marine beds in Headon Hill. The Survey identify them, and correlate the freshwater beds immediately above and below as Upper and Lower Headon respectively in both localities. This succession, however, is stigmatized (op. ct. p. 142) as a “ mistake ” of which the ‘‘ primary cause ” is considered to be an “ assumed” existence ‘“ of a great anticlinal fold of which the summit is supposed to be seen in Totland Bay. The manner in which this supposed anticlinal is regarded as having affected the strata is illustrated in Prof. E. Forbes’s memoir, pl. vii. fig. 1, and also in Sheet 47 of the Horizontal Sections published by the Geological Survey. And yet further on we read (op. ert. p. 146), “ at Totland Bay there is undoubted evidence of the presence of a slight anticlinal fold having its summit near Widdick Chine, to the westward of which the beds for a short distance have a slight dip to the south;” so that after all the only mistake the Survey could have made would have been to exaggerate the dip. We are next told of the E. and W. flexure, which causes a slight * On the Tertiary Fluvio-marine Formation of the Isle of Wight, by Prof. E. Forbes, F.R.S., 1856 [Memoirs of the Geological Survey of Great Britain]. + The Geology of the Isle of Wight, by H. W. Bristow [Sheet 10], 1862. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 87 dip to the W. in Headon Hill; and it is implied that the Survey section is false, owing to the neglect of this consideration. We must point out, however, that this Section, Sheet 47, fig. 2,in which alone the anticlinal is shown, does not go through the summit of Headon Hill at all, and passes about a half mile inland at the latitude of Widdick Chine : init the Upper Bagshot beds are made to appear a little above the sea-level at that spot; and we have great confidence that the calculations on which this is grounded are correct*; they would be brought up by the anticlinal of which the existence has just been acknowledged. So far then, we may observe, no reason has been shown why the Survey Section should not be received as correct. Prof. Judd’s Section (op. cit. pl. vii..—We do not find any indications of the direction in which this section is drawn ; but, from the names of chines which occur in it, we presume that it is intended to start from Alum Bay in a N.E. direction through Headon Hill, and after that to follow the coast-line; in this case it is not precisely comparable with the Survey Section, whose direction is indicated on the map as passing inland and crossing from sea to Solent. It will be seen at once why it does not correspond to nature, and agree with the views of other observers. It will be noticed that the Marine bed of the Headon Group (3 of fig. 3, pl. vii.) is made to exist at the sea-level near Widdick Chine; and in the letterpress we read (op. cit. p. 147) “it is admitted on all hands that at the north-east angle of Headon Hill the marine band [ Middle Headon beds | makes its appearence just above the sea-level.” On the contrary, we cannot imagime any one putting the bed in this position. At the spot indi- cated the top of the Middle Headon is about 105 feet above the sea- level; so that the section, in our opinion, is erroneous: the dip thereby is exaggerated ; and 105 feet of beds are intercalated which do not exist. We shall prove this presently by a detail-section at this point of the hill; at the present moment we wish to point out that, with the correction of this error of 105 feet, the argument against the accuracy of the Survey section entirely breaks down. - Thus, we are told that the height of the Bembridge Limestone above the sea-level at this point is 250 feet (op. ct. p. 147); then, the marine band being put at the sea-level, it follows that 250 feet of strata must intervene between that and the Bembridge Limestone ; “but the Geological Survey [vertical] section shows less than one half of that thickness of beds, and in Colwell Bay the distance between the Bembridge Limestone and the marine band is 120 meew yy SS ss But 250 feet of strata is precisely the thickness required by my interpretation.” Since we have to subtract 105 feet from * The greater thickness of Upper Bagshot beds above the sea-level in the same line of section in the old edition of Sheet 47, and in the Plates of the Memoirs, seems due to the outline of the ground at the S. end being raised too high above Ordnance datum; probably the accurate height of the Beacon Hill was not obtainable till 1870, when the revised edition of Sheet 47 was published ; other- wise the sections are practically identical. t It would be 133 feet, according to Mr. Bristow’s estimate (Forbes’s Mem. p. 142), to the top of the Bembridge Limestone in Colwell Bay. H2 88 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT the estimate in the paper above referred to, we shall certainly have to abandon this section in favour of that of the Survey; for the thickness left, viz. 145 feet, more nearly corresponds with the thickness of the Bembridge, Upper Headon, and Osborne beds, which are stated by the Survey to exist at this spot, and whose thickness would be 144 feet *. It would appear, therefore, that there is no necessity for supposing “that, in a distance of little more than one mile, a mass of beds 120 feet thick has expanded to 250 feet, and, further, that the beds have been entirely changed in their mineral character.” We do not understand the warning (op. ct. p. 146) against “trusting to the general impression which is produced by viewing these beds from a distance,” nor the purport of the following statement :— “ The strata of How Ledge and Warden Point are seen in such a true- scale section to be clearly continued in precisely similar beds ap- pearing underneath the gravel of Headon Hill.” The section offered to us is on rather too small a scale to show detail ; but, in our opinion, the beds are inaccurately laid down in Warden Cliff, and no such bed as the Brockenhurst bed occurs at all in Headon Hill. Before we commenced drawing our section, we traced the beds along the cliffs, measured their thickness, and obtained their height above the sea-level at various points, but found it possible only to show general results in the horizontal section ; the details are embodied in the vertical sections. Vertical Section at North-east Corner of Headon Hill (fig. 1, p- 91).—We may now proceed to a more detailed account of the beds. We will begin near the N.E. corner, where the Bembridge Limestone is seen, indicated on the section (op. cit. pl. vii. fig. 2) with an asterisk, and lettered 250 feet altitude. (The quarry there is not now at work; but it is the place at which one of us has ob- tained most of the finest specimens of Palwothertum which have been found in the Isle of Wight.) We take the thickness from the Survey Memoir as 25 feet. Beneath this, in the section (fig. 3, pl. vil.), we notice a blank space with the legend “slopes covered by gravel and landslips.” We think this scarcely a correct description. Landslips exist in that the clays and marls tumble and form taluses ; but we saw no gravel covering the slopes between the Bembridge Limestone and Widdick Chine ; nor indeed does it entirely conceal the beds between the Bembridgeand Upper Headon Limestones all the way to Heatherwood Point in the other direction; at intervals tumbled gravel conceals a limited portion of them. The gravel west of Widdick Chine is not accurately depicted: the thickness is exaggerated, while its horizontal extent is overestimated here t. It does not appear in the cliff certainly beyond 80 yards * Taking the Bembridge Limestone at 25 feet, the Upper Headon and Osborne at 119 feet, as read off their vertical section by scale, after altering their lower boundary a few feet to correspond with our own. t Mr. Bristow (Forbes’s Mem. p. 105) gives the entire horizontal extension of the Lacustrine beds (Post-Tertiary) on both sides of the chine as 350 yards ; the section under review makes the gravel extend about 720 yards. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 89 from the chine on the west side (private grounds interfere with examination nearer the chine). We should say, from our examination of the ground which inter- venes between the escarpment of the thick Upper Headon Limestone and the Bembridge, that there is no difficulty in seeing what beds exist there. It is true they are sufficiently interrupted by local taluses to cause trouble in making a continuous measurement ; but the tumbled portions are partial and affect only a few feet of beds at a time, so that by moving the observer’s position laterally it is possible to see all the bedsinturn. ‘This we claim to have done; we do not pretend that our measurement of the beds is any thing but rough, though controlled by repetition, because we had no levelling-instruments with us, and in shitting horizontally from one spot to another there might be frequently a slight error in picking up the next bed to be measured. We might describe the beds immediately below the Bem- bridge Limestone, in descending order, as follows:—yellowish and ochry marls, red and grey mottled marls, marly clays with nodular bands, greenish-grey clays, pale greyish clays, grey and ochry clays, stiff pale or whitish clays with calcareous lenticular bands, red and green mottled marls, ‘“ cherry marl” with calcareous bands. These are the Osborne beds of the Survey ; and they come in the precise position assigned to them in the Survey Memoirs. They are well characterized by the ‘cherry marl”—the mottled red and pale greenish marl which distinguishes the group from the Upper Headon ; and we consider the subdivision a very useful one in the classification of beds. Their thickness here, by our comparatively rough way of measurement, is 70 feet. The vertical section of the Survey Sheet 25 gives 71 feet, reading off by scale down to the bed which we have taken as our boundary; their measurement seems to have been taken near Heatherwood Point, where this series contains a thick limestone of 18 feet; as noticed by previous observers, this limestone thins out to the east, and is only represented by nodular calcareous bands at the east end of the hill ; its loss will probably be compensated by an increase in the clays. Our results are perhaps suffi- ciently near to those of the Survey to prove that the same series of beds has been examined in both cases. Weshould remark that from the lower red beds it is perfectly easy to draw a continuous vertical section to the beds below; and from here our measurements to the Lower Headon are uninterrupted in a vertical line. The Osborne beds yield few interesting fossils; Lomnea is abundant in the calcareous bands ; but, as noticed by Forbes (Mem. p. 81), the shell- substance is not preserved. These beds are identical lithologically with the mottled red and pale green series at Cliff End. The beds next below the red series are grey and ochry clays, very rich in Potamomya gregaria and Paludina lenta beautifully pre- served: we place these in the Upper Headon; they are about 15 feet thick *. * Tn the Survey Section no. 5, at Headon Hill (sheet 25, Vert. Sect.), these clays are included in the Osborne series; but in the Vert. Section. no. 4, at Col- rh the boundary is so drawn that analogous claysare put into the Upper Headon. 90 ON THE BEDS AT HEADON HILL AND COLWELL BAY. This brings us to the vertical escarpment of the thick limestones, so conspicuous a band along Headon Hill cliffs that it is indicated on the Ordnance maps, both on the 25-inch and the 6-inch. We pause awhile to draw attention to the fact that we have accounted for about 110 feet of beds from the top of the Bembridge to the top of the great limestone (Upper Headon) ; and the Brockenhurst series does not exist here. There is not a single marine fossil to be found am that interval; nor is there any bed with the faintest resemblance either lithologically or palceeontologically to the Colwell-Bay Venus-bed. This is in opposition to the view (I. ¢. p. 176 e¢ passim) that the Colwell-Bay series exists here “‘entirely concealed” by some supposed gravel talus*; yetit is upon the existence of such a second marine series, thus supposed to be added above the Venus-bed that the presence of a Brockenhurst series at the west end of the island is inferred. Next we turn to consider the thick limestone of the escarpment, the Limneea-limestone of the Upper Headon. It is in several beds, of which details are given by the Survey; we measured it by sus- pending a tape, and found it 27 feet (fig. 1, p. 91). The only difficulty in correlating the Headon-Hill beds with those of Colwell Bay is centred in this limestone: it might be a difficulty to those who would have expected a priori that the limestone would have maintained its thickness in direction of dip for a mile or two to the north; for we identify it with a limestone at Cliff End not above 1 foot 8 inches thick. It would be equally a difficulty aecor- ding to the correlation which identifies it with the How-lLedge limestone (op. cit. p. 144); im this case the 27 feet has thinned to 3 feet at How Ledge, a distance of 14 mile in a straight line, while in Warden Cliff (only 926 yards distant from the Headon-Hill bed) it is about 5 feet; so that it must have thinned very rapidly at the first stage. The limestone in the Upper Headon at Cliff End, with which we identify it, is distant 1 mile 926 yards. In either case it thins out considerably to the north, as noticed by E. Forbes (Mem. p. 84). We shall be able to prove that it does not occupy the same position as the How-Ledge bed; for we have recognized that bed, which forms the summit of the Lower Headon, at a lower position in Headon Hill and in its natural position, viz. below the marine series (Middle Headon), as in Warden Cliff, where it was last seen. As a paleontological distinction between the Upper Headon limestone of Headon Hill and Cliff End + and the Lower Headon limestone * With respect to the “inextricable difficulties and confusion” (7. ¢. p. 144) in which the Survey is supposed to be involved by their not allowing the Col- well Marine bed to come where the Osborne beds are placed, and which is sup- posed to be shown by 48 feet of strata being classified in the letterpress (Forbes’s Mem. p. 81) as Osborne, while in the plate they are classed as Upper Headon, this is merely a, question of classification and the drawing of a boundary-line, matters entirely subjective and not affecting the total thickness ; their vertical section shows almost the same thickness of beds consistently, notwithstanding certain irregularities in the boundaries and classification. t In the legend to the Survey Vertical Section, sheet 25, no. 4, this limestone in the Upper Headon is said to form How Ledge; this is plainly an oversight or clerical error, as is also the statement in Forbes’s Memoir, p. 132, that the How-Ledge bed is faulted in Warden Cliff. The fourteen faults mentioned affect the Upper Headon limestone at Cliff End: their section is fairly correct; but there seems to have been some confusion between Warden Point and Cliff-Hnd Point in the letterpress. Fig. 1.— Vertical Section of Beds at the North-east corner of Headon Hill. 63-8 0 UppEerR HEADON. bs oOo ® 8 9 in.-2ft 6 in 6 in ft." 1m: 1 10 MIDDLE HEADON. ~~ A ow f=) seen. LowER HEADON. CO for) (Scale, 8 feet to the inch.) BEMBRIDGE LIMESTONE ...... a0 feet. OSBORNE BEDS SO esac r esos ersorccras WPPER HIBADON ....0:0.0-0s00008 9 a (total). al & aye 36 CO Part of thick Limnea-limestone. Limnea fusiformis, &c. Laminated greenish clay, with broken Paludina. Whity-brown to buff sands, with layers of lignitic matter. Greyer sands below. Potamomya, Melania muricata, Unio, Paludina lenta. > Lignite. * Greyish-green oye. C. ven- \ Vicarya concava, Marginella vittata, Neri- tricosum-bed With ...........026 tina concava, Melania muricata, &e. --- Limnea-limestone, soft and crumbling, with a thin lignite at top. -*-Verdigris-green clay, with rootlets. Limnea-limestone. Stiff green clays with conchoidal fracture in drying. Oyster-bed towards the base. Fusus labiatus, Mel. fasciata, M. muricata, Clay becoming greyer below. ) Nerita aperta, Cer. variabile, C. pseudo- IM OSSIUS |i cccestcseerstessseccceees cinctum, Ostrea velata, Mytilus affinis, Cor- bicula obovata, Lucina colvellensis. Alternating grey and ochry clays. Cyth. incrassata, Mactra fastigiata, “‘ Venus-bed,” richest portion, con- Mya angustata, Corbicula obovata, tains scattered flints, brown sandy | Wucula lissa, N. headonensis, Trig. - clay becoming green clay and sand deltoidea, Fusus labiatus, Cancell. Ibelow-m Ue ROSSILSSscsceseccs-0s.-sanenescess elongata, Melanopsis fusiformis, / Voluta spinosa, Vic. concava, Natica Thin grey sandy clays, Studert. weathering brown. Cytherea incrassata &c. scattered throughout. Mya angustata, especially near base. Trig. deltoidea, Cer. pseudocinctum, Chocolate-brown or blackish sands. { Natica labellata, Melan. fusiformis, Trigonocelia-bed. Blackish-brown sands, Weritina-bed: WV. concava, M. fusiformis, C. obovata. Very stiff tenacious clay. Limnea-limestone, ‘‘ How-Ledge limestone:” ZL. longiscata, fusiformis, &c. Whity-brown or yellow sands and sand-rock, with layers of Paludina and Potamomya. [The base concealed by tumble and undercliff. | 92 H, KEEPING AND E. B. TAWNEY ON THE BEDS AT of How Ledge, we may adduce that the former is very rich in Planorbis, while in the latter- these shells are comparatively rare, the fossils being chiefly Limnea. As the Upper Headon limestone is the strongest bed in the section and forms an escarpment through the greater part of Headon Hill, in it we may obtain a datum line. We take as a point of reference the spot where the top of the escarpment cuts the surface of the ground or outline of the cliff; this 1s seen on the 6-inch or 25-inch map to be about halfway between the 100 feet and 200 feet contour-lines. From these points of known altitude, by the aid of the barometer, we obtain the height of our point of reference: ibis about 140 feet above Ordnance datum. The beds below the Limnea-lmestone are green clay, 1 foot, with broken Paludina, then pale butt or whitish sands, varying from 63 to 8 feet, with occasional layers of lignitic matter, Potamomya, and Melana muricata; below, where it is sometimes grey, Unio Solandri and Paludina lenta may occur. We take this bed with the 3-inch lignite below as the base of the Upper Headon. The boundary chosen is, of course, arbitrary ; but the fact of the next bed being decidedly brackish inclines us to draw the division from the Middle Headon here ; the Survey vertical section, sheet 25, places it a few feet lower. Our estimate for the Upper’ Headon at Headon Hill amounts to a thickness of 50 feet; the thickness on the Survey vertical section is given as 37 feet ; but if we read off the distance between the beds which we have taken as boundaries, it becomes 48 feet. The com- bined thickness of Osborne and Upper Headon beds, according to the Survey section, is 119 feet, 2.¢. adopting the top of the C. ven- tricosum bed as the boundary ; our estimate, taken at the north-east end of the hill, is 120 feet. The agreement is sufficiently close to render it probable that a thickness above the average of the cal- careous portion is accompanied at the same spot by a diminution in the clays; so that the balance of average thickness is maintained at both ends of the hill. As we have said, we think it convenient to retain the name “ Osborne Series ” for the red and greenish mottled clays and marls and pale greenish-white limestones, since these ‘physical characters distinguish them at once along this side of the Solent. We must decline to accept the statement (op. cit. p. 169) that ‘‘under this name beds lying below the Brockenhurst series, as at Headon Hill, have been confounded with others on a totally diffe- rent horizon, above the Brockenhurst series.” We next come to the Middle Headon. HH. Forbes relates (Mem. p. 85) that the uppermost and lower portions at Headon Hill are brackish-water beds abounding in Cerithium ventricosum, C. pseudo- cinctum, C. concavum, Neritina concava, and Nemature, the condi- tions being less purely marine than at Colwell Bay. This is, no doubt, true of the series as a whole ; for below the C. ventricosum bed there are two freshwater Limnea-limestones. But it appears to us that too much has been made of this ; for instance, the lower Neri- tina-bed is identical in Headon Hill with the similar bed at Warden HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 93 Cliff, and again in Colwell Bay, both physically and as to its fossils ; when we come to the portion of the series richest in Cytherea incras- sata known as the ‘“ Venus-bed” proper, we find identically the same fossils as in Colwell Bay; we cannot detect any difference, so far as our researches go. The uppermost bed is dark greenish clay, varying from 9 inches to 2 feet in thickness. When at its thickest the sands above are proportionally thinner. It is extraordinarily rich in fossils for the depth of one foot, chiefly Meritena concava, Cerithium (Viearya) concavum, C. ventricosum, Corbicula obovata, Marginella vittata, Melania muricata, Melanopsis fusiformis. We may callit the C. ven- tricosum bed; for this fossil seems almost confined to this horizon, while the Vicarya concava is found here all through the Middle Headon, though especially plentiful in this bed; its underside is occupied by an impure lignite band, with freshwater shells (Limnea and Planorbis), and rootlets, lying on a thin crumbly buff Limnza- limestone. Both together are 4 to 6 inches. Below is verdigris- green clay, 6 inches; next a buff freshwater limestone with Limnea, 1 foot 8 inches to 2 feet, the shells frequently blackish in colour. These beds betoken certainly a recurrence of freshwater conditions after the brackish bed above. Next follow truly marine beds—first stiff green and grey clay, about 9 feet, with a conchoidal fracture when dry. ‘The abundance of oysters, O. velata (Wood), is the chief fossil feature ; these and the other fossils occur mostly towards the base. Such fossils are Musws (Pisania) labiatus, Nerita aperta, Me- lana fasciata, M. muricata, Cerithiwm variabile, C. pseudocinctum, Mytilus affinis, Corbicula obovata. The fossils cited are merely the most common, such as may be found in a few minutes’ search; but these being the most charac- teristic, seem to us precisely those required for the identification and correlation of beds. Comparing the beds noticed so far with those at Colwell Bay, we observe practically identity of fossils; this last, which we may call the oyster-bed, is identical with the oyster-bed of Colwell Bay ; in both localities OC. ventricoswm occurs above this bed, and there only, so far as we know. Next below follow alternations of grey and ochry silts, 3 feet, in which we observed no fossils. Below are 11 feet of beds, brown sandy clay above becoming green clayey sands below, and then grey sandy clays: this is the ‘ Venus-bed ” of collectors ; Cytherea imcras- sata occurs near the top and is scattered throughout the whole bed, but is most abundant for the space of one foot. The fossils obtained from the Venus-bed in a few minutes’ search were Cytherea incras- sata, Mya angustata, Mactra fastiqiata, Psammobia rudis, P. westu- arina (Hid. MS8.), Nucula headonensis, N. nudata, Trigonocelia del- tordea, Corbicula obovata, Vicarya concava, Ancillaria buccinordes, FPusus labiatus, Cancellaria elongata, Natica Studeri, N. labellata, Voluta spinosa, Melanopsis fusiformis, Nematura parvula, Limnea longiscata, Planorbis obtusus, crab-claw (Callianassa), coprolite. 94 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT Towards the base Mya angustata and Corbicula obovata were parti- cularly abundant. Next below the Venus-bed is the Trigonocelia-bed, chocolate- brown sands, 34 feet, with sometimes a blackish tint ; we so call the bed from the principal fossil which occurs at precisely the same horizon in Warden Cliff and Colwell Bay; other fossils are Ceri- Ehium pseudocinctum, Melanopsis fusiformis, Natica labellata. Below are blackish-brown sands, 6 inches; at base is enough carbonaceous matter almost to amount to a lignite band; this may be called the Neritina-bed, the chief fossils are Weritina concava, Melanopsis fust- formis, Corbicula obovata, the latter in perfect condition, very large, and showing concentric colour-bands. The Nertina-bed occurs in the same position towards the base of the series in Warden Cliff and Colwell Bay. This can only be explained by admitting that the marine series in Totland Bay and Colwell Bay are identical ; the Ventricosum bed at the top, and the Neritina and Trigonocela beds at the base, identical in physical and fossil characters, are strong presumptive proof of this. Below is very stiff dark-grey clay, 1 foot to 1 foot 3 inches; fossils occur in patches, Neritina concava, Cerithium pseudocinctum, Melana muricata, Limnea, Corbicula obovata. This is the lowest bed of the Middle Headon here. Summing up, we obtain a thickness of from 31 feet 9 inches to 33 feet for the Middle Headon of Headon Hill at the N.E. end. Reading off the Survey vertical section by scale, we obtain 35 feet for it between the boundaries adopted by us * for the thickness towards the west end. The height of the base of the Middle Headon above the sea at this point, viz. about 120 yards in horizontal distance west of our reference-point, is by subtraction 72 feet. Direct barometric obser- vations gave about 70 feet. We have already used these figures when alluding to the position assigned to this series. The first bed of the Lower Headon is a Limnea-limestone of the usual buff colour: it is 2 feet thick at this point; but a little further west we obtained a measure of 4 feet. This is in our opinion the well-known bed which forms the top of the Lower Headon in Warden Clif, where it is quite a marked feature. It has there and in Col- well Bay precisely the same position in the series, supporting the Middle Headon—recognized by the Neritina-bed with all its charac- teristic features, the Trigonocelia-bed, and so on. From Warden Cliffit is traceable uninterruptedly to How Ledge, where it disappears beneath the sea-level ; we therefore speak of it as the ‘“‘ How-Ledge limestone.” It is correctly drawn on the Survey vertical sections, sheet 25, nos. 4 and 5, where in the legend is a clerical error, to which we have already alluded. This bed is so distinctly lacustrine, * Some irregularity in the boundaries of the Survey vertical section is to be noticed : viz. in the Headon-Hill Section, no. 5, the boundary is placed below the How-Ledge Limestone ; in the Colwell-Bay section, no. 4, the boundary is placed at one bed above the How-Ledge Limestone; this seems an error of the engraver, and of course does not affect the thickness of the beds. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 95 being almost made up of Limnea shells and their débris, that. it seems most convenient to include it in the Lower freshwater group ; and it makes an especially good boundary. The beds which follow next below the Limnza-limestone at this spot are whity-brown or yellowish sands, and sand rock with Paludina and Potamomya in layers ; 8} feet are seen; this is all that is exposed of the Lower Headon here. At this spot all the eround below is tumble or undercliff; and for a more complete section of the Lower Headon in the present state of the cliffs we shall do best to take the one exposed in Warden Cliff and under Totland Bay Hotel, where the fresh road-cuttings to the New Pier have been of service to us. We consider that we have already sufficiently proved the identity of this lower limestone in Headon Hill with the How-Ledge bed of Warden Cliff; but if any objection be raised that its relations to certain specified bands in the (Middle Headon) Marine series being found identical in both localities is not conclusive—even though it has been shown that no other marine bed exists in Headon Hill— we have further means at our command. ‘There are, however, only two beds in Headon Hill with which the How-Ledge bed could be continuous: viz. either it is the same as the thick Upper Headon limestone (which we consider impossible, as the beds both above and below would then just be reversed, viz. freshwater above and marine below, instead of vice versa), or it must be identical with the one to which we assign it. It is possible, however, to settle the point by ocular demonstration*. Though the How-Ledge limestone is denuded from the top of the curve between Weston and Widdick Chines, some of the lower beds are traceable the whole way. Ac- cordingly we can join on the section in Headon Hill to that in Warden Cliff. We account thus for a continuous section of beds from the lowest seen beds of the Lower Headon, through the Middle and Upper Headon of Colwell Bay, to the Bembridge Limestone on the north, and again from the same base of the Lower Headon, through the succeeding Lower Headon beds in the cliffs between Weston and Widdick Chines, to the sand below the How-Ledge limestone at the N.E. corner of Headon Hill, and thence up to the Bembridge limestone on the south ; and we find that the section is identical in both cases. There is only one Marine (Middle Headon) | series, lying between two freshwater ‘series, the Lower and Upper Headon. Of course, all this has been done before by the Geological Survey, and our work is nothing but a confirmation of results already suffi- ciently established by E. Forbes and Mr. Bristow. On our horizontal section of the coast we have endeavoured ‘to represent the position of the beds in the cliffs and the extent to which * We are indebted to the Rev. O. Fisher, F.G.S., who specially visited Tot- land Bay this autumn, for the information that the Venus-bed is found in the Totland-Bay brickyard some little way above and behind the top of the cliff between the chines. He points out that since this is the only part where it is missing from the cliff, itis the link needed to prove that the bed is continuous all through. 96 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT they are traceable ; but though the vertical scale is more than twice the horizontal, it is not possible to show the details; for these we must refer to the vertical sections. Lower Headon Beds of the cliffs between Weston and Wraddick Chines.—We left off with the Potamomya-sands, 83 feet below the How-Ledge limestone ; these are easily traceable through the grass- slopes, exposures of a few feet occurring at intervals all the way to Widdick Chine, about 230 yards distant. There good sections are seen on each side of the roadway; here the sands have in the upper part more clay mixed with them, as in Warden Cliff. Whiter sands are below; above are alternating whity-brown sands with bluish silts. Melania turritissima occurs in the latter, a shell which occurs, indeed, in the Bembridge Marls at Hamstead, but which, in this district, we only know at one horizon, viz. the Unio Solandri bed, and one above that, at Warden Cliff and Colwell Bay; we remark its analogous position at Widdick Chine. Below the sands, again, is pale greenish clay, 3 inches ; in descending order, soft buft Limnea- limestone, 1 foot; brownish sands with Potamomya and reptile dermal ossifications : these occupy the position of the Crocodile-beds in the Lower Headon at Hordwell; they continue along the cliff as we walk northwards. Below is a carbonaceous band or impure lig- nite, 6 inches, then a repetition of clays with carbonaceous layers, fol- lowed by another Limnza-limestone, 10 inches, Potamomya-clays, 4 feet, another Limnea-limestone, 1 foot, greenish clays with Palu- dina lenta, Potamomya plana, Melanopsis brevis, Limnea &e. ; another Limneea-limestone, 8 inches, full of Gyrogonites, below that clay with carbonaceous layers passing to drab sands, about 2 feet ; then a lignite layer and impure Limnea-limestone, soft and crumbling at the outcrop. The limestone full of Gyrogonites is noteworthy, as it occurs only low down in the Lower Headon, and serves to mark our position in the series at this spot. Now this Chara-limestone is exposed again at the back of the Reading Rooms, where there are | the same five limestones seen as near Widdick Chine, the Chara-bed being the lowest but one; this is well seen behind the Reading ltooms, where there has been a cutting through the Lower Headon beds for a new pathway. It will be noticed that we have passed five thin Limnea-limestones in the lower part of the Lower Headon in the cliffs immediately north of Widdick Chine, and again behind the Reading Rooms; we see them again as they rise from beneath the sea-level beyond the new pier under Warden Cliff; they are seen also in the recent scarping under Totland Bay Hotel. The continuity, then, of the section from the five lower limestones under Warden Cliff through Weston Chine to Widdick Chine is undoubted ; and from there we continue through the sands above to the How- Ledge limestone in Headon Hill. The beds in the cliff here belong entirely to the Lower Headon*. * The top of the cliff at the back of the Reading Rooms has a capping of about 7 feet of Post-tertiary sand ; at the base of this is a layer of flints and derived marine fossils, Cytherea incrassata, Ostrea velata, &e., showing that the marine Middle Headon series existed here above this level. This Post- tertiary sand lies on the (Lower Headon) Warden sands. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 97 Lower Headon of Warden Cliff.—The lowest beds seen are below the Totland Bay Hotel at Weston Chine. We therefore start from here in making a measure of the Lower Headon series, meeting higher beds as we go north towards Warden Point. ‘The details of this measurement are given in the vertical section, fig. 2(p. 98); we need here allude to only a few of the beds. The lowest bed seen in Totland Bay is a few yards south of the Coast-guard boathouse ; here are visible, in ascending order, whity-brown sands, 13 foot, a hard purple sandy ironstone, 6 inches, greenish clay, 4 feet. In our vertical section, fig.2, we have drawn an interruption in these clays, because we are not absolutely sure that they are identical with the similar clay in the road-cutting at the pier-head on the other side of the boathouse; there is little doubt of it, however. At the pier-head are seen 3 feet of clay in the road-cutting (after our section was drawn, on a later visit, a drain-opening showed 41 feet) ; and from this point the section 1s continuous without interruption. There is therefore only a possibility of error of a few feet. Five thin Limnea-limestones will be noticed. The two lowest contain Chara-seeds at the pier-head; on following them to their outcrop along the shore it is found that the fourth of them also has Chara-seeds at that point; the fifth, 1 foot 3 inches thick, is 50 feet. below the top of the How-Ledge limestone, or top of the Lower Headon. From the top of the cliffs these limestones may be seen at low water, forming five submarine ledges parallel to the great ledge at Warden Point; they strike N. 36° W. As there is great uniformity in these freshwater beds, and their fossils mostly occur in the Upper Headon also, we pass over many beds to notice the sand-rock bed, which is a conspicuous feature in Warden Chiff. This sandstone, somewhat calcareous at places and friable at others, forms Warden Ledge, and runs out at the top of the cliff _close below the flagstaff of the coast-guard station. About 11 feet above that comes the Unio Solandri bed* with Melama turritissema, a bed of which we have already noted a portion in the cliff at Widdick Chine; the MW. turritissima occurs through a greater thickness of the 114 feet of clays than does the Unio. Next comes the How-Ledge limestone, 5 feet, in two beds containing Limnea fusiformis, with a carbonaceous or lignitic band at the base. The shells are more crowded at the base; and the lower surface of the blocks fallen from the cliff is a sight pleasing to the collector of fossils. The dip of this bed in Warden Cliff, calculated from its horizontal extension on the 25-inch map, between its position at How Ledge at sea-level and a point of known elevation near the flag- staff, is a slope of about 1 im 45. This How-Ledge bed crops out at the top of the cliff, a little north of the coast-guard flagstaff; it is * 'The Unio-bed with Melania turritissima occupies an analogous position at Hordwell Cliff, being at about the same distance below the Lymnza-limestone (which is a diminutive representative of the How-Ledge bed), where it has pre- cisely the same lithological characters with the same abundance of black seeds (Carpolithes) as at Colwell Bay and Warden Cliff; it crops out again with the same fossils on the shore immediately south of Milford. 98 H. KEEPING AND E. B, TAWNEY ON THE BEDS AT Fig. 2.—Vertical Section of Lower-Headon (freshwater) Beds eaposed between Weston Chine and Warden Cliff. (Scale, 12 feet to the inch.) Limnea-limestone, in two beds, buff colour. How-Ledge bed; extends from How Ledge to Coast-guard Station.—Z. fusiformis. = Lignitic or carbonaceous band. Pale grey clays, rather mottled. Paludina at intervals. but more so above. WMelania turritissima and other fos- {batons greenish and greyer clays, rather sandy below, sils in layers at intervals, Unio-bed. Laminated grey sandy clay, with Unio Solandri, Mel. turritissima. Stiff clays, grey or brown in places, =| Sandy grey clay, more sandy above, with Paludina lenta, Mela- nopsis brevis, Potamomya plana, Unio Solandrt. >:| Pale whity-brown crumbling sand-rock, - . ‘| with strings of clay. } Cyr ena depen | Ferruginous sand-rock, variable in thickness, with Limnea, &e. Concretionary calcareous sand- rock, | 7; forms Warden Ledge. } wer alp BMSEIOCIL LG: LowER HEADON. eee on _ bo bo lor} ler) oO nn = (S) (=) Ne) (>) (=) fer) (=) =) = White and whity-brown sands, Potamomya in layers at intervals. 10 0 20 Stiff grey or sandy clay, with Potamomya crowded together. =, Lignitic band. S150 Greenish clay. Potamomya gregaria. 1 3 Buff-coloured limestone, with Limnea. Lignitic or carbonaceous band. 33-4 0 = Greenish-grey clays. 0 6 Soft crumbling buff limestone, with Limnea longiscata. Grey marly clay. 2 0 0 8 ==| Bulf limestones, with Limnea. 50 SS Greenish clay. [bed. = Buff compact limestone, with Chara-seeds (Gyrogonites) Chara- 2:27] Whity-brown sand-rock, with lignitic band below. “| ,..Butt limestone, with a few Chara-seeds. -*-Carbonaceous layers. Green clay (behind wall at Pier-head). ...Green clay, a little south of Coast-guard boathouse. =| ---Nodular purple-red ironstone band. <| Pale yellowish or whity-brown sands. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 99 denuded away in the centre of Totland Bay. It cannot, however, have been many feet above the present cliffs near Widdick Chine, while a little south of that it exists in the cliff. At the N.E. corner of Headon Hill we noted it at a height of 70 feet above the sea ; it is well seen also in the cliffs near Heatherwood Point, where it has a somewhat higher elevation. There is evidence of an anticlinal in Totland Bay, as indicated in K. Forbes’s sketchy diagram ; the summit of the anticlinal we infer to be near the old wooden pier. If we add up, we obtain a thickness for the Lower Headon exposed in Warden Chiff of 714 feet (supposing no interruption at the point specified above). It remains to settle the relation of the lowest bed here seen to the Upper Bagshot Sands. The junction of the Lower Headon with the Upper Bagshot is well seen north of Alum-Bay Chine; immediately above the Upper Bagshot sands come greenish-grey clays ; then, in ascending order, alternation of clays and sands ; next, pale greenish-grey sands ; then a stiff marly clay: total 16 feet. Above is a purplish red clay-ironstone band; and succeeding that is the first thin Limnea-limestone. If we consider this red iron-band to be at the same horizon as that noticed under Weston Chine, as is extremely probable (though it is some 4 feet nearer to the lowest Limnza-limestone), then we must add 12 or 16 feet to the 71 feet obtained for the Lower Headon in Warden Cliff, making a total of 83 or 87 feet before we reach the yellowish sand of the Upper Bagshot. Knowing now the full thickness of the Lower Headon, we are able to test the argument as to the position of the Upper Bagshot, or Headon sands, as they were once called by E. Forbes (Mem. p. 34-6), in Totland Bay. It is stated (op. at. p. 147) that the Survey actually represent these as occurring, in both the Vertical and Hori- zontal Sections, near the summit of the anticlinal in Totland Bay*. There seems a little inconsistency in the Survey Vertical Sections concerning both boundaries of the Lower Headon; if we may classify these green clays in the Lower Headon, and then read off by scale from the top of the How-Ledge bed, the Survey Section would give a thick- ness of 85 ft. for the Lower Headon in Totland Bay. It is urged “that the Headon-Hill sands do not occur in the position indicated by the Geological Survey ;” and the crucial test of * We pointed out above that the Survey Vertical Section [edition 1870] shows them only just above the sea-level at a point some way inland. The Vertical Section we are inclined to interpret in this way—that the beds denoted Upper Bagshot in Totland Bay are what we have classed in the Lower Headon ; for the legend states, below the “‘clay-ironstone ” are 6 inches sand in “ Totland Bay,” then “ green clay with lenticular patches of sand” 15 feet, so that the de- scription agrees fairly with our lowest bedsat Heatherwood Point. The Survey Section does not state that these were exposed in the centre of Totland Bay; but as Heatherwood Point is the western point of the bay, we may assume that they may have been seen anywhere short of that point along the base of Headon Hill, where it is quite certain that they exist, and must have been open at that time at more points than one; for the white glass-house sands were then being actively worked in Headon Hill, and the yellower sands above them may still be seen about a mile from the N.E. corner of Headon Hill. 100 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT the excavations at the new Reading Rooms is supposed to prove “‘ that beds are [there] found which have their exact counterpart in the Headon-Hill section, not at the base, but at a much higher part of the series.” We can readily understand that the Upper Bagshot sands were not found in the excavation; we could even have predicted the fact beforehand ; from a rough calculation, we estimate that the sands are above a dozen feet below the foundation. In favour of this view, we may add that, when the old wooden pier was being made, one of us heard from those engaged in the work, that the piles were driven with difficulty because of their piercing solid sand! *, But putting aside the question as to their exact depth here below the Reading Rooms, we are able to refute the notion that any thing higher than Lower Headon beds exist here. As mentioned above, we recognized immediately behind the Reading Rooms the five Lymnea- limestones which come below the Potamomya-sand, both at Widdick Chine on the south, and Weston Chine on the north, between which places the sands may be traced almost continuously. The whole of the cliffs between Weston and Widdick Chines are occupied solely and throughout by Lower Headon beds (neglecting the cap of Post- tertiaries). All this part of the section (op. ct. pl. vu. f. 2) is in- accurate, in consequence of the Middle Headon being placed too low in Headon Hill. Middle Headon of Warden Cliff—Again, we cannot agree with that part of the section between Weston Chine and Warden Point. Here no marine bed is indicated ; for in fig. 3, the section drawn to true scale, the Colwell bed is made to die out before the Warden battery is reached, which is occupied by an exaggerated thickness of gravel 7. There is no fact more patent to any observer than that the Col- well-Bay marine bed extends all through Warden Point and Cliff, where it is supported by the How-Ledge limestone. We made a measure in detail of the Colwell marine bed (Middle Headon) at a point about midway between Warden Battery and Weston Chine ; it is here 344 feet thick ; we noticed there the Neritina-bed + with its characteristic features and fossils below the Venus-bed as at Col- well Bay and Headon Hill. It would be wearisome to give all the details ; but the Colwell bed here is easily recognizable as identical with the Middle Headon of Headon Hill, both physically and pale- ontologically. At Warden Battery, above the Middle Headon, comes some Upper Headon § besides the Post-tertiary cap. In the section drawn to true scale (pl. vu. fig. 3) the Colwell- Bay marine bed has its horizontal extension curtailed by almost one * The iron columns of the new pier are stated to pierce a bluish clay; we should interpret this to indicate the clays immediately above the Upper Bagshot Sands, and which are described as greenish in the Survey section. t We cannot reconcile this with fig. 2 of the same plate, where the marine bed is more nearly correctly drawn. { Previously noted as bed 16 by Dr. Wright, Proc. Cotsw. Club, i. p. 95, and Ann. & Mag. Nat. Hist. s. 2, vol. vii. § Described by Dr. Wright in 1850 as bed 5, Proc. Cotsw. Club, i. p. 90. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 101 half; the result is that the dip is considerably exaggerated. Again, in Headon Hill an exaggeration of dip is produced by the error of 105 ft. in plotting the marine beds. 'The effect is that the identical bed (Middle Headon) is split into two beds separated by 120 ft. of beds. From our point of view this could only be done by counting more than 100 ft. of beds twice over. In the legend attached to the diagram section fig. 2, “‘ new interpretation,” the beds g are, in our opinion, the Lower Headon; f is the Middle Headon ; e the Upper Headon ; d and ¢ are the Lower Headon; 6} is the Middle Headon again. According to the old view, which we certainly should prefer, this last 105 ft. has no existence in fact*. Lower Headon of Colwell Bay.—The section in Colwell Bay is continuous with that of Warden Cliff; but in the bay, as we go north, a few lithological changes occur in the marine beds 7, as noticed below, and which cause the marine beds at one part of Col- well Bay to differ far more from the same beds at the centre of the bay than the latter do from the marine beds of Headon Hill. On rounding Warden Point, beyond the sea-wall, is a small rifle-target ; and from here the beds are fairly well exposed throughout the bay, though tumbled portions or a diminutive undercliff may conceal some of the beds in places, sufficiently to give considerable trouble in measuring the beds. ° The Unio-Solandri bed with Melania turritissima has been fre- quently worked by one of us below the How-Ledge limestone here but this summer we could only find tumbled portions of it. The said limestone rises from beneath the sea-level at How Ledge +, whence its appellation, and crosses Colwell Chine; here and at the target it has the same lignitic band and clays beneath it as on the south side of Warden Cliff; it thins down to 8 ft. north of the chine. We mention these upper beds of the Lower Headon to show that the Colwell-Bay marine bed, as at Warden Cliff, reposes on the same succession of beds as in Headon Hill. Middle Headon of Colwell Bay.—The lowest or Neritina-bed at the S.W. end of the bay, by the target, is now covered by tumbled matter, but is well seen a little further on about 50 yards short of Colwell Chine. Here, 7. ¢. between the target and the chine in the * The diagram in Forbes’s posthumous work is so schematic that it omits the higher part of Headon Hill, and, perhaps for clearness sake, the effect of the anticlinal is exaggerated. It is rather severe to treat it as if it were drawn to scale, and, because the Upper Bagshot Sands were brought up too much in the centre of the roll, to say that H. Forbes was “‘mistaken in his interpretation” (op. cit. p. 176) of the beds. Forbes’s diagram, in fact, with this qualification, represents the beds in their right position ; thus the Lower Headon, no. 6, occu- pies the summit of the cliffs in the centre of Totland Bay, while no. 7, Middle Headon, is denuded from above it—all which is perfectly correct. Tt Since Warden Battery has been built it is forbidden to search for fossils on the slopes at Warden Poimt. Many years ago one of the authors was in the habit of frequently exploiting the beds here for fossils, and many of them are incidentally described by Dr. Wright, Proc. Cotsw. Club, i. pp. 91, 92 (1850). The engineers, however, have not succeeded in grassing all the slopes, and fallen fossils may still be picked up at the base. _ ¢ Bed 18 of Dr. Wright, who describes it correctly in Warden Cliff, but appears to have mistaken its position in Headon sub (2b. p. 95). Q.J.G.S8. No. 146. I 102 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT Marine series above the Neritina-bed,is some foxy-coloured sand- rock, while lenticular layers of white sand are seen thinning out in the grey clays, which also contain red-clay ironstone nodules, a dif- ferent lithological condition from what occurs further north in the bay, or to the south of Warden Point. Above comes the richest part of the Venus-bed with abundant Cytherea incrassata ; these fossils not only strew the tumbled clays, but with other fossils are commingled by the waves with recent shells on the strand. Atthis point of the bay Ostrea velata is very abundant above the part richest in Cytherea. In the centre of the bay between Colwell and Bramble Chines | this oyster is in extraordinary abundance; at one place it nearly crowds out most of the other fossils, and forms a massive oyster- bank about 21 ft. thick *, of which the lower 12 ft. is almost made up of these shells. The ordinary character of the Venus- bed is quite altered here, though its fossils occur mixed up with the oysters; we may notice as especially abundant Murex seaden- tatus, Pisania labiata, Natca labellata, Nerita aperta, Cerithium variabile, Ostrea velata, Nucula headonensis, Cytherea merassata, &c. On either side of this spot, near both How Ledge and Bramble Chine, the excess of oysters has disappeared, and they are chiefly abundant in a zone above the richest part of the Venus-bed, though they do occur sparingly throughout. Cytherea merassata occurs through several feet, but the richest part of the Venus-bed consists of about 9 inches of bluish-green sand ; in this the shells are found in the best state of preservation. Above the grey and greenish-grey sandy clays richest in marine forms are some pale bluish-green clayey sands, between Bramble and Linstone (or Lynchen) Chines ; in these, at the level of about 5 feet from the base, is a band very rich in Ceri- thium ventricosum, C. variabile, Melania muricata, Corbicula obovata, with, occasionally, Verita ; this bed is seen just beyond the spring at Linstone Chine, it is only a few feet below the base of the Upper Headon. We wish to draw attention to the first-named fossil, as it occurs also near the top of the marine beds at Headon Hill, 7. ¢. in quite an analogous position. Upper Headon of Colwell Bay.—The slate-coloured grey clay with Potamomya ~ (ammediately succeeding the bluish-green sands) we take as the base of the Upper Headon. A detail vertical section is given (fig. 3), in order to show the lithological differences existing between the series here and equivalent beds in Headon Hill. We do not, however, think them greater than the differences already noticed as occurring in the marine series in different parts of Col- well Bay, while the resemblances are sufficiently great to allow of their perfect correlation—not to mention their position between the Osborne beds above (so identical with the Osborne beds of Headon Hill) and the marine series below, which we have shown, on strati- graphical grounds, is most certainly identical with the Middle Headon of Headon Hill. The differences consist here in a greater develop- ment of sand and a reduction of limestone: the sands at the base * Bed 12 of Dr. Wright (zd. p. 92), who, however, much underrates its thickness. t Base of bed no. 5 of Dr. Wright (2d. p. 91). HEADON. UPPER HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 103 Fig. 3.— Vertical Section A Upper Headon Beds in Colwell Bay. (Scale, 8 feet to 1 inch.) Paludina lenta, Pota- Light-grey clays, with layers of shells. ; momya gregaria, 8c Pale brownish-grey and light-brown clays in alternations. ft. in 15 0 5 : _ 3. § Corbicula obovata, Cy- 0 6-9 Nodular band offerruginous sand-rock.) “7 Ty» Sete 2 x Whitish sand above, yellow and ORE NYRI, I MAE 24 eee eee ears z lenta, Corbicula obovata, Torn y ; Melanopsis fusi ormis. Th al a " { Mi) | Buff-coloured TZimnea-limestone, with carbonaceous layer li th ali | below. 16 > Pale greenish-grey clays. SS Zone of Cerithium trizonatum. ———————=| Pale greenish clay. ew == Ferruginous sand layer. == Paler sandy clays. Cyrena Wrightii, Corbicula obo- io Stiff slate-coloured grey clay.4 vata, Melanopsis fusiformis, = Paludina lenta. 5 6 Varying alternations of paler and darker grey clays. = : Pale buff-coloured sands, with lignitic or carbonaceous bands. 15 0 2 rede ssene ts re, ee 2 ee or | Greyer sands below. Potamomya, a roonn ee. aa: AO ee otal’... 463 feet Bo 104 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT are considerably thicker here than at Headon Hill; their character is very variable even in different parts of Colwell Bay. The Limnza-limestone has here a much reduced thickness, as mentioned previously. Among the other beds we may specially notice the horizon of Cyrena Wrightw; it occurs in the nodular ferruginous sand-rock with Corbicula obovata, also in the sand immediately above the Limnza-limestone, and again in some of the clays below. Another fossil that we have only observed at one horizon is the Cerithium trizonatum (Morr.); it occurs in the pale greenish clays beneath the limestone, occupying only a narrow band in these clays. Equally characteristic is Serpula tenuis (Sow.), which occurs at the same horizon both here and at Headon Hill, viz. in the Upper Potamomya- clay just above the limestone. The Paludina-clays at the top are identical with those of Headon Hill at the top of the thick limestones. Measurement by tape here gave 15 feet ; total of the Upper Headon near Cliff End 46} feet. Osborne Beds of Cliff Hnd.—The red and greenish mottled marls of the Osborne series follow. These beds show for afew yards only, and then become hidden under the grass which the engineers have grown on the artificial slopes below the battery. One of us well remembers the numerous little faults (14 are enumerated by the Survey, as cited above) which repeated the Limnza-limestone. Beyond the battery the Osborne beds form the tumbled cliff; a measurement is no longer to be made with profit. Mr. Bristow gives 62 feet for the series here. III. PanxontToLogicaAL EVIDENCE. The question now arises, Does the distribution of fossils bear out the separation of the Colwell-Bay and Headon-Hill marine beds and their reference to different horizons ? and does it sanction the notion of the Brockenhurst bed being equivalent to the Colwell-Bay bed ? Two lists of fossils are laid before us, viz. one which mixes up the fossils from the Brockenhurst, Whitecliff Bay, and Colwell-Bay localities, and the other which gives those from the marine beds of Headon Hill and Hordwell Chiff; of the hundred forms (in round numbers) which occur in the latter list, it 1s said (op. cit. p. 150) ‘less than one half occur at the other three places.” We may urge at the commencement that it comes rather near to begging the question to mix up Colwell-Bay with Brockenhurst-series localities. We conceive one of the main points in dispute to be whether the Colwell-Bay bed has any more affinities with the Brockenhurst fauna than has the Headon-Hill bed; and to this sub- ject we shall address ourselves after we have first compared the fauna of the Colwell-Bay and Headon-Hill marine beds. The first thing to be done is to separate the faunas of all the localities which are to be compared together; this we have done in HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 105 the lists at the end of this communication. These have been drawn up from an inspection of the Edwards collection of Tertiary fossils in the British Museum ft, while we have added many occurrences from our researches this summer, and we believe our list to be fairly correct. We do not admit that the Edwards collection is sufficient for a full knowledge of the distribution of English Tertiary fossils: ¢.g. if a certain shell does not exist in the Edwards collection from Colwell Bay, it is no proof that it does not occur there, but simply that Mr. Edwards had better specimens of it from Headon Hill, which was considered, as, indeed, it is, an equivalent bed. The greater part of the Edwards collection was made by the hands of one of the authors; but we do not think it possible to establish from this col- lection that Colwell Bay contains so many marine forms which do not occur in Headon Hill, and therefore that the marine beds at these localities are not on the same horizon. The following reasons induce us to say this:—It was the habit of the local collectors to exploit the Colwell-Bay bed far more than the Headon-Hill locality, because it was more accessible and showed a larger extent exposed ; the fossils were better preserved, and in every respect it was easier to collect from. Again, some species which one of us supplied to Mr. Edwards from Headon Hill were not incorporated in this collection, probably owing to his having better ones from Colwell Bay; and doubtless they were exchanged with foreign correspondents or given away. These two con- siderations would serve to account for his Colwell-Bay collection being richer than his Headon-Hill one. Corroborative of this is the fact that in a few days’. search this summer we have found several species in the marine bed at Headon Hill which do not exist in the Edwards collection from that locality. Moreover we hold that the best test as to the contemporaneity of these beds is not to be obtained from the rarer forms, which may be evidenced only by a single example, but from a comparison of the commoner, which we should consider the more characteristic forms. Accordingly we add a list of the species obtained by us this summer from the Middle Headon of both localities: H indicates Headon Hill; C the equivalent bed at Colwell Bay. List of all the Shells obtained on the ground by the authors for the purposes of this paper (Aug. 1880). (Those with an asterisk pass up from Barton beds.) *Lamna contortidens (4g.) ... C,H. | *Borsonia sulcata(fouw.)......... C. Marginella vittata (Hdw.)...... H. | *Rimellarimosa (Sol.) ......:.. OVE *Voluta spinosa (Lam.) ......... C, H. Murex sexdentatus (Sow.)...... Opal or otoma headonensis (Hdw.).C, H. BPs Seen eee Cees on awetces lek denticula (Bast.), var. Pisania labiata (Sow.) ......... CO, H. odonbellat(Hd.) os. 5.00.42 C,H. | *Cominella flexuosa (Hd. MS.) O, H. t We are greatly indebted to the Keeper of the Geological Department, Dr. H. Woodward, F.R.S., for his courtesy in allowing us access to the collection, even during the laborious process of packing up and moving the national col- lection to the new buildin g at South Kensington. 106 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT List of Shells (contenued). *Ancillaria buccinoides (Lam.) C, H. Planorbis euomphalus (Sow.) C, H. *Natica Studeri (Bronm)......... C, H. obtusus (Sow.) .........00. C,H. labellata (Lam.) ......... @), dal Paludina lenta (Sow.) ........- C,H. Cancellaria muricata (Wood) C, H. Ostrea velata (Wood) ......... C, H. elongata (Nys?)..........+ C,H. | *Mytilus affinis (Sow.) ......... C, H. Odostomia gracilis (Hd. MS.) C. *Trigonccelia deltoidea (Lam.), OC, H. Scalaria tessellata (Hd. MS.)... C. Leda propinqua (Wood) ...... C, H. *Cerithium yariabile (Desh.) ... C, H. Nucula headonensis (forbes) C, H. we pseudocinctum (d’ Ord.) CO, H. nudata (Wood)... <2... C, H. duplex GSow:) ae. sees H. Lucina (Strigilla) colvellensis ventricosum (Sow.) ...... OC, H. (Ed « MS.) 05 occ eeceeeeeee C,H. varians (Hd. MS.) ...... C,H. | Cytherea suborbicularis —— (Vicarya) concayum (Od MIS. Vins Sine eee C,H. (SOU) ca tn owilnvecturboaecmees Cons incrassata (Sow.) ........- C, H. Melania fasciata (Sow.)......... C, H. Tellina headonensis (Hd. MS.) OC, H. muricata (Sow.) ......... C,H. | [*]Psammobia compressa (Sow.), Hydrobia bulimoides (Hd. MS.) C, H. var. estuarina (Hd. MS.)... C, H. Melanopsis subfusiformis rudis: (Lane) eee eee C, H. (GUGIS anecion waoaatocadoasec SG, Ee Mactra fastigiata (Hd. MS.).,. C,H. fusiformis (Sow.) ......... C,H. | *Mya angustata (Sow.)=pro- Acton dactylinus (Ed. MS.) . C. ducta (2d MiS)) ieee eee C,H. Nematura parvula (Desh.) C,H. | *Corbula cuspidata (Sow.) ...... C,H. Nerita aperta (Sow.) .........5 55) (O4dab Corbicula obovata (Sow.) ...... CoE zonula) CHW 00d) <...2-2ke-- C,H. Cyrena cycladiformis (Desh.). C, H. Neritina concava (Sow.) ...... C,H. | *Balanus unguiformis (Sow.). C, H. Bulla estuarina (Hd. MS.) ... CO, H. Callianassa Baylu (Woodw.). C,H. Limneea longiscata (Sow.)...... C, H. Faunas of Middle Headon from Colwell Bay and Headon Hill compared. The above list contains only the commoner forms, such as may be found in a few days’ search. Out of a total of 58 species it will be observed that all but 7 were found by us in both localities *, and all but three are known to be common, or, again, a proportion of 94 per cent. of commoner Colwell-Bay forms occur.at Headon Hill. Surely from this we may presume a very close .affinity if not idence of these beds. It is stated (op. cit. p. 150) that less than one half of the Headon-Hill and Hordwell species occur at Colwell Bay—a result, it seems to us, only obtained by mixing up fossils from Brockenhurst and Whitecliff Bay in the same list with the Colwell forms. We shall show below that the fossils cited from these two latter localities belong to a lower zone. Next as to the statement that at Colwell Bay “ the strata are of purely marine origin” while “ the so-called Middle Marine beds of Headon Hill and Hordwell Cliff are of totally different character ” * Some of these do not exist from both localities in the Edwards collection, val have not found their way into the Headon-Hill.and Hordwell list in the paper referred to, though previously cited by Dr. Wr ight: from Hordwell ib. p. 124). ~ Comparing the whole known fauna from the Middle laid of Colwell Bay and Headon Hill, we obtain the following result, viz..74-per cent. of the Colwell-Bay marine forms have been found at Headon Hill. ‘This is counting as separate species many names in the Edwards collection which are founded on imperfect and single specimens. As we have said, we think a surer guide in comparing faunas is to take only the characteristic and less rare species. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT, 107 &e. (l.¢. p. 148). Of the brackish-water genera which are supposed to be found in Headon Hill only, we may remark that we found Cerithium, Cyrena, Hydrobia, Liamnea, Paludina, Planorbis, Melania, and Melanopsis fully as plentiful m the marine series of Colwell Bay as at Headon Hill; e.g. in a quarter of an hour we turned out half a dozen specimens of Limnea longiscata* from the richest nine inches of the Venus-bed, the best zone for Voluta, Cancellaria, Murex, Cytherea, &c. It has always been the opinion of one of us, who has worked these beds for so long, that these freshwater forms were either drifted down by flood-waters or were dead shells washed out of lacustrine or brackish deposits. They cannot have lived in the waters depositing the marine bed at Headon Hill any more than at Colwell Bay. Another argument brought forward in opposition to the views of the Geological Survey is, that certain species of Cerithiwm are confined to Headon Hill and do not occur in Colwell Bay; and by this means have been “ detected the serious errors which have crept into our classification and correlation of the strata we are now con- sidering” (op. cit. p. 149). Cerrthiwm ventricosum and C. concavum are said to be entirely confined to the Headon Hill and Hordwell localities. We cannot agree with the statement as to the distri- bution of C. ventricosuwm in the Headon-Hill beds and its “* prodigious abundance.” It is there, as far as we have observed, found only in one bed; moreover, it is equally abundant in a bed in a precisely similar position at Colwell Bay, viz. at the top of the Middle Headon. Its analogous position in these two localities we consider as fossil evidence confirmatory of the stratigraphical. Nor do our observations confirm the statement of “ prodigious abundance ” of C.(Vicarya) concavum at Hordwell Cliffin the Middle Headon. One of the authors who worked that bed when a special excavation was made for the purpose‘, considers that V. concava was extremely rare in the Hordwell bed; but, as is well known, it occurs abundantly in the Upper Bagshot sands further west at Long Mead End. As to the supposed absence of V. concava from Colwell Bay, we remark that we had not been many minutes at work on the richest portion of the Venus-bed before we found a specimen, subsequently followed by a dozen more. It can scarcely be maintained, therefore, that the Colwell-Bay bed does not belong to the C.-concavwm zone.» This species is here, however, not so common as at Headon Hill +. * Also noticed by Mr. Bristow, F.R.S. (Mem. 10*, p. 61), as well as by previous writers. t The Middle marine or Middle Headon bed at Rook Cliff, Hordwell, has not been exposed for the last twenty-eight years ; it is covered up by a great thick- ness of gravel, and its precise position is known but to few geologists. It was quite a thin bed, but rich in fossils, especially minute forms. Fossils in existing collections were all obtained about a quarter of a century ago. { This species exists, however, in the Edwards collection, labelled as from Colwell Bay. The absence of a shell in the Edwards collection from Colwell Bay is no proof that it did not occur there; the local dealers might not have thought of picking up V’. concava at Colwell Bay. For this species they went to Headon Hill, where it was more abundantly found and in better preserva- 108 H, KEEPING AND E. B. TAWNEY ON THE BEDS AT Summary.—There is but one marine bed here, namely that in the Middle Headon ; for the Colwell-Bay bed can be traced strati- graphically into the Headon-Hill Venus-bed, and the paleontological evidence 1s in harmony with the stratigraphical. The place of the Brockenhurst bed is at a lower horizon in the Middle Headon ; but it does not appear anywhere in the west end of the island. TV. Watreciirr Bay anp New Forest. Middle Headon of Whitechff Bay—We next have to raise a more serious objection to the way in which the Whitecliff Bay section has been interpreted. In mixing up all the beds in the marine series there together and calling them Brockenhurst series, it seems to us that the question has been obscured, if not begged. The statement is that the 100 feet of marine beds at Whiteclifi Bay are the equivalents of the 25-feet of marine beds at Colwell Bay and of the beds in the New Forest with the Brockenhurst fauna (op. cit. 148). Hence the Colwell-Bay bed is placed in the Brockenhurst series, which is said to occupy a higher horizon than the Headon- Hill and Hordwell marine bed; and this view is indicated by dotted lines in the vertical sections on p. 170. Since the 100 feet of marine beds are classed together and called *‘ Brockenhurst Series,” we suppose that the Brockenhurst fauna is imagined to occur throughout them. Asa matter of fact, that fauna is confined to one zone, and that the very base of the series. Though we have worked over this part of the section bed by bed, we need not here give all the details, but will refer to the description of it on the Survey Vertical Section on Sheet 25. This series is there justly referred to the Middle Headon, since it lies between the freshwater Lower and Upper Headon, its total thickness read off by scale being 90 feet. At the top are clayey sands and yellow sands about 19 feet; then the “ Venus-bed” clays &c., 15 feet; next, below, are compact sands with nodules about 42 feet, said to contain Sanguinolaria Hollowaysii*; then come 14 feet of brown clays, the base ‘‘ greenish and brownish clay, very fossiliferous.” Now there is no doubt at all about the bed above and the Venus- bed here being any thing but the Colwell-Bay and Headon-Hill marine bed—its position and its fossils prove that; the characteristic Brockenhurst fossils are absent from it, and it is therefore certain that it differs entirely from the Beeeonheret beds. The Sanguinolaria-sand differs lithogically from the lower part of the marine series, both at Colwell Bay and at Headon Hill. Such few fossils as we observed in it are distinctive, not of the Venus- tion. Moreover these collectors sought to obtain as many forms as possible, but were not concerned in finding the same species in both localities; and if they found only a few stray examples at one locality of a species of which they had a great number from the other, they were liable not to put a separate label for the odd few, but to mix them up with the larger parcel. * A wrong determination; the shell is Psammobia compressa, var. estuarina, Kd. MS. It occurs in the natural position of life, 7. ¢. across the bedding. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 109 bed, but of Brockenhurst beds; but though we have not sufficiently worked out this fauna, we may say that we are satisfied that they belong to a lower zone than any of the marine beds at Colwell Bay or Headon Hill, the Middle Headon being more fully developed at Whitecliff Bay than elsewhere. Brockenhurst Zone at Whatecliff Bay.—The succeeding 14 feet are the equivalents of the Brockenhurst beds; the lowest two feet we shall call the Brockenhurst zone ; the remainder of the thickness is not nearly so rich in species, and their grouping, as well as the litho- logical character, is more lke that of the Roydon beds. At the time the Survey section was made, the interesting bed at Brockenhurst had not been discovered nor its fauna described; hence such Brockenhurst fossils as were found in this zone here were not rightly determined (thus in the Survey section we must read Cardita deltoidea, Sow., for C. acuticosta), or specific names were withheld from them. Subsequent observers* have recognized the Brockenhurst fauna in this lowest bed. As we have obtained more fossils from it than previous observers, we have embodied our results in a separate column in the lists at the end of this essay ; that column contains nothing except what we have collected with our own hands this summer from the lowest two feet +, lying on an eroded surface of the freshwater Lower Headon. Comparison of this list with the fauna from Brockenhurst itself will convince most, we think, of the perfect equivalence of the zone in the island and in the forest, while its position at Whitecliff Bay shows that it is at the base of the Middle Headon. Brockenhurst Zone in the New Forest.—The greater part of the fossils from Brockenhurst were collected by the hands of one of the authors, and thence were dispersed into various public and private collections. They were obtained during the doubling of the line and widening of the cutting at Whitley Ridge, near Brocken- hurst t, about twenty-three years ago. During this work he had - the advantage of seeing more of the beds than any other geologist. He found the rich Brockenhurst zone (which varied from a few inches to nearly a foot) lying immediately upon the freshwater Lower Headon ; while about half a mile up the line, near the bridge by Lady-Cross Lodge, the Middle Headon Venus-bed was seen, followed by the freshwater Upper Headon beds above, the beds having a very gentle dip up the line oreasterly§. It is evident that the succession * Videlicet Von Konen, Quart. Journ. Geol. Soc. vol. xx. p. 98; Rev. O. Fisher, Quart. Journ. Geol. Soe. vol. xviii. p. 67, footnote; Mr. T. Codrington, Quart. Journ. Geol. Soc. vol. xxiv. p. 519; Dr. Duncan, Pal. Soc., ‘ Fossil Corals,’ i. p- 40 (1865). t In the Edwards collection the label “ Whitecliff Bay” includes many Venus-bed forms, indeed Lower and Upper Headon, or it may be any thing from the London Clay to the Bembridge Marls; there is therefore good reason for not allowing this collection to stand as evidence of what is found in the Brockenhurst zone at Whitecliff Bay. { The railway-cutting at Brockenhurst (op. cit. p. 152) refers to the same spot as Whitley Ridge. § We visited the New-Forest localities together this summer, and found the Whitley-Ridge cutting entirely grassed over (the rich zone was below the level 110 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT here is, in ascending order, freshwater Lower Headon, Brockenhurst zone, Venus-bed, then freshwater Upper Headon, which agrees with the succession in Whitecliff Bay. Yet in the New-Forest section (op. cit. p. 170) we find the Brockenhurst bed placed above the marine band or Middle Headon of Totland Bay—in other words, the natural succession is inverted. Again, in Headon Hill as we have seen, an imaginary Brocken- hurst bed (of which the Colwell-Bay Middle Headon is stated to be the equivalent) is placed above the Upper Headon, in ground which is really occupied by the Osborne beds. If the Brockenhurst bed is at a higher horizon than the Middle Headon of Headon Hill, then where is the marine Middle Headon at Whitecliff Bay ? We can scarcely adopt a theory which makes the Colwell-Bay bed occupy a higher horizon than that of Headon Hill, because it is supposed to contain more Brockenhurst fossils, when the latter fauna is found below the zone with (Colwell-Bay or) marine Headon fossils both at Whitecliff Bay and near Brockenhurst. When once the position of the Brockenhurst fauna is recognized (and it has been correctly described by previous observers), the inconsistency of the theory is apparent, Affinities of the Brockenhurst Fauna.—Seeing that the Brockenhurst fauna, if different in age from the Marine Headon, is older, instead of being younger, it would be rather anomalous to find that “ while nearly one third of the Hordwell and Headon-Hill marine shells are Barton forms, not more than one fifth of those occurring at Brocken- hurst, Colwell Bay, and Whitecliff are found at Barton.” We have already mentioned one feature in the lists on which this statement is based by which the question is almost begged. We must next allude to what seem to us clerical errors, in order to justify the very different statistics which we have obtained by inspection of the Edwards collection, supplemented by our own researches. In the Headon-Hill list we observe nine species* that are said to pass down into Barton beds, while in the Brockenhurst list this range is denied to them; and besides these nine, the range into Barton, as proved by the Edwards collection, is omitted in the Brock- enhurst list in the case of twenty-two other species. Discrepancies of this sort must seriously detract from the value of any statistics based on such lists. of the rails, and will never be seen again here); the upper beds were yellowish clayey sands, poor in fossils. Sufficient characteristic Brockenhurst fossils may still be seen, however, on the old spoil-banks of the date of the making of the original single line, about forty-two years ago. By Lady-Cross Bridge the cutting is also grassed over ; but evidence can still be found of the Venus-bed in the side drains and of the Upper Headon in the slopes above it. * These species are—Borsonia sulcata, Nematura parvula, Mytilus strigil- latus, Cardium obliquum, Trigonocelia deltoidea, LIncina obesa, L. concava, Panopea subeffusa, and Scintilla angusta. On the other hand, an error on the opposite side, omitting the range into Barton beds in the case of Marginella simplex and Corbula cuspidata, goes only one quarter of the way towards redressing the balance. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. pa So far from the Brockenhurst zone having fewer forms common to the Barton than the Headon-Hill marine bed, we consider that it has rather more, as its position at the base of the Middle Headon at Whitecliff Bay would lead us to expect. Thus, if we take the whole Brockenhurst fauna (including the 15 corals which are mostly special to the zone), we obtain a total of 151 species, of which from 74 to 81 pass up from Barton, or a proportion of about one half. Summing up the Headon-Hill forms in the same way—out of a total of 79 species, 23 pass up from Barton beds, or a proportion of 29 per cent. But, instead of including the rarer forms, if we take only the more characteristic and abundant species of the Brocken- hurst zone, it would be perhaps a preferable course. The following list is a catalogue of the fossils in the Woodwardian Museum from the Brockenhurst zone, obtained by one of us many years ago at Whitley-Ridge railway-cutting, New Forest; and it may be taken to include the chief characteristic fossils of the zone. We have found all, except two, in the 2-feet bed at Whitecliff Bay this summer. xInfundibulum trochiforme = obli- quum (Sow.). *Hippocrenes (Rostellaria) ampla. *Rimella rimosa. Murex hantoniensis (Hd. MS.). xTyphis pungens, *Strepsidura armata. Cancellaria muricata. Pisania (Fusus) labiata. *Clavella (Fusus) longeeva. Leiostoma ovatum, xCassis ambigua. Ancillaria buccinoides. Pleurotoma transversaria. cymezea. headonensis. ¥ denticula. % pyrgota. oo decora (Beyr.) = maga é). * spinosa. suturalis. geminata. xActezon simulatus. Marginella estuarina. xNatica hantoniensis. * Studeri. x— , var. grossiuscula (Hd. MS.). x—— labellata. Chenopus Margerini, var. speci- osus, Phorus eretifer (Ed. MS.). Ostrea ventilabrum = prona (S. Wood). xAnomia tenuistriata. Pecten bellicostatus, Modiola Nysti. xAvicula media. xLucina bartonensis (Ed. MS.). x Cardium porulosum. Protocardium hantoniense (Hd. MS.). Cardita deltoidea. xCytherea incrassata. suborbicularis (Hd. WS.). Solandri. Cyprina Nysti. * [x]Crassatella Sowerbyi, var. hanto- niensis (Hd. MS.). xCorbula ficus. x cuspidata. [x]Psammobia compressa, var. arcu- ata (Hd. MS.). Panopeea sulculosa (Hd. MS.). Madrepora anglica. Dendrophyllha. Lobopsammia cariosa. Balanophyllia granulata. Solenastrza cellulosa. Of this shorter list a proportion of about 50 per cent. pass up from Barton or Bracklesham beds; so that, taking the whole fauna or the more characteristic members of it only, in either case nearly one half pass up from Barton beds—a very different thing from one 12 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT fifth, which was the proportion given in the paper referred to (op. cit. p. 150). If we compare with this the ratio of Barton forms in the list of commoner Headon-Hill given above (supra, p. 105), we find that 30 per cent. pass up from Barton beds (while in the complete list of the Headon-Hill fauna the porportion is also about 29 per cent.). Moreover, if we first eliminate the forms that occur also at Brocken- hurst, so as to obtain what we may consider as specially Headon species, the proportion becomes even lower. On all hands the paleeontological evidence seems perfectly in accord with the strati- eraphical. Relation of Colwell Marine to Brockenhurst Fauna.—In order to see whether the Colwell bed is more nearly related to the Brocken- hurst than is the Headon-Hill bed, we will first take the proportion of Barton forms in it for comparison with similar treatment of the Headon-Hill catalogue. Examination of the list gives us a pro- portion of 29 per cent. of Barton forms in the Colwell-Bay bed; we saw above that, in the Brockenhurst bed, the ratio was about 50 per cent. and in the Headon marine bed 29 percent. An inspection of the list of more characteristic Colwell and Headon marine fossils (supra, p. 105) showed that these faunas are practically identical Now we see that their proportion of Barton forms is nearly equal, and far lower than in the Brockenhurst bed. To complete the proof from fossils, if any such is needed, we may inquire whether there are more Brockenhurst forms peculiar to Colwell Bay than to Headon Hill. Examination of the lists shows that only the following Brockenhurst species occur at Colwell Bay and not at Headon Hill, viz. Scalaria tessellata and Tellina affinis, the latter passing up from Barton beds; while those occurring at Headon Hill and not at Colwell Bay are Marginella estuarina and Cardita paucicostata—two only in each case, which amounts to perfect equality. If, on the other hand, we count those common to the Colwell and Headon marine beds, ail not occurring at Brockenhurst, we find twenty-six species in this category. We are at a loss to understand how any one could imagine that the Brockenhurst fauna is identical with that of the Colwell-Bay bed and newer than that of Headon Hill. That the Colwell-Bay bed is stratigraphically identical with the Headon Middle Marine we hope that we have sufficiently proved ; and the fact is confirmed by fossil evidence. The same twofold proof has been brought forward to demonstrate that the Brocken- hurst bed, where present, lies at the base of the marine Headon beds and immediately above the Lower Headon. This bed is absent at Colwell Bay and Headon Hill, but is seen at Whitecliff Bay, Brockenhurst, and Lyndhurst. Since the Middle Headon includes every thing between the fresh- water Upper and Lower Headon, it must be allowed to include fhe Brockenhurst beds, though that special fauna was not known when names were given to these groups of strata. It would cause the ereatest inconvenience to abandon the term Middle Headon, as it HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. Lis would entail the abandonment of the names Upper and Lower also. There is therefore no room for the term ‘“ Brockenhurst Series” in the sense proposed in the paper referred to (op. ct. p. 168)—a classification which would be in conflict with the best authorities, and founded, as we have endeavoured to show, on a defective appre- ciation of the beds. We may urge that no new facts have been discovered concerning the succession of the strata to make any correctionary classification or nomenclature in the main groups of the Upper Eocene at all necessary or desirable, and we should prefer to retain the local groupings which have been so long familiar to geologists. We may denote as the “ Brockenhurst Beds ” the lower part of the Middle Headon with the rich Brockenhurst fauna. It is not always developed; the absence of admixture of fresh water was evidently the necessary condition of its abundance of marine mollusca and of the existence of its corals. It may perhaps be subdivided into the “‘ Brockenhurst zone” and the ‘‘Roydon zone”*. The correlation of this fauna was justly made * We found the Roydon brick-yard pit in a good state for examination this summer, and obtained from it twenty-eight species. They all came from the sandy clays with bands of iron-ore septaria; the lowest beds were below the level of the standing water. ‘The section is as follows :-— Gravel, Post-Tertiary. 2-3 feet. Bluish to yellow-grey clay. ( Murex -sexdentatus, Cardita ob- | longa var. transversa, Pisania Gr hies { ‘“Shell-bed ;” clay very full of 4 labiata, Trigonocelia deltoidea, 5 shells. | Ostrea velata, Cytherea incras- | sata, Cyrena obovata var..sub- \ regularis (Ed. MS.). Voluta geminata, Voluta spinosa, Strepsidura armata, Pleuro- ( Grey clay. | toma transversaria, Pleuro- | Two nodule-bands of iron-ore 4 feet. | septaria separated ° Sune: ae 7 feet 4 1 pare by grey toma hantoniensis, Natica } sandy clay. ; 1 , . eee 9 epiglottina, Bulla Lamarckit, | Stiff bluish clay for the lower 2 . Sete leet Protocardium hantoniense, : Cytherea suborbicularis, Psam- mobia estuarina, Corbula pisum, &e. 7 feet. Greenish-grey clayey sands. — Reposing on Lower Headon fresh- water clays. The shelly bed, we consider, represents part of the Venus-bed or Headon- Hill marine zone, since it contains the characteristic oyster and Murex sexden- tatus, &e. The clays and clayey sands below, of which we examined 74 feet, while, according to the statement of the men employed, the remaining sandy beds below are another 7 feet, we propose provisionally to term the “ Roydon zone.” Tt is characterized paleontologically by the abundance of Voluta geminata, differing from the “ Brockenhurst zone” by the absence or great rarity of Voluta suturalis, Pleurotoma cymea, and Cytherea Solandrz, for the latter shells are 114 H. KEEPING AND E, B. TAWNEY ON THE BEDS AT by Von Konen in 1864, and by Dr. Duncan for the corals in 1866, whose work is not in any way affected by any thing in the present essay. We have merely striven to prevent the beds in the Isle of Wight and the New Forest being thrown again into confusion, and the accurate work of EK. Forbes and the Geological Survey being rejected on such insufficient grounds as have ‘been recently put forward. abundant in that rich zone, a few inches thick, in the Whitley-Ridge railway- cuttin At the time of obtaining the Roydon fossils in the Edwards collection, one of the authors sank a pit to the base of these beds, and they were found lying immediately on the freshwater Lower Headon clays, the Brockenhurst zone being absent; the latter has apparently thinned out here, as the Roydon zone itself thins out a little further west. Of the White-Cliff Bay beds we are disposed to place i in the Roydon zone all those between the ‘‘ Venus-bed ” clays of the Geological Survey Vertical Section and the lowest two feet of sandy clays lying immediately on the eroded surface of the freshwater Lower Headon, which said bed we have described above as the “ Brockenhurst zone.” The Roydon zone will thus include the 42 feet of yellow and green sands with ironstone-nodules, of which the chief fossil is Psammobia estuarina (=Sanguinolaria Hollowayst of the Geological Survey Section), also the remaining 12 feet of beds described in the legend as “brown clay,” but which in their unweathered condition are slate-colour to greenish grey. We had unfortunately not sufficient opportunity to work out the fauna com- pletely; but such fossils as we found induce us to parallel these beds with those of the Roydon brick-pit. The lithological character of these lower beds, as seen below low-water mark at equinoctial tide, is singularly like the clayey sands of the Roydon brick-yard, while their chief fossils are the Psammobia, lying in the natural position of life, and Cardita deltotdea. At some future time we hope to work out the fauna more completely. At Cutwalk hill, Lyndhurst, both the Brockenhurst and Roydon zones occur. In the pits which one of the authors sank in obtaining fossils for Mr. Edwards and others from this locality, the Brockenhurst zone was found lying, as usual, immediately on the freshwater clays of the Lower Headon. Hence we may divide the Middle Headon into three zones, distinguishable easily by fossils (though, of course, many species are common to all three), viz. the Brockenhurst, the Roydon, and the Venus-bed or Headon-Hill zone. The percentage of Barton forms diminishes as we ascend in the series. HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. aly Fossils of the Middle Headon series, including the Brockenhurst beds *. Mippie Heapon. UprEr Brockenhurst |Baagsuor. beds. Be : . os} ~ See es Loe, SCRA rk se iene lees ® qd]jc sl | SS 5 Si ly. = S) a 5 2ijf}] 4 lines mS | = | SO = S 3) ceo Ei slaielel/e12| #3 ee] fe tS Ge) ee tase | ea | 9s xLamna contortidens (Ag.) ......... ee mae eg « |K.T. K.T. Marginella simplex (Hdw.)......... shew lien ec Mi ctestet torte 50 * estuarina (LdW:) 2 ssccss- dae | pe * * —— Vittata (Hdw.) .........sceeseees con ||: Voluta geminata (Sow.) ............ noo ees x |K.T.) x — var.tereticosta(Hd.MS.)| ... |... | ... | * | * % decors @Bey7,.)—maga (Hd.).| ... ||... |... |... (KT) x | x suturalis (Mysz) = contabu- [EUG (JEG) ih ORO ie Aa ee eee A iy Oem eae Glico at Gee Rn 2 lia * SIMON CHAI)! Vesetaesacko ces x |K.T.) « | x |K.T.) «x | * Mitra gracilenta (Hd. MS.)......... Sisteeall meted esate | ie apmcenmiatan (HO neVLS. yess ectesclcoae || veces |v eines. vooes (Woelyl) eae |) 96 — polygyra (Hd. MS.) ......... Woe hrceiseahlpravo sis Wacetece litsione [itevictant| thee [¥ |Conorbis dormitor(So/.), var. semi- TUTOR CLUGAVE Vcasescce tase odasecsases ser ea semlinatisedl mend lieteate ta 9a! (UE [x] procerus (Beyr.)=alatus, var. Ienatllissey CHA ac. ce ces sede ecumese x |K.T.) x | x Pleurotoma transversaria (Lam.) x |K.T.| x | x Gyameear(HG.)) cli ecbecc noses ees Soe eo Ren pean eae ome ea De * — var. nana (Hd.)......... Gon ates aay re * YGC OA CHG.) ccs vepecsceset ec Shou epee toes es x [ayes Dio kes (27000 Meee ee eee Bel ee Ae vse ate % Pa NVOOUU CHA.) 20. ccececsesoseee Paes le oe headonensis (Hdw.)............ x | « | « |K.T. i le as * denticula; (Base.) 0.20.2. a | to =| z ee elelelel2|sla| we Sele ele ie |e] ee pal eel tS) | ee |e es) | el ns Brought forward ............ AL SGP Gl toon aoninclenieon 14 Nucula nudata ( Wood)...........066. weer [Fe Se le * BUMS (SOW)! 6.2 jcoeacte nsec tordle cea lavaren ol aoe EET * Lng! (17/12) aan eec ae Ws [baceat | notes Goods Leda propinqua (Wood) ............| «+ KET ioe Ke Ke: ¥—— minima (Sow.)...ccecceccceceene att 5 OSE Be * Cardita simplex (Hd. MS.)......... Bad * paucicostata (Sandb.)=nodi- Celeb HAWS) ses cy vas ocoed sede ee ens Xe tele oi |i 9e deltoidea (Sow.) .........c0008. ae oy ESSAI HERG ASSN) Pee alee orbicularis (Goldf.) ......... Lil Selteta lias «Re [+] oblonga (Sow.), var. trans- SENAY MIS.) atcseoes ee stoacsc: Se GOIN Wicca [has alee oe % ——— [*|Crassatella Sowerbyi, var. hanto- MINEMSIS A CHA s)) osc secvdeerccesss ens xLucina obesa (Hd. MS.) CONCAVACDELT.) sss )\iacsecceasascess wos: |: wales | Me se 0 eet ENGIN s IRterereial HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 121 Notes relating to Divergences from Prof. Judd’s lists of Fossils, with Observations on the Edwards Collection of Middle-Headon Fossils. Marginella simplex is not in the Edwards collection as from Barton ; this and the following ten species, viz. Mytilus strigillatus, Borsonia sulcata, Nematura parvula, Trigonocelia deltoidea, Lucina obesa, L. concava, Cardita oblonga var., Cardium obliquum, Scintilla angusta, and Panopea subeffusa, are given in one list as occurring in Barton beds; and in the other list this range is denied to them. The range into Barton or Bracklesham beds (as shown by the Edwards collection) has been overlooked in the case of many species (viz. twenty-two) in the lists of Colwell-Bay and Brockenhurst fossils (op. cit. pp. 153-156). Voluta depauperata (Sow.) has been cited in error by Forbes from Colwell Bay ; it occurs only in Barton or Bracklesham beds, and is therefore omitted from our list. V. spinosa (Lam.), type, occurs in Barton and Bracklesham beds; the form from Brockenhurst and Middle-Headon localities might be recognized as a distinct variety. V. tereticosta (Kd. MS.) is plainly only a variety of V. geminata (Sow.) in which the coste are a little less spiny; all intermediate degrees occur. Clavella longeva, var. egrequt (Beyr.). Von Konen mentions this from Brockenhurst; but we find the absence of ridges as rare as in Barton examples, and therefore omit the varietal name. Hippocrenes ampla (Sow.) is not in Edw. coll. from Headon Hill or Hordwell ; and we consider it does not occur there. Murex seadentatus, var. cinctus (Kd. MS8.), is labelled in Edw. coll. as from Barton; but we suspect this to be inerror; it appears to be from Colwell Bay, as in Prof. Judd’s lists. Natica obovata (Sow.) occurs at Bracklesham, as shown by the Edw. coll. WV. grossiuscula (Kd. MS.) is probably only a variety of NV. Studer, as transitions exist between them. NV. dubia (Kd. M8.) we consider only a large variety of NV. lamellata (Lam.). NV. epi- glottina (Lam.) is in Edw. coll. labelled as from Hordwell; we found it at Roydon. Cancellaria elongata (Nyst) is in Edw. coll. from Headon Hill and Hordwell, and C. muricata from Hordwell. C. roydonensis (Kd. MS.) seems a doubtful species. : Cerithiwm pyrgotum (Ed. MS.) we consider a Lower-Headon form, and omit it therefore. C. varians (Kd. MS.) isin Edw. coll. from Headon Hill. C. cavatum (Kd. MS.) may be only a variety of C. con- cavum ; both it and C. speculatum (Kd. MS.) occur only at Long Mead End, and should be omitted from the list. C. ventricosum (Sow.) is in Edw. coll. from Colwell Bay; C. subventricosum (Kd. MS.) and C. deperditum? (Lam.) in Edw. coll. seem to be worn specimens of C.. ventricoswm; we agree with Prof. Judd in omitting them. C. marginatum (Kd. MS.), var. recentius (d’Orb.), occurs as in our list; its title to a distinct specific appellation seems doubtful. O. (Vicarya) concavum (Sow.) is in Edw. coll, labelled as from 122 H. KEEPING AND E, B. TAWNEY ON THE BEDS AT Colwell Bay. C. trizonatum (Morr.) is in Edw. coll. from Hordwell &c.; it would therefore seem to be Lower Headon as well as Upper. C. subconordeum (Kd. MS.) seems doubtfully a distinct form. C. estuarinum (Kd. MS.) as from Hordwell, in Edw. coll., is founded on a minute fragment, and we do not insert it. C. headonense (Hd, MS.) is in Edw. coll. from Headon Hill; but we do not insert it, as it may be Lower Headon only. C. pulchrum (Kd. MS.) is in Edw. coll. as from marine Headon beds of. Hordwell. C. pseudo- cinctum (d’Orb.) is in Edw. coll. as from Barton. Lurbonilla plicatella (Kd. MS.) is in Edw.:coll: from Roydon only ; T. plicatilis (Kid. MS.) is there as from Colwell Bay.and Barton. Melania brevicula (Kd. MS.) is in Edw. coll. as from Hordwell and Barton; M. Woodi (Kd. MS.) from Hordwell seems a Lower- Headon shell, and is not inserted; the names M. conica, M. polygyra, M. minima (Sow. ),we could not ‘find represented in the Edw. coll., and omit them. fiissoa carinata (Kd. M8.) occurs at Barton, but. not the var. cden- ticulata, according to the labels in Edw. coll. Conorbis dormitor (Sol.) and C. procerus (Beyr.) occur at Barton, but not the special varieties to which MS. names are anes in the Kidw. coll. Hydrobia polita (Kdw.) from Headon Hill is Upper Hesnal only, and therefore omitted; H. anceps (Wood) is in Edw. coll. as from Hordwell, Colwell Bay, Long Mead End, and Barton; H. Dubuis- soni, var. rimata (Ed. MS.) is labelled as from Hordwell marine bed. Trochus pictus (Ed. MS.) is in Edw. coll. from Hordwell ; it is in the Woodwardian museum from near Setley Common, Lymington. Melanopsis ancillaroides (Desh.) was in Edw. coll. subsequently labelled MZ. subfusiformis (Morr.) ; the former name may be omitted. M, subulata (Sow.) occurs twice in Prof. Judd’s list, and with diffe- rent ranges. MZ. swbcarinata (Morr.) is in Edw. coll. as from Bem- bridge and Hordwell marine bed; at the latter locality it probably came from Lower Headon (freshwater) to judge from the aspect of the shell. The species of Adeorbis in Edw. coll. are founded on single indi- viduals ; A. estuarina (Kd. MS.) we could not find. Orthostoma crenatum (Sow.) is in Edw. coll. from Brockenhurst and Barton. Tornatella altera (Desh.) is determined from a single individual in bad preservation, and seems to us very doubtful. 7’. hinnefor- mis (Sandb.) [ste] of the Brockenhurst list 1s the same shell as Acicon limneiformis (Sandb.) of the Headon-Hill list; it is in Edw. coll. as from those two localities. A. stimulatus is inserted on the faith of our own researches. A. dactylinus (Desh.) is in Edw. coll. from Colwell Bay ; we have found it there also. Rtingicula ringens (Lam.) occurs only in Barton or lower beds. Trochita obliqua (Sow.) was described originally as “a small but perfect specimen from. Brakenhurst [sc], in Sussex ; the species is found much larger in the: cliff at Barton ;” the type was only 1 ineh long, and was recognized by Sowerby as occurring at Barton, “while HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. I 5S at Brockenhurst it would have been found in Bracklesham beds probably ; we consider it a young example of 7. infundibuliformis (Lam.). We do not know why Edwards should have referred Brockenhurst forms to a different species from the common Barton species: we have compared numerous examples both from Brocken- hurst and Whitecliff Bay, and consider them identical with Lamarck’s species ; we therefore rank 7’. obliqua as a synonym. Nerita estuarina (Ed. MS.) seems identical with 1. zonula (8. Woo Be pha neritopsidea (Kd. MS.) is from the Upper oe we therefore omit it. Ostrea flabellula (Lam.) is in Edw. coll. labelled as from “ Col- well Bay or Headon Hill” [sic], but in error; the specimen has evidently come from the Barton Clay. O. ventilabrum is not in Edw. coll. from Colwell Bay, and it does not occur there; we are not sure that the distinctive differences given by Mr. Searles Wood between this species and O. flabellula are constant ; we have com- pared examples of this oyster, so common in the Brockenhurst bed at Whitecliff Bay, with other examples from Barton beds; and some we consider perfectly identical with Barton and Bracklesham forms. Avicula media (Sow.) is not in Edw. coll. as from Hordwell or Headon Hill, but from Long Mead End, 7. e. probably from the Beacon Bunny (Barton) beds. Dreissena Brardii (Fauj.) is in Edw. coll. from Hamstead and Hordwell (Long Mead End), ranging thus through all the fresh- water series. It occurs in social groups. We found a single derived specimen just above the Lower-Headon boundary at Whitecliff Bay. Anomia tenuistriata (Desh.) is in Edw. coll. labelled as from Barton, Brockenhurst, and Hordwell; if the Brockenhurst example is to be identified with A. Alcestiana (Nyst), probably the Barton ones are so also. We follow Edwards in considering them all one species. Mytilus strigillatus (Wood) in Edw. coll. is only from Barton beds; we therefore omit it. M. affinis is abundant at Colwell Bay, and ranges up from Barton beds. Nucula semilis (Sow.) is in Edw. coll. only from Barton beds ; we found one imperfect specimen, however, at White Cliff Bay in the Brockenhurst bed. NV. lissa (S. Wood) is said by Mr. Wood to oecur at Brockenhurst; but it is not in Edw. coll. as from there; the Hordwell examples are possibly from the Upper Bagshot Sands. Arca appendiculata (Sow.) and A. levigata (Caill.) are in Edw. coll. as from Barton and Bracklesham beds. . Cardita deltoidea (Sow.) is not in Edw. coll. from Colwell Bay ; and we doubt the fact of its occurring there. Incina Menardi (Desh.), as so determined, is not in Edw. coll. L. gibbosula (Lam.) and LZ. pratensis (Kd. MS.), in Edw. coll. as from Long Mead End, are not from Headon beds: we therefore omit them. JZ. obesa (Kd. MS.) and Z. inflata (Kd. MS.) are in Edw. coll, as from Barton beds; the former is not distinguishable 124 H, KEEPING AND E. B. TAWNEY ON THE BEDS AT from those labelled Z. concava; indeed Edwards admitted it as doubtfully distinct. Diplodonta obesa (Kd. MS.) exists in Edw. coll. from Barton and Bracklesham beds. D. dilatata (Sow.), determined as such, is not in Edw. coll. ; Dixon cites it from Bracklesham. Strigilla colvellensis (Kd. MS.) we found to be not unfrequent at Headon Hill. S. pulchella (Ag.) is in Kdw. coll. determined from a single imperfect valve: it is impossible to say whether it is a second species ; and we therefore omit it. Cardium Edwards (Desh.) is not in Edw. coll. from Brocken- hurst; it is'a Bracklesham shell. Cytherea suborbicularis (Kd. MS.) we found at Colwell Bay and also in the Brockenhurst zone at Whitecliff Bay. C. swessonensis (Desh.): under this name we find in Edw. coll. Barton and Lower- Kocene shells usually referred to C. tenwstriata (Sow.); we have it from the Brockenhurst zone in Whitecliff Bay. C. partemsulcata (Ed. MS.) is from Long Mead End only; we therefore omit it. Cyprina scutellaria (Desh.) is solely from Lower-Hocene locali- ties. OC. Nystt (Héb.) is the only species from Brockenhurst in Edw. coll.; we have it from Whitecliff Bay. Psammobia compressa (Sow.) is in Edw. coll. from Barton; the var. arcuata (MS.) is from Roydon and Brockenhurst; and var. estuarina (MS8.) is, according to the labels, from Hordwell, Colwell Bay, and Roydon. We have compared these examples, and cannot see any valid differences; we consider them all one species. Sanguinolaria Hollowaysuv (Sow.) is not in Kdw. coll. from Lynd- hurst, but is a Bracklesham shell. The Geological Survey cite it in error from Middle Headon beds of Whitecliff Bay. Tellina corbuloides (Kid. MS.) in Edw. coll., from Colwell Bay, is a crushed specimen, undeterminable, but probably not the Hamstead species; we omit it. 7’. ambigua (Sow.) is in Edw. coll. labelled as from Hordwell, but is probably not from Middle Headon beds, although Forbes cites it from Colwell Bay. 7. reflewa (Kdw.) is a Bracklesham shell, and does not occur above the Upper Bagshot of Long Mead End. 7. headonensis (Edw.)is in Edw. coll. from Hea- don Hill and Colwell Bay. TZ. sphenoides (Kdw.) is from Colwell Bay. 7’. affinis (Kd. MS.) is in Edw. coll. from Brockenhurst and Barton ; we found it also at Whitecliff Bay. Syndosmya colvellensis (Kd. MS.) seems founded on a single minute valve. Scintilla angusta (Kd. MS.) is from Colwell Bay, Hordwell, and Barton beds. Mactra fastigiata (Kd. MS.) which occurs abundantly at Headon Hill, is from Hordwell in Edw. coll. M. jilosa (Kd. MS.) from Colwell Bay and Long Mead End (Upper Bagshot) we are unable to consider a separate form, so rank it as a synonym. Solen gracilis (Sow.) is in Edw. coll. from Brockenhurst and Bar- ton beds. _ Mya angustata (Sow.) was originally found by Prof. Sedgwick at Colwell Bay; Edwards has an apparently identical shell which he HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT. 125 at first identified with Sowerby’s species, but subseqnently altered to M. bartonensis (Ed. MS.) ; we cannot, however, see any distinc- tion between these and M. producta (Kd. MS.) from Colwell Bay ; we consider them all one species. Mr. Searles Wood doubts the Hamstead form M. minor, Forbes, being a distinct species either. Corbula pisum (Sow.) is not in Kdw. coll. from Headon Hill or Hordwell; and we doubt its occurrence there; it is, however, fairly abundant in the Brockenhurst zone of Whitecliff Bay. C. mtida (Sow.) is in Edw. coll. from Roydon and Long Mead End; it was originally described from Prof. Sedgwick’s specimens brought from Middle Headon beds. C. fortisulcata (Kd. MS.) we consider merely a variety of C. pisum; we omit it, since it comes from Long Mead End, probably from Barton beds. C. jicus was found at Brocken- hurst by one of the authors, and is now in the Woodwardian Mu- seum. Panopea corrugata (Sow.) is, according to the Edw. coll., only a Barton and Bracklesham shell ; but quite possibly P. subeffusa (Hd. MS.) is not really separable from this species; in either case one of the names must be omitted. Cyrena subreqularis (Kd. MS.) seems to pass into C. obovata (Sow.). C’, deperdita (Lam.) is in Edw. coll. from Headon Hill and Barton ; it is probably from the Lower Headon. C. arenaria (Forbes) in Edw. coll. is scarcely separable from the preceding; it is from Headon Hill and Hordwell, but apparently from Lower Headon. Mr. Searles Wood figures a different form as Forbes’s species. C. gibbosula (Morr.) is scarcely a distinct form. C. estuarina, C. alti- rupestris, C. obliquata, MS. names in Kdw. coll., are not inserted ; ' they may probably be Lower-Headon shells. Balanus wnguiforms (Sow.) we found as frequent at Headon Hill as at Colwell Bay. Pollicipes reflecus (Sow.) is cited by Forbes from Colwell Bay. Callianassa Baylii (H. Woodw.) we have from all the zones of the Middle Headon. EXPLANATION OF PLATE V. The figure represents the coast-section from Headon Hill east of Heather- wood Point to Cliff End, Colwell Bay ; it passes inland at Warden Point to avoid the projecting promontory. An attempt is made to represent where beds may be seen iz situ, and where they are concealed by fallen material. Notwithstanding that the vertical scale is more than double the horizontal, it is impossible to show much detail; and for this refe- rence is to be made to the accompanying vertical sections (pp. 91, 98, and 103). The bends of the coast-line are approximately indicated by the com- pass-bearings given, DIScuUssIon. The Prestpent remarked that the paper was one of great impor- tance. The question at issue was one sharply defined but difficult to come to a conclusion upon without visiting the sections. 126 H. KEEPING AND E. B. TAWNEY ON THE BEDS AT Rey. O. Fisuur said he had visited the locality with Mr. Tawney’s paper and sections in manuscript, and that he agreed with the authors’ conclusions. He thought the error on the part of Prof. Judd might have proceeded from the fact that at the N.E. corner of Headon Hill the Middle-Headon fossils were found by the sea-shore; these, however, were not truly im situ, but had been brought down by aslip; this he thought possibly the key to the erro- neous interpretation. In confirmation of the view that the Colwell- Bay and Headon Venus-beds are one stratum, he had found it with its fossils in Totland brick-field, near the Hotel, exactly where it should occur to connect the disjoined portions. He had worked personally at the Brockenhurst locality at Whitley Ridge, and had identified the bed at the base of the Middle Headon in Whitecliff Bay. Consequently, if the Brockenhurst bed is to be called Oligo- cene, the Middle Headon can no longer be called Eocene. Prof. Jupp said that the paper rested largely on assumptions. His method of work in the field and in the museum had been made a matter of assumption. He had not hastily arrived at his conclusions, but for twenty years he had worked on these British beds, and for twelve years had studied their continental equivalents and collec- tions of fossils made from them. ‘The series of Kocene and Oligo- cene strata in Western Europe is perfectly clear; but when we come to Britain a difficulty has always existed. This confusion was removed by distinguishing the zone of C. concavum from the Broc- kenhurst series. ‘The authors’ sections were supposed to support those of the Survey; he thought on examination they would not do so. The mistake had really originated from using Cytherea encras- sata to fix the so-called “ Venus-bed” of fossil-collectors—the fact ~ being that that shell has a wide range, and there is more than one ‘‘Venus-bed.” In asserting that the different ‘‘ Venus-beds” are upon the same horizon the authors begged the whole question. This autumn he again visited the island, and found that an excavation had been opened by the authors in a Venus-bed in Totland Bay, but in one quite different from that in Colwell Bay. Had the authors searched the Headon cliff they might have found other Venus-beds. The authors had confirmed his own conclusion that the Headon-Hill sands do not occur in Totland Bay; and this is fatal to their reading of the section. As regards the paleontological evidence, he thought that the method of comparison of most abundant fossils was often misleading, as might be shown in the case of the Cornbrash and Ragstone of the Lower Oolite. The authors say that the Brocken- hurst bed is not above but below the Venus-bed. Now the former is the equivalent of the Tongrian beds of Belgium; and foreign geolo- gists all regard the zone of C. concavwm as the top of the Bar- tonian—that is, of the Kocene. Hence the result of their interpre- tation of the section was to place beds with an Upper-Hocene fauna above those containing a Lower-Oligocene fauna. Mr. Starxrz Garpner said he had always thought that in the particular section under discussion there was only one Venus-bed : the section of Headon Hill till last year had been fairly clear; and HEADON HILL AND COLWELL BAY IN THE ISLE OF WIGHT.” 127 he had never seen more himself. He thought ‘“‘ Upper Eocene” and “Oligocene” equivalent terms, and the question, which should remain in use? one of priority. Mr. Wuitaxsr said that cliff-sections in soft beds were apt to vary from time to time, so that observers who saw them under different conditions of exposure were likely to differ in inter- preting them. The examination of other parts of the island, and especially the mapping of limestones or other well-marked beds, might. partly help to settle the question in dispute. Perhaps the Geological Survey map had been constructed rather too much on theoretical grounds. The PresrpEnt said that on the one hand we had the minute measurements of Messrs. Keeping and Tawney, and, on the other, the wider views of Prof. Judd. At any rate these views were now on both sides well laid before the Society ; and the question, although a difficult one, as he had himself found in working over the ground 25 years since with Dr. Wright, would be now carefully examined by many others. Mr. Tawney said that he thought Von Konen, in 1864, had rightly correlated the German and English beds. As for Mr. Whitaker’s remarks, he thought a person who was puzzled by a clitf-section would make but little of a drift-covered country where no sections were to be seen. He still denied what Prof. Judd had said about there being more than one marine series. The Certthiwm concavum zone of Prof. Hébert at Hordwell did not occupy the position attri- buted to this zone by Prof. Judd in Headon Hill. He maintained that there was but one Venus-bed. The 6-inch Ostrea vectensis bed in the Bembridge beds could not be confused with the Middle-Headon Venus-bed. (QS0a0" P. H, CARPENTER ON TWO NEW CRINOIDS FROM 10. On two new Crrnorps from the Upper Cuarx of SovTHERN Swepen. By P. Herpert Carpenter, Esq., M.A., Assistant Master at Eton College. Communicated by Prof. P. Martin Duncan, M.B. Lond., F.R.S., F.G.S. (Read February 2, 1881.) [Puare VIL] Tux large work* of Prof. Geinitz on the fossils from the valley of the Elbe in Saxony contains a description of a small stalked Crinoid from. the “ Pliinerkalk’’ of Strehlen that has long been known to contain stem-joints of the type to which d’Orbigny gave the name Bourgueticrinust. Although no calyx was ever met with, the characteristic stem-joints were supposed to be sufficiently indi- cative of the presence of the common JB. ellipticus. Some ten years ago, however, a singularly perfect specimen was discovered, con- sisting of a complete stem with radicular cirrhi, and a calyx on the top of it (Pl. VI. fig. 1). But this calyx (fig. 2) proved to be totally different in its characters from those of the species of Bourgueticrinus described by @’Orbigny. In B. ellipticus the calyx is widest round the basal circlet, and tapers gradually downwards into the stem, while the outer surface of the radials has a considerable slant from above downwards and outwards. In @Orbigny’s other species, B. equalis, both radials and basals, especially the latter, are relatively narrower and higher, and the whole calyx, together with the top stem-joint, is almost uniformly cylindrical ; but in each case the top stem-joint is very large and relatively higher than that of any Apiocrinus, while in B. ellupticus it widens considerably from below upwards. Its height may be as much as or more than that of the basal and radial circlets together ; and the joints immediately below it gradually diminish in width until they resemble the ordinary stem-joints. In the Strehlen fossilt, however, the calyx is widest at its upper end, around the upper and outer edges of the radials, so that its diameter diminishes gradually from above downwards (Pl. VI. fig. 2a). The broad external faces of the radials slope downwards and inwards; and the basal circlet narrows still more, so that the diameter of its lower face is less than two thirds that of the upper surface of the radials (fig. 2b). But the top stem-joint, on which the calyx rests, only expands a little from its lower to its upper margin, and its increase in thickness over the one below it is far less marked * “Das Hlbthalgebirge in Sachsen,” Palzontographica, Band xx. Theil 2, Ds 1S Sh oS Histoire naturelle générale et particuliére des Crinoides vivans et fossiles (Paris, 1840), pp. 95, 96. ¢ Thanks to the kindness of Prof. Geinitz, who has made a second exami- nation of this specimen, I am enabled to give a slightly more accurate figure of it than that published by him in the ‘ Elbthalgebirge’ (Taf. vi. fig. 9a). In the older figure only two joints are represented between the calyx and the enlargement on the stem. In that given here three joints are shown instead of two, this later interpretation of the markings on the upper part of the stem being considered by Prof. Geinitz to be the more accurate one. O AN Soe ay JOUrT. or + U Mintern. Bros imp ith. Berean del et Cc ;OUS RE TAC = Sat @ deal if SWE.D THE UPPER CHALK OF SOUTHERN SWEDEN. 129 than is the case in Bourgueticrinus. Both it and the joints imme- diately below it are decidedly smaller than those forming the lower part of the stem, which is just the reverse of what we find in Bour- gueticrinus. Prof. Geinitz was unable therefore to refer this specimen to Bourgueticrinus, despite the resemblance of its stem-joints to those of that type; but he supposed it to belong to the genus Antedon on account of the resemblance of its calyx to that of Ant. Sarsw, as represented by M. Sars in his well-known ‘ Mémoires pour servir a la connaissance des Crinoides vivants.’ In the specimens figured by Sars*, ‘‘the upper end of the stem is not thickened, and the calyx widens from its base to the upper end of the first radial,” just as in the fossil from Strehlen (Geinitz Tt). This resemblance is but an imperfect one, however; for Sars’s specimens were merely the stalked larvee of Ant. Sarsei, not more than an inch long. In the later stages of these larvet the uppermost stem-joint or future centro-dorsal piece of the mature and free Antedon not only becomes considerably enlarged, so as entirely to conceal the basals, but it also bears cirrhi. This, however, is not the case with the corre- sponding joint of the Strehlen fossil. Prof. Lundgren § has already pointed out that this specimen is too completely developed to be a larval Antedon like those figured by Sars; but neither he nor any other paleontologist has referred it to any definite position among the Crinoids. The ‘‘Mucronatenkreide” (= Upper Chalk) of Kopinge, near Ystad, in Southern Sweden, contains a quantity of stem-~joints (Pl. VI. figs. 3-6) which have been considered as belonging to Bourguetierinus ellipticus, though no calyx of this species has yet been found associated with them. Some years ago, however, a singular calyx was met with (Pl. VI. fig. 7), which was presented to the Geological Museum of the University of Lund by Herr Rector Bruzelius, of Ystad, into whose hands it had come. Its discovery was announced in the ‘Neues Jahrbuch ftir Mineralogie’ by Prof. Lundgren, who at once recognized its resemblance to the so-called Antedon Fischert of Geinitz (Pl. VI. fig. 2). It is, however, con- siderably larger and much less conical; but it has even less resem- blance than the Strehlen fossil has to the cylindrical Bowrgueticrinus equalis. During my recent visit to Lund for the purpose of examining the Comatulw of the Retzian collection, Prof. Lundgren showed me this fossil, and was kind enough to intrust it to me for description. For this and for many other acts of kindness I offer him my heartiest thanks. For these two fossils, so similar in their general characters though differing in points of detail, I believe that not only a new genus, but also a new family must be established. I propose to call the genus Mesocrinus; for while allied to Bourgueticrinus in the cha- racters of its stem-joints, it is quite as closely allied to the Penta- crinidg in the characters of its calyx. The only real resemblance * Op. cit. pls. v., Vi. T Op. cit. p. 18. t Op. cit. pl. vi. fig. 24. § Neues Jahrbuch fiir Mineralogie, i876, pp. 180-182. 130 P. H. CARPENTER ON TWO NEW CRINOIDS FROM to the Apiocrinide is in the nature of the faces of the stem-joints, which resemble those of Bourgueticrinus and Rhizocrinus. But similar joints occur in the larval Antedon and in Platycrinus * ; so that this resemblance does not go for much, especially when it is remembered that d’Orbigny speaks of the articular faces of the stem-joints in the Apiocrinide as being most frequently marked with radiating striee 7. On the other hand, Mesocrznus differs from the Apiocrinide as defined by d’Orbigny and by Zittelt, in the presence of verticils of cirrhi on the stem, and in the small size of its upper joints. Further, the central funnel of the calyx (Pl. VI. fig. 7a) is small, not wide and patent; while the articular faces of the radials are large, and not separated or barely in contact, but meet one another along the whole length of their sides, and bear distinct fossee for the attachment of muscles and ligaments. All these characters are more or less distinctive of the type of the Pentacrinide ; but Meso- crinus cannot be referred to that family, as it lacks the petaloid markings on the faces of the stem-joints. MzsocrInvs, 0. &. Calyx more or less conical, composed of five basals forming a complete ring, and five radials with high articular faces which bear distinct muscular and ligament-fosse, and are in contact for the whole length of their sides. Upper stem-joints the smallest. Lower ones with elliptical faces, the long axes of which are occupied by articular ridges. ‘The planes of these ridges on the two faces of each joint are more or less inclined to one another. The joints may bear single cirrhi, or two may combine to form a node for a verticil of two cirrhi. MRadicular cirrhi at the base of the stem. MusocRINvs suEDIcUS, n. sp. (Plate VI. fig. 7.) The lower part of the calyx is formed by five pentagonal basals, which are in complete contact laterally, so as entirely to cut off the radials from the top stem-joint. The lower surface of the basal pentagon is very concave, with a large central perforation ; but it is quite simple and devoid of any kind of ornamentation (Pl. VI. fig. 7b). The radials have very high outer surfaces (nearly twice as high as the basals), which slope very slightly upwards and outwards. Their articular faces are also rather high and trapezoidal, with large muscle-plates standing up around the centralfunnel. The thickened edges of the two muscle-plates on each radial are separated by a slight intermuscular notch. The transverse ridge expands into a large articular surface perforated by the opening of the central canal, to which it forms a thick rim on all sides. From this surface short ridges proceed upwards and outwards, separating the large muscle- fosse from the small ligament-fosse. The dorsal fossa for the elastic ligament is relatively small, being represented by little more * A Natural History of the Crinoidea (Bristol, 1821), pp. 34, 75. t Op. cit. p. 1. { Handbuch der Palaontologie, Band i. p. 388. THE UPPER CHALK OF SOUTHERN SWEDEN. USE than the pit beneath the transverse articular ridge, around which is a simple smooth surface. Size. Height 5 millim., greatest diameter 5 mm., least diameter 4 mm., greatest height of basals 1:5 mm., least height 1 mm., width 3 mm. As already pointed out by Prot. Lundgren, the stem-joints which occur associated with this calyx are referable to three principal types. It is, of course, possible that they belong to anothe™ species altogether ; but, like Prof. Lundgren, I am inclined to refer both calyx and stem-joints to one and the same species. ‘The question cannot, of course, be decided until the Swedish collectors are fortunate enough to meet with a perfect specimen. Type 1. Thin circular disks, about 3 millim. in diameter, with faces perforated in the centre, but without markings of any kind. They probably belong to the upper part of the stem immediately beneath the calyx. Both in some forms of Bourgueticrinus* and in Rhizocrinus the upper stem-joints have simple faces without any of the characteristic sculpture which occurs lower down the stem. The thin penultimate joint of Mesocrinus Fischert (Pl. VI. fig. 1) would seem to have been of this nature; and one can readily imagine that in the larger WM. suedicus the number of such simple stem-joints was larger, as is actually the case in Bourgueticrinus. Type 2. The joints of this, the commonest type (PI. VI. fig. 3.), are higher, with oval articular faces, the long axes of which are occupied by transverse ridges. ‘The planes of the ridges at the two ends of each joint are inclined to one another at angles of from 60° to nearly 90°; and the centre of each ridge expands considerably around the opening of the central canal into a well-marked articular surface. A median groove extends along each half of the ridge from the central opening towards the margin of the joint-face; and short shallow branches proceed from it on each side so as to cut out the upper portion of the ridge into a double row of small teeth. According to Prof. Lundgren these joints vary in size from 3 to 8 millim. in diameter, usually 5 or 6 millim.; and their height is about equal to their diameter. Type 3 (Pl. VI. figs. 4-6). Wider but lower joints, the oval faces of which are much more pointed than in those oftype 2. They differ very much in the proportion of height to diameter. In the thicker ones (fig. 4) the expansion of the transverse ridge around the opening of the central canal is very distinct, and there is a crescentic pit on either side of it. But these features are much less marked in the thinner joints, the faces of which are flatter (fig. 5), while the transverse ridges scarcely expand at all around the central canal. Some of these joints bear portions of the cirrhus-sockets, as already pointed out by Prof. Lundgren. As in some species of Pentacrinus, * This is certainly the case in several specimens that I have examined, some of which, contained in the University collection at Berlin, were kindly shown to me by Prof. Beyrich. On the other hand, Quenstedt figures a top stem- joint of B. ellipticus with a distinct transverse ridge and articular facet on its under surface (Petrefactenkunde Deutschlands, Band iv. tab. 104, fig. 76). 132 P, H. CARPENTER ON TWO NEW CRINOIDS FROM two joints contribute to the formation of a socket, and the sockets are at the ends of the long axes of the two apposed faces, so that there were two cirrhi at each node, and at least three nodes in im- mediate succession (fig. 6). The two grooves in the respective transverse ridges form by their apposition a canal which lodged the vessels proceeding to the two cirrhi from the central vascular axis of the stem, and opened at the bottom of each cirrhus-socket*. In the absence of complete specimens of Mesocrinus suedicus it is, of course, impossible to determine where these nodes or, rather, groups of nodes occured in the stem, or even whether a stem of this character was associated with the calyx under description. I suspect that the verticils of cirrhi were limited to the wide lower part of the stem ; but it is not possible to form any opinion as to whether they were simple cirrhi like those of Pentacrinus, or irregularly branched radicular structures like those on the lower stem-joints of Rhizocrinus lofotensis. Failing direct evidence to the contrary, it certainly seems to me most probable that the calyx and stem-joints from Kopinge all belong to one species, which would then differ from dMesocrinus Mischert in other characters than those of the calyx; for the lower joints of M. Fischert (P1\. VI. fig. 1) are all much longer than wide, as in Rhazocrinus, Bathycrinus, and the larval Antedon, and they rarely bore cirrhi(Pl. VI.fig. 1). In this latter respect Z. swedecus must have been related to MW. Fischert much in the same way as Lhizocrinus lofotensis with abundant radicular cirrhi on the lower part of its stem is related to &. Rawsoni f which had “ very few radicular cirrhi.” The wide lower stem-joints of MW. swedicus (Pl. VI. figs. 4-6) have much less resemblance to the corresponding joints of 1. Fischert than to those of Bourgueticrinus ellipticus, as represented by @Orbigny ; and they further resemble these last in having the transverse ridge continuous across the articular face. In B. con- strictus, however, the ridge is interrupted in the centre, and the more or less marked excavations in the lateral portions of the joint- face are connected with one another round the central canal, very much in the same manner as the two cornua of the grey matter unite around the central canal of the spinal cord +. ‘This feature is also distinctive of Rhizocrinus and of the Antedon-larva$; and in both of these types the transverse ridge is cut up into a double row of minute teeth. So far as I know, this character has never been described in any species of Bowrgueticrinus; and Geinitz neither de- scribes nor figures it in Mesocrinus Fischert, though (as mentioned above) it occurs in J. swedicus. * Quenstedt (op. ciz. p. 368) has described cirrhus-sockets on stem-joints from the White Chalk of Rugen, and has remarked that the grooves in the trans- verse ridges of these nodal joints are often very distinct; but there seem to have been more than two cirrhi at each node (tab. 104, fig. 63). t ‘Zoological results of the Hassler Expedition. I. Hchini, Crinoids, and Corals,’ Ilustr. Catalog. Mus. Comp. Zool. at Harvard College, no. viii. p. 27. { This seems to be also the case in the so-called B. ellipticus from the Hocene of Traunstein (Quenstedt, op. cit. iv. tab. 104. fig. 82). § Sars, ‘ Crinoides vivants,’ pp. 5, 6, pl. ii. fig. 27, and pl. vi. fig. 17. THE UPPER CHALK OF SOUTHERN SWEDEN. 133 One of the stem-joints of WM. Fischeri which is figured by Geinitz is very singular. There isa thick articular rim around the opening of the central canal of the oval-oblong face, as in the lower stem- joints of MW. swedicus; but the ridge proceeding from this to either end of the oval is Y-shaped and not simple. Some of the stem-joints of B. ellipticus and also of Platycrinus levis which were figured by Miiler * present very similar characters; but I have not found them in any of the joints of J. suedicus. The genus Bourgueticrinus is rather in confusion just at present, all those stem-joints occurring in the Mesozoic rocks being referred to it that have elliptical articular faces with transverse ridges upon them, which are in different planes at the two ends of each joint; and we are unable to classify these joints properly in the absence of sufficiently perfect specimens of whole individuals. In any case, however, the so-called Antedon Fischeri and its Swedish ally have stem-joints closely approximating to the Bour- gueticrinus type (Pl. VI. figs. 3-6). Nevertheless the calyx is quite different (Pl. VI. figs. 2 & 7). Seen from the side it has a certain resemblance to some forms of Millericrinus Miinsterianus, d’Orb., and of WM. Nodotianus, dOrb. But this resemblance dis- appears altogether when the upper surface of the calyx is examined ; for the central funnel of Millericrinus is very large, and the articular faces of the radials which surround it are wide and low, barely meeting laterally; while Mesocrinus has quite a narrow central funnel (fig. 7, a) and relatively high articular faces, which diminish considerably in width towards their upper ends, but are in contact for the whole length of their sides. | On the whole the calyx most resembles that of a Pentacrinus, or rather of that section of the genus with a closed basal circlet, which is referred by de Loriol ¢ to Cawnocrinus. Had I either calyx alone before me, I should certainly refer it to Pentacrinus, among the recent species of which there is a considerable variation in the proportions of the different parts of the calyx t. In Carnocrinus, Millericrinus, Bourgueticrinus §, and Rhizocrinus, the composition of the calyx is the same. In all these genera there is a closed circlet of five basals, on which the five radials rest, just as they do in Mesocrinus. But there are considerable differences among these five types in the nature of the articular faces of the radials. In Afllericrinus these faces are very wide and low, and are nearly or quite separated laterally, while the fossee for the muscles and interarticular ligaments are usually but shghtly developed ||. In Bourgueticrinus the articular * Op. cat. pp. 34, 75. t ‘Monographie des Crinoides fossiles de Ja Suisse,’ Geneva, 1877-1879, p- 111. See also ‘On the Genus Solanocrinus, Goldfuss, and its Relation to Recent Comatulz,” Journ. Linn. Soc., Zool. vol. xv. { Compare figs. 21 and 23 on pl. xi. of Journ. Linn. Soc. Zool. vol. xv., with fig. 3 on pl. vi. of the Popular Science Review, new ser. vol. iv. § In some specimens of “ 6. ellipticus” in the national collection at South Kensington the basals appear externally as small triangular pieces which are not in contact laterally. || According to de Loriol, Millericrinus differs from Apiocrinus in the presence of an articular facet on the first radials. In the latter genus “leur surface 134 P. H. CARPENTER ON TWO NEW CRINOIDS FROM faces are much reduced and quite insignificant, while in Rhizocrinus they are wide, low, and separate, and more like those of Milleri- crinus, enclosing a wide central funnel. In both species of Mesocrinus, however, and especially in 7. suedicus, the articular faces are much higher relatively to their width, are in contact for the whole length of their sides (figs. 2 a, 7 a), and have distinct muscle-plates, which stand up around the opening of the small central funnel. These are largest and best developed in MW. suedicus, but are distinctly recognizable in Geinitz’s figure of I. Fischer. These features are eminently characteristic of the Pentacrinide and Comatulide, though I have seen no calyx of Pentacrinus which could be said to be precisely like that of Mesocrinus. But then no two Pentacrinus-calices that I have seen are precisely like one another; for they differ very much among themselves, not only in the relative development of the basals, but also in the shape and proportions of the outer surface of the radials and of their articular faces. At the same time it must be remembered that we are only just beginning to become acquainted with these modifications ; for neither species nor individuals of living. forms are at all abundant in collections, much less are they available ;for anatomical investi- gation. The calyx of Mesocrinus Fischer. (fig. 2) finds its nearest ally in that of P. Wyville-Thomsoni*. ‘It is in a side view that the resemblance is most evident, the chief point of difference being the greater height of the radials and of their articular faces in Mesocrinus. These features are more marked, however, in P. asteriay ; but in this species the basals are small, and do not meet laterally as they do in Mesocrinus. Except in this point the calyx supérieure ne servait pas de point d’attache” (oc. cit. p. 32). I regret that I cannot altogether agree with this opinion of the distinguished Swiss paleeon- tologist, who must have been unfortunate in some of the specimens he examined. The large concave surface in which the second radials of Apiocrinus rest is some- times an expansion of what in recent Crinoid#is the dorsal fossa lodging the elastic ligament. ‘This is well shown in de Loriol’s own figures of the ventral aspect ot the first radials of A. Meriani (pl. ii. figs. 4a, 5a), and also in d’Orbigny’s figure of A. Murchisonianus (pl. vi. fig. 7). The central end of this large fossa is bounded by the transverse articular ridge pierced by the opening of the central canal ; and rising up from this ridge so as to form a part of the rim of the central funnel are larger or smaller plates for the muscle- and ligament-fosse. In some specimens of Aptocrinus in the national collection these fossz are re- latively large and are separated by a well-marked ridge; but in most species they are greatly reduced in size, as is also frequently the case in Millericrinus. In the latter genus the dorsal fossa never reaches the enormous size that it does in Aptocrinus, and, though it is sometimes relatively large (as in some Comatule), it is occasionally comparatively small. I am not prepared to say, however, that no Apiocrinus had the first and second radials united otherwise than by muscles and ligaments. They must, in some cases, have been joined by a syzygy, e.g. Ap. Parkinsonti (VOrbigny, pl. v. fig. 6) ; but such cases are very anomalous; for in the other Articulate Crinoids syzygial union, though common enough between the second and third radials, never takes place between the first and second. Hyen in Marsupites there is a distinct articular facet on the first radials. * Journ. Linn. Soc., Zool. vol. xv. pl. xi. fig. 23. t Journ. Linn. Soc., Zool. vol. xy. pl. xi. fig. 21. THE UPPER CHALK OF SOUTHERN SWEDEN. aa of P. asteria is not unlike that of M. suedicus (Pl. VI. fig. '7), which has much more resemblance to it than to the calyx of any species of Millericrinus. Despite the resemblance of their calices to the Pentacrinus-type, the so-called Antedon Fischeri and its Swedish ally obviously belong to another genus than Pentacrinus. They lack the typical stem of this genus, with its verticils of cirrhi from the top downwards and the petaloid markings on the joint-faces. Neither does Mesocrinus possess the typical stem of Apiocrinus or Millericrinus ; but while chiefly resembling Bowrgueticrinus in the character of its stem, it differs from that genus and approaches /thizocrinus in the compara- tively small size of its upper stem-joints; and it especially resembles ft. Rawsoni in the relatively slight increase in the diameter of its calyx from below upwards. In neither case does the calyx pass gradually downwards into the thickened upper end of the stem, as it does in the typical Apiocrinide and in Bourqueticrinus, to which, in other respects, Wesocrinus has a considerable resemblance. The similarity of the articular faces of the stem-joints of the young Antedon Sarsii to those of Rhizocrinus, and the imperfect radicular processes that proceed from the lower part of its stem, are characters which connect the Comatulide with the Apiocrinide. The Pentacrinide, however, are sharply marked off from the latter family by the striking differences in the character of the stem. It is therefore of no small morphological interest to find a type in which the characters of a Pentacrinus-calyx are combined with those of a Bourgueticrinus-stem. It is quite possible that some of the stem-joints now referred to Bourgueticrinus may belong to other intermediate forms, the calices of which are still unknown to us, as that of Mesocrinus was but a few years ago. Prof. Lundgren has been kind enough also to intrust to me the description of a new Antedon, two specimens of which were found by himself and by Mr. J. Chr. Moberg in the Ignaberga Limestone at Balsberg, in the province of Scania, 8. W. Sweden. ANTEDON IMPREsSA, n. sp. (Plate VI. figs. 8, 9.) Centrodorsal a thin convex pentagonal disk with very indistinct traces of a small five-rayed impression at the dorsal pole. In the larger specimen the dorsal surface is somewhat flattened and tolerably free from cirrhi (fig. 8), but in the smaller and younger specimen there is but a very small cirrhus-free space (fig. 9,a). There are about 40-50 sockets arranged in three rather indistinct rows, with occasional traces of a fourth; but they are too much worn to show any structural details. The ventral surface is entirely obscured by matrix in the larger specimen, and is only partly visible in the smaller one. It is markedly concave; and its angles are somewhat produced upwards and outwards. In the middle line of each radial area is a median erooye, the central end of which is deepened and forms a radial pit. The outline of the central opening was probably decagonal; and L 2 136 CRINOIDS FROM THE UPPER CHALK OF SOUTHERN SWEDEN. from its interradial sides there proceed five linear oblong basals, which do not quite reach the circumference. The outer end of each is marked by an oval oblong impression. Diameter, larger specimen 8 millim., smaller 5 mm. Height, larger specimen, 2 millim., smaller 1°5 mm. Remarks. I do not know of any Comatula, either recent or fossil, with a centrodorsal at all like that of this species, except Antedon Tourtie. Schliiter * figures five slight grooves radiating outwards from the radial pits of this last species ; but he makes no mention of them in the text. The ventral surface of its centrodorsal is slightly concave, with the angles raised; but this is far less distinctly the case than in A. impressa ; and the centrodorsal is half as high as it is wide, and bears four vertical rows of cirrhus-sockets, characters which distinguish it sharply from A. impressa. . The persistence of the basals in connexion with the centrodorsal rather than with the radials is also interesting. They are probably the original embryonic basals (or orthobasals), as in A. Lundgreni and a few other speciest. If they were merely basal rays connected with a rosette, some trace would have remained of such a connexion ; but I can find none. | EXPLANATION OF PLATE VI. Fig. 1. Mesocrinus Fischeri, Geinitz, sp., natural size. Copied from an amended drawing of the original, kindly lent by Prof. Geinitz. 2. Calyx of Mesocrinus Fischeri, enlarged: a, side view; 0, dorsal surface. 3-6. Stem-joints from Kopinge, 8. Sweden, probably belonging to Mesocrinus suedicus, all X 2: a, face views; 0, side views. 7. Calyx of Mesocrinus suedicus, nov. gen. et sp., X 6: a, veutral surface; 0, dorsal surface; c, from the side. 8 & 9. Centrodorsals of Antedon impressa, n. sp.: fig. 8. Dorsal sur- face of the larger specimen, X 4; fig. 9. Smaller specimen, x6: a, dorsal surface; 6, from the side; c, ventral surface. Discussion. The Prestpent expressed his sense of the value of the author’s communication, and hoped he would continue his work. Prof. Duncan expressed the same view. Prof. Srrrey said the variations of Bowrgueticrinus were very remarkable, both in form and in structure of the calyx. He had never seen one with the structure of that described, but had seen some approaching it. He eulogized the description given by the author, and expected from him important contributions to the knowledge of the Cretaceous Crinoids. * “Ueber einige astylide Crinoiden,” Zeitschr. d. deutsch. geol. Gesellsch. 1878, p. 41, Taf. i. figs. 4-6. t Quart. Journ. Geol. Soc. 1880, vol. xxxvi. p. 550; Journ. Linn. Soe., Zool. vol. xv. p. 213. ON A BOULDER OF HORNBLENDE PICRITE. 137 11. On a BovtvER of HorneienveE PrcritE near Prn-y-Carntsioe, Anetzsry. By Prof. T. G. Bonney, M.A., F.R.S., Sec. G. S. (Read January 5, 1881.) Last summer, through the kindness of Professor Rosenbusch, I was able to examine several specimens of picrite in: the Geological Museum at Heidelberg, and to study the rock in the field, near the village of Schriesheim, a few miles to the north of that town. In September I was walking with some students along the road which leads out of the village of Pen-y-Carnisiog northwards to Bwlyn (Anglesey), when I observed, in a field on the left, the fractured face of a boulder, in which a number of large crystals resembling augite, glittering in the sunlight, in a dull dark matrix, recalled at once the characteristic aspect of the Schriesheim picrite. The boulder had been broken, apparently rather recently, into three pieces, one much smaller than the other two ; and its volume must have been not much less than a cubic yard. In its weathered sur- face and toughness under the hammer it also resembled the Schries- heim rock. In both, the larger crystals (which are often about two thirds of an inch long) contain a number of dark serpentinous- looking enclosures, giving to the cleavage-faces an interrupted lustre somewhat resembling (except in the absence of a metallic gleam) that of bastite. The Pen-y-Carnisiog rock looks a little more de- composed ; but macroscopically the resemblance between my two specimens is so great that one could believe them to have been broken from different parts of the same mass. Part of a Slice from a Boulder of Hornblende Prerite near Pen-y- Carnisiog, Anglesey. (Magnified 30 diameters.) sa ATTAY 4 ip \pbind “ , ym, Tin BY UALR \ NaN INS fF SON PAu) — +S D2 WP IS 1 Mast SON ; Td < \ a. One of the grains of altered olivine. 6, 6, b. Aggregated small crystals of hornblende, probably of secondary origin. 138 PROF, T. G. BONNEY ON A BOULDER OF HORNBLENDE When the Pen-y-Carnisiog rock is examined microscopically, the difference between the two is rather more marked. In it the pre- dominant mineral is undoubtedly hornblende. This occurs under three forms :—(a) inumerable small acicular or blade-like crystals, in irregular tufted groups, forming a kind of ground-mass; these vary from a pale green tinge to almost colourless, and are generally very feebly, if at all, dichroic; the comparatively small extinction- angle shows them to be hornblende (actinolite) ; and there can be no reasonable doubt that they are of secondary origin; (6) small crystals, exhibiting often characteristic cleavages and even external forms (combinations of oP and oP), green-coloured and strongly dichroic; (c) large crystals (those mentioned above as supposed augite), including grains &c. of more than one other mineral, to be presently noticed. Augite occurs not unfrequently in almost colourless grains and crystals, some of which show a characteristic cleavage, and, in one or two cases, the characteristic outline (section of the combination of oP, oP‘, aP%s). The extinction-angle of the longitudinal sections of these crystals is large, ranging com- monly from 30° to 40°. They usually occur interspersed in a dull olive-green serpentinous or chloritic mineral. No olivine can now be recognized with certainty in the slides; but there are a number of irregular grains associated in them with the other minerals, and abundantly included (these being of smaller size) in the large horn- blende crystals, which there is every reason to believe are pseudo- morphs after the former mineral. Opacite and rounded crystal- line grains resembling magnetite abound in these, often clustered together or lining roughly parallel cracks, which remind us of the uregular cleavages of olivine; from these a brown staining often extends inwards for some little distance. ‘The pseudomorphic con- stituents vary considerably: sometimes, as described above, they are a brownish or yellowish green, of filmy granular structure, act- ing upon polarized light, but not greatly changing their colour; sometimes they exhibit a finely-speckled aspect, as though aggre- gates of extremely minute grains or folia of a mineral that acts strongly upon polarized light; sometimes they are aggregated small folia of a mineral resembling tale; and sometimes a clear isotropic, or nearly isotropic, mineral, such as is common in ordinary serpen- tines. Two or three of the inclusions in the larger hornblende crystals exhibit a radial aggregate structure with the usual black crosses. Though most of the above microlithic products are not the most usual replacements of olivine, I have seen them occasionally in my studies of peridotites, and have no doubt that this mineral was formerly present in the Pen-y-Carnisiog rock (probably a rather ferriferous variety). Magnetite is not uncommon as an in- clusion in the larger hornbiende crystals; the slide also contains a little altered brown mica, and a few small crystals of apatite. With regard to the large black crystals already mentioned, although, from their optical properties, one cannot but regard them as a brown hornblende, I doubt much whether this is not due to subsequent paramorphic change, and believe that they were for- PICRITE NEAR PEN-Y-CARNISIOG, ANGLESEY. 139 merly a true augite. The extinction-angles are generally less than 20°; but in one slide are two crystals which, though dichroic, give angles of about 30°. The larger crystals of the Schriesheim picrite agree with augite in their feeble dichroism and general appearance ; but the results of several measurements of the extinction-angle are less than 20° in the case of the largest crystal, while in a smaller one they are over 30°; and a very considerable quantity of horn- blende, similar to the varieties (a) and (6), is present in the body of the slide. A little olivine has escaped change; and the structure of the rest is rather more characteristic. Still, except that a mica is decidedly more abundant in the Schriesheim than in the Pen-y-Car- nisiog rock, the main difference appears to me to be, that the latter has undergone more alteration than the former, so that it, too, has once been a true picrite. Another rock is very abundant in boulders in this district of Anglesey, which, at first sight, has some resemblance to the picrite, though less porphyritic in structure. Closer examination, however, shows that felspar is always a constituent of this, though it is often not very conspicuous. I have examined a specimen microscopically, and find it to consist of a green hornblende, an altered felspar, a brown mica more or less changed, apatite, and a chloritic mineral. The hornblende and felspar are rather irregular in external form: the latter is almost wholly replaced by microlithic products; but one or two grains still retain sufficient traces of their original structure to show they have been plagioclastic. It is somewhat singular that a rather similar rock, but with more brown mica and better-preserved felspar, occurs at Schriesheim, within a short distance of the picrite, also intrusive in the granite. This is named a Labrador-diorite by the German petrologists. The only other instances known to me of the occurrence of picrite in the british Isles are two in Fifeshire, described by Prof. Geikie in his excellent monograph on the Carboniferous Volcanic Rocks of the Firth-of-Forth basin*, and one described by myself, to which, as I had at that time never examined a typical picrite, and had failed to obtain a very clear notion of the rock from such descrip- tions as I had seen, I did not venture to attach the name. In this, however, olivine (still very fairly preserved) is the dominant mi- neral; so that it comes nearer to a normal peridotite. The rock was collected many years ago by Professor Sedgwick, near Penarfynydd, in the Lleyn peninsula; and Mr. HK. B. Tawney, who lately visited the locality expressly to search for it in sitw, failed to find it, and believes that the specimen must have been taken from a boulder. Here also, as described by Mr. Tawney, are olivine-diabases and hornblendic diabases. I have ventured to draw especial attention to this Anglesey spe- cimen, in the hope that some geologist may succeed in discovering a like rock in situ. As the picrite just described is so uncommon and of so marked a character, we might assume with much confi- * Transact. Roy. Soc. Edinburgh, vol. xxix. p. 437. + In a paper by Mr. E. B. Tawney, Geol. Mag. Dee. 2, vol. vii. p. 208. 140 . ON A BOULDER OF HORNBLENDE PICRITE. dence (supposing no physical difficulties presented themselves) the Pen-y-Carnisiog boulder to have been derived from that parent rock. This would afford us most valuable evidence as to the direction in which the agent of transport (doubtless ice) had formerly moved. The south side of the Lleyn peninsula, even if the rocks corresponded (which they do not), seems excluded by physical considerations ; but I may remark that, though I have examined many boulders in Anglesey, I have failed as yet to identify any of them with rocks . from the Lake District or from Scotland, and think that we must look to the mountains of North Wales for the home of those which cannot be found in Anglesey itself. ON THE GEOLOGY OF CENTRAL WALES. 141 12. The Gronoey of Cenrran Wares. By Watrer Kezerine, Esq., M.A., F.G.S., Lecturer on Geology in the University of Cam- bridge. With an ApPENDIX on some new Species of CLADOPHORA, by Coartes Lapwortu, Hsq., F.G.S. (Read June 23, 188U.) [Puate VIT.] ConTENTS, Part I. More distant sections. Introduction and Historical. (1) Llanbrynmaer area. Illustrative section from Aberystwyth (2) Dovey ~wallsy area. to the Devil’s Bridge and Plyn- (3) Corris area. laniaion. (4) The Cardigan area. (1) The Aberystwyth Grits. General results. (2) The Metalliferous-slate Group. Part IIT (3) The Plynlimmon Group. art 1h. Paleontological evidence. Part II. General summary. Brief notes upon other sections. (1) Aberystwyth through Pont APPENDIX. Erwyd to Builth. Appendix by Mr. C. Lapworth, F.G:S., (2) Llandeilo to Aberaeron. on new Species of Cladophora. (3) Rhyader to the Teifi Pools. General results. Part I. Tue following communication contains some of the results worked out in frequent field-excursions during three years’ stay at the University College of Wales, Aberystwyth. In many of these ex- cursions I was accompanied by my students of the College; and to them I am indebted for much help both in the museum and the field*. Our principal field of work was, naturally, within a radius of some fifteen to twenty miles around the town of Aberystwyth ; and most of this country we have carefully searched and traversed through and through. The more distant areas to the south of Cader Idris, at Llanbrynmaer, Llandovery, Llandeilo, Cardigan, &c., have also been visited with a view to the determination of the extent and variation of the Cardiganshire rock-groups, and the dis- covery of their stratigraphical relations to other and better-known geological horizons. Little has yet been done by geologists to elucidate the structure of this part of Wales: while the most careful labours of our greatest authorities have been devoted to the study of the northern counties and eastern borders of Wales, and also, of late years, to the south in Pembrokeshire, this barren and chaotic area of Mid Wales has been always neglected, and is even now very rarely touched with the geolo- gical hammer. For the bibliography of our subject we have therefore but little to say, the most important contributions being :—the work * In particular, I have received much assistance from my former pupils Mr. T. Roberts, now of St. John’s College, and Mr. E. Evans of Sidney College, Cambridge. 142 WALTER KEEPING ON THE of our great leader Sedgwick, published in the Society’s Journal in the year 1847 (vol. ii. p. 150); a slight account of the Dol Fan conglome- rate, and a description of the Lampeter worm-trails in Murchison’s ‘Silurian System, pp. 316, 317; some scattered remarks by Mr. Salter and Sir Roderick Murchison in ‘ Siluria, and by the former in the Cambridge ‘Catalogue.’ The Pont Erwyd district is referred to in Symonds’s ‘ Records of the Rocks,’ pp. 132,133; some Graptolites from Aberystwyth are recorded by Mr. J. Hopkinson, F.G.S., in the ‘Journal’ of the Quekett Microscopical Club, vol. 1. p. 151; and the mineral veins of Cardiganshire are described by Mr. Warington Smyth, M.A., F.R.S., in the ‘Memoirs of the Geological Survey,’ vol. ii. part 2, p. 485. The present communication, although worked out in a close net- work of foot-routes in parts of the district, and cross traverses over the rest, is offered only as a first reconnoissance report, which must be followed by many labours betore the structure of this great and complicated district can be thoroughly mastered. Some general order amidst the whirl of contortions, however, is now made clear, such as the general succession of the Aberystwyth, Metalliferous, and Plynlimmon groups, and the great Plynliimmon synelinal. The enormous apparent thicknesses of the rocks are shown to be in part due to a series of inversions; and, further, the interpretation and correlation of the beds are greatly helped forward by discoveries of fossils, mostly Graptolites, in a number of localities, which afford excellent data for comparison with the more distant Scotch and Cumbrian Silurians. Even Sedgwick’s work (1846) was but of a very superficial kind ; for he tells us, ‘“‘ I profess not to know well this most contorted and perplexing country.” He makes four rock groups, namely the (1) Aberystwyth, (2) Plynliimmon, (3) Upper South-Wales Slate with the Rhyader Slate, and (4) Cambro-Silurian groups, which appear in ascending order from west to east; but he adds, “the sections are singularly contorted, the groups ill defined, and the actual order of superposition obscure.” His first group I still main- tain under the name of the Aberystwyth grits; but nearly all the remaining rocks in the line of section might, I believe, have been included in his second great series—the Plynlimmon ‘group. On the other hand, I have considered this great Plynlimmon group of Sedgwick under two distinct headings, namely the (inferior) slate series or Metalliferous-slate group, and the overlying Plynlimmon grits, so that we now have the following succession of deposits :— (3) The Plynlimmon grits, forming a line of high country in the centre of Wales, including Plynliimmon. (2) The Metalliferous-slate group, forming a broad zone of Cardiganshire contorted country on each side of (8). group. (1) The Aberystwyth grits, best developed between Aberyst- wyth and Aberayron. Nos. (1) and (2), being closely bound together by their fossils, are known together as the Cardiganshire group. For an illustrative section of our area the best I can offer is that from Aberystwyth, through the Devil’s Bridge, to the Plynlimmon 143 GEOLOGY OF CENTRAL WALES, Fig. 1.—Section from Aberystwyth to the Devil’s Bridge. (Horizontal scale, $ ch to a mile.) Ww. BE. Aberystwyth. Devil’s Bridge. The Arch, RAS Zz PP pos Vi S = PUG Aberystwyth grits. Metalliferous-slate group. The continuous dotted line indicates the probable course of the Cefn Hendre and the Devil’s-Bridge fossil-zone. . Fig. 2.—Continuation of the Section in fig. 1, from Owm Ystwyth, through Craig Lluest, to Rhyader and Gwastaden. (Horizontal scale, 1 inch to a mile.) W.<—e@ m—>T, W.N.W.<—= Onc Popty. Cwm Ystwyth. Craig Lluest (about 1200 feet), Pont ur Elan, Pen Rhiev Wen. Gwyn Llyn. Rhyader, Gwastaden, Za A (j---~---+--=- i Large-llag series. Grits and paper shales. Pale slate, thin Shaly slates. Hard pale Conglo- The Plynlimmon Synclinal. grits and shales. slates. merates. 144 WALTER KEEPING ON THE Mountains (figs. 1 & 2). The lowest beds form the cliffs at the coast, while the highest go to form the central mountains. All these rocks are amazingly contorted ; and we find the intermediate or Metallifer- ous group especially thrown into frequent and violent twistings, with, in many places, actual inversion. A continuation of the section eastward, beyond Plynlimmon (fig. 2), shows a similar appearance of the rock groups in reversed order of outcrop and with contrary dips, thus showing that Plynlimmon, like its more ancient and greater fellow Snowdon, stands in a geological valley or synclinal. We proceed to the detail of this section. 1. The Aberystwyth grits consist of beds of hard, compact, dark grey grit or greywacke, and dark shales, rabs, and imperfect slates in strikingly regular alternation, as may be seen in the ordinary photographs of the Aberystwyth cliffs. The grey grits or greywackes are of great sameness and regularity both in structure and composition around Aberystwyth, being hard, grained rocks, often felspathic, regularly and sharply jointed. A cross fracture often shows a remarkable contortion in the lines of lamine, this being, I believe, mostly of subsequent “‘ concre- tionary” origin; many of the beds themselves are also, in part, of the same concretionary growth. ‘The beds measure very con- stantly about 4 to 6 inches in thickness; and their under sur- faces exhibit an abundance of raised markings, which are irregular or tortuous, branching, net-lke or worm-like, these being also, in part, of concretionary origin. The argillaceous partings are usually of about the same thickness as the greywacke-beds, sometimes thicker (especially to the east and south), sometimes thinner (as In many places around Aberystwyth). Most of their varieties are the result of subsequent metamorphic changes acting differently upon the rock according to a slight diversity of original constitution, or depending upon slightly dif- ferent mechanical conditions. Thus have been produced the various forms of shivery shale, large platy shale, rubbly rab* of various forms and sizes, soft shaly slate, and even very well-marked regular slate in the Aberystwyth district. They are uniformly of dark colour, and never greatly indurated. Lenticular nodules with ‘‘ cone- in-cone”’ structure are of frequent occurrence both in this and the ~ following (or Metalliferous) series. Fossils—Fucoidal and worm-like markings are of frequent and wide-spread occurrence throughout this series, appearing for the most part in the form of raised markings upon the under surfaces of the grits. I have also found Graptolites in several localities. Quarry at Cwm, on the south side of the Clarach Valley. Monograptus Sedgewickii, Portt. Monograptus tenuis, Portl.? —— Clingani, Carr. Buthotrephis, small species. lobiferus, M/‘Coy. * A rab is a fine-grained rock, usually argillaceous and not indurated, which readily breaks up into a rubble of cuboidal or prism-like fragments. GEOLOGY OF CENTRAL WALES. 145 From the Bryn-y-Carnau Quarry near the old Water-reservoir, Aberystwyth. Monograptus tenuis ? Dictyonema delicatulum, Lapw. (n. sp.). But the most important zone of fossils is found somewhat higher up in the series, exposed in a quarry formerly worked for building- stone and road-metal, in a field below Cefn Hendre, where we have collected Monograptus Sedgewickii, Porél. , var. distans, Port. —— tenuis, Portl.? —— Clingani, Carr. Oxthoceras, sp. lobiterus, M*‘Coy. Calymene. turriculatus, Barr. | Monograptus Sedgewickii, Port. crenularis, Lapw.? lobiferus, M/‘Coy. Monograptus Hisingeri, Carr. Many of these fossils were found in a set of thin, dark-grey, mica- ceous flags (the large-flag series), which are sometimes to be obtained of large size (4-6 feet square). Dips and Foldings.—Many rock-foldings, some of them very violent and accompanied by fractures, are seen in our line of section, good examples being exposed near the second milestone from Aberystwyth ; so that for a mile anda half along the Upper Devil’s Bridge-road we appear to have only the same set of beds, repeated again and again by a number of rock-foldings (see section, fig. 1). But beyond this the easterly dip becomes more constant, and we appear to be traversing the outcrop of a continually ascend- ing series. 2. The Metalliferous-slate Group.—In the Cefn-Hendre quarry, only about a mile and a half east of Aberystwyth, we already find a larger proportion of shales to grits than in the coast-section around Aberystwyth ; and this change becomes still more marked in another quarry on the road-side further on down the hill towards Gogerddan. ° The same gradual though not perfectly regular disappearance of the grits to the east may be seen in our present section along the Devil’s-Bridge road, as, indeed, in any other of the east and west roads from Aberystwyth*. A change in the character of the argillaceous rocks appears in regular correlation with the loss of the grits. They become more and more indurated and cleaved, until, as the boundaries of the grit-series are reached, the normal cleavage of the district (striking N.N.E. and 8.8.W.) is found even in the partings between thick erit-beds. Thus gradually do we enter into the territory of our second series—the Metalliferous Slates. The hills of this district are barren and desolate, even more so than in the Aberystwyth-grit country ; but they are decidedly more rounded and regular in their contours. Here, in the uplands, is a * The limit of the grit series may be placed, as indicated by the yellow dots upon the Survey Map, at about three miles and a half east of Aberystwyth. In my pocket-book I find “‘at three miles from town, grits fewer and thinner than at Aberystwyth,” at four miles ‘‘a few grit-beds as much as 6 inehes thick,” and at five miles “still a few thin grits.” 146 WALTER KEEPING ON THE vast dreary mountain-region of bare sheep-walks scantily marked out into districts by poor stone walls and wire fences, with much waste bog and peat land. The rocks are of uniform pale blue and grey colours, varying from small papery shales (rarer) to large irregular platy shale or regularly cleaved slates, also (in some areas) much indurated small slate rock cross-cut into fragments by frequent bedding- and joint-planes. Some zones of softer rab, like that of the Aberystwyth group, and a pale mudstone rab are sometimes found, especially near the junction with the grit series. As we reach the central mining district, some seven miles east of Aberystwyth, beds of hard, pale, indurated slate-rock with frequent bed-bandings occur; and in the immediate neighbourhood of the mineral veins such induration is nearly always well marked. Occasionally a bed of grit, 2-6 inches thick, occurs in this series ; but such occurrences are very rare, so that the building- and road- stones of the central ‘‘ Metalliferous” country have to be carried for miles, either from the Aberystwyth grit quarries in the west, or from the Plynlimmon group further east. Thin ferruginous gritty bands, about 4 inch thick, however, are more frequent, some of them being highly micaceous. Many such are seen between the fourth and fifth milestones and around the ninth milestone on the Devil’s-Bridge road. Rock-foldings.—Several excellent exposures of the contortions in this slaty series are seen along our line of section. There is a neat little synclinal at seven miles, an anticlinal towards the eighth mile- stone; and several folds may be detected about nine miles east of Aberystwyth; but overriding these foldings the prevailing dip is seen to be clearly and determinately to the east. Fossils are rare in this series. The curious branching structure, Nematolites Edwardsii, Keep., occurs at Ty Llwydd; and I have found the Fan Algal (Buthotrephis major, Keep.), the Nematolites, and worm-trails (Werettes) in several other places. But in other areas, especially at Cwm Symlog and near Machynlleth, out of our present line of section, a rich Graptolite-fauna has been discovered, and will be described later on. The Devils Bridge, great Inversion of the Rocks (fig. 1).—As we de- scend the hill to the Devil’s Bridge we pass over pale, hard, shaly slates* with thin gritty bands about 1 inch thick, marked with worm- tracks (Nereites Sedgwickit), impressions of the Fan Angal (Buthotrephis major, Keep.), and Nematolites Hdwardsi, Keep. Coarse roofing- slates have been worked about half a mile west of the hotel. These rocks belong to the Metalliferous-slate series; but coming to the waterfall we meet with a set of alternating thin grits and large pale shaly slates, together with some large flags of laminated grit, with fossils exactly resembling those already noticed from Cefn Hendre. Many of the grit bands are thin and little jointed, so that large * Although the splitting of these rocks is so irregular and shale-like, yet the planes of division are of subsequent cleavage origin ; therefore I call them shaly slates rather than slaty shales. GEOLOGY OF CENTRAL WALES. 147 flags may be readily extracted, from which feature I have called this zone the ‘‘ Large-flag series” (fig. 2). The dip is to the west, running under the slaty series. Fossils may be found here in the open quarry above the Devil’s Punch-bowl, and along the sides of the waterfall itself; but all the Dendroid Graptolites (Cladophora) are from a band above the iron bridge in the waterfall-grounds ; while some other beds below this bridge are so crowded with Crinoid ossicles as almost to merit the name of Limestone. Fossils from the Devil’s Bridge :— Monograptus Sedgewickii, Port/.? Odontocaulis Keepingii, Lapw. —— spiralis, Geinitz. (n. sp.). turriculatus, Barr. Chonetes lzevigatus, Sby.? Dictyonema venustum, Lapw.(n.sp.). Orthis, sp. —— corrugatellum, Lapw. (n. sp.). Other fragmentary Brachiopods. Calyptograpsus (?) plumosus, Lapw. Phacops, n. sp. n. sp.). Fragments of Encrinites. Rhizograptus (?) digitatus, Lapw. Nereites Sedgwickii, Murch. (nu. sp.). Myrianites tenuis, I‘ Cov. —— ramosus, Lapw. (n. sp.). Now in this list of fossils the species of Graptolites are seen to correspond with those of Cefn Hendre, the occurrence of Mono- graptus turriculatus in both places being an especially important fact, this being a species of very limited range. And we have seen that they occur in identical rocks in the two places; I therefore cannot doubt that these are really one and the same set of beds, seen in the upper part of the Aberystwyth grits at Cefn Hendre and reappearing at the surface in an anticlinal at the Devil’s Bridge. | This conclusion, however, is in direct antagonism with the strati- graphical appearances ; for, notwithstanding the numerous folds, the predominance of the easterly dip is most determined and impresses itself strongly upon the mind. Being convinced of these appearances, it was determined to test the thickness of the beds by actual measurement; and our exact observations and calculations, made at more than 100 exposures along the Devil’s-Bridge road from Aberystwyth, strikingly con- firmed our earlier impressions, giving, indeed, a result of nearly four miles thickness of strata (3 miles 1612 yards). In such apparent conflict of evidence, and in the absence of large faults, the only reasonable explanation is that the original natural order of the rocks has been destroyed by the formation of a great inversion, or rather, as I believe, a series of inversion folds in the Metalliferous- slate series (see fig. 1). And indeed this interpretation, in con- formity with the fossil evidence, is independently almost demanded to explain away the enormous apparent thicknesses of similar rock- beds as measured from their present dips. We may then safely conclude that in the Devil’s-Bridge rocks we are again upon the upper part of the Aberystwyth-grit series. Continuing our section eastward over the hill through the Arch to Cwm Ystwyth (fig. 2), we still traverse a series of shaly slates of the “‘ Metalliferous” type, with here and there thin grits apparently 148 WALTER KEEPING ON THE belonging to the Large-flag series. The presence of Fucoidal mark- ings, including the Net Algal (Reteofucus extensus, Keep.) and Nema- tolites at one mile and a half from the Devil’s Bridge, also serves to indicate our proximity to the Aberystwyth-grit series. The general apparent dip is clearly eastward, with some foldings. Descending beyond the Arch towards Kglwys Newydd the only matter of special interest is the occurrence of a thin seam of rotten- stone (a decomposed limestone) which may perhaps correspond with the crinoidal zone in the Devil’s Bridge. Such rocks are of very rare occurrence in Mid Wales, so that it is a popular saying that ‘‘ there is no lime in Cardiganshire.” The rocks around Popty, near Eglwys Newydd, are again seen to be very much of the Devil’s-Bridge type; and I am inclined to think that a set of rock-folds with reversals have here brought that fossil zone near to the surface again. We now reach a region of manifest great rock-foldings; and as we ascend Cwm Ystwyth a number of exposures display an important westerly fold in the rocks. The beds appear still to belong to near the junction of the Aberystwyth grits with the Metalliferous-slate group. Next we pass through the rich metalliferous district of Cwm Ystwyth with its network of mineral veins in the Metalliferous- slate group, beyond which, at Blaen y Cwm, thicker grits (about 2 feet) with east and south-east dips come in amongst the slates; and these latter become replaced over a considerable area by dark shaly slate, rab, pencil-rab, and other softer forms of the argillaceous rocks. Thick grit-beds, which are cleaved, next form a prominent feature at Craig Lluest (fig. 2); and these I regard as belonging to the lower part of our upper grit-series—the Plynlimmon group. This series is better developed in Plynlimmon itself, to which we must refer (infra, p. 156) for a more detailed description. The dip is eastward. Some three quarters of a mile beyond the top of the pass*the dip changes, a westerly slope setting in; and this remains well sustained, though with some east foldings, on to the hills west of Rhyader. Here, then, we see the eastern half of the great axial synelinal of this part of Wales—a great fold in which Plynlimmon lies; ‘but its apparent magnitude is greatly exaggerated by the phenomena of reversed dips. The lithological details of this part of the section in its frequent slight variations, but general monotony, till we reach Rhyader, would occupy much space and be of little value. Some pale shaly slates approaching the character of the Tarannon shales, and thin gritty bands, are seen on the dreary bog- and moor- land of the higher plateau and on the east of the pass, where also some of the grits exhibit irregular fucoidal markings and the Net Algal (Retiofucus extensus, Keep.) upon their under surfaces. One good anticlinal fold in “ Metalliferous slates” with some erits has its axis about one mile and a half east of Aber Gynwy; and a synclinal in the slaty series is indicated lower down on the road to Pont ur Elan. ‘The well-sustained general westerly dip of the GEOLOGY OF CENTRAL WALES. 149 great series of thin-bedded slate, pale- or dark-coloured, seen in the ascent from this bridge to the top of the high part of Rhyader, appears to show these rocks as in regular serial continuity with the beds passed over on the great slope next to the west, the whole forming the vast westerly fold complementary to the great easterly dip of the Metalliferous-slate series between Aberystwyth and the Devil’s Bridge. And just as the latter is now proved to be not one continuous series, but reallya much smaller group affected by fre- quent inversion, so we are, I think, justified in adopting a like inter- pretation for these beds to the east of the Plynlimmon axis. The correspondence of the rocks on the two meridional sides of this axis is further illustrated as we reach the eastern edge of the great plateau above Gwyn Llyn (fig. 2), where we find a zone of pale banded and indurated slate rock identical with the Strata Florida slate rock used in the Aberystwyth stone pier. These pass under a series of grits and conglomerates in the beautiful valley of Cwm Elan. A little further south these latter beds reappear in the Gwastaden grits and conglomerates of the hills east of Rhyader. Thus our accumu- lated facts support the view that there are two distinct series of grits in Cardiganshire, separated by the Metalliferous-slate group. The Rhyader grits and conglomerates appear to be nearer to the Plynlimmon than the Aberystwyth group. The Aberystywth Grits, extent, varrations, Jc-—A glance at the yellow-dotted area upon the Geological Survey Map will show that the Aberystwyth grits form a crescent-shaped patch of country in the centre of the western Welsh border, extending from near Borth, some five miles north of Aberystwyth, to Traeth Bach, south of Llangrannog, a distance of thirty miles, and with a maximum breadth, at Mynydd Bach, of nine miles. A set of characteristic surface-features marks off its boundaries with tolerable distinctness, its barren or gorse-covered ridgy hills, elongated along the line of rock strike, and with minor sets of parallel crests and ledges (sometimes forming sets of small step- like structures or inclined ribs in the lines of more durable grit-beds), giving to the group an almost unmistakable appearance*.. The general characters of the rocks have already been described in our Deyil’s-Bridge line of section; and the variations from those types are neither numerous nor very striking. The greywacke and erits are very constant in grain, never becoming conglomeratic; but some coarser varieties are found in a few localities—for example, around the Hiddwen lakes, Mynydd Bach, and at Pen Craig, north of Llanilar. In these places the greywacke is almost granitoid in appearance, the quartz grains being large, flakes of mica common, * It was pointed out long ago by Sir Henry de la Beche that the same type of physical features appears in the area of the grits of Penrhyn ddu in the Lleyn peninsula, which are marked Lower Cambrian upon the Survey Map. To me this resemblance also appeared very striking. The Penrhyn-ddu beds are totally unlike any of the Longmynd group known to me; and I cannot believe they are of such an age; nor do I think with de la Beche that they are of the same age as the Aberystwyth beds, but regard them as an exceptionally developed type of the Tremadoc series. Q.J.G.8. No. 146. M 150 WALTER KEEPING ON THE and crystals of felspar, sometimes perfect but more often fragmen- tary, very numerous and conspicuous. These constituents are held together in the dark argillaceous greywacke matrix. The rock often closely simulates a volcanic ash*. In thin sections, examined under the microscope, the Pen-Craig rock exhibits irregular angular and subangular fragments of quartz and felspar, the interstices filled with the dark, opaque (argillaceous) matrix. The quartz is somewhat cracked, and includes numerous cavities and minute spicular microliths; these latter are grouped into wavy stratoid zones, between which the principal cracking of the quartz runs. ‘The felspar is mostly in the form of.very angular crystal-fragments, usually much decomposed, and of powdery ap- pearance; but some of the better-preserved fragments show the characteristic ribbon banding of the Plagioclase group when ex- amined with polarized light. The ordinary finer-grained beds are usually of darker colour than these, and contain less felspar; they are, however, for the most part quite similarly constituted, the felspar crystals often being readily recognizable when decomposed into irony spots over the weathered portions of the rock. Paler thin bands, very compact in texture, occur in the southern part of the district: around Llangrannog. The presence of crystals of iron-pyrites occasionally gives a marked feature to the grits; and at Pen Craig, Llanilar, and in the Garthen valley, Eglwys Fach, some beds occur beautifully studded with these brassy cubes. The ordinary thickness (4—6 inches) of the beds is very constant throughout the area of the Aberstwyth grits; but more massive beds (1-13 foot) occur under Allt wen and at Aberaeron, in the Garthen valley, and at Llangwyryfon (24 feet). At Cefn Coch, Pen Pegwyns, and along the coast south of Aberystwyth still thicker beds are seen (3-4 feet); they are worked for building-stones. At the southern limit of the series, near Llangrannog, the grits become thin, irreguler, and inconstant, thus gradually dying away, to be replaced by the ar- gillaceous slate-rocks. | Two structural peculiarities are very characteristic of the Car- diganshire grits—namely, the remarkable contorted lamination seen on a fractured surface, and an abundance of fucoid markings, to- - gether with strange-looking ripplings, ridgings, volutings, and other raised forms of structure found upon the undersurfaces of the grit- beds?. The rock is jointed, sometimes into large blocks good for building- and flag-stones, but often much more closely into regular oblong or rhomboidal fragments. At Allt-wen cliff and elsewhere it is divided up into oblong, prismatic, or rudely lozenge-shaped fragments a foot long, forming a kind of coarse grit-rab. * This abundance of felspar crystals, so general in the Silurian rocks (Upper Silurian of North Wales, South Wales, and the Lake-district), points to the neighbouring presence of a vast mass of early, perhaps primeval, igneous rocks as the great source of sediment-supply in Silurian times. + It is proposed to give a more particular account of the rock-structure of Cardiganshire in a separate paper. GEOLOGY OF CENTRAL WALES. 151 Little need be added to what we have said on the alternating argillaceous beds. The rab-type is best seen upon tho Allt-wen cliffs, south of Aberystwyth, where also au elongated type, serving very well as rough pencils, forms the ‘“pencil-rab;” these beds commonly appear much more solid in the heart of the rock, the fragmentary structure only becoming developed after exposure to weathering. The cleavage phenomena here are often of peculiar interest from their frequently incomplete, half-developed character. The dark shaly- . looking rock of Constitution Hill, Aberystwyth, would not at first be suspected of a slaty structure ; but Specimens may be gathered with the cleavage seen distinctly cutting across the stripe- lines ; and many of the dark, soft, shaly rock-beds of the neighbouring cliffs, when seen in situ, will be found to have their shetsme) ies running across the bedding. A pale, homogeneous, hardened mud, or pale mudstone, splitting with curved, conchoidal surfaces into 2-3-inch blocks, is of additional importance, as probably marking a zone; I have found it some half a mile north-east of Nant Kos, at Pen Craig, Llanilar, on the road-side N.N.E. of Llanbardarn, and at five miles east of Aberaeron. This appears to indicate a distinct zone near the top of the Aberystwyth grits. The same kind of rock occurs at Llyn Fyrddyn Fawr, and at Rhos Rhydd, near Llantrisant, where itis again not far removed from the thin-flag series—another zone at the top of the Abery- stwyth grits. The Large-flag series.—This series, which belongs to the upper part of the Aberystwyth group, is best marked to the east, in the neighbourhood of the Devil’s Bridge and Llantrisant, where its presence is soon made manifest in the construction of piggeries and other low huts, whose sides and roofs are covered with the flags. These are grit-beds, about 1 inch thick, but little jointed, so that large slabs, frequently 4 and 6 feet square, are commonly extracted for use in the neighbourhood. ‘The rock is fine-grained, and usually exhibits a complete contortion in the lines of lamine; altogether it much resembles some of the grit-beds of the middle Lingula- flags of North Wales. The surfaces are gencrally undulated, and often show true ripple-marks, over which we find Graptolites and other fossils spread out: the Devil’s-Bridge Dendroid Graptolites occur in this series. The further extensions of these beds, north and south, are seen around Pont Erwyd and south of Llantrisant; and the same series seems to be brought up to day by folds, on the east at Hglwys - Newydd, and to the west near the head of Cwm Symlog. Bands of rottenstone (decomposed limestone) occur at the Devil’s Bridge, near Eglwys Newydd, and again in the hills some four miles east of Tal y bont. Around Aberystwyth these flags are not so well developed; but they may be recognized at Cefn Hendre, as already described, where they are put to the usual purpose of hut-making for the work- men’s shelter. M 2 12 WALTER KEEPING ON THE Fossil localities, other than those already given, are :— Parson’s Bridge, and Temple Mine, south of Pont Erwyd. Monograptus lobiferus, MM‘ Coy. Myrianites M‘Leayi, Murch. Climacograpsus scalaris, His. Buthotrephis major, Keep. Nereites Sedgwickii, Murch. Cefn Coch, near Aberystwyth. Monograptus Sedgewickiu, Pordd. | Retiofucus extensus, Keep. Buthotrephis major, Keep. Paleochorda tardifureata, Keep. minor, Keep. | Constitution Hill, Aberystwyth; Wern Grug, Llanilar; Mynydd Bach; Plas Crag, Llanbardarn, and other places have also yielded a few fossils, mostly Alovw,—Buthotrephis major, B. minor, B. minimus, Keep., Paleochorda tardifurcata, Retiofucus extensus, and Nema- tolites Hdwardsu, Keep. Thickness.—We have seen that much of the apparent thickness of the Aberystwyth grits is deceptive, being explained away by great folds. Still a great series remains, well attested even by the simple heights of many hills which are built up entirely of this group. Mynydd Bach appears to be over 1000 feet high (by aneroid); and I should estimate the group at certainly not less than 1000 feet of maximum thickness. Our detailed measurements along the Devil’s- Bridge road show a thickness of 1639 yards, or nearly one mile. The Metalliferous-slate Growp—A very large part of Cardi- ganshire is constructed of this series. In the special area of our work it forms a broad semicircular zone, some eight to ten miles across, sweeping around the dome of the Aberystwyth grits to form all the slaty country from north of Cardigan up to Machynlleth. At the latter place it forks into two, the main curve being continued out to sea by the Dovey valley, while a broad but rapidly attenu- ating arm runs northward to Dinas Mowddwy, at which place its much diminished representative is found between the Bala series and the Tarannon shale. The above area, however, is not unmixed metalliferous slate, the upper part of the Aberystwyth group being brought up by folds, along north-and-south lines, running through Ystrad Meurig, and the Deyil’s Bridge; and possibly some outliers of the Plynlimmon group may be folded in. Still the general absence of arenaceous beds is marked over miles of country without a single grit-bed to serve for road- or building-stone. ‘Tin Llidiart village uses the hard pale slate of Goginan for building ; and at Llanafon the numerous boulders serve for road-metal. Further south, in the line between Cardigan and Llandovery, the slate series appears in still greater extent, which is due to its there representing also the Aberystwyth group, and perhaps also the Plynlimmon grits, these groups not being developed as such in that area. A marked lithological character distinguishes this group as a whole, the terms “‘ indurated shaly slate,” ‘irregular slate,” « flagey slate,” ‘ pale blue slate,” from my note-book, serving to indicate the usual character; I also find “hard pale slate-rock, often much GEOLOGY OF CENTRAL WALES. 158 jointed,” and “pale slates with ribbon bandings.” ‘The prevailing colour of the rocks is a pale blue-grey ; and they generally break up into moderately small fragments, shale-fashion. Slate.—In many places where the slates are better cleaved, and less frequently jointed, they are worked in small quarries for inferior local purposes, as around Llancynfelin, in the hills south of Machyn- lleth, in the island mounds of the Gorsfochno flats, and near the Devil's Bridge. These rough slates are pale-coloured, hard, and coarse; but around Machynlleth thoroughly good slates are worked in anumber of large quarries. One of the largest of these is at Pont Faen, where large slate-flags are extracted ; but the most interesting to the geologist is that at Morben, two and a half miles south-west of the town of Machynlleth. These are thick-hedded slate rocks, with a few thin grits, dipping at high angles to the W.N.W. This, however, is an inverted dip, a8 is proved by the position of the prominences on the grit-bands, which are upon the upper sur- faces of the grits, the worm-trails being also manifestly in inverted position. ‘The cleavage-dip diverges only slightly from the bedding-. dip, so that the Graptolites usually run for some distance along the surface of the slate, and then gradually skim under it. The slates are very good, dark and soft. Nodules of iron-pyrites occur; and nearly all the Graptolites are converted into this bright mineral. This quarry has yielded the finest of our fossils, the Monograptus Sedgewicki being especially magnificent. Fossils from Morben Quarry, Machynlleth. Rastrites peregrinus, Barr. Monograptus Hisingeri, Carr. maximus, Carr. -—— tenuis, Portl. Monograptus Sedgewickii, Pord/. myolutus, Lapw. — , var. spinigerus, Wich. Diplograpsus Hughesu, Nich. —. , var. distans, Pordd. Climacograpsus normalis, Lapw. ? —— intermedius, Carr. Myrianites M‘Leayi, Murch. ? spiralis, Geinitz. Lapworthii, Keep. ——- lobiferus, M‘Coy. Buthotrephis major, Keep. Below this in present position, but by inversion really above it, is a zone of pale blue and olive shaly slate, pointed out to me by Mr. J. EK. Marr, F.G.S., as being similar to certain zones in the Coniston mudstones. The great slate district of Corris, usually referred to the Llandeilo age, is in this series, as will be shown later on; also the excellent slates of Dol y Mynach and Cwm par Adwys, west of Rhyader. Other rock-varieties, such as dark paper-shales, rab, and mud- stone, approach near to the Aberystwyth types; and some of the occurrences of such rocks doubtless indicate the uprising of that group as inliers in the metalliferous-slate area. In the Devil’s-Bridge road such rocks appear in several places. Grit-beds as thick as 6 inches may occur; and thin zones, about 4 inch to 2 inches thick, are frequent; these are generally red- stained at the surface, and micaceous. ‘They are trustworthy guides to the dip of the beds, which in their absence is often not easy to discover. 154. WALTER KEEPING ON THE At Cwm Symlog, nine miles east of Aberystwyth, pale soft shales. and slates with thin grits, rich in fossils, are found. Fossils from Cwm Symlog. Rastrites peregrinus, Barr. Monograptus palmeus, Barr. Monograptus Sedgewickii, Port. Climacograpsus spiralis, His. crenularis, Lapw. FORAMINIFERA. cog ee ape _Dentalina. eregarius, Lapw. : intermedius, Carr. Textularia. —— Clingani, Carr. Rotalia. spiralis, Gezn. ANNELIDA. lobiferus, M*Coy. Nereites Sedgwickii, Murch. runcinatus, Lap. Myrianites M‘leayi, Murch. —— Hisingeri, Carr. , Nemertites Olivantii, Murch. tenuis, Portl.? involutus, Lapw. ALG. Diplograpsus tamariscus, Nich. ? Buthotrephis major, Keep. Hughesii, Nich. minor, Keep. sinuatus, Nich. —— minimus, Keep. At Ystrad Meurig, forming the high ground twelve miles south- east of Aberystwyth, and in near association with a fossil zone, is a series of hard pale grits with thin indurated slate partings, firmly bound together, to form hard massive rocks (a); we refer to this as the “Strata Florida rock.” Another well-marked rock demanding attention is (6) a pale indu- rated slate rock, usually cross-banded by thin arenaceous stripe- zones. It is well seen at Machynlleth, to the east of Kglwys Fach, and east of Tal y bont. Again, the same type of rock occurs (c) in the hills east of the Rheidol gorge; I find blocks of it around Llan- trisant; and it is also well developed at Goginan, to the east of Llanddewi Brefi, and in the hills west of Rhyader. These rocks are essentially similar to those of Ystrad Meurig, only less arena- ceous; and I include them all under the name of “Strata Florida rock.” Now, classifying these various localities of the Strata Florida rock, we find they fall into a set of definite lines having similar relations to neighbouring rock-beds, and indicating, as I believe, definite zones. First, the more arenaceous rock of Ystrad Meurig is in the same north-and-south line with Cwm Symlog and with Llantrisant ; and in the first two of these places it is in near relation with a rich fossil zone. Also it is about in the same line with the Large-flag series, as seen in the Devil’s-Bridge road, nine miles east of Aberyst- wyth, also found at Llantrisant aud Cwm Symlog. In these places we have, then, evidence of the occurrence of this hard pale slate rock along a definite line of country in near association with the Grap- tolite fossil-zone and the Large-flag series. Next we find that the group b also falls into a line east of Aberystwyth; and here, again, it is associated with the great fossil zone at Machynlleth, while near Tal y bont it is connected with a band of rottenstone like that of the Large-flag series at the Devil’s Bridge. Altogether, then, the result of these lithological correlational is to support our theory of four different lines of north-and-south foldings GEOLOGY OF CENTRAL WALES. BUS, between Aberystwyth and Plynlimmon—namely, (1) the principal axis of Aberystwyth, (2) the principal axis of Plynlimmon, and (3) the two minor axes running (a) from Ystrad Meurig through Llan- trisant and Cwm Symlog, ‘and (5) along the gorge of the Rheidol about Pont Erwyd and the Devil’s Bridge. The further northward extension of (a) is probably indicated by the grits of Yr Carreg and Moel y Llyn. These minor folds have not been recognized further south ; and, indeed, we might expect them to disappear as the lateral spread of the beds becomes restricted. At Lampeter the well-known olive shales and grits with Werettes &c. are exposed in quarries. The grits are in thin beds (2-4 inches), hard and laminated, with partings of shale either light (olive) or dark in colour; no cleavage is seen. The fossils are :— Nereites Sedgwickii, Murch. Nemertites Olivantii, Murch. -— cambrensis, Murch. Palzochorda tardifurcata, Keep. Myrianites M‘Leayi, Murch. Retiofucus extensus, Keep. also other worm-like and algoid markings. -Altogether this is a pretty typical metalliferous slate, though more arenaceous than is common, Fig. 3.—Railway-cutiing between Traws Coed and Caradoc Waterfall, 1877. Contortions.—The woodcut (fig. 3), representing a railway-cutting east of Traws Coed, illustrates some of the frequent and sharp contor- tions of this series. All the perceptible foldings, however, numerous and striking as they are, are far from adequately showing the real complexity of the infinite twistings and inversions of this series, the existence of which is proved principally by considerations from the fossil evidences. Without fossil evidences it is, as a rule, impossible to detect inversions; but still, in those places where grits with fucoidal markings &c. occur, we have another key to the true rela- tions of the rocks—namely, in observing the positions of the curious prominences upon the surface of the grit, these being normally always upon the under surfaces; but where the beds are inverted (as at Morben, Machynlleth, &c.) they are found upon the upper sur- faces, the true worm-trails (Nereites &c. ) being in this case Sapresed upon the under surfaces, in inverted position. These inversions explain away in great degree the enormous apparent thickness (miles of strata) of this series; but a very great vertical series still remains well attested in actual steep and preci- 156 WALTER KEEPING ON THE pitous mountain-sides, from 500-1000 feet high, and I cannot esti- mate the group at less than 2000 feet. Fossils are almost entirely wanting in this series, except, I be- - lieve, near the base, where I would place the rich beds of the Morben Quarry, Machynlleth, and Cwm Symlog. The fauna is intimately related to that of the Aberystwyth group, as indeed was to be expectec ; for the fossils of this latter group occur in its higher part, not far removed, in serial order, from the Metalliferous-slate fossil-zone. The worm-like and fucoidal markings, Nereites, the Fan Algal, and the Nematolites are widely distributed throughout the series, worm-trails being very characteristic. Other fossil localities are :— Llantrisant, with the large Fan Algal (Buthotrephis major) and Nematolites tubularis, Keep.; the Llyfnant valley, with Climaco- grapsus scalaris, His. ; Dyffryn Castell, with Climacograpsus scalaris, worm-tubes, and the Nematolites, also some of the shells recorded by Sedgwick ; and Steddfa Gurig (west of), with Monograptus Sedge- wickri, “Portl., M. tenuis, Port. ?, Climacograpsus scalaris, His., and Orthoceras. ‘Above Taliesin, frasments of dendroid Graptolites and Nemutolites tubularis occur ; sm the latter is also found at Llantri- sant, Ty Llwydd, on the Rheidol, and elsewhere. Richer localities are :— West of Lisburne Mine, Ystwyth Valley. Climacograpsus scalaris, His. Nereites Sedgwickii, Murch. Buthotrephis, Rastrites ? Monograptus Sedgewickii, Pordl. spiralis, Gein. -—— lobiferus, MM‘ Coy, Garthen Valley, Melin Newydd. Monograptus Sedgewickii, Por?/, Orthoceras, sp. Climacograpsus scalaris, Hs. The numerous metalliferous veins (lead and silver) in the rich mining-districts of Cardiganshire are almost confined to this set of beds; hence the name here given it. 3. The Plynlimmon Group.—As we ascend the mountain-slopes along the valley of the Rheidol, above Pont Erwyd, we meet with the rocks of the upper grit series, which form the upper part of the Plynlimmon mountain. Around Garn Fach some of the grit-beds are seen, of rather coarse type, sometimes assuming a very grani- toid aspect, there being a profusion of felspar crystals and large blebby quartz grains. The felspar crystals may be of large size; and some of them are perfectly preserved and glassy. Doubtless the rock is the direct result of the degradation of an ancient acidic igneous rock. Conglomerate beds, consisting of pebbles of white quartzite and vein quartz in a grit matrix, are also met with; but they are few. The group here is still, in the main, a slate series, some of the slate being pale and papery, of the “ pale-slate type.’ Iron pyrites is GEOLOGY OF CENTRAL WALES. 157 common in both grits and slates. The main dip is decidedly easterly. Mounting the higher ground above Nant y Moch the grits become more conspicuous, forming a terraced structure by their outcrops around the hill; and near the summit a rugged steplike structure is produced by the regularly bedded fine-grained grit series. The finest exposure of these Plynlimmon grits is in the magnificent pre- cipices fronting the Rheidol lake (Llyn Llygad Rheidol), where many of the beds are very thick (7-10 feet) and of moderately coarse grain. A particular feature in them is their regular columnar jointing; this is quite as regular as in many columnar felsites, numerous columns, 7 feet long, being seen running right across the beds. Conglomerates are uncommon; but some three or four beds of beautiful “pudding stone,” with white quartz pebbles in a fel- spathic sandstone matrix, are seen upon the western slope of the mountain. Coarser beds of this group are probably indicated by a large boulder upon the hill-top above Gogerddan House, near Aberystwyth, which has pebbles as big as cannon-balls. These grits form a crescent zone from north of Plynlimmon through the high ground of Mid Wales, running east of Tregaron and east of Lampeter; but they die out beyond this to the south. They constitute the genuine backbone of South Wales, lying in a synclinal. The highest beds form the upper ground of Plynlimmon itself. In the Teifi area the beds are remarkable for being coarsely eleayed into large tiles about 4 inch to 1 inch thick, a structure which graduates into jointing on the one hand and true slaty cleavage on the other. Similar cleaved grits occur around Llyn Fyrddyn and on the hills east of Lampeter, all of which I would place on the same horizon. There is, however, some doubt in my mind as to the exact position of this horizon in the geological series. Foldings.—As a group, the Plynlimmon grits are characterized, not by violent contortions or sharp bendings with fracture, but by broad and gentle foldings. These are beautifully seen around the Teifi lakelets and Llynodd Ieuan; also at Llyn Fyrddyn Fach and Llyn Fyrddyn Fawr, and Llyn Bugeilyn, where the ridges and minor | hog-backs of the surface correspond with the structural anticlinal and periclinal foldings of the rocks with great regularity and clearness. Amongst the rock-varieties we may notice blue shaly slates, blue slates in 6-inch beds, cleaved into oblong flags about 3 inch thick (Llynodd Ieuan), and arenaceous mudstone, broken up into coarse rab, the fragments measuring 1-3 feetx3 inchesx1 inch, also smaller rab. The lithological differences between this series and the Aberyst- wyth group are the greater thickness of the grit beds, the presence of quartz conglomerates, and the rarity of curious rock-surfaces and fucoidal markings. The slates also are, as a whole, of paler type. Fossils have not been discovered in the grit series of the Plyn- 158 WALTER KEEPING ON THE limmon mountain; but large boulders of coarse conglomerate upon Gogerddan Hill, Aberystwyth, which surely came originally from this mountain, contain casts of various fossils, of which I can men- tion crinoid ossicles, Nebulipora?, Petraia, a large cup coral, and Meristella. The thickness of this group must also be very great; for Plyn- limmon rises some 1500 feet above Garn Fach, the level where the Plynlimmon grits first appear. A thousand feet is probably an under-estimate. Part Il. Brief Notes upon other Sections. (1) From Aberystwyth through Pont Erwyd to near Builth (fig. 4),— From Aberstwyth to Pont Erwyd the section is similar to that from Aberystwyth to the Devil’s Bridge; but the western dips are more important towards Pont Erwyd, so that the inversions required are here smaller or less numerous. Beyond Pont Erwyd is a great slate and shale series running for many miles, nearly to Rhyader, and forming by the apparent dips an important synclinal under the highest ground south of Plynlimmon. Fossils occur at Pont Erwyd and Dyffryn Castell, also near Rhyader. The hard pale-slate series comes in at Goginan and east of Pont Erwyd, and there is a similar rock west of Rhyader; but the Plynlimmon grits do not actually appear as such, unless the grits of Llynodd Ieuan belong here; they are, however, well developed in the mass of Plynlimmon further north. Some of the paler slate along our line of section may represent this series. Grit beds similar to those of the Aberystwyth group are found in two places near the ‘Glansevern Arms,’ and again further on east of Langwrig. ‘They are to be regarded as special local de- velopments inthe Metalliferous Slates. Other grits and conglomerates seen at Gwastaden are probably the representatives of the Plyn- lmmon grits. (2) Section from Llandeilo to Aberaeron.—Three principal grit- areas are met with in the line of this section—namely, at Talieris (west of Landeilo), east of Lampeter, and at Aberaeron. The Talieris conglomerate beds are surmounted by dark shales and shaly slates of the Metalliferous types ; and beds of the Rhyader pale-slate type next underlie the grits of the country east of Lampeter. By an exceptional appearance the Metalliferous slates of the west seem to underlie the Aberaeron grits. No pale slates are seen beneath the ~ Talieris conglomerate beds. The principal axial fold in the section seems to be a synclinal in the Teifi valley by Lampeter—the mountains to the east and west of this being (that is, if the rock- dips have any truth in them) great anticlinals in the Metalliferous- slate and Rhyader Pale-slate series. The beds of grit and con- glomerate at Talieris are not of the nature of basement beds of a stratigraphical group, but indicate nothing more than such slight physical variations as a shallowing of the sea-bed or a change in the direction of the currents, resulting in the formation of sand and pebble banks ; for the pebbles are not fragments of the underlying rocks, and there is no trace of any physical break or even change of lithological character above and below them. 159 GEOLOGY OF CENTRAL WALES, "SOJEIS OV "PF ‘oyvys ATOATYS pure ‘ssvy pur sqias uy, *z “SOIVIOTLO[SMOD PUL SII *G “SOTVYS PUL SeqVIS B[eq “ ‘goye[S pu ‘soreys ‘s}las yaeq *T ¢ F | g I , i NON esses 20/20 Sa S J 4, RENE Ze : Ss oe bn 5 ce Q “ropedyy ‘jeyon | “UCT WMO ‘qo1o £ went gE *Lapeg Jury "TIeT, WOAIOeO "B[00g FOE Bao Sopvyueyy f Patel fee . S m "a e “M Copu 8 07 TOUT + “oTVos TeyUOZTIOF,) “uapmisvmy pun sapohyar 07 spoog afiaT 22 pun nprwopst VID) Wouf W0LNDI2gG—'G *SIyT ‘jeurpouds tomutpudTg °9 "SOIVIOULO[SUDD) “F "A1ZUNOD P219A00-JJIICL *Z ¢ SUOISIOAUT “9 "SOJLS OMT ‘ “S}IAG UIY} MOF V YIM ‘SOTLOS O}VIS-SNOIOFIT[VIONL *T SS Soxt AD D SN { J ! I 1 ‘ ! 1 t i x "Stamp VF pperg = Sos nfs Topryseay ‘ropesyyy -SramBavyy *[00 *ppAMAFT 40 “O'S$<—ax PpAMIT FUOg m—>"M'N ‘<—as ep—>" M ‘wapoliyy puw biumbunyy 07 svn vomunuliyg oyn ybnouy phnag wwog wort wy0y—'F “Stir 160 WALTER KEEPING ON THE (3) Rhyader to the Teift Pools (fig. 5)—In this section three main folds are indicated, namely a central anticlinal fold at Disgwylfa in the Rhyader pale slates, with a synclinal on its east and west. The former is a comparatively simple basin of pale slates beneath the grits and conglomerates of Cwm Elan and Gwastaden; the latter, corresponding with the Plynlimmon synclinal, is a much more complicated structure, pro- bably much affected by reversals. General Result of the Hast and West Traverses (see fig. 6).—The sections just indicated show that the great area of Mid Wales is made up of a thick series of im- perfect slates, pale slates, shales and grits, having a general re-. semblance and intimate connexion throughout, as of one continuous group, but divisible into the sub- groups indicated in the beginning of this paper. These rocks form one great pri- mary synclinal extending from the Aberystwyth axis to the east of Rhyader, and from the west of Llandeilo to Aberaeron. For in the west of Cardiganshire we find a great and continuous course of easterly dips running inwards from the coast; and in the eastern half of the area, in Radnorshire and Caermarthenshire, is a correspond- ing grand set of westerly dips. Subordinate but still very great anticlinal folds along north and south axes bring up on the east tee Metalliferous slates between Plynlimmon and Rhyader, and on the west the upper part of the Aberystwyth grits, in the line of Pont Erwyd and the Devil’s Bridge. Another, smaller fold with the same result as the last, runs from Ystrad Meurig north- wards towards Machynlleth, thus Rhyader. 3. Rhyader pale slates. f. Plynlimmon grits Plynlimmon. rous-slate group. Fig. 6.—Diagram of the General Structure of Central Wales. yth grits. 1. Aberystw Metallife 2. Aberystwyth. { GEOLOGY OF CENTRAL WALES, 161 greatly swelling out the surface-area of the Metalliferous-slate series. ~ Minor foldings, also with north-and-south axes are infinitely abundant over the whole district, with frequent inversions which obscure the original order of the rock-beds. Included in these foldings are four principal crescent zones of grit, greywacke, and conglomerate, namely :—1Ist, the western or Aberystwyth grits; 2nd, the eastern grits—an imperfect line of grits seen at Talieris, west of Llandeilo, and perhaps continued in other grits Kast of Gwastaden. These I regard as the diminutive representatives of the Aberystwyth grits. 3rd. The Plynlimmon grits and conglomerates, with much associated pale slate; and Ath, the grits and conglomerates of Cwm Elan and Gwastaden, which are also associated with, and, indeed, included in, a great pale-slate group, the Rhyader pale slates. The last two are upon the same stratigraphical parallel, occupying synclinals. These grits I do not regard as necessarily holding exactly the same horizon over wide areas, they having been probably shingle banks over the old sea-bottom. The pale slates seen in so many sections appear to belong to two principal horizons—(a) the lower or Strata Florida pale slates in the lower part of the Metalliferous-slate group, and (6) the Rhyader pale slates, associated with the Cwm Elen, Gwastaden, and Plynlimmon grits. More distant Sections (see Table of Vertical Sections, p. 164). In order to work out the relations of these rocks to the other lower paleeozoic groups, I have visited various districts where the Cardiganshire series might be studied in association with rocks of some definite and decided horizon—namely, amongst other places, Llanbrynmaer, Dinas Mowddwy, Cardigan, N. Wales, &c. In the Llanbrynmaer area we find (1) an upper series of grey grit and greywacke, often very felspathic and micaceous, cleaved, and with fragments of fossils (the Denbighshire grits). These pass down by gradual transition into (2) a series of pale blue and green, grey and purple shales, with some darker shale, rab, and green grits, which make up the true Tarannon shales; some obscure fragments of Graptolites have been found here. Again, there is no break, but a simple passage between this series and (3) the lower series of grits, greywacke, and dark shales which belong to the Cardiganshire type, and contain the fossils of our Metalliferous group. To the west of the Tarannon plateau we find the Metalliferous slates fully developed in the Pennant valley, with the characteristic fossil species, and the usual associated lead-mines of that group. The whole character of this Metalliferous slate is different from that of the Tarannon shale; and our evidence points uniformly and decisively to the Metalliferous slate being a more ancient series, the two groups being separated in this area by a zone of grits. The Plynlimmon grits and associated pale slates, as seen in Plynlimmon and around Rhyader, appear to be wanting in this area, there being no well-developed grit series between the Llandovery group and the 162 WALTER KEEPING ON THE Denbighshire grits. We shall find reason to believe that they are. represented, feebly it is true, by the thin grit-beds above described in the Tarannon shale. The Dovey Valley : Section across the valley near Llany Mowddwy (fig. 7)—This section gives, in a narrow area (measuring less than 2 miles across), the complete series from the Lower Bala slates tothe Denbighshire grits. The Denbighshire grits pass down into the Tarannon shales, which are pale-slate rocks resembling the Rhyader pale slates. There is, again, no break between these and the series of greywackes and dark shales and slates beneath. ‘This latter group (which is in part the Lower Llandovery of the Survey) agrees both in its lithological characters and in its fossils, so far as known, with the Aberystwyth grits. There is no distinct basement bed of conglomerate or grit to this series, such as might indicate the existence of an important strati- graphical break ; but, on the contrary, there is every appearance of a passage from the underlying great slate group up into the true Silurian series. This great slate group corresponds in its general lithoisgical characters and its “ cone-in-cone” nodules with our Metalliferous slates; but I have not detected any pale slate in them. In the absence of fossils, it is not possible to determine their exact age; but they should probably be regarded as belonging for the most part to the Upper Bala slate group. I should, however, expect some of its upper beds to be the representatives of our Cardiganshire Group, the lower part (Upper Bala) corresponding with much of the slate group between Cardigan and Llangrannog, about to be de- scribed. It is further noteworthy that im this line of section, as in the Cardiganshire Group, the newer rocks are much better cleaved, and appear altogether more highly metamorphosed than the more ancient rocks to the west. Corris Avca.—The great slate district of Corris, usually referred to the Llandeilo age, belongs to our Metalliferous-slate group, as is proved by the following fossils from Corris and Taren y Gesail :— Monograptus Sedgewickii, Portl.? Climacograpsus scalaris, His, —- tenuis, Portl.? Orthoceras, sp. (same as the Cefn- Hendre shell). The Cardigan District.—The pale felspathic grits and black slates of Newport Bay and Cardigan, hitherto placed in the Lower Llan- dovery series “‘b 4,” are not of this age, but belong to the Middle Bala or Caradoc Group ; as to this the fossil evidence is conclusive. Above these come rolling beds of pale slates and shales, then darker shaly slates, with a zone of pale felspathic grits; these also I refer to the Bala period. The overlying shaly slates and rab are passage- beds of Caradoc-Llandovery age, presenting the gradual incoming of the Aberystwyth grits. Reviewing the several rock-groups and their distribution, we find:— (A) The Plynlimmon grits, seen upon Plynlimmon and around Rhyader, also probably in the hills west of Tregaron and Lampeter ; 163 GEOLOGY OF CENTRAL WALES. Dark slates (Lower Bala). Bala ash-bed. Bala limestone. - Bala phosphate-bed. Calcareous slates and flagey shales. ig. 7.—Section through the Dovey Valley at Llan y Mowddwy. (Horizontal scale, about 23 inches to the mile.) Lian y Mowddwy. Cwm Maenydd. Waterfall. 6. Imperfect slates, dark shaly and 9. Hard pale slates, banded slates, well rubbly slates and shales. cleaved (‘Tarannon group). 7. Dark-coloured irregular slates, with 10. Dark rab and coarse gritty rab, and “‘eone-in-cone” nodules. grit bands (Passage-beds). 8. Grits, dark shaly slates, and rab. 11. Denbighshire grits, with dark rab Fossils (Lower Llandovery). and slates. 164 | WALTER KEEPING ON THE but they die out to the north and south of these localities. No fossils have yet been detected in this series in Cardiganshire, except, perhaps, in the boulder-block at Gogerddan, near Aberystwyth. Stratigraphically they lie above the Metalliferous slates; and these latter are proved, in the Llanbrynmaer, Dovey, and other areas, to be inferior to the Tarannon shales. Therefore, with reference to the Noedd Grug section, they must either represent the uppermost part of the Llandovery rocks or a special gritty development in the Taran- non shales. I adopt the latter view for the following reasons :— because of (1) their dissimilarity to the grits (Llandovery grits) which underlie the Tarannon shales at Llanbrynmaer, Llan y Mowddwy, and elsewhere ; (2) their association with pale slate rocks similar to the Tarannon shales of the Corwen area; and (3) the oc- currence of such a group of rocks, the Gala grits*, in this position in the south of Scotland. Moreover, in the Tarannon district the first appearance of such a development of grits in the Tarannon shale is already seen in the numerous thin grits, with contorted structure, of Tarannon Hill, Llanbrynmaer. The Rhyader pale slates thus appear also to belong to the true pale slate or Tarannon- shale series. (B) The Metalliferous slates, which are enormously developed, spreading over a very wide area of Mid Wales. ‘They maintain one general lithological character of hard, pale, shaly slates, also containing a zone of pale slates. The group is generally plumbiferous. (C) The Aberystwyth grits may be taken as an arenaceous de- velopment of the Metalliferous slates, in its lower part. Like the Plynlimmon grits they die out to the north and south, their southerly attenuation being well exhibited around Llangrannog. They dip persistently under the Metalliferous slates ; and the truth of this dip is proved by the position of the contorted raised structures upon the under surfaces of the gritst. In the Dovey valley the Aberystwyth Grits are represented by eritty beds of a similar character; but here they appear to belong to a slightly different horizon, namely to the upper part of the Metalliferous slates; and the same is true of the grits exposed in the deep Talerddig cutting, Llanbrynmaer. The grits and conglome- rates of Talieris, west of Llandeilo, very probably belong to the horizon of the Aberystwyth beds. In the Noedd-Grug section these beds, together with the Metalliferous slate, are represented by the Llandovery group of the Geological Survey, probably the Upper and part of the Lower Llandovery ; but from the fossil evidence we can- not recognize the lowest part of the Llandovery Group (Lower Birk- hill) in Central or West Cardiganshire. Part of the lower set of slates beneath the grits, described in the Cardigan and Dovey areas, * Mr. Lapworth, F.G.S., writes me that “all the paleontological affinities of the Gala beds are with the Tarannon shales, with which, and not with the Upper Birkhill beds, they must eventually be connected.” + Ihave confirmed this test by examining the beds of Lingula-flags, Trema- doc and Bala rocks of South Wales, where these curious worm-like, fucoidal, and irregular prominences are found uniformly upon the wader surfaces of the grits. les and othe 6. Abery, and Plyn Pale |. Shales, | ~~ Mudstone Group. Bala Group. || a ae a we wit. ore Fr 2 South Scotland (Lspworth). 3. South-west Cardiganshire. Comparative Vertical Sections of Silurian Rocks in Central Wales and other Localities. 5. West of Llandeilo, 6. Aborystwyth 4, Rhyader, Plynlimmon | Grits. | Grits. alee peers Pale Slates. Metalliferous Slates. MUM WOME CMM TT) WMUMLLA Aberystwyth [= Gnits. ee MM PTL LYM | HMMM WW and Plynlimmon. 7. Lianbrynmaer, Denbighshire Grits. Tarannon Shales. (AF LZ V7 8, Llan y Mowddwy, | ie Bala - Limestone. [Vo fuce puge 164, 9, Clwydd Valley (Prof, Hughes). Corwen === Grits. | GEOLOGY OF CENTRAL WALES. 165 probably belong here and complete the series ; for we have seen there is no break of continuity down into the Bala formation. Part III. Paleontological Evidence. Although the order of succession of the forms of life have in all eases been first determined from the results of stratigraphical re- search, yet, that order once established, the organic remains, as considered in groups, become the supreme test of the age and rela- tive order of rock-masses—superseding, and in some cases actually overthrowing, the apparent evidence of succession in the rock-beds. Such we find to be the case over a large area of Cardiganshire, where, as in the section from Aberystwyth to the Devil’s Bridge, the stratigraphical series, seeming to be continuously ascending through some three or four miles of thickness to higher and higher beds, is proved by a handful of Graptolites to be radically misleading ; for, the fossils being identical at almost the extreme ends of the section, the true reading is shown to be a series of inversions, by which the original order of the beds is obscured. Thus, over a large part of Mid Wales, it is to the organic remains alone that we can look with confidence for unravelling the apparently inextricable, and often delusive, tanglement of the strata, and establishing their true succession. The most important element in our paleontological evidence is the group of Graptolites, these fossils having now become, through the most successful work of Mr. C. Lapworth, F.G.S., of the highest value in Cambrian and Silurian geology. All my work in this branch of study is based upon the published results of that geologist : and I am also particularly indebted to him for examining many of my Graptolites, as indicated in the Table (p. 170), and for valu- able notes upon them conveyed to me in letters, from which I have quoted below. A study of the Table brings out a number of important facts. At the first glance it is clear that all our fossil lists exhibit one and the same general geological fauna. Our richest localities are Cwm Symlog (eight or nine miles east of Aberystwyth) and Morben (near Machynlleth), also Cefn Hendre, Cwm, and Bryn y Carnauin the neighbourhood of Aberystwyth). A single one of these localities (Cwm Symlog) yields every species, except four, of the true Graptolites known in Cardiganshire ; and this, together with the intimately allied fauna of Morben Quarry, includes all our spe- cies except one (Monograptus turriculatus). The other forms of life show no less clearly the unity of our fossiliferous Cardiganshire rocks. The Orthoceras is wide-spread in the Metalliferous Slates, and occurs also amidst the Greywacke- Flags of Cefn Hendre; and the various forms of Alge and worm- trails are also characteristic of the whole range of the Cardiganshire group. For the typical area of reference for Graptolitic deposits we must unquestionably look to the rich beds in the south of Scotland, of Llandeilo to Wenlock age, which have been so thoroughly worked Q. J.G.S. No. 146. N 166 WALTER KEEPING ON THE out by Mr. Lapworth. And here the reading is decisive: our Cardi- ganshire series belongs not to the Glenkiln (Llandeilo) or the Hartfell (Bala) periods, when the Didymograpti and Diplograpti flourished, but to that part of the newer or Silurian era when the Monograptide were predominant, and in their greatest development, namely the Llandovery period (Birkhill Shales). The more detailed affinities of our fauna will appear by a study of the last five columns of the ‘‘ Table of Distribution.” Every one of our species of true Graptolitide (excluding the Cladophora) occurs in the Upper Birkhill beds of Scotland. Eight species are common to the Lower Birkhill Shales, and nine (as also two species of Cla- dophora) to the overlying Gala and Tarannon groups. Of the fossils of the two principal Graptolitic localities, Cwm Symlog and Morben, Mr. Lapworth writes that they ‘‘lie within the same general Graptolitic zone.” .. . “‘ Of the nine species recognized in the slates of Morben, five (namely Rastrites maximus, Carr., Monograpius spinigerus, Nich., MW. distans, Portl., M. Hisingere, Carr., and Diplograpsus Hughesiz, Nich.) make their first appearance in the Upper Birkhill Shales of 8. Scotland. astrites maximus is confined to the highest zone of that formation, in the Moffat area, and, together with its common associate, Dipl. Hughesii, seems to have become extinct before the deposition of the overlying beds of the Gala group. “¢ The second fossil locality (Cwm Symlog) is even more strikingly marked by its Upper Birkhill Graptolites. Sixteen forms have been identified from this spot. Of these, seven species (viz. Monograptus crenularis, Lapw., MV. pir Ge Carr., M. Clingani, Carr., MW. runcinatus, Lapw., M@. Hisingeri, Carr., Dipl. Hughes, Nich., and D. sinuatus, Nich.) are forms which are known for the first time in the Upper Birkhill of S. Scotland, and its equivalents in Girvan, Treland, and the north of England. J. intermedius, M. runcinatus, and M. crenularis are peculiar to the Upper Birkhill, as also is M1. Clingani, which is confined to a small seam in the very centre of the Birkhill beds.” Of the nine species of true Graptolites common to our Mid-Wales rocks and the Gala and Tarannon, not one is a special Tarannon form. Monograptus turriculatus, however, is most frequent in those beds; and the two species of Cladophora Dictyonema venustum, Lapw., and Rhizograptus ramosus, Lapw., are only known, else- where, in the Gala group (Scotland). From the fossil evidence, therefore, there can be no hesitation in referring our Mid-Wales rocks to the same age as the Upper Birk- hill of Scotland; and beyond the occurrence of the three species just mentioned in the rocks of the Devil’s Bridge, there is nothing that conveys the slightest hint that any of our strata are newer than Upper Birkhill. Lake District.—Fifteen of the Mid-Wales Graptolites are known from the Coniston or Graptolitie Mudstones of the Lake District; so that, in Mr. Lapworth’s ,words, ‘‘ the general facies of this Mid- Wales fauna is distinctly that of the Coniston Mudstones.” These GEOLOGY OF CENTRAL WALES. 167 species are nearly equally distributed in the two zones (tenwis- and argenteus-zones) of that area. Also, amongst the other fossils, Mr. Marr recognizes the fragment (tail) of a Phacops, found by him at the Devil’s Bridge, as a new species also occurring in the “ Mud- stones” of the Lake District. Lithologically we find some of the Graptolite-bearing beds not unlike part of the Coniston Mudstone series; and some paler shales and slates, associated with the Graptolite beds at Morben, Machynlleth, and occurring again at Cwm Symlog, have their repre- sentatives also in the same series. Still, regarding them as a whole, our Cardiganshire rocks decidedly do not resemble their representa- tives in the English lake country. We have seen that our verious fossil lists indicate that all the fossiliferous beds belong to one general horizon. Certain minor differences, however, are to be observed, of which the most important are the presence of Monograptus turriculatus at Cefn Hendre and the Devil’s Bridge, and the species of Cladograpt: of the latter place, which are common to the Gala group of Scotland. From the Graptolitic evidence Mr. Lapworth would arrange the beds in the following order :— D. The M. turriculatus beds of the Devil’s Bridge. C. Next below (or perhaps identical with (D)) the Cefn Hendre erits. B. Then comes the rich Graptolitic zone of Morben and Cwm Symlog. A. Lowest of all lie jointly the Diplograpsus beds of Taren y Gesail, Corris, and Steddfa Gurig. It will, however, be seen that this order does not correspond with our well-established succession of the greater groups as given in the first pages of this :paper; for it would make our Metalliferous slates newer than the Aberystwyth grits. Now, with all my general confidence in fossil evidence, | consider that the paleontological data are here insufficient to maintain this order against the very strong stratigraphical evidence to the contrary., The principal fact is the occurrence of a single species (Monograptus turriculatus) in certain localities ; and this does occur, only less commonly, in the Birkhill © shales as well as the Gala group. Moreover the similar (more arenaceous) physical conditions common to our Devil’s-Bridge bed and the Gala group may have to do with their having some fossils im common. Mr. Lapworth’s researches have established the fact that, in the Sulurian and Cambrian periods, zones of Graptolites characterize defi- nite stages, just as do the Ammonites in the Jurassic rocks. But these latter have their exceptions, and the minor zones are only trust- worthy within local limits; much more so should we expect this to be the case with the more lowly organized group of Graptolitide. The occurrence of Monogr. colonus with Birkhill species in N. Wales is, indeed, an example of such. On the other hand the stratigra- phical succession of the Metalliferous slates over the Aberystwyth nN 2 168 WALTER KEEPING ON THE erits is clearly marked over a large area, extending from near Borth, through Aberystwyth, and southwards to Llangrannog*, a gradual transition between the two groups being everywhereseen. And that the rocks here occupy their original relative positions is proved by the position of the raised rock-markings, which are uniformly found upon the wnder surfaces, as is normal in the lower paleozoic rocks, whereas when the beds are inverted these structures are found upon the upper surface. The only plan of exact harmonization would be, it seems to me, to suppose that the Aberystwyth grits had died out to the east before we reach the fossiliferous beds of Morben and Cwm Symlog, — and older beds had been somehow brought up in the confusion of folds and synclinals. This may be considered possible; but it again increases our difficulties in the eastern area (Devil’s Bridge &c.), where the usual order of the strata persists. I therefore adhere to my original classification as given in the earlier pages of this work; for, indeed, the paleontological difficulty is very insig- nificant. Comparing the faunas of the Aberystwyth grits and Metalli- ferous slates, we find that these do not afford any important data for their separation, but rather demonstrate their intimate connexion. ‘Twelve of our species of Graptolites are known only from the Metalliferous slates—namely Rastrites maximus, R. pere- grinus, Monograptus spinigerus, cyphus, gregarius, intermedius, runcinatus, and mvolutus, and all the Diplograpt:: but this signi- fies probably little more than the less favourable conditions for Graptolitic life in the Aberystwyth grits; for the two rock-groups, are bound together by the common possession of Monograptus dis- tans, Clingam, Sedgewickir, crenularis, spiralis, lobiferus, Hisingeri, and tenuis, and Climacograpsus scalaris. The Orthoceras, Nereites, Fan-Alge (Buthotrephis), and Nematolites(?) tubularis are also common to the two groups, and serve further to demonstrate the paleontological unity of the Aberystwyth and Metalliferous series. But there is another group (the Plynlimmon group) whose exact age has not yet been fully considered. ‘These beds contain no fossil evidence in themselves; but still, from their relations to the under- lying metalliferous slates and to the more eastern sections, there is little room for hesitation as to their stratigraphical position. The underlying rocks are shown to belong to the highest part of the Lower Llandovery group, or Upper Birkhill series, some of the species being actually of Gala types. Paralleling, then, our rocks with those of the south of Scotland, this overlying group of grits must either be close upon or actually in the Gala and Tarannon group. We have already seen how the stratigraphical and lithological con- siderations support the view that they really are an arenaceous development of the lower part of the Tarannon shale, thus being the true representatives of the Gala group of Scotland. The asso- ciated pale slates of Plynlimmon and around Rhyader (Rhyader pale * There is a difficulty, as before mentioned, at Aberaeron, which I have not been able thoroughly to investigate. : GEOLOGY OF CENTRAL WALKS. 169 slates) thus fall into line with the great pale-slate group of Britain —the Tarannon group. N.E. Wales.—The Silurian rocks of Corwen, Llangollen, and the Vale of Clwydd have been described by Prof. T. M‘K. Hughes * ; and fossils have been collected from the Tarannon shales of the Conway area by Mr. C. Lapworth, F.G.S. Those species which are common to our Cardigan rocks are indicated in the Table; but, besides these, other species, characteristic of newer groups, are found—namely fetrolites, and Monograptus priodon, colonus, and galaensis. From these fossils Mr. Lapworth concludes that the Conway beds “ correspond with the earlier portion of the Gala beds” of Scotland. Thus the results of paleontology correspond well with the stratigraphical evidences, which show, in my opinion, that the great mass of our Cardiganshire series is unrepresented in the N.E. of Wales (as in the English border counties), or rather that it is represented by the break which Professor Hughes has worked out in that district at the base of his Corwen Grit. The black bands in the Clwydd valley, from which Prof. Hughes obtained Graptolites, do appear to lie in the parallel of the higher part of our Metalliferous group. My friend Mr. J. HK. Marr has just published the details of the sections at Cerrig y Druidion, N. of Balat. He there finds beneath the Tarannon shales the equivalent of the Graptolitic mudstones, containing fossils of the same species as our Cardiganshire series, but also including Monograptus colonus, which is characteristic of much higher beds. ‘The section appears to show a greater deve- lopment of Graptolitic “‘ mudstones ” than in the Clwydd valley, and so far to represent more fully our Cardiganshire group; but here, again, as at Corwen, the series appears to be incomplete, because of the existence of the Silurian unconformity. General Summary. Central and West-central Wales is made up almost entirely of a great series of imperfect slates and Greywackes belonging to our Cardiganshire group, together with the overlying pale slates and grits of Rhyader and Plynlimmon. ‘The Cardiganshire group is sub- divided into the (1) Aberystwyth Grits, and (2) Metalliferous Slates ; and part of the underlying slates may, perhaps, hereafter be proved to belong to the same group. Some minor subdivisions are also distinguishable. The arenaceous rocks are not constant over large areas, but die out both to north and south. The rock-beds are astonishingly folded into violent contortions, with frequent inversions, especially in the Metalliferous series, so as often to produce the misleading appearance of a regular and continuous ascending series exceeding five miles in thickness, All the important axes of elevation in the country havea common N. and S. direction, two of the main folds being the Aberystwyth anticlinal * Quart. Journ. Geol. Soc. vol. xxxiii. p. 207, vol. xxxv. p. 694. t Ann. Nat. Hist. 1880, vol. ix. p. 49. $ Quart. Journ. Geol. Soe. vol. xxxvi. p. 277. 170 ON THE GEOLOGY OF CENTRAL WALES. and the Plynlimmon synclinal. Secondary axes of upheaval bring up the lower beds of the series at the Devil’s Bridge and near Cwm Symlog. Innumerable minor foldings preserve the same north-and- south strike. The included fossil remains, especially the Graptolites, prove the — Aberystwyth Grits and Metalliferous Slates to belong to the same general geological horizon—namely, on the parallel of the upper Birkhill series of 8. Scotland and the Coniston “‘ Mudstones” of the English Lake district. The Plynlimmon Grits are probably an arenaceous development of the Tarannon Shales; and the Cwm-Hlan conglomerates and Rhyader Pale Slates belong to the same series. Following up our Cardiganshire series to the lines of junction with older and newer groups in parts of Montgomeryshire, South Cardiganshire, Caermarthenshire, and Merioneth, we find no evidence of a break in any part of the rock-groups; but there 1s concordant evidence of lithological passage from the Bala to the Llandovery groups, and from these upwards into the Tarannon Shales and the Denbighshire Grit series. On the contrary, in Denbighshire and N.E. Merioneth, Professor Hughes has shown that, just as in the English-Welsh border districts, there is an important stratigraphical break at about this Llandovery period. Now the beds below that break in N. Wales are the Bala group, i. e. inferior to our Cardigan- shire series; those above it are basement-grits and conglomerates (the Corwen Grit), covered by the pale slates, with some black Graptolitic bands. These latter contain some species In common with the Cardiganshire group, but also others characteristic of a higher horizon than our fossiliferous beds. Thus all the facts are harmonious in pointing to a continuity of the Silurian and Cambrian deposits in West and Central Wales, while an important break exists in the east and north-east. Our Cardiganshire group is only partially developed (its upper part) in the latter district, being, in fact, there represented by a great strati- graphical break—Sedgwick’s original May-Hill unconformity. I conclude that, while in the latest Cambrian times(Sedgw.) and the dawn of the Silurian era the elevatory forces, acting in a north-and- south direction, lifted up the sea-bed to form a land-surface over the west of England and the Welsh borders, these forces influenced the greater part of Wales only in a less degree—producing, it may be, the shallower water in which the Aberystwyth grits were laid down, but not interfering with the continuous deposition of sediment and the unbroken sequence of the geological record from the Cambrian to the Silurian eras. te ceee . : N Lance showing Distribution of Fossils in Central Wales. [fo fitce page 170. Specimens marked thus (+) have been identified by Mr. O. Lapworth, F.G.S, ee : Scotland. Goniel | 4 $ 7 E 3 5 : o a 2 ack , es = = loniston a =) =) . = BS |e sc bb gb gE 2 = F Mudstones. | = SS SE eS Me Sle | ES d S/ = |da/ 2/3 |ae| 2 | 2 ae = (=) y s = i el =| BD 5 iS) is} & Er Gi |f dg a ] fs = = f is} e |Ce § 5 a: S 2 [2] A 5 3 6 Sst | el is] Se ee ep eta 2 Bs SRS elle eee a a. WE Ale SS eee ee hee 2 (ee |e ca |e E/E) e] 2 lee} | & [te] 6 | 2] es | Bele 3 o S$ = 5 2A gy] SSPE ES ads (Euston ae | ae (ee 8 |e ee ee ee se ie = we eres pi saesee: [eeeeee fiimesene [i CHES {| SE eeseee [ii seuwen || cuncne | cvauen f susece | savese | (senses | srene * * * mf | feces iceewee | Pf YP ceecce | ceceee | | wevene | ceeeee fi ceeeee | ccceee | ceceen | ceesee | teneee | cesses * S| eee | aaa | A eseeadl|eseeeod lessees Sef I cee leanne |i eecea eeeno lll Seeeee [ic osewn) || wecurn | vecees ||| eeene= | uunmac'G ccvses | eevee | secone | cevene * 1} Be * ; J rere) * MH cecece Ho] wacees * MO oceeece * * * 2 = * . * | « | * | HO osneene * . en l(encceceul| Gastipe Hf ewaee | cervee * * * el | ee ee eee . an eter Ts * 2 * * * * * * * . — << i * ee ‘ conece: | herrecoe |) ceooce | acetce: |) atacos 1} cocece ? Nemeriies Obvantii, Murch, > * Tertoerrs. Calymene. sp. coerey ¥ | | Bs ass |. serene Wiesarecd|it-cssmillsrenee | teraction | [tensor || tsaseas oh lWhecerxo * oP ecco * esacee toe A peek | recap Woseree 2 WN eeeceadd| nccee \Weeeepe | ecacce || cone |) Goch * | Gala Group. Z * Confined to the highest 1 nari Pageant Dae pata qu she Br Birkhill of South Scotland eae ° A few fraginents only in the highest seams of U, per Birkhill (Lapw.), 2 Known only in ee Birkhill bed Shale and its equivalents (Lapw,), ° M, runcinatus is peculiar to Upper Birkhill Shales Lapw,). * Restricted to a single zone in the Te Moffat and Girvan (Lapw.), 7 Makes its first appearance in middle of U: per Birkhill (Lapw.) oper Birkhill Shales (Lapu.). * Diployrapsus Hughesié is confined to the Upper Birkhill Shales (Lapw.). Quart. Jour. Geol. aha Volo ea ae Mintern Bros. imp. INES Hoomdeliti WELSH CLADOPHORA. ON SOME NEW SPECIES OF CLADOPHORA. 171 APPENDIX. On the Cravopnora (Hopk.) or Drenproip GRaprorirss collected by Professor Kuxpine in the Llandovery Rocks of Mid Wales. By Cuas. Lapwortu, Esq., F.G.8. &e. [Puate VII.| Tse forms of Cladophora collected by Professor Keeping from the Llandovery rocks of Cardiganshire are, regarded collectively, of a type almost new to British paleontology. “Although intimately allied to the well-known dendroid species of the Quebec and Arenig formation, they are very distinct in their minor features. They are essentially of a Silurian (Upper) facies, and they remind us strongly of a group recently made known to us through the researches of Mr. Spencer, of Toronto, which characterizes the Clinton and Niagara Groups of New York and Upper Canada *. None of our examples shows the complete polypary, or affords dis- tinct proof of the presence of all the more minute classificatory features ; but there is satisfactory evidence of the presence of at least four distinct genera and seven species. The genera repre- sented are Dictyonema, Hall, Calyptograptus, Spencer, -Acantho- graptus, Spencer, and Odontocaulis, Lapw. Dictyonema is a well- known British genus; the remainder are new to British paleon- tology. Calyptograptus and . My observations would lead me to limit the attachment of Paleocoryne to one species of Menestella, viz. Fenestella nodulosa, and to the pore-face generally, rather than the margin of the polyzoarium. ‘The frequency with which I noticed this association of Palwocoryne with Fenestella nodulosa, led me to go carefully over my collection, and ascertain definitely the particular species of enestella with which it was most fre- quently allied. The result was, that, out of ninety-seven specimens * Hixplan. Sh. 23, Geol. Surv. Scotland, p. 96. Tt Quart. Journ. Geol. Soc. vol. xxix. p. 413. THE CARBONIFEROUS FENESTELLIDZA. 185 of Paleocoryne, eighty-five are either attached to the polypite- face of Fenestella nodulosa or associated with it, while twelve only are free and unattached. In no one instance is there a sus- picion that the species is any other than the one mentioned. Simi- lar evidence to this I get in another way. I have more than one hundred specimens of /. nodulosa from the shales of Halkin Mountain, but out of them not one example showing the poriferous face ; all without exception present to view the reverse side. This I explain by supposing that, as it is on the poriferous face of Penes- tella that Palcocoryne finds its seat, that face has in consequence held the more firmly of the two to the shale. ‘This is really so; for the base of it may be detected by a slight bulging of the polyzoary of the enestella, and its presence demonstrated by removing a portion. There is no similar difficulty in obtaining the obverse face of the other species from the same beds; nor do the other species present the appearance on the reverse face which I have noticed in /’. nodulosa. ‘The concurrence of these two species may be accidental; but I scarcely think so; the extent of its occurrence is not in favour of that idea; so that I think we may conclude that we have good evidence that Palwocoryne in the upper beds of the Mountain Limestone is exclusively confined to the polyzoary of Fenestella nodulosa, Phill. I may mention that I have found Palewocoryne associated with . nodulosa in the middle beds of the Mountain-Limestone series of North Wales, and also directly seated upon the poriferous face of the same from the Calciferous Sandstones of Scotland. I have previously pointed out that several inferior organisms are parasitic upon or incrust the polyzoaria of /e- nestella, from Silurian times upwards—such as Hemitrypa, Aulopora, Alveolites, and Diastopora ; and now Palcocoryne has to be added to the list. In the case of the previous incrustations there is no doubt that the parallel branches of the Menestella afforded suitable base- lines for the attachment of the incrusting coral. Whether there was any thing more than this in the preference shown by Palwo- coryne, is one of the problems to be worked out; and it will, IL have no doubt, receive due attention at the hands of Prof. Martin Duncan, to whom I have handed over my specimens of Palcocoryne for further elucidation. FENESTELLA PoLYPORATA, Phill. Geol. Yorks. pl. 1. figs. 19, 20. Fenestella multiporata, M‘Coy, Syn. Carb. Foss. Ivel. pl. xxvii. fig. 9. intermedia, Prout, Trans. Acad. St. Louis, p. 231. variabilis, Prout, Trans. Acad. St. Louis. p. 231. Sp. char.—Polyzoarium foliaceous at first, arising from a stem, ultimately becoming a flat circular network. Jnterstices large, broad, rounded, keeled. Dissepiments thin, one third the thick- ness of the interstice, somewhat irregularly placed, not expanding at junction with interstice. enestrules large and elongated, three times as long as wide. Four fenestrules in the space of two lines 186 G. W. SHRUBSOLE—FURTHER NOTES ON measured transversely, and two fenestrules in the space of two lines longitudinally. cel rounded and well developed, marked along its course with numerous spiny processes. ores small and round, their own diameter apart, prominent when preserved, from five to nine in the length of a fenestrule. Obs. This is the largest species of Carboniferous Lenestella, as regards the size of its interstice and fenestrules, although not so as to the ultimate growth of its polyzoary, which is considerably less than that of F. plebea, M‘Coy. This peculiarity at once dis- tinguishes it from any other species. It is subject to considerable variation, and will be met with both larger and smaller than the one described. Prof. M‘Coy, in his arrangement of the Fenestelle, assigned the smaller type to Phillips’s /. polyporata, and the type with the larger development and greater number of pores he de- scribed as I”. multiporata; but as both conform so well to the type in other respects, there is no reason for this division. This species often, on the reverse, attains to the size and character of some of the Polypore, from which it may readily be known by having only a double row of pore-cells. Lenestella polyporata never occurs very freely in any locality, but seems generally distributed throughout the Carboniferous strata. FryEsrecta crassa, M‘Coy, Syn. Carb. Foss. Irel. pl. xxix. fig. 1. Fenestella laxa, Phill. Geol. Yorks. pl. 1. figs. 26-30. My previous notice of this species was limited to the extent of showing the identity in character between the species respectively described by Phillips and Prof. M‘Coy as VFenestella laxa and Fenestella crassa. Since then a wider acquaintance with the Car- boniferous Polyzoa, and more especially with the Irish species in my possession, has caused me to hesitate about including it among the Fenestelle, since I have good reason for believing that the frag- ment which has been described as such will be ultimately found to belong to another fenestrate genus of Polyzoa. The original drawing of this species by Phillips clearly included two species of Polyzoa—one a Polypora with three rows, and the other apparently a Henestella with two rows of pores—the latter from Ireland, be it remembered, where in certain localities Jchthyorhachis Newenhamu, M‘Coy, and Glauconome grandis, M‘Coy, are not uncommon. Specimens in my possession lead me to say that I have little doubt of being able to show that the fenestrated form described as Frenestella crassa, M‘Coy, is likely to prove to be the network or polyzoary of one or other of the above species, or some kindred form. A glance at the drawings of Fenestella crassa given by Prof. M‘Coy would seem to confirm this view. The coarseness of the interstice, and irregularity of the dissepiment and growth generally, are not characteristic of the Fenestellide. Its true affinities have yet to be ascertained. or the present it is enough to say that its claims to be considered a Fenestella are very doubtful. THE CARBONIFEROUS FENESTELLID ZA. 187 FrnESTELLA HALKINENSIS, Sp. Nov. Sp. char.—Polyzoarium a flat, oval, or circular expansion, folia- ceous in its early growth, having a stem and expanded root-base. Interstices flat, broad, very slightly rounded, keeled, the sides often showing a flap or fringe. Dissepiments irregularly placed, very thin in early stage, gradually thickening, rarely more than one third of the thickness of the interstice. enestrules elongated, becoming oval with increased growth, four in the space of two lines measured longitudinally, and six in the same space transversely. Pores small, round, often twice their diameter apart, three in the length of a fenestrule, four and sometimes five in the later and larger growth. cel a narrow round wavy line, having three nodes in the length of a fenestrule. This species at first sight would seem to have a strong resem- blance to Fenestella plebeia, which, however, is soon dispelled by careful scrutiny ; for it will be found to possess scarcely a feature incommon. Its broad flat interstices at once distinguish it from the round or more often angular stem of J”. plebera ; while the greater distance between the cell-pores, and the thread-lke nature of the keel, complete the distinction. Again, the growth of /. plebeva is noticeable for its regularity and smoothness, while 7. halkiensis is by comparison coarse and irregular. The root-base of F. plebeia is secured in position by numerous rootlets; while /. halkinensis has a broad, expansive and adherent base, which is further strengthened by the interstices which, in favourable positions, directly adhere to the rock by a similar flat calcareous base. Its expansion was not only much smaller in size than that of F. plebeia, but the last details would seem to indicate a species of dissimilar habits, by attaching itself to rock-surfaces which the rootlets of /. plebeia would fail to enter. The affinities of this new species are, on the other hand, clearly with Fenestella polyporata, Phill., rather than with F. plebeia. It has very much the appearance of being a diminutive variety of enes- tella polyporata, especially on the obverse face. Locality. Fairly abundant in the Cement-stone Quarries near Pen yr Wylfa, Halkin Mountain, North Wales. It will be noticed that I have described only five species of Carboniferous Fenestelle. I by no means wish it to be understood that I do not believe in the existence of species other than those I have described. Of these F. plebeia, M‘Coy, F. nodulosa, Phill., F, polyporata, Phill., and F. membranacea, Phill., are really the principal and prevailing forms, and will be found in more or less abundance in association with other Polyzoa, from the Calciferous Sandstones of Scotland to the topmost beds of the Mountain- Limestone series of North Wales and elsewhere. The only partial exception to the rule is Fenestella membranacea, Phill., which is either absent or not so abundant at certain points and localities. Having proved the existence and association of these species at various horizons in the Carboniferous series, I speak confidently as to the fact that the leading species of the Carboniferous Fenestelle are few in number, while the reputed species are made up from the 188 G. W. SHRUBSOLE—FURTHER NOTES ON mutilated and altered fragments of the polyzoarium in all stages of its growth, not only of Fenestella, but of Ptylopora, Glauconome, and Ichihyorhachis, all of which presented a reticulated character, which under certain conditions might have been (and, indeed, has been) mistaken for that of Fenestella. Hitherto it has been the practice to assign any free fenestrate fragment of a polyzoan to Wenestella. With the discovery to which I have alluded, that Fenesteila was not singular in possessing a fenestrate polyzoary, the error of that method of determining polyzoan fragments becomes apparent. The few species to which I have reduced the British Carboniferous Fenestellide is, I find, in perfect. agreement with the results arrived at on the American continent. Prout has published * a list of eight species of /enestella from the Carboniferous rocks of North America ; although no plates are given, the details of the species are so minute as to leave nothing to be desired in the way of description. Of these eight species there is only one (Lenesiella Norwoodiana) that has any claim to be considered a new species ; they correspond exactly with the types of our various English species. Prout did his work without the aid of Prof. M‘Coy’s book on the Carboniferous Fossils of Ireland, in which the Fenestellide are for the first time adequately described. The result of Prout’s independent research, apparently with ample material, is that only five good species are made out. These American synonyms I have placed for the first time under their respective English species. There are possibly a local species or two of Fenestella which I have not described. Ltnestella halkinensis is one of these local forms. I have not met with it elsewhere than at Halkin Mountain. These local species require to be very carefully worked out before they can be definitely pronounced to belong to the Fenestellide. I would suggest, with a view to the prevention of a needless mul- tiplication of species of Fenestella in the future, that, before descri- bing a new species, the nature of the attachment of the fenestration should be definitely ascertained, whether to a root-stalk with root- lets, as in Fenestella, or to a midrib or stem, as in Péylopora and Glauconome. Nor do I consider this standard too high. I have adopted it with the five species which I have described. Owing to the complicated surroundings which I have shown to be connected with Henestella and the forms allied to it, I am strongly of opinion that some such course of procedure is necessary. The following table of the measurements of the leading features in the several Fenestelle will be found of considerable service in es- tablishing their identity. * Trans. Acad. St. Louis, vol. i. p. 228-286. THE CARBONIFEROUS FENESTELLIDA, 189 N ie NT | Number ne mee of fenes- | of fenes- aaa gees aren Shape of | trulesin. | trules in co Lissep;. | fenestrule.| two lines | two lines | Seer : trans- | longitu- ; versely. | dinally. | | Fenestella plebeia, M‘Ooy..., 3-4 Oval. 6 ce membranacea, Phill...., 3-4 Opreng, 8 | : nodulosa, Phill.......... | 3-4 | Square. | 6-8 noes polyporata, Pile 6-9") | | Elongate. 4 | 2 halkinensis, Shrubsole| 3-4 | Elongate. | 6 ay 4 | Discussion. The Presrpent stated that this group, which ranged from the Llandeilo to the Permian, is often represented by very imperfect specimens, and that there is great difficulty in discriminating the forms. Nowhere are they so well preserved as in the Carboniferous rocks of Flintshire and Scotland. Dr. Mucrre spoke of the value of the method adopted by the author in studying the varieties of growth and development in each species. Prof. Snetey regarded the principle on which the author had worked as a very sound one, and he thought that such revisions of certain life-groups were calculated to be of the greatest service to geology. The PrestpentT remarked upon the difliculty that arises from the fact that many type specimens are inaccessible for reference. He deprecated the creation of new specific names before a rigid com- parison had been made with the old ones. 190 PROF. P. M. DUNCAN ON THE 14. On the Corarttirerovs Suriss of Srnp, and ets Connexion with the last Upuuavat of the Himatayas. By Prof. P. Marty Duncan, M.B. Lond., F.R.S., F.L.8., &. (Read February 2, 1881.) ConrTENTS. I. Introduction: the History of the Geology of Sind; Questions involved. II. The Stratigraphical Position of the Series and of the Ossiferous Manchhar deposits. III. General Results regarding the Alliances and Peculiarities of the Corals of the Series. The Prenummulitic, Nummulitic, Oligocene, and Miocene Coral-faunas of Sind. IV. The Equivalence of the Mancbhar and Sivalik deposits. V. General Considerations regarding the Age of the last Himalayan Uplift. I. Introduction §e. A memoir by Grant, illustrated by Sowerby, which appeared in the ‘Transactions’ of the Geological Society (series 2, vol: v. 1837), first brought the countries west and east of the Indus under the notice of European geologists. Fossils from Sind, Baluchistan, and Cutch were therein described ; and the types were presented to the Society. But it is to Mr. Vicary that science owes the first attempt at a com- plete description of the geology of Sind*. Written in 1847, the fossils which should have illustrated his paper, and some others collected by Lieut. Blagrove in Cutch, were handed over to MM. d’Archiac and Jules Haime for examination and publica- tion. Their fine work, the ‘ Description des Animaux Fossiles du groupe Nummulitique de Inde, was published in 1852. They only recognized one geological horizon, the Nummulitic, although Grant had expressed an opinion that there was more than one fossili- — ferous series. Messrs. Cook and Carter added to the knowledge of the Baluchistan area in 1860; and the last-named naturalist felt it necessary to give a Miocene age to some fossils which Sowerby had figured for Grant. On the other hand, MM. d’Archiac and Jules Haime severely criticised M. d’Orbigny for stating that some of their Sindian species were of Falunian age, and decided against Dr. Carter’s grouping of some of the marine Tertiary beds as Miocene x. In 1863 Mr. Henry M. Jenkins, F.G.S., at that time Assistant- Secretary of this Society, and myself were endeavouring to learn something about Tertiary deposits situated as remotely as possible from European types. A collection of Mollusca and Corals from Java had been sent to the Society by M. de Groot; and we proceeded to examine them; and in order to determine the affinities of some, which seemed to be of younger age than the Nummulitic, it became necessary to study the work of MM.d’Archiac and Haime on India and to examine their types. We found that there was a species in Java which my friend called * Quart. Journ. Geol. Soc. 1847, vol. ui. p. 334. t Cook, Trans. Med. Phys. Soc. Bombay, vol. vi. pp. 1-45; Carter, Journ. Lombay Royal Asiatic Society, vol. vi. p. 184. t See Carter, Geol. Papers on Western India, pp. 628-776, and the general résumé in their work already noticed. CORALLIFEROUS SERIES OF SIND. 191 Vicarye callosa, and that it was closely allied (the distinctions being only of ornament) to Vicarya Vernewlit, D’Arch., from Sind. This Vicarya of Java was associated with species which are closely allied to those of the Miocene of Western and Eastern Kurope. A Coral (Heliastrea Herklotzi, Dunc.) from Java resembled the Miocene Heliastreeans more than those of the Kocene*. Following up the subject carefully, 1 examined all the collections submitted to MM. d’Archiac and Jules Haime; and to my surprise I found that they had neglected many specimens which had not only a Miocene but even a Pliocene facies. The identity of some species with members of the West-Indian Miocene Coral-fauna was tolerably evident; and whilst admitting the similarity of the mineral con- dition of all the fossils, I divided the genera into those which else- where are found in Hocene, Miocene, and Pliocene formations. During the following yeart I examined and described twenty- six species of Corals from Sind, from unknown geological horizons near Karachi. Some had been previously described by D’Archiac and Haime; but the bulk had not. ‘he evidence of the existence of three Tertiary deposits instead of one seemed overwhelming. Subsequently the Geological Surveyors of India, Mr. W. T. Blaniord, F.R.S., and Mr. Fedden, made an elaborate survey of Upper and Lower Sind, and Messrs. Wynne and Fedden of Cutch.: In the last-mentioned district fossils were found which proved the equivalency of the deposits with those of Sind, and that there was a higher Tertiary horizon than the Nummulitic. The Corals collected in Sind by the Survey, under the superinten- dence of Messrs. Blanford and Fedden, and carefully marked with figures denoting their geological horizons, were sent to me by Mr. Medlicott, F.R.S., the present Superintendent of the Survey, so that they might be described in a volume of the ‘ Paleeontologia Indica.’ But simple description was not all the requirement of the Survey and myself. Certain questions of great importance had arisen in the geolcgy of India; and it was probable that the determination of the age of the coralliferous strata would assist in deciding the ages of the Himalayan upheaval, of the Sivdlik deposits, and of certain olive shales underlying a trap at the base of the Nummulitic series— questions which had been attempted and had been answered by Messrs. Medlicott, Blanford, and Lydekker, and which required some confirmatory evidence. There was another reason why I should be honoured by the Geo- logical Survey of India asking me to undertake the description of their splendid collection of fossil Corals from Sind. In my Presi- dential Address for the year 1878 I criticised the opinions of my friends Messrs. Medlicott and Blanford regarding the age of the Himalayas, and inferred that their decision regarding the Post- pliocene date of the so-called Nerbudda fauna, was influenced by the discovery of a human implement in the containing deposit. They took the very philosophical course of affording me the means of converting myself to their opinions. * Quart. Journ. Geol. Soe. vol. xx. p. 45. t Ann. & Mag. Nat. Hist., April 1864 (Fossil Corals from Sind). 192 PROF, P. M. DUNCAN ON THE J The description of the Corals has been published in a volume of the ‘ Paleontologia Indica ;’ and although the evidence regarding a Pliocene marine fauna failed, there is no doubt about the former ex- istence of Lower and Middle Tertiary coralliferous deposits in Sind. Il. The Stratigraphical Position of the Series and of the Ossiferous Manchhar Deposits. The detailed description of the geology of Western Sind is to be found in the Memoirs of the Geological Survey of India, vol. xvii. part 1 (1879), by W. T. Blanford, F.R.S. &e., Deputy Superintendent of the Geological Survey of India; andit is therefore only necessary - to explain those parts of it which relate to the general succession of the stratigraphical series and the position of the strata yielding Corals. The mountain-ranges west of the Indus run nearly north and south, in irregular parallel series. The longest range, the Khirthar, is slightly curved, the concavity being to the east; and it extends from slightly south of 26° N. lat. to close to 28° N. lat. The Laki range, more to the east, is nearly coincident with the sixty-eighth parallel of east longitude, and extends from the Indus north of 26° N. lat. to nearly a degree to the south. Other minor but nearly north-and-south ranges occur; and the whole were com- prised by MM. d’Archiac and Haime under the title of the Hala range (a name unknown to the natives). The following is the list of geological formations in Western Sind, and which are to be recognized in one or other of the moun- tain-ranges * :— | amantell | Group. Subdivisions. pe Age. Remarks. feet. ALYUN EA RG oe |eae eeleneeeene ? Post-Tertiary. i (Uipperyic sce OOO shacsonae Plioceneteeee Unfossiliferous. MESSCHEAB 00. { mae eno 3000-5000 ...| Old Pliocene or} Vertebrate remains. Upper Miocene. GAS PRR SEER WDC ER RE at 1000-1500 ...| Miocene ............ Coralliferous; no Nummulites. | Wiper meeeeacs 4000-6000 | scl brad epee eee Unfossiliferous. NARI wecrccree Oligocene. ( Coralliferous, with | Tower:.......: 100-1500 | VaR ile nnn game Nummulites garan- Sensis. { Wipperyeeees.: 500-3000...| Nummulitic ...... Nummulitie lime- KHIRTHAR .. stone. I OWers acces GOOO'? a..36 sacle erence Unfossiliferous. IRANIKOT cuca digester cee D000 eae Lower Nummu-| Fossiliferous. litic. Corals and Nummu- lites. RAP ys so AO | gr Mea Rr Cana AD-90 ewesea oe Decean Trap. (| Cardita Beau-\ 350-450 ...... Transition beds ...| Fossiliferous. || sandstones ..| 700 © Hori t det andstonesi.|) (OO meesccnecs si. retaceous. orizon not deter- ee racrous 1 Limestones | 320. mined. || with Hip- (| purites. * From the Memoir by Blanford, p. 32, slightly modified. CORALLIFEROUS SERIES OF SIND. 193 Explanatory Sections.—In the Laki range, south-west of Amri on the Indus, are dark-coloured hills which contrast with the cliffs of grey and white Nummulitic limestone behind them. A section close to the hill called Barrah is given by W. T. Blanford, F.R.S. ; and it shows that the range consists of three parallel ridges (see fig. 1, p- 194). The outer, to the east, is composed of Tertiary rocks, while the intermediate one consists of Cretaceous beds faulted to the eastward against the Lower Eocene strata, and dipping under them io the westward. This section shows the normal sequence of the groups of strata from the Cretaceous to the Khirthar inclusive. Above the Nummu- litic limestone of this last group the Nari and Gaj series are wanting, and the Manchhar succeeds*. On the Gaj river, in the Khirthar range, W. T. Blanford remarks that a thickness of at least 25,000 feet of strata is exposed (see fig. 2,p.194). The succession from west to east 1s :—unfossiliferous strata, probably of Cretaceous age, followed by Khirthar strata (the Ranikot series, the lowest Eocene, being absent); then the lower and upper Nari series come in, and are followed by the Gaj and the Manchhar deposits. The lowest coralliferous deposits occur in the soft olive shales and sandstones with volcanic ash, belonging to the Cardita Beaumonti Series below the trap. The accompanying remains are those of Amphiccelian Crocodilia and Echinodermata; and the deposit was neither a reef-structure nor a deep-water one. The lower part of the Ranikot series, resting immediately on the trap, consists of soft sandstones, shales, clays with gypsum and lignite, and pyritous shale. A few fragments of bones and some dicoty- ledonous leaves occur. ‘These freshwater strata are succeeded by highly fossiliferous marine limestones, often brown in colour, inter- stratified with sandstones, shales, clays, and ferruginous bands. Nummulites appear for the first time, and there is a grand develop- ment of Corals, Echinodermata, Gasteropoda, and Cephalopoda. It was not a very deep-water formation. Hrosion of the surface of the Ranikot strata occurred before the deposition of the next series. The Khirthar series includes in its highest portion a massive, pale or dark grey, hard, compact Nummulitic limestone whose extreme thickness is 3000 feet. It thins out to the south-west, and disap- pears within a distance of twenty-five miles of its greatest develop- ment. Other Nummulitic limestone-beds are found, which may be lower in the series ; and they and the main group are represented elsewhere by shaly limestones and sandstones with calcareous bands. In some districts flint occurs in a limestone with Alveolinw. The lower members of the series are often wanting, and are well repre- sented by shales, marls, and sandstones and where these are present ; unconformity with the underlying .Ranikot beds is not seen; but where they are absent the Nummulitic limestone (as in the Laki range) rests unconformably. The compact limestone is of course x Memoirs Geol, Survey of India, vol. xvii. pt. 1, p. 151. PROF. P. M. DUNCAN ON THE 194. ) —— WENT ToMOT Sf pus ‘xeddn ‘9 {fey ‘p ‘savyYouepy comoy ‘9 ‘soddn ‘9 fureyd snpuy ‘» *IVYYOuB ‘Teo "118 NT . *TVY JALIL "s00084019 —_— A eee 1 oe creme con neers eee recess SN ee SE SS re Q---— 4 oI Se Rane eee ane wee wee ee ee---— —-— snpuy “gH eney9 “rey ANIL (ozis yey “proyurrg 104zV) ‘ony (ny ayy fo yung yptou ayn v0 ohuny mypnyy fo woyoay onoumunibyrp [n.10U09—% *S1 yf ‘g ‘oUOJSOTUT] AVYZATY, OYJ pur ‘ssouyoryy yvoasd Jo st yor ‘9 ‘dnosd yoyravy ey} wsy} pure ‘sMor[oy ‘p ‘deay oyy, ‘a “yuownvog vpplnD YIM SeTLYS OATTO oA’ STIL OT} Jo ysoAA ‘sovoq ueTIYdoa ATyuoIvdde jo syuoutseay pure stoysho SuLureyUOD poq ¥v st STII oY Jo doz oy} UG ‘ST[TY pecmopoo-yavp oyy Jo doz oT surtoy tf ‘soulojspuLs Jo SSOUWOIY] yvors YW ‘sopranddizy sureyuoco pur [IPT Yyerawg oy} Jo osuq oy} sumtoj ‘/ ‘ouoysouny, sip, “yuri v suroq Arwpunog oyy ‘ouojsoutty Aors puv oy Jo ype gsurese Apdnaqe puso soyy, ‘crip pestoAod v ITM osye “no do.to ‘9 ‘dnoas yoyrury, oy] JO saoquIEut aaddn oy} yor ut ‘punoas MoT oUTOS st OSPIA OY} JO JSoAA ‘OSPIA Wtoysva OY} SUIAOJ oUOJseETMUTT STT, “480A\ oY} 0} Sutddrp speq ouy ‘9 ‘oUOYSOUTT] ONTNWUINA AvyyAIYYy UO UeY} pue “Va2vawharTy sapynuwun\y YA ‘ose avqzaIyyZ Jo sAvypo oped uo ysou “VyvaysS TeTOULTL “# PARE. e W SA RSS| <=, Is VG} TS |! ee Ne qm “TE Were (4903 QOGY 0} Your T ‘epeog) “W2FT young ybnowy, WOLVIG—'T *SLy | 1 { hs i | ‘ “4923 006T TINH ang “HOATE [B) 293 000 “JooF Q00e “wear [vQ ‘YB “AA CORALLIFEROUS SERIES OF SIND. 195 highly fossiliferous ; but it has not yielded very good specimens of Corals. The indications of a fringing-reef-building fauna, or a bank of coral, are distinct; but the species when compared with their modern analogues do not indicate a great reef-development. Amongst other fossils, Lamellibranchiata prevail, and the Echini are numerous; but the Nummulites and Orbitolites, Alveoline, and Patelline are the most important organic remains. Amongst the Nummulites there are WV. Ramondi, N. biaritzensis, N. Beaumonin, NV. granulosus, and N. Leymerier. The Nari series rests conformably on the Khirthars; and there is occasionally an apparent passage from the Khirthar limestone into a yellow or brown rock of the Nari group. There is, however, a biological break; for the Nummulites of the upper group are characteristic, and differ from those of the lower. The Khirthar forms are not found, and Nummulites garansensis appears with Orbi- toides papyracea, in the Nari series. The rock is a limestone with intercalations of sandstones and shales. Asa rule, shales, fine sand- stones, and occasional bands of limestone form the base of the Nari group, and pass upwards into coarse, massive, thick-bedded sand- stones, attaining a thickness of from 4000 to 5000 feet on the eastern flank of the Khirthar range. A local break occurs to the south, and the Upper Nari beds rest unconformably on the denuded edges of the Lower Nari brown limestones; and still further south, fifty miles east of Karachi, there is a well-marked distinction between the upper members of the group with Orbitoides papyracea and the lower with Nummulites garansensis; and in one locality the upper member overlaps the lower, and rests on Khirthar limestone. To the east of the Laki range the Nari beds are wanting, and the Mancbhar series rests unconformably on the Khirthar, with some pebble-beds of the Gaj series intervening. But to the west of the range the Nari and the Gaj series are found in their normal sequence ; and towards the coast the exact distinction which can be drawn elsewhere, stratigraphically and petrologically, between the Ter- tiary series is not possible: this is mainly due to the disappear- ance of the limestone element of the Khirthar and Nari series, and to the prevalence of sandstones and shales. The fossils, however, distinguish the groups: but the horizons of the zones of Nummu- lites and Orbitoides vary, being higher or lower in their proper series according to locality. The upper sandstones of the Nari group have not yielded marine fossils, and in Upper Sind they contain the remains of plants. The Nummulites become fewer in their species at the base of the Nari group, and cease to be found in the lower beds of the suc- ceeding series, the Gaj. The Gaj group, with a base of highly fossiliferous limestones and calcareous beds, more or less shaly and stratified, overlies the softer shales and sandstones of the Nari series. The development of the calcareous series is great; but it is subordinate to an arenaceous element. ‘The sandstones are intercalated with clays with gypsum; 196 PROF, P. M. DUNCAN ON THE and shales and bands of limestone, highly coralliferous, are very constant. The Gj series rests conformably on the Nari series; although there is a mineralogical break, the passage is so oradual that cu careous bands of the Gaj series are found interstratified with the uppermost Nari sandstone. The Gaj series overlaps the Nari to the south and rests on the Khirthars, and is wanting in Lower Sind to the eastward of the Laki range North-east of Karachi the series is highly developed, and massive limestones occur ; they reach as far as the coast. The uppermost beds of the Guaj series are variegated clays and grey sandstones, which form a passage into the everlymg Manchhar strata, and contain species of Ostrea, Corbula, Arca, Scalaria, Buccmum, and Turritella. A crab of the genus Typilobus and the Vicarya Vernewliw already noticed have also been found there. The oyster is like Ostrea multicostata. 'The Echinoderms are of the genera Maretia, Meoma, Breynia, Clypeaster, &e. The Corals found, principally associated with the lmestone-beds, are very different in their aspect from those of the underlying Nari group. They belong for the most part to reef-building genera, and some are represented in modern reefs by allied species, which grow in the surf and in the most exposed parts. ‘The rest were dwellers ; in quiet water. Great masses of the coral limestone consist of Stephanocena maxima in casts, and the presence of the genera Madrepora, Hahastrea, and Porities is very significant of s shallow-water conditions. Resting on the Gaj strata is the Manchhar group; and where the Gaj beds are wanting, and even where the Nari series is deficient, this upper group rests on the Khirthars. Of ereat thickness (10,000 feet on the flanks of the Khirthar range) the Manchhar series is divisible into an upper and a lower group. In the lower group much grey sandstone, soft and fine-grained, and composed of quartz with some felspar and hornblende, is found ; and red sandstones and conglomeratic beds exist towards the base, as well as red, brown, and grey clays. The conglomerates do not contain pebbles of older Tertiary rocks; but cream-coloured clay, soft sandstone, quartzite, and micaceous shale are found in them. The conglomerates near the base are ossiferous; and the vertebrate remains are teeth or bones more or less rolled. There is no satis- factory distinction to be made on the Gaj river between the estua- rine beds at the top of the Gilj series a the lowest beds of the Manchhar; and osseous remains have been found in the upper Gaj series. The upper part of the Manchhars has a conglomerate in it with stones derived from the older tertiaries ; below it are red, brown, or buff sandstones, with some clays; and there ure no fossils. Although usually conformable to the Gaj series, in places the Manchhars rest on the older rocks ; ; and there are proofs of the Gaj strata having been greatly denuded before the deposition of the Manchhars. In the Laki range the Manchhars have an ossiferous conglo- CORALLIFEROUS SER1ES OF SIND. 197 merate near the base. In the Vero plain, running southward along the eastern side of the Laki range, south of Ranikot, large frag- ments of silicified wood are common, and some trunks of trees are upwards of 30 feet in length and 10 in girth. In the south, near Karachi, the Manchhars pass into Gaj strata, and marine fossils are associated with the lowest beds. The general conformity of this great tertiary series is evident— although local unconformities occur, and there is evidence ‘in fayour of there having been some disturbance of the older rocks before the deposition of the Lower Manchhar group. The sub- sidence which took place during the deposition of these thick shallow- water deposits was vast; and there were occasional slight upheavals. The final epoch of the mountain-formation occurred after the depo- sition of the Manchhars. The strike of the chains is in the main north and south; and the thrust came from the west to the east, and from east to west. The date of the grand mountain-formation is subsequent to the deposition of the sedimentary strata forming the Upper Manchhars. III. General Results of the Study of the Corals of the Series. The Corals of the strata with Cardita Beaumonti below the trap. are shallow-water forms; and their development indicates conditions untavourable to vigorous coral-growth. There was no reef. The species are new ; and there are no characteristic Cretaceous or Hocene forms present. The Smelotroch: and the Lhabdophyllic are Secondary genera; but the first-named had species inthe European Tertiaries, and: may be represented in the shallow seas of the present day. The Cary- ophyllice of the deposit have four cycles of septa only ; and this gives. them somewhat an ancient look ; and the Litharwa would pass either as a Cretaceous or as an Eocene form. The fauna as a whole is deficient in characteristic forms, and may be considered transitional. It has been stated by the Geological Surveyors that the trap re- sembles the Deccan and Malwa trap; and the position of that vast outburst is anterior to the Nummulitic age, and subsequent to some Cretaceous strata in Western Central India. Certainly there are no- Nummulites in the sands and shales which contain the fessils; and these are the remains of a Cretaceous crustacean. The Corals of the Ranikot group are numerous in genera and species ; and the fauna is remarkable for the number of simple forms: it contains, and for the predominance of the family Fungide. The compound Corals of the family Astraide, which so largely enter and have entered into the composition of reefs, are but feebly represented ; and the few species which have been described did not ageregate to form massive limestones, but were seated on small flat circular epithecate bases. Fine as were many of the growths, yet it was a stunted Coral-fauna; and the shape of the majority of the forms would rather indicate that they lived in still water, and not in the rush of the waves. Certainly the Corals were beyond the Q.J.G.8. No. 146. P 198 PROF. P. M. DUNCAN ON THE rolling or scouring action of the sea; for any such movement would have displaced most of them. As a generic assemblage, the Ranikot Corals are Kocene in facies; and of the twenty-seven genera three are new, and oneis known in the Miocene of the West Indies. Out of fifty species seven are identical with European Eocene types, which are not found, however, as a whole, on one geological horizon in Europe. Some of them are found in the Lower Eocene; and the others are to be noticed in the deposits of San Giovanni Ilarione and of the Oberburg, in Styria, and in the higher horizon of Castel Gomberto. These forms are located in one series in Sind, and at the bottom of the Nummulitic group. Five species are closely allied to Kuropean forms out of the same great vertical series; and about seven others have a distinctly European facies. These 19 species give the Eocene charaeter to the fauna. The most distinctive genera are Stylocwinia, Plocophyllia, Stepha- nophyllia, Stephanocenia, Prronastrea, Reussastreea, Cyclolites, and Litharea. It is remarkable that five well-marked species of the genus Turbinoseris (nobis) should have lived on the Indian area; they are allied to those of the Eocene of St. Bartholomew in the West Indies.* ‘The genus is Lower Cretaceous in England ; and, under another name, it has lately been introduced into the N attheim oolitic Coral- fauna. The presence of a species of St yline and of Thamnastrea recalls the Jurassic and Cretaceous ages: ‘the first is found also in the next or Khirthar series of deposits; and'thé last probably lived longer in the southern seas. Only one genus, Placocyathus, is a Miocene type ; and its distribution is West-Indian. There is one very remarkable species, Stylocenia maxima, nobis, in the Ranikot group; and it is so called from the great size of the intercalicular projections, which are little monticules in the British Stylocenia emarciata, Lamk. sp. ae emeniun of the Cages base of the coral is 3} inches; an A concentric epitheca is on the base; and the upper surface i is @ mass of calices, and tall columns between them, some reaching 75 inch in height. The minute columns in the British | ‘species are proved by the examination of these gigantic growths to be stunted or aborted calices. Some of the processes of Btylocoenia maxima have a perfect calice on the top, and the costs come down the side; others have ill- developed calices, and a mere representation of a columella and septa. It is acommon form, and is characteristic of ‘the Ranikot series; its nearest ally is Stylocania macrostyla, Reuss, which is associated with Nummultes planulatus and Cerithium ¢ Uh iganteum in the district of San Giovanni ilarione. Young specimens, of course, greatly resemble the European Stylo- coenia emarciata of Bracklesham age. ‘Two other species of Stylocoenia are very common: one, S. Vecaryt, was described by D’Archiac and Haime ; and the other is new. It has a large circular epithecate base ; and the calices are large. Astr ocoenta, that very widely distributed genus, whose range, ver- * P. M. Duncan, Quart. Journ. Geol. Soc. vol. xxix. p. 554, CORALLIFEROUS SERIES OF SIND. 199 tical and horizontal, is enormous, has several species in the Ranikot eroup: and Astrocenia ramosa, Sowerby, a Cretaceous type from the Turonian of Gosau, found also in the Hocene of St. Bartholomew in the West Indies, is present in Sind, and has the usual tuberose branchlets so characteristic of the form. Large solitary Corals of the genus Montliwvaltia abound, and also a characteristic series of simple forms which I have had the oppor- tunity of naming after Mr. Fedden, to whom the geology of Sind owes so much. The genus Meddenia had its young forms growing on small shells, which, with the growth of the base of the Coral, were enclosed, and the Coral became free. This is not an unusual habit in some two or three species of Corals of another family, now living; and an elongate base is produced. One of the results of this change of condition was that the coste did not radiate upwards from the peduncle, as is seen in the larger forms, but from the whole length of the enclosing base. Another pecu- harity is that the epitheca is like a broken mosaic. There are three well-marked species of this genus; and one has two good varieties in the Ranikot series. It is near Montlivaltia as a genus; and, admitting the great variability of all simple or solitary Corals, it is very possible that the unusual method of early growth may have produced correlative changes in the general shape and septal arrangement. The genus was restricted to this Lower Eocene horizon. Amongst the great series of Fungide found in the Ranikot group there are two European forms of Cyclolites ; and the other seven species of the genus are well differentiated. Finally, the genus Stephanophylliia, amongst the Kupsammines, ought to be represented, to give the full early Tertiary facies to the Cyclolitic assemblage. Stephanophylla indica, from Jhirk, in the Ranikot group, maintains the character of the genus for beauty, and is an exquisite gem. Its affinities are rather with the Cre- taceous species; and it differs but slightly from S. Bowerbanki of the Lower Chalk in its main characters, and from S. discoides of the London Clay more decidedly. Like most of the Ranikot corals, the individuals of the new species began life by settling down on a ~ Nummulite; this formed the permanent base of the coral; and the usual radiating costal arrangement of Stephanophylha either does not exist or cannot be seen for the Nummulite. Khirthar Serives.—Sixteen species of Corals were found in this series; but ten of them were taken from beds so high up that it is not satisfactorily proved whether they are at the top of the Khirthars, or form the base of the Nari group, next in vertical succession. By separating these ten species, an indubitable Lower-Khirthar fauna is decided to have existed: its biological conditions were unfavourable to vigorous coral-growth ; and there were no littoral or reef-building forms. From the analogy of recent forms, which are identical generically with those of the Khirthar series, or whose shape resembles that of the ancient species, a sea-floor at the depth of from 20 to 200 fathoms or more would be indicated. RP 2 200 PROF. P. M. DUNCAN ON THE The species belong to the genera TZ'rochocyathus, Leptocyathus, Stylophora, Montlivaliia, Calamophyllia, and Astrocema. The Trochocyathus is one of the discoid group, and the Leptocyathus is of course a low flat Coral. The Stylophora is a minute Coral; and so is the Astrocenia. But the Montlvaliia is a finely grown form ; and so is the Calamophyllia. Of these, Stylophora contorta, Leymerie, has also been found in the Kocene of the West indies and at La Palareain Europe; and Astrocenia nunusma, Defrance, sp., is a Nummulitic form at Gap, in the district of Nice. This little assemblage is thus clearly of Hocene age. No distinction can be made between this and the preceding deposit, as regards age, from the Corals. ; Now the ten species, the horizon of which may be of Upper Khirthar or Lower Nari age, belong to genera whose morphology in- dicates the presence of totally different physical conditions from those which environed the stunted, simple, and the few large Corals of the lower horizon. They were shallow-water and reef-building forms ; and the genera are Stylina, Latimceandra, Hydnophora, Favia, Isastrea, Pterastrea, Plesiastrea, and Porites. Thereis nota simple Coral amongst them ; and the facies is singularly mixed, old and new, Mesozoic and Cainozoic genera existing together. The species Hydnophora malirrensis, nobis, is allied to H. venusta, Catullo, of the European Nummulitic. Porites Pellegrinit, D’Achiardi, is found in Kurope at San Giovanni Llarione; and Porites indicus, nobis, is allied to a species from Crosara, Oberburg. The distinctness of this fauna from that of the Ranikot group is evident enough ; but it is still Nummulitic in facies and character. The Narv Series—From the remarks already made upon the localities of some of the presumed Khirthar Corals, it is as well to admit that some upheaval took place at the close of that period, and produced fringing-reef-building or bank conditions. But they did not prevail throughout the accumulation of the whole of the Nari Series; and it is somewhat remarkable that no species passes up from the reef-building horizon into that of the Nari proper. The genera of the Nari group above the base, indicate rather vigorous coral-growth, and both deep- and shallow-water conditions. The genera Trochocyathus, Trochosmilia, Montlivaltia, Cycloseres, and COyclolites are the simple forms; and the vigorous compound forms are included in the genera Dasyphyliia, Rhabdophylhia, Lep- toria, Meandrina, and Prionastrea. | . The Trochocyatht of the Nari series are remarkable for having commenced their growth in a discoid shape. Some retained that shape; but one in particular, Trochocyathus nariensis, nobis, grew upwards from its disk-shaped base in a perfect cylinder, reaching 2 inches in height. Trochocyathus cyclolitoides, Kd. & H., is a widely distributed Huropean Eocene coral, and is found in the Nari series of Sind. Stylophora pulcherrima, D’Achiardi, from the Hocene of Friuli, Trochosmilia varicosa, Reuss, from Crosara, Stylocenia taurinensis, Kd. & H., from European Eocene and Miocene strata, and Cycloseris CORALLIFEROUS SERIES OF SIND. 201 Perezi, Ed. & H., from the Nummulitic of Nice, are also amongst the Nari corals. Hence, out of the twenty species of fossil Corals in the Nari series, one is found at a higher horizon, and one fourth of the number are found in the Upper Nummulitic and Oligocene deposits of Hurope. Taking the Ranikot, Khirthar, and Nari Coral-faunas as a whole, there are 76 species of Corals in them and severai varieties ; and of these 16 are identical with European forms which are found in strata yielding Nummulites to the top of the Oligocene. There are also eight species in the Sind series, closely allied to those of the European fauna; and whilst some species are common to the West- Indian Eocene, an important genus is also common to both localities. The absence of species passing up from one series to another is very remarkable. The Gdj Series——This series of strata contains a large number of Corals in bands ; but although most of the forms are massive and compound, and suited for reef-building, the majority are peduncu- late instead of covering a large surface and incrusting. Some are very massive ; and the presence of species of Madrepora, an essentially reef-building genus, with Porites, Agaricia, Echinopora, Prionastrea Plesiastrea, Brachyphyllia, Leptoria, and Dasyphyllia would indicate, at the present time, very active shallow-water coral- growth. The majority of the Gaj genera still flourish; and most of those which are extinct had the structures requisite for reef- buildmg. The Astreide as a family preponderate, and simple Corals are rare, in the collection. But although the facies of the Gaj coral-fauna is very recent, and there are very few extinct genera, still the modern Coral-fauna of the Eastern and Red Seas is not represented by a single species. Two species are identical with Miocene West-Indian forms; and one is found in the Nari series. Tt is interesting to find the West-Indian Miocene and recent genus Antillia represented. Nevertheless the evidence afforded by the Corals is in favour of a mid-tertiary age being given to the 41 species from Gaj. In the communication to the ‘ Annals and Magazine of Natural History’ (1864, vol. xiii.) I described several species of fossil Corals from unknown geological horizons in Sind; and it is not difficult, with the lists of the genera and species which have been collected by the Geological Survey of India, to decide whence the majority eame. Thus Oculina halensis, nobis, Antillia dentata, nobis, An- tillia ponderosa, Ed. & H., Cladocora haimei, nobis, Hydnophora rudis, H. plana, H. hemispherica, nobis, point, with Cyathoseris val- mondorisiaca, Kd. & H., and Agaricia agaricites, Ed. & H., to a late Miocene horizon. They were picked up in the neighbourhood of Karachi, and doubtless came from the highest beds of the Gaj series. Finally, many specimens of Isidinee have been found in the Gaj series; and some of them closely resemble modern forms. No Num- mulites occur in the Gaj series ; and the Zoantharian eviderce indi- cates a Miocene age, and not an early one. 202 PROF. P. M. DUNCAN ON THE The Sindian formations, including the transitional beds with Car- dita Beawmonti beneath the trap, and the Gaj series at the top, contain 136 species of Corals and many varieties. Of these, 9 belong to the transitional series, and 56 to strata in which Nummu- lites exist in profusion. The Nari series, with few species of Num- mulites and with WV. garansensis, a Kuropean type characteristic of the Upper Nummulitic or Oligocene, contains 20 species. The Miocene strata of Ga} contain 41 species. I omit from these numbers all species from doubtful localities. Thus, as only one species transgresses, there are five Coral-faunas ; and if the separa- tion I have proposed at the base of the Nari series is right, there are six in the same area, included in a prodigious depth of sedimentary strata. The study of the Corals does not indicate that there was ever a sea there with a depth of many hundred fathoms; on the contrary, shallow-water conditions (within 20 fathoms) usually pre- vailed during the slow oscillation of the area, in which subsidence on the whole predominated. IV. The Lquivalence of the Manchhar Strata of Sind and the Swalik Group of the Himalayas. The Geological Survey of Sind by Blanford and Fedden has proved that in some places the Manchhar strata succeed transgressively to the Gaj series, whilst in others there has been erosion of the marine deposit before the deposition of the freshwater one, or else the lower group is absent. Bones of Rhinoceros siwalensis have been found in the uppermost Gij strata. There is no doubt that the greater part of the coralliferous strata of the Gaj series accumulated in shallow water, and yet beyond the reach of the wash-down of a coast-line. But the lower Manchhars were deposited in shallow water within the scope of terrestrial denudation. There must have been considerable general changes in the physical conditions of the area: and they persisted ; for coral life has never since prevailed there. The first change probably was one of slight general up- heaval; and subsequently a slow and progressive subsidence occurred, during which the vast vertical development (8000 to 10,000 fect at least) of the lower and upper Manchhars accumulated. Conglomerates, coloured sandstones (grey, green, and red in tint), and clays, are the prevailing deposits on the inner flanks of the vast. mountain-system which surrounds Peninsular India, from Sind to Burma inclusive. Made up of stone brought down by tne rivers of the extra- peninsular mountain-system, before it became of very ereat height, and when its breadth was probably much greater than it is now, these deposits are to be traced to the north of Sind in the . Suleiman Mountains, in the Salt range, covering much of the surface of the Northern Punjab, and then, forming part of the sub- Himalayan range, as far east as the Brahmapootra. Similar deposits are found, in diminished thickness, in the Assam range ; and they became important in the Burmese territories. The Sindian suc- cession of the strata is the normal one; and it extended to a certain distance northwards ; moreover it is exemplified in the Burmese CORALLIFEROUS SERIES OF SIND. 203 area. ut in the great central area of the Himalayas proper there was no marine Miocene. No marine beds, the equivalents of the Gaj series of Sind and Burma, underlie the Sivalik deposits in the Sub-Himalayas. There was open sea during the Nummulitic age where the great semicircle of extra-peninsular mountains now exists ; and marine con- ditions persisted on the west and east during the Miocene period ; but a post-Nummulitic upheaval developed a land-surface and_hill- tracts from Kashmir to Assam. This upheaval commenced before, but culminated after, the close of the Eocene age; and a considerable thickness of purple sand- stones, red clays, and grey and purple sandstones containing plants accumulated in the swamps on the plains of that age, near the mountains on the edge of the Nummulitic sea-floor. These plant-bearing strata and the underlying Nummulite- bearing strata have a minimum thickness of 2000 feet, and they form the Sirmuir series of Indian geologists. This series underlies unconformably the vast freshwater sedi- mentary formation comprising the Nahun and Sivalk strata, which attain a thickness of about 15,000 feet, and, except where buried beneath recent deposits in one locality, extend, with a varying development, along the south of the great mountain mass, and are found on one of the great tablelands to the north of the Central Himalayan axis. The lower, or Néhun unfossiliferous, series con- sists mainly of grey lignitiferous sandstones. On it the upper, or fossiliferous Sivaliks, accumulated as sandstones and clays; and on the top of all are conglomerates. A great fauna is represented in all parts of the Sivalik deposits above the Nabun beds. From the lie of these sedimentary strata, it may be very reason- ably inferred that the Nahun and Sivalik deposits are the equivalents of the Lower and Upper Manchhars of Sind; and the inference may be extended to the Upper Tertiaries of Burma. There is an outlier of this series and of its lower member in the Gulf of Cambay or Perim Island. In attempting to establish exact parallelism between the Sindian and the Himalayan deposits called Manchhar and Sivalik, it must be noticed that the vertical development of the last-named rocks is the greatest, and that whilst the lowest beds of the Sindian series are fossiliferous, those of the Nahun beds of the Sivdliks are not*. On the other hand, osseous remains are found throughout the Sivaliks proper (above the Ndhun beds), but not in the Upper Manchhars in a recognizable form. The Manchhar and Sivdlik series have been upheaved, uptilted, and in the last instance greatly contorted. Both series were the youngest implicated in the great orographical development; and although they are on different lines of strike, they were affected during the same geological period. The denudation of their exposed edges has been great. Both are * Possibly the ossiferous deposit at Kushalghar, near Attock, is of Nahun age. See further on. 204 PROF, P. M. DUNCAN ON THE covered in some places with high-level gravels; and the Sivaliks of the great tableland of Hundes underlie deposits referable to the glacial period, whose effects were not felt so far south as Sind. In considering the paleontology of these deposit it may be stated that the osseous remains are found on several horizons. First, the Rhinoceros-remains which were found included in the Marine Miocene series of Gaj in Sind, lead to the inference of the existence of neighbouring contemporaneous land-surfaces and shallow seas—that is to say, of Miocene land washed by a Miocene sea. The Rhinoceros, according to Mr. W. T. Blanford, F.R.S., is Rhinoceres sivalensis. Secondly, the fragmentary bones and teeth found in the conglome- rate near the base of the Lower Manchhar formation in Sind accu- mulated later on than the Gaj series. The following is the list of the vertebrate remains which were collected by Mr. Fedden, and named by him and Mr. Lydekker in the publications of the Geological Survey of India :— Amphicyon paleindicus. Anthracotherium silistrense. Dinotherium pentapotamiz. Hyopotomus palzeindicus. indicum, Hyotherium sindiense. , Sp. nov. Dorcatherium majus. Mastodon perimensis. minus. latidens. Falconeri. EDENTATA. Rhinoceros paleindicus. Manis sindiensis. , sp. near R. deccanensis. Acerotherium perimense. Rerriita. Sus hysudricus, Crocodilus, sp. Hemimeryx, sp. Chelonia, sp. Sivameryx, 2 sp. Ophidia, sp. Chalicotherium sivalense. The conglomerate containing the bones was composed of the wash-down of the Lower Manchhars themselves. Thirdly, the fossil bones which have been discovered in consider- able quantities in a conglomerate on Perim Island (although their place in the geological series cannot be determined, from the absence of a succession of rocks) appear to have belonged to a fauna allied to that of the Manchhars and to that about to be noticed. The species which have been determined by Falconer and Lydekker are :— Dinotherium indicum. Brahmatherium perimense. Mastodon latidens. Camelopardalis sivalensis. perimensis. Capra perimensis. sivalensis. Antilope, sp. Rhinoceros, sp. Sus hysudricus. Acerotherium perimense. The remains are included in sandstone blocks. The same, or a slightly higher, horizon is recognized far away to the north, not, however, by its geological position, but by the fossil remains, in the neighbourhood of Kushdlghar, forty miles south of Attock, in the Punjab. The deposit contained, according to Falconer and Lydekker, the following genera and species :—Dznotherwum CORALLIFEROUS SERIES OF SIND. 205 pentapotamic, Mastodon, Listriodon pentapotanie, Rhinoceros, Merycopotamus, Dorcatherium, Sanitherium Schlagintweitii, and Amphicyon palemdicus. It is remarkable that all these horizons should contain Dzno- therium. Fourthly, the Sivalik clays, sands, and conglomerates above the Ndahun series present an assemblage of genera and species second to none in importance, and much grander generally than the existing fauna, which is very slightly represented. A number of genera which had lived in the Miocene elsewhere were associated with genera not known in mid-Tertiary deposits, and which have lasted either into Post-Pliocene or into Recent times. The Dinotheria are absent. Omitting the common Miocene genera and stating the others, an African element is noticed, as in the instance of Pikermi in Greece. The characteristic genera are Paleopithecus, Macacus, Semnopithecus, Stegodon, Elephas, Lowodon, Hippopotamus, Camelo- pardalis, Camelus, Bos, Bubalus, Bison, Cervus, Equus, Canis, Ursus, Mellivora, Meles, Lutra, Enhydriodon, Tapirus, Hystrix, Mus, Rhizo- mys, Crocodilus, Gavialis, Varanus, Ballia, and Emys. Moreover many of the freshwater Mollusca are identical with recent forms. It appears that these wide horizons are separable into an older serles, with Dinotherium, Hyotherium, Hemimeryx, Sivameryx, Hyopo- tamus, Anthracotherium, Acerotherium, and Manis, but without Elephas and the later bovines; and into a younger series, in which genera of an African type, such as Hippopotamus and Camelo- pardalis, are found, and without the older types. The first series is in relation with the Lower Manchhars, and the second with the Sivalik deposits above the Nahun beds. But the interesting fauna from Kushalghar appears to be probably of Nahun age; and if that should be proved eventually, the succession of two fairly distinct faunas, linked together by some species, will be evident. Finally, the Post-Pliocene deposits of the Jumna area, and also of the Nerbudda and other peninsular rivers, contain some species of the Sivalik horizon. V. General Considerations regarding the Age of the last Himalayan Upheaval. It might be supposed, considering the ready manner in which European mammalian faunas are placed in the scale of geological succession, that the contemporaneity of some of them with the Sivalik-Manchhar assemblage of vertebrates would be easily deter- mined, and a geological age given for the latter without doubt. But the critical examination of the positions in which many of the European mammaliferous deposits have been found, indicates that the manner in which the Tertiary deposits have been classified by means of the fossil Mammalia is open to exception. Two well-known examples of European mammaliferous deposits which resemble, in their succession, the Sivalik-Manchhar series are 206 PROF. P. M. DUNCAN ON THE those of Eppelsheim and Heppenheim and of Pikermi. At Eppels- heim sands in small patches are at the top; and they contain a few specimens of fossil Rodentia, Insectivora, and Carnivora. A con- glomerate is beneath, and contains the Dinotherian remains, and also those of Rhinoceros and Mastodon; and all are remanic¢. At the base of the series there is a clay with freshwater shells; and it rests conformably on a marine limestone; and this is of late Miocene age. : At Pikermi the deposit, the bones from which have been so ably described by Gaudry and commented on by W. T. Blanford, rests on a freshwater stratum of Phocene age, and a Miocene lacustrine series underlies the whole unconformably. Now at Sind as at Eppelsheim the underlying marine beds are Upper Miocene in age, and freshwater conditions prevailed subse- quently, during which the osseous remains were depoisted. What is the age given to the Dinotherium-conglomerate at Eppeis- heim? Carl Vogt, influenced apparently by the presence of Dinotherum, and regardless of any stratigraphical arguments, and not considering the important changes which had occurred in the area, in the relative level of the land and sea-floor, decided that the overlying freshwater beds are of mid-Tertiary age. But Credner, dealing with the subject more philosophically, and not being so much impressed with the presence of a genus which elsewhere is represented in Pliocene strata as by the evidence of the considerable mutations which had occurred in the physical geography of the district, and which had brought a marine deposit above the original sea-level, places the bone-bearing conglomerates in the Pliocene age. Thus one geologist associates the land and marine elements together, and the other separates them. | The comparison which can be instituted between the Manchhar and Sivalik deposits and those at Pikermi is very close; and, accord- ing to the ordinary rules of stratigraphy, if the osseous remains at the last-mentioned locality overlie Pliocene deposits, the animals which left their bones could not have been of Miocene age. Yet Prof. W. Boyd Dawkins, in his contributions to this Society, asserts that the Pikermi fauna flourished in the Miocene. Evidently there is a great diversity of opinion regarding the age which should be given to land-surface remains overlying marine strata, freshwater and estuarine deposits intervening. And if this subject is studied it will be found that great discrepancies of opinion have existed in regard to similarly placed deposits in many of the great formations. About the relative age of some superincumbent terrestrial remains there is no discrepancy of opinion. ‘The Coal-measures are associated with the underlying grits and limestones; and the land remains of the Inferior Oolite are similarly connected in classification with the marine deposits beneath them. When the lateral extension of the strata can be traced, and distant marine equivalents of the overlying series can be proved to contain fossils representative of or identical with those of the underlying series, the land-surface is classified with the formation in which the marine strata are placed. Or when there is CORALLIFEROUS SERIES OF SIND. 207 a succession, on the same area, of marine beds over the terrestrial and freshwater series, if the fossils of the upper strata resemble those of the deeply seated marine ones, the whole belongs to one great aspect of nature. It will be found that there can be no exception made to placing the Woolwich and Reading series out of the formation which includes the white chalk; but the propriety on any grounds of linking the Purbecks on to the marine Portland is open to exception. Except im instances similar to those of the Carboniferous and Inferior Oolite, and where there is also decided unconformity, the question can only be answered after a careful consideration of the amount and extent of area implicated in the changes in the physical geography which may fairly be assumed to have occurred since the underlying marine deposit was completed. The upheaval of limited marine deposits recognizable as belonging to a particular formation, and the accumulation upon them of freshwater deposits and conglo- merates, would hardly necessitate the belief in such a change in the aspect of nature as would warrant their classification under two great geological ages. But when the same phenomenon is witnessed over widely separated areas, and a conglomerate is followed by some thousands of feet of fluviatile and other freshwater strata (the wreckage of high land close at hand), it becomes certain that the physical change has been wide enough to admit of an alteration in the geological nomenclature. With regard to Sind, the Lower Manchhars usually rest conform- ably on the Gaj marine Miocene, and marine, estuarine, and fresh- water intercalations exist at their base and before the conglomerate is fully developed. In one locality the Manchhars rest on a greatly denuded surface of Gaj strata. Upheaval (slow, irregular, and on a very grand scale) occurred subsequently to the deposition of the Gaj series; and ‘a marine tract became estuarine, fluviatile, and a region of wearing of high land. Subsequently the enormous subsidence took place, doubtless almost synchronously with the deposit of the thousands of feet of the Mafnchhars ; and yet the sea never broke in: it was far away. The area of change was vast; and it appears to be unreasonable to associate all these deposits under one geological formation. The disassociation of the Manchhar and Gaj series is a necessity ; and the nature of the fauna, so singularly allied to that of Pikermi, necessitates its relegation to the early Pliocene time. In following up this subject it must be remembered that the Sivalik strata, the horizon of which is above the Lower Manchhars, have a vast vertical as well as horizontal development, and that osseous remains have been found in them throughout their height. The fauna as a whole has a later facies than that of the Lower Manchhars, and resembles, even in its African element, that of Pikermi. : On the ground of inferred equivalency with the Upper Manchhars, and of faunal alliance with the assemblage at Pikermi, it must be credited that the Sivalik strata are of Pliocene age. 208 PROF. P. M. DUNCAN ON THE There is another argument which has not hitherto been employed and which favours this theory. The Sivdlik strata rest on, or are in contact with (along a line of © fault), certain plant-bearing strata of the Sirmur series which are associated with the underlying Nummulite-bearing Eocene rocks of Subathu. The nature of the flora, so far as it has been examined, is not very foreign to that of India at the present time; and by the same kind of reasoning which asserts the separation of other terrestrial and underlying marine strata, these beds, called Kasaoli and Dugr’hai, may well be the remains of the Miocene land. Whence were the vast thicknesses of the sands, clays, and con- glomerates of the Manchhar-Sivaliks derived? ‘They represent a ruined mountain-chain in bulk; and they are found not only on the flanks of, but also within, the orographical systems of which they form parts. It appears from the study of the history of the Himalayas by Strachey, Stoliczka, Medlicott, and those Indian Geological Sur- veyors who have laboured so industriously of late years, that a low mountain-chain existed on the area after the Nummulitic age—the results of a post-Nummulitic earth-movement. It was probably a broad chain, and not a sufficient barrier to prevent the roaming of the animals on and over it in the subsequent geological age. This chain appears to have had an axis of old rocks; and the whole suffered denudation during the age of the Sivalik Mammalia. In Hundes Strachey found the great tableland (now at an altitude of from 14,000 to 16,000 feet) to consist of sediments filling up a basin in old rocks to the depth of 3000 feet—sediments which included osseous remains of animals that could not have traversed high and difficult mountain-ground. . Part of the Sivalik sediments were formed out of low mountain- ground by fluviatile denudation; but to account for the vast deve- lopment of the rest on the flanks of the present mountain-system, it is necessary to admit that the upheaval, and much of the north-and- south crush of the Himalayas, occurred part passu with the slow accumulation. Elevation and diminution of the breadth of the area progressed ; and doubtless much of the great crush which folded and often reversed the flanking strata was final. The Sivalik strata rest on the flanks of the chain and on old rocks within the range; and they were the youngest deposits affected by the mountain-making. Hence the Himalayas, as a grand system, culminated during and subsequently to the collection of these strata, which have been pronounced to be Pliocene in age. The Sivalik deposits in the tableland of Hundes are overlain by relics of the great glacialization of the Himalayas. Hence, before the vast glaciers of the glacial period accumulated, valleys had been Worn out and denudation had proceeded. So it is necessary to recognize that the culmination of the movements which developed the height of the Himalayas occurred in preglacial times and during the Pliocene age. CORALLIFEROUS SERIES OF SIND. 209: To complete the serial changes, it is to be remembered that a Postglacial fauna was found in the old alluvium of the Jumna, a wash-down of the Himalayas. With regard to the Upper Manchhars, they were implicated in the great orographical movement, which was contemporaneous through- out the extrapeninsnlar area of India; and they are of Pliocene age. The details regarding the succession of strata and of many of their organic remains are to be found in the publications of the Geological Survey of India, and in the communications to this Society by Strachey, Grant, Falconer, and others. The ‘Manual of Indian Geology, by Medlicott and Blanford, an abstract of the labours of the Survey, contains them in a condensed form. I am under great obligations to all those writers, and also to Messrs. Med- licott and Blanford for much unpublished information ; and in thus heartily acknowledging my obligations, I am glad to have the opportunity of expressing my assent to their conclusions regarding the age of the Himalayas. Discussion. The Presrprent remarked on the persistence of Mesozoic types in the Tertiary strata of the Himalayas. Mr. Branrorp expressed the obligations of himself and other Indian geologists to Dr. Duncan for his researches on the Corals. which they had collected. The base of the series of Sind consists of a limestone containing Hippurites; and above these Cretaceous beds are strata partly unfossiliferous and the representatives of the Deccan traps, the whole being overlain by the Lower Eocene and Nummulitic. This succession is shown both in the Laki and the Khirthar ranges. The Nari or Oligocene group is 5000 or 6000 feet thick ; its upper subdivision, which is much thicker than the lower, is of freshwater origin, and contains imperfect plant-remains. This is overlain by the Gaj and Manchhar. The unconformities which occur in the series are purely local. He was gratified to find that Dr. Duncan had arrived at the same conclusion as the Geolo- gical Survey of India as to the age of the Sivdlik beds. He replied to the opinions expressed by Prof. Boyd-Dawkins and Mr. Bose on this subject. Lieut.-Col. Gopwiy-Auvsten remarked upon the greater contortion of the Nummulitic strata in the Western Himalayas as compared with those of Assam. The Autor stated, in reply to the President’s remarks, that while the genera of Corals are remarkably persistent, the species are not. He bore testimony to the great value of the volume published by the Geological Survey of India. He doubted the value of the terrestrial Mammalia as fixing the age of the strata overlying the Kocenes. 210 C. CALLAWAY ON THE ARCH XAN 15. The Arncoman Guotocy of Anetusny. By C. Carraway, Hsq., M.A:, D.Sc., F.G.S. With an Appenprx on the Microscopic SrRuctuRE of some ANGLESEY Rocks, by Prof. T. G Bonney, M.A., F.RB.S., Sec.G.8. (Read January 5, 1881.) [Prats VIIT.] Introduction (p. 211). A. Description of Areas and Sections. i. Menai Anticline (p. 211). a, Gmeissic series. (1) South of Pentraeth. (2) Mynydd Llwydiarth. (3) Gaerwen to Menai Bridge. b, Slaty series. If. Llangefni “Syncline” (p. 215). a. Slaty series. (1) Llangristiolus slates and grits. (2) Llangefni conglomerates and shales. (3) Cerrig-Ceinwen slaty and calcareous group. 6. Gneissic series. (1) Coast section from Porth Nobla to Aberffraw Sands. (2) Railway-section from Ty Croes to Bodorgan. (3) Distribution of the subdivisions in localities to the north-east. IIY. Central Zone (p. 218). a. Slaty series. 6, Gneissic series. (1) Bodafon Mountain. (2) Section between Llangwyllog and Llanerchymedda. (3) Area south-east of Paris Mountain. (4) Structure of the Zone. ITV. Northern Area (p. 221). (1) Voleanic group of Paris Mountain. (2) Chloritic schists of Mynydd Mechell. (3) Llanfechell grits. (4) Rhosbeirio shales. (5) Sharply contorted group south-east of Amlwch. (6) Slates and limestones of Amlwch and Cemmaes. Y. North-western Area (p. 224). Section from Porth y defaid to Pen bryn ’r Eglwys. VI. Western Area (p. 225). (1) The Mainland. . (2) Holyhead district. (3) Rhoscolyn district. B. Summary of Results. 1. Distribution of the Rocks (p. 227). a, Gmeissic series. (1) Geographical. (2) Stratigraphical. b. Slaty series. Ii. Evidence of Age (p. 229). a. Relations to Palzozoic groups. 6. Relations to each other. c. Relations to other areas. (1) Caernarvonshire. (3) Shropshire. (2) St. David's, (4) Charnwood. Conclusions (p. 232). Appendix (p. 232). eee ee ee Quart. Journ. Geol, Soc. Vol. XEXVIL. Pl. Vill Fig. 6. Section ir lower Quarry, Nebo. Fig. 5. Section in upper Quarry, Nebo, jes = yy en ob s U Yf S S Uf) Yi Yi Fig.3. Section across lynydd Llyydiarth . we a a OT Oy Mi , ' . ‘ ‘ ‘ as v , he 4 i Quart. Journ. Geol. Soe. Vol. AMV Pl. Vi Fig. 6. Section in lowe Quarry, Nebo. Fy. 1. MAP of ANCLESEY Fig.2.; Generalized Section trom Holyhead Mountain to Menar Strats. ESE. 5.5.W. EE. Malllraeth Tlangefne Marsh Werrrre EMAL A= Ss with a, overlying Fred* Donyerlield Lith Lorudon 18 GEOLOGY OF ANGLESEY. PU: INTRODUCTION. [I wave to submit evidence in support of the conclusion that there are represented in Anglesey two Precambrian or Archean groups, each distinguished by well-marked lithological characters. They may be named respectively the Slaty and Gneissic formations. The ‘“‘ Pebidian ” schists, as recognized by Dr. Hicks, I include in the eneiss group ; but, in other parts of the island, I have worked out a great assemblage of fragmental felspathic rocks, which I believe to be of true Pebidian age. I shall also attempt to prove that the quartz rocks of Anglesey are included in the gneissic series. ‘The microscopic notes, kindly furnished by Prof. Bonney, are based upon the specimens only, and have all the value of independent evidence. The growing importance of these ancient rocks would seem to re- quire that the provisional term ‘“ Precambrian” should give place to a word which can be permanently used; and I have ventured to adopt the term “‘ Archean,” so widely employed by Continental and American geologists. To the words ‘“‘ Azoic” and “ Hozoic” there are obvious objections. A. Description oF AREAS AND SECTIONS. I. Menat ANTICLINE. For convenience I accept this name for the rocks lying between Menai Straits and the great fault which throws down the newer Paleozoic rocks of Malldracth Marsh, though the term is not strictly applicable. It is true that the strata on the south-east side of the area dip to the south-east, and those on the north-west side to the north-west; but between these two extremities the undulations are frequent, and some of the dips are probably overthrows. a. Gineissec Series. _ wo Varieties of Gnerss—The structure of the area is rendered much clearer by the recognition of two prevailing gneissic types— a dark micaceous or hornblendic variety, associated with chloritic schists*, and a greyish or light-red rock clearly foliated into an or- dinary ternary gneiss. in the Craig-yr-allor anticline ft the two varieties are also recognizable, the grey gneiss passing up through the dark type into the granitoidite. In the Menai district the dark schist holds the same relation to the grey variety, and the passage between the two may be seen at many points. 1. Gneiss south of Pentracth.—Next to the fine section at Gaer- wen #, the rocks south and east of Pentraeth require notice. The band coloured “ greenstone,” running from Tai hirion north-north- east to Plas-gwyn lodge, appears to be a highly-altered hornblendic gneiss, with a north-west dip. In the field to the south-east of the lodge is a small quarry of typical grey gneiss; and in the plantation. * Throughout the paper the terms “schist ” and “schistose ” are strictly con- fined to foliated rock. t Geol, Mag., March 1880, p. 119. + Geol. Mag., March 1880, p. 121. Dal C. CALLAWAY ON THE ARCH HAN a few yards to the north-west we find the dark-green schist in its usual position above, both dipping to the north-west. Following the strike to the north-east, the gneiss is seen to con- tain angular pieces of quartz, some of which are flattened in the foliation-planes, so as to appear like ordinary folia. 2. Mynydd Llwydiarth.—This elevated ridge, trending to the north-east for nearly a mile and a half, is composed of highly con- torted gneiss. The general dip is north-west; but in some places the beds undulate to the south-east ; and the thickness cannot be great. The rock is more quartzose than usual; and the quartz frag- ments become much more numerous,—the two facts suggesting the proximity of quartz land. The gneiss, rolling in frequent contor- tions, is finely exposed in low vertical cliffs ; and the angular pieces of white quartz studding and projecting from the surface produce a very striking effect. Both types of gneiss appear to be present;. but the felspar being less abundant, the rock is brought into closer resemblance to the micaceous and chloritic schists of Holyhead. On the north-western slope the gneiss is extensively brecciated, and the presence of a fault is further inferred from an abrupt transition to younger formations. At the foot of the slope, near the fault, the eneiss becomes somewhat granitoid, suggesting the commencement of the passage into granitoidite, which is seen so clearly in the Craig- yr-allor anticline. At the extreme north-east of the ridge, on Red Wharf Bay, the green schist is exposed for some distance across the strike to the east ; but south of Wern we come onto the grey gneiss, all the dips being north-west. Further to the east, at Bryn y Castell, the grey type is again seen, but with the dip reversed; and at Hafodty, still further to the south-east, the dark schist comesin again. This is a clear anticline, the grey gneiss exposed at the apex throwing off green schist in opposite directions. The structure of this district is shown in Pl. VIII. fig. 3. 3. Gaerwen to Menart Bridge.—Crossing the anticline from Gaer- wen, we find at the railway-junction that the dark schist has rolled over to the south-east; but half a mile to the south-east, near Llanddaniel, the series, represented by contorted quartzose chloritic schist, dips north-west. Exposures then are rare till we come to the Anglesey column, where dips are in opposite directions, but usually to the north-west. This rock is of the ordinary dark green type, occurring everywhere above the grey gneiss in both the Menai and Craig-yr-allor areas. It has been described by Prof. Bonney *. On the Straits, near the west end of Menai Bridge, green schists occur of the ordinary varieties. From the above facts it is clear that the rocks of the anticline belong to the gneiss series. Well-foliated gneiss occurs at numerous localities; and the green schist 1s so intimately associated with it, as an upper band, as to remove all shadow of doubt that the two types form one unbroken group. In this area I have not found a trace of any thing but true crystalline schists. The thickness of the group is not great. The foldings are nume- * Quart. Journ. Geol. Soe. vol. xxxv. p. 308. GEOLOGY OF ANGLESEY. 213 rous, the lower gneiss being brought up to the surface at many points. Probably the entire series is exposed west of Gaerwen, where the total thickness can hardly reach 1000 feet. An isolated mass of schist occurs south-west of Pentraeth. It is about a mile long by half a mile broad, and is bounded by faults on all sides. Carboniferous Limestone surrounds it on three sides. On the east it is separated from the main mass by a faulted strip of the formation now to be described. b. Slaty Serves, This wedge is composed of rocks of a newer Archean group. At Tan y graig, south-west of Pentraeth, is a greenish, coarse, slaty variety. In the fields to the north-west, a jaspery rock projects in small knolls through the turf; but its relation to the bedded type is obscure. Crossing a hollow to the east, we come to a ridge, the western slope of which is occupied by massive breccias. The fragments and the matrix are similar, and are composed of a grey and purplish dolomite (Note 40, p. 235). Further to the north-east, above Wugan bach, is a light-purple slaty rock, which in places is literally smashed. Between this point and Plas ewyn, along the ridge, the prevailing type is a purplish ashy rock, sometimes brecciated, and in places altered almost to a hornstone. The dip is high to the north-west. The breccias of this ridge are obviously connected with faulting. North of Plas gwyn, on the same strike, is a puzzling rock, which, under the microscope (Note 41, p. 236) appears to be a limestone rendered impure by volcanic mud. Slaty beds are again seen to the north-east of Pentraeth, dipping south-east. For the greater part of its length this mass is separated from the Menai anticline by a faulted strip of Paleozoic shale; but north- east of Rhiwlas it comes up against the gneiss, the fault between the two formations running along a hollow. The heavy breccias which fringe both groups point to excessive fracture and crushing. Il. Luaneernt *‘ SyNcLINE.” Crossing the Carboniferous strata of Malldraeth Marsh, we come to an altered and contorted group of rocks which bear a superficial resemblance, especially in their colour, to the Menai schists, and have hitherto been regarded as a part of the same group. But on a careful examination of the lithology of this series, fundamental differences are observable. Even my first day’s work in the district convinced me of the existence of Pebidian rocks of the St.-David’s type; and fuller working-out of the area brought to light analogies with other Pebidian localities. Slaty rocks occupy only the south-eastern side of the apparent syncline lying between the Carboniferous on the south-east and the granitoid band on the north-west, the north-western side being composed of gneissic strata. The two groups are brought together by a fault or faults; but as they dip towards each other, an appa- rent basin is formed (see section near Aberffraw, Pl. VIII. fig. 2). Q.J.G.8. No. 146. f°) aa C. CALLAWAY ON THE ARCHAZAN a. Slaty Serves. 1. Llangristiolus Grits and Slates—The lowest beds I have ob- served are exposed near the Llangristiolus turnpike. The rock is clearly bedded; but its lithology is obscure, and the jointing is so close that clean fractures are hard to get. It appears to be a quartzo- felspathic grit, altered into a sort of hilleflinta. Some bands are coarser, grains of quartz and ielspar being visible to the naked eye. A similar rock is common in the Pre-Cambrian rocks of the Chureh- Stretton district (Shropshire). A little further to the north, near Cerrig ddwyffordd, a quarry- section is very interesting. At the base is a purplish grey felspathic grit (Note 42, p. 235). This is overlain by a pale-green slaty rock of Charnwood type, associated with, and apparently passing into, a sort of porcellanite (Note 43, p. 234), singularly like a rock at Caer bwdy, St. David’s. ‘The Charnwood facies of these varieties is also noted by Prof. Bonney. The section is capped by a repetition of gritty bands. The dip is northerly, at from 40° to 50°. Similar rocks are exposed in neighbouring fields. 2. Llangefni Conglomerates and Shales.—TYo the west and north- west of the Llangristiolus sections are several exposures of a greenish conglomerate. ee is seen on the Holyhead road, near Waen She and north of Cerrig ddwyffordd, at the arrow mark, but is ver y fully ex- posed in the railway- -cuttings north-west of Llangefni, the enclosed fragments, weathering white, being clearly visible even from the rail- way carriage. The matrix is green shale. The pebbles are very varied, the following being the principal types :—quartzite, common ; | pinkish grit (Note 38, p. 235) of quartz and felspar; greenish fel- spathic grit; and green and grey hornstone. Of these, quartzite is the only ingredient which is certainly foreign to the slaty series ; the others are such as are found in some part of the group. The con- glomerate would then seem to have originated in contemporaneous denudation, such as is common in yolcanic rocks, together with the wearing-down of a preexisting land composed of quartzite. Interstratified with the conglomerate are beds of purple and green felspathic shale, with some bands of hornstone.. The resem- blance of these rocks to the St.-David’s types is unmistakable. The road-sections in this district display a prevailing north-west. dip; but the more complete exposures in the railway-cuttings reveal frequent undulations, so that the thickness is not great. A similar group, probably on the same horizon, is well displayed in the cuttings east of Bodorgan station, 6 miles south-west of Llangefni. The prevailing rocks are purple, green, and grey ashy shales, with some felspathic breccias. Conglomerates occur con- taining pebbles of felsite, in which microscopic examination (Note 39, p. 236) reveals a very interesting resemblance to modern lavas. Associated with the shales at Llangefni station is a thick bed of quartzose grit, such as might have been preduced by the denudation of an ancient quartzite. Still following the strike to the south-west, we find the slaty series well exposed at numerous points on the coast from Bodowen GEOLOGY OF ANGLESEY. 215 round to the Aberfiraw sands. On the shore near Bodowen, Llan- gefni conelomerate is underlain by grit. On the east side of Porth twyn mawr, it overlies slaty rock, and contains large unrounded pieces of it. The same conglomerate, repeated by folding, occurs at intervals as far as the Aberffraw sands. Quartzite occurs in lenticular bands in pale-green slate west of Porth Cadwaladr. This association is also very common in Northern Anglesey. : Near, or at, the base of the slaty series of this district are several masses of quartz rock of obscure origin. The most prominent of these is the craggy hill* called Craig fawr, above Llangefni. It 1s a massive white rock, in which the quartzite structure is still apparent. It is surr ounded by Carboniferous strata, and it probably formed an island in the Carboniferous sea. The same rock may be traced at intervals along the strike to the south-west. One conspicuous crag of it is seen at Bethel, north-east of Bodorgan station. I can only suggest that we may have in these bosses the remains of an ancient land, from which the Pentraeth gneiss derived its angular fragments, and the Llangefni conglomerate its rounded pebbles and its quartz grits. 3. Cerrig-Ceinwen Slaty Group.—These rocks lie to the north- west of the last group, with the same (north-west) dip ; and as they present important lithological differences, they can hardly be repe- titions. North east of Cerrig Ceinwen, a little south of the Holyhead road, are green and purplish slaty beds. Some bands are calcareous ; and the formation on the whole resembles the rocks forming the ridge at Wugan bach. Large nodules of jasper are abundant. Their origin is not absolutely clear; but 1 am disposed to regard them as included pebbles. The beds have a high dip to the north-west. On about the same strike to the south-west are pale green ashy bands. Close to Cerrig-Ceinwen church, to the north, are prominent ridges, composed of a greenish rock, so tough that it was difficult to obtain a specimen. It has a superficial resemblance to a greenstone, but is undoubtedly an indurated sedimentary rock, presumably an ash. It suggested the “‘ greenstone” of Clegyr Foia, St. David’s. A fine exposure in a quarry west-south-west of Cerrig Ceinwen, on perhaps a little lower horizon than the jasper conglomerate, deserves attention. The rock is typical of the slaty series of Anglesey. In the field it appears asa pale green slate or indurated shale, sometimes faintly banded. Comparing it with some of the Charnwood slates (Note 44, p. 234), the only difference I could detect was that the Anglesey type displayed a slight lustre indicative of incipient metamorphism. It sometimes passes into a rock in which the alteration has been carried further, sometimes into a sort of hornstone. ‘The dip is still to the north-west. At Bod enlli we are apparently on the horizon of the jasper conglomerate. The conglomerate itself is not exposed; but we have limestone, purple slates, and green breccias and shales, not unlike those associated with the j jasper. * Coloured ‘‘ greenstone” on the map. 216 C. CALLAWAY ON THE ARCH AN Over a mile to the west-south-west, at Cerrig engan fawr, there is a considerable exposure of a very tough, green, ashy-looking rock, similar to the variety north of the church. All the above rocks dip to north-west ; but a little further to the west, at Ty’n-y-buarth, the beds dip to south-east. The strata at this locality are highly contorted on a small scale, and are very similar to the gnarled rocks east and south-east of Amlwch, the alteration, as in that district, having proceeded further than at Cerrig Cemwen. In all other respects these altered rocks resemble the Cerrig-Ceinwen slates. ' b. Gnerssice Serves. The band of gneissic rocks is hardly a mile in width at Bodwrog, where it is separated from the Pebidian by a faulted strip of Pale- ozoic shale ; but it gradually expands to a breadth of over two miles in the coast section on the south-west. The fault separating it from the slaty series runs from near Bodwina, north of the Holyhead road, in a south-south-west direction, losing itself under the sands of Aberffraw. On the west of the sands are nothing but schists ; on the east are nothing but rocks of the slaty series. It will be necessary to describe two sections across this zone. 1. Coast Section from Porth Nobla to Aberffraw Sands.—On the coast, west of Porth Nobla, is a considerable exposure of greenish felspathic rock similar to the halleflinta seen on the strike to the north-east, near Ty Croes. It appears to form an anticline. On Porth Nobla it clearly dips south-east, and passes gradually up into grey gneiss. To the west it presents but slight indications of Stratification ; but as we proceed eastward planes of separation grad- ually appear, which, at first obscure, become more and more distinct, till the rock is undistinguishable from a true gneiss. I have long since noticed an unbroken passage between hilleflinta and gneiss in the Wrekin; and Dr. Hicks has recorded the presence of an incipient foliation in the hilleflinta (‘‘ Arvonian’’) of St. David’s. The grey gneiss soon passes up into quartzose gneiss, quartz-schist, and quartzite. These siliceous types, which are of much theoretical interest, occur in great force, and form the main part of the headlands of Mynydd baen and Pen y cnwe, which project between Porth Nobla and Porth Trecastell. On the west side of Porth Trecastell is a band of calcareous quartz-schist, or quartzose foliated limestone. The silicate appears to be tale or chlorite. ‘There is a break in the section at the centre of the bay; but as the cal- careous schist reappears on the east side, there can be but a slight break, if any, in the succession. Overlying the last-named rock is a considerable exposure of decomposed grey gneiss interstratified with bands of quartz rock, the whole being much contorted and broken. A greenstone dyke appears here. After a few undulations, the de- composed rock disappears, and ordinary white and grey gneiss, some- times rather quartzose, dips steadily to the south-east for some distance. Then comes in the well-known green schist of the Menai type, which is continued to the Aberfiraw sands. It is well GEOLOGY OF ANGLESEY. ATE seen on Porth gwyfen with the normal south-east dip. On the shore, near the church, an undulation brings up grey gneiss. Then the south-east dip is resumed; but at Trwyn du, just before reaching the Aberffraw estuary, the beds turn up to the south-east, and a syncline is formed. The sucgession described is shown in fig. 4. 52. Section on the Railway from Ty Croes to Bodorgan.—tThe suc- cession is substantially the same as the last; but it furnishes one or two additional facts. At Felin bont is the hilleflinta first noticed by Dr. Hicks, who rightly conjectured, as the last section proves, that it passed to the south-east beneath the schist at Ty Croes*, though he was in error in placing it above the granitoidite, which les at the very summit of the gneiss series. West of Ty Croes quartzose schist is exposed in the road. Coming to the railway-cuttings, we find to the east of the station a good section of grey quartzose gneiss (Note 34, p. 233) dipping south- east. This rock is one of the most prominent types seen west of Porth Trecastell. Grey and white gneiss are seen for more than a mile with the same dip. In the cutting from near Bodgedwydd to near Graig bach, the grey gneiss grows gradually darker and darker, and passes without a break into the dark green Menai type, which is continued to the end of the exposure. The average dip is 40° to south-south-east. Leaving the line, we find in a road-section north- east of Tre’ Iddon that the dip of the dark schist has changed to the north-west. Hast of this point we come almost at once to the slaty beds and conglomerates of the newer series. The railway-section is thus seen to agree precisely with the suc- cession on the coast. In some points it is less complete; but it dis- plays very clearly the passage of the grey into the dark schist, so well seen in the Menai anticline. 3. Distribution of the gneissic Subdivisions to the North-east. Halle- jinia.—This band gradually narrows towards the north-east, being cut out by the fault that brings the lower part of the gneiss series against the granitoidite ; so that at Gwalchmai it is hardly visible, and at the section; north of the Holyhead road the grey gneiss is in immediate contact with the granitoidite. Some of the gneissic rocks of Gwalchmai display a transition towards the halleflinta types. Quartz-schist. This zone is not well exposed near ‘'y Croes. To the north-east it is seen at Melin Ddrydwy andGlan’rafon ; but it may be best studied at Gwalchmai near the church, on the road to the south- west of the church, and on the Holyhead road. In the last locality the granitoidite is brought against it by the fault. The dip is to the south-east, except on the Holyhead road, where it is reversed. There is probably a broken anticline at this point. The prevailing * I was at first disposed (Geol. Mag. March 1880, p. 128) to associate this halleflinta with the granitoidite, since similar rock, together with quartz- felsites, is included in that group; but the coast-section has cleared up several difficulties. t Geol. Mag. March 1880, p. 124. 21 oa) C. CALLAWAY ON THE ARCHAAN type is a true quartz-schist, sometimes passing into a quartzose eneiss. As it was important to critically compare this rock with the quartz rocks of Holyhead and Bodaton Mountains, I have had four specimens cut; and Prof. Bonney appends full descriptions (Notes 25-28, p. 233). Grey Gneiss. From the section east of Ty Croes station, this band may be traced to the north-east, forming the low ridge from Tal y llyn to Gwalchmai. Passing to the east of Gwalchmai church, it 1s well exposed about Bodkin tae. and on both sides ef the marsh (Cors Bodwrog) and pool. At Bodwrog church it is in contact with granitoidite, the quartz-schist having in its turn been cut out. In the ridge to the east of the church the gneiss is interstratified with a highly crystalline limestone, or calcite-schist (Note 51, p.236). From the microscopic description 14 will be seen that this rock re- sembles the quartzose limestone of Porth Trecastell ; and it is pro- bably on the same horizon. The normal south-east dip is preserved as far as the old Holyhead road, north of the marsh ; but near here a change takes place, and black Paleozoic shales come up to the eranitoidite. Dark Schist. This zone is not well seen north of the Holyhead railway. The fault which limits it on the south-east gradually cuts it out, so that east of Gwaichmai the slaty series approaches the grey gneiss. The gneiss is exposed at Pen bryncele, dipping south- east ; and half a mile to the south-east, near the +2 milestone on the Holyhead road, we find green slaty rocks with a north-east dip. IIL. Centrrat Zone. This band extends across the island from sea to sea. It is the most complicated part of this broken and contorted area; but by steady attention to the lithology, assisted by some stratigraphieal indications, clear results may be secured. a. S laty Series. In following up the gneissic rocks towards Bodafon Mountain in order to ascertain the relation between the gneissic and quartz groups, I was surprised to find an area of slaty rocks between the two. At Plas Llanfihangel dark gneiss and granitoidite dip to the south-west; and a little further to the north, at Ma’n-addwyn, felspathic shales dip as if they would pass below the gneiss. The rock is of a typical St.-David’s type, and unquestionably pelongs to the younger series. High up on the south-east slope of the moun- tain are similar rocks dipping south-east. The dips in both cases being away from the quartz rock, it was natural to infer that the shales rested immediately upon it. Further examination proved that such was not the case. At the south-west end of the quartz ridge north-west of Tyn— ‘Wlidiart the flanking rock is a sort of hornstone, banded pink and green, of a common Pebidian type. Approaching the hill, the horn- stone is seen to dip away from it to the south-west; but surmounting the low ridge made by the. hornstone, the beds are observed to roll GEOLOGY OF ANGLESEY. 219 over and plunge directly towards the quartz at the very junction. A little further to the north, above the farm, the evidence of dis- cordance is, if possible, still clearer. In the quarry is a good section of a greenish slaty rock, with the beds striking directly at the quartz forming the ridge at a distance of about 20 yards. It is thus evident that the slaty series is faulted against the quartz group of Bodafon Mountain. b. Gneassic Series. 1. Bodafon Mountain.—This mass cannot be brought into direct stratigraphical relation with the gneiss group, for the reasons just explained; but the lithological resemblance of this rock to the quartz-schist of Gwalchmai and Mynydd baen is so close as to render the correlation highly probable. In hand specimens some of the varieties are undistinguishable. Prof. Bonney, who was not aware of my views when working on the specimens, groups a typical variety (Note 29, p. 233) from the summit of the mountain with the Gwalchmai schists. The rock of the western ridge, which is a little more schistose, is also closely represented in the south-western localities. The inlier, surrounded by newer Paleozoic rocks, which lies to the south-east is precisely similar to the eastern part of the hill. 2. Section between Llangwyllog and Llanerchymedd.—The sue- cession between these points is similar to that of the coast-section between Ty Croes and Aberfiraw; but the granitoidite is wanting in the latter, and the dip is reversed, so that the two sections repre- sent the two sides of a broken anticline. The granitoidite occupies a large area south of Llanerchymedd. Round Coedana is a grey variety, rather small-grained. To the south, at Rhydgoch, it is much coarser, consisting of fragments of pink felspar in a matrix of smaller bits of quartz, felspar, and chlorite. The older geologists would have called it a porphyritic granite. The dark: schist is exposed about half a mile further to the south, at Glanrhyd, being the band of “ greenstone” of the Survey. Prof. Ramsay originally suspected the metamorphic origin of this rock ; and, after microscopic examination,: Prof. Bonney is disposed to consider it a hornblendic gneiss, containing felspar, epidote, horn- blende, and some quartz, the felspar being decomposed, and the whole a good deal “‘ messed.” Any obscurity arising from such an altered rock is removed if we follow the strike to the north-east. To the south-west of Graig Ilwyd there is a good exposure of hornblendic gneiss interstratified with granitoidite. The gneiss sometimes passes into hornblende-schist. The hornblende is well erystallized ; and the alteration is very slight. The grey gneiss is well seen about two furlongs south of Glanrhyd, in a quarry north of the stream. The dip is high, to the north-west ; and the rock is of the ordinary type. No rock is seen on the road for two or three furlongs to the south ; but north of Llangwyllog church there is a slight exposure of hallefinia too small to display dip. There is room for the 220 C. CALLAWAY ON THE ARCH EZAN quariz-schist between this rock and the grey gneiss; but I cannot affirm its presence. 3. Arca south-east of Paris Mountain.—Vhe gneissic rocks of this district are on the north-easterly prolongation of the strike of the central zone; and there is no doubt that they belong to the same series: the dip, the succession, and the prevailing rock-types are the same. On the coast between Dulas Bay and the fault north of Porth lygan the succession is not satisfactory, owing to great disturbance. South of Porth lygan the shore is mainly occupied by sandstones and conglomerates, presumably of newer Paleozoic age, deposited on an eroded surface of gneiss, and sometimes wrapping round little promontories of the older rock. ‘The prevailing dip is easterly. On Porth lygan the gneissic series is well exposed. Grey gneiss, dark- green schist, granitoidite, and quartzite (all the types which occur on the strike inland) lie in confused, contorted and shattered masses. At the north of the bay the rocks are less disturbed, and dip to the north-west. Black Ordovician (Lower Silurian) shales are faulted down to the north, resting, in a clear coast-section, against the granitoidite. Between Llanwenllwyfo and Nebo the succession is more dis- tinct. Thin-bedded gneiss is seen at several points north-west of Llanwenllwyfo and north of Plas uchaf; the overlying green schist is well exposed around Rhos manarch; and the capping grani- toidite, with associated schist-bands, forms the ridge between Nebo and ‘'y-Newydd, the whole dipping to the north-west. In two quarrics south-east of Nebo there is well exposed a strong band of quartzite and quartzose conglomerate, similar to some of the masses on the coast. I have elsewhere* given reasons for believing that this is the equivalent of the quartzose conglomerate of Twt Hill; and I stated that in these quarries the rock is unconformably oyver- lain by black shales of (at least) Bala age. If the shales are Tremadoc or Arenig, as Prof. Hughes has since maintained, the fact is still more strongly confirmatory of the Archsan age of the quartzite. As the point is important, I submit a section (figs. 5 & 6) from each of these quarries. Since this quartz-rock is Precambrian, is conformably underlain and overlain by gneissic strata, and occurs on the coast in the closest association with gneiss and granitoidite, there can be no hesitation in referring it to the older Archean series. It must not. however, be confounded with the quartz-schist of Bodafon Mountain, which is lower in the succession. Granitoidite and dark schist are found along the ridge to the west, and at the base of Paris Mountain. I have not examined the latter locality ; but Mr. Allport kindly allows me to use some notes made by him in May 1868. The numbers refer to specimens now in the British Museum. “443, Fine-grained granite? South-east flank of Paris Mountain. “444, Granite? 100 yards from last. * Geol. Mag. March 1880, p. 118. GEOLOGY OF ANGLESEY. DN © 338. Hornblende-schist, passing into a more compact rock (339), not schistose, or very slightly so; and this, again, into hornblende and felspar rock, very similar in appearance to that at Malvern (340). ‘«‘This is an interesting case, as the above varieties were taken from different parts of the same large block. As in Malvern rocks, epidote occurs near the joints. No.340 might be called hornblendic eneiss; the felspar 1s well crystallized in many parts of the mass. In this place there is also hornblende and mica rock; and in some parts hornblende and felspar occur in separate layers. The granite has a gneissose structure; and to my mind the evidence is clear that the whole form a series of metamorphic rocks.” I am glad to quote confirmatory evidence by so competent a ithologist, and trust that it will be unnecessary to multiply retuta- tions of the old “‘ granite-vein”’ hypothesis. A, Structure of the Zone.—-For structural purposes it is necessary to include in this band the gneissic area between Ty Croes and Aberfiraw. The whole forms a complex and shattered anticline. A central fault margins the granitoidite on the east from Caernarvon Bay to a mile north-east of Bodwrog; then 16 curves a little to the east, and, passing near Llangwyllog, leaves room for the underlying gneisses. ‘The strata dip in opposite directions on each side of the fault. This dislocation thus appears to pass along the summit of the anticline; but it does so obliquely, permitting the south-east side of the arch to open out towards the south-west, and the north- west side to expand towards the north-east ; so that we have an almost complete section between Llangwyllog and Llanerchymedd as we have between Porth Nobla and Aberfiraw, but with the oppo- site dip. The large area occupied by the granitoidite may seem to require explanation. As this rock rarely exhibits bedding, its relations are frequently obscure; but, fortunately, in the district between Gwalchmai and Llechyn farwy the dark Holyhead schist, which immediately underlies it, is brought up by repeated undulations. Near Gwalchmai, for example, the granitoid rock dips north-west ; but to the west 1t is thrown off in opposite directions in the Craig- yr-allor anticline. The granitoidite is, then, a band of no great thickness, repeated by contortions. ‘There is also evidence of repe- tition by faults. The generalized section (fig. 1) represents the relations of the principal rock-groups in Anglesey. LV. NorrHern AREA. This district is bounded on the south by the curved fault which, starting from near Carmel’s Point on the west, passes round by way of Llanfilewin and Paris Mountain to the east coast at Porth-y- corwg. The rocks are very much contorted and broken; but there appears to be, on the whole, an ascending series from the fault to the north coast. The dip is generally to the north; but on the west it is to the north-east. 1. Volcanic group of Paris Mountain.—The chief types of this 922 C. CALLAWAY ON THE ARCH AN ridge are a felspathic breccia and a compact felspathic rock like hornstone; but both varieties are quite distinguishable from any other rocks in Anglesey. At the east end of the hill, by the Pen- sarn chapel, the breccia is well exposed. The fragments, some of which are a yard in diameter, are a very compact felstone; and the matrix is a sort of greenish ash. The rock, as a whole, reminded me of the compact breccias at Clegyr Bridge, St. David’s. Prof. Bonney has since examined a shde; and he is of opinion that it exhibits both perlitic and fragmental structure and is in all proba- bility a rhyolitic ash. The dip of the series is to the north, agreeing with that of the adjoining slaty group. About 100 yards to the east of the chapel, and therefore on the same strike, is a knoll composed of a pale-green and purple rock, very compact, almost like jasper. Under the microscope (Note 53, p. 236) it is seen to be a trachyte, probably from a lava-flow, and is, in Prof. Bonney’s judgment, as ~ modern in its appearance as the rock composing the enclosed pebbles near Bodorgan. The purple variety is not unlike some of the fiinty felstone common in the Precambrian of Shropshire. From litho- logical characters the Archean age of this group appears not improbable, though it must be conceded that it does not resemble very closely the volcanic group south of Bangor. Field work throws little ight upon the problem. On the north side the ridge is bounded by afault; and at the base of the southern escarpment black shales seem to pass conformably under the volcanic group. The two rocks appear within three or four feet of each other, dipping in the same direction into the steep face of the hill. There are, however, no signs of a passage between the black shales and the compact felspa- thic rock. The appearance of conformity might be produced by faulting, accompanied or followed by lateral pressure; and there are abundant proofs of both in the district ; but, under the circumstances, I hesitate to include this group in the slaty series. 2. Chloritic Schists of Mynydd Mechell.—This group is the “ foli- ated grit” of Ramsay. In the field it has the appearance of a quartzose grit, with chlorite covering lamination surfaces ; but Prof. Bonney, after microscopic examination, regards it as a true schist (Note 45, p. 234) of the Holyhead type, though he admits that the chloritic constituent is ‘rather minute.” These rocks are much contorted, and are frequently penetrated by dykes of felstone and dolerite running with the strike. At a higher horizon, at Cas Clock and west of Rhos-y-pill, the rock is an undoubted schist with silvery lustre (Note 46, p. 234). Some of the strata north-east of Paris Mountain are not unlike this schist, as may be seen at Cerrig, on the Amlwch road. It is important to observe that Prof. Bonney agrees with me in - recognizing (Notes under C, p. 284) a difference between these schists and those of the gneissic districts ; but in such a broken country it is difficult to prove that these rocks should go with the slaty series. My reasons for so associating them are the following :— (1) The rocks are not uniformly foliated. To the west of Cas Clock, for example, chloritic schist is overlain in the same quarry GEOLOGY OF ANGLESEY. 995 by hard grits. In the same neighbourhood are bands of hornstone, similar to ordinary Pebidian types, associated with schist. (2) These rocks appear to pass up through less altered varieties into slightly metamorphosed grits and shales, as seen south of Llanfechell and in localities to the east. (3) The alteration is not carried so far as is usual in the older series. (4) In the undoubted siaty districts, as west of Cerrig Ceimwen, the rock frequently undergoes partial metamorphism, and bands occur which might almost pass for true foliated schists. (5) This area is scored with intrusive dykes. I do not suppose that these are the cause of metamorphism; but they indicate the proximity of a source of heat, which may have been concerned in the change. 3. Llanfechell Grits—The altered rocks of Mynydd Mechell appear to pass to the north into a series of green grits, well exposed round Llanfechell. This rock has undergone much alteration, to which probably the colour is due. Microscopic examination (Note 47, p. 234) brings to light the important result that it is “ almost certainly derived from the older gneissic and schist rocks of this region of North Wales.” These grits are also seen on the strike to the south- east near Bodewryd. 4, Rhosbeirio Shales.—W est of Rhosbeirio church, and in a quarry near the farm, are very interesting exposures of felspathic rocks of the true St.-David’s type. These shales are soft, well laminated, and fine-grained; but here and there are thin seams of grit like the variety just described. While the body of the rock is but slightly altered, the gritty bands have a very schistose look. The prevailing colours are pale green and purple; but there are also some beds of a soft yellow shale, like a common variety in North-western Anglesey. The dip is to the north at 30°. Under this group, at Nant-y-cyntin, slaty beds of a more altered character, intermediate between these shales and the Cas-Clock schist, are seen. 5. Sharply contorted Group south-cast of Amlwch.—At Crogan goch and on towards Llaneilian are exposed the remarkable rocks described and figured by Prof. Ramsay. The foliated structure (Note 49, p. 234) of these slaty-lcoking beds is not very evident to the naked eye. Similar rocks compose the tongue-shaped promontory of Point A‘lianus; but sometimes contortion is wanting. At the base of the promontory the beds are gritty and partially altered (Note 48, p. 235). 6. Amlwch Slates and Cemmaes Limestones.—South of Amlwch, near Crogan goch, the last group, with the intervention of an ashy band, is succeeded by uncontorted pale-green chloritic slaty beds (Note 50, p. 235), which are continued to the coast and all along the west side of Bull Bay to Ogo’ goch. The dip is steadily to the north at a moderate angle; and the thickness must be considerable. Towards the west, on Porth wen, bands of limestone and quartz- conglomerate come in; and further west, on Cemmaes Bay, limestone and quartz rock predominate over the slaty beds. I am not certain D4 C. CALLAWAY ON THE ARCH.BAN that the Amlwch and Cemmaes groups are on the same horizon ; but the evidence appears to point in that direction. The limestone (Note 52, p. 236) is very similar to a rock at Llanfaethlu; and the associated slaty beds are alike in both localities; but it would be rash to correlate the groups. The northern area (omitting Paris Mountain) is thus seen to consist of chloritic schists, felspatho-quartzose grits, felspathic shales, chloritic slates, quartz-conglomerates, and grey limestones, with some limestone bands on more than one horizon. In such a shattered district a true succession.can hardly be indicated. North of Paris Mountain some of the horizons are probably faulted out of sight, and it is difficult to correlate the rocks of this district with those further west. Most of the groups appear to pass into each other, either vertically or laterally ; and I cannot avoid the conclusion that they all belong to the same epoch. V. NortH-wESTERN AREA. In Western Anglesey are two well-marked groups of strata, which I regard as the equivalents respectively of the Gneissic and Slaty series. They are brought together by a fault. Porth-y-defaid Fault.—This dislocation is situated on the west coast, about midway between the northern and southern extremities. It 1s well seen on the shore, striking inland to the east. Greenstone is erupted along the line of junction, and alters the strata for a short distance on each side. South of the fault are thoroughly crys- talline quartzo-micaceous and chloritic schists dipping north-west ; while close at hand on the north side green ashy shales dip north- east. This fault is the boundary between the two series ; for true schists with a south-west strike occur everywhere to the south, while a comparatively unaltered group, with a prevailing northerly dip, stretches right up to the north-west corner of the island. Section from the Fault to Pen brywv Eglwys.—Proceeding north- ward, we find at Trefadog ashy shales dipping north-east. In the quarries north of Llanfaethlu church is a good section of grey lime- stone and pale-green ashy slate with easterly dip. The former is com- pact, and apparently as unaltered as any Paleozoic limestone. A careful search revealed no trace of fossils. The slaty rock is of a common Anglesey type; and the whole group is like the Cemmaes series. About Llanrhyddlad exposures are numerous. At Porth Swtan, on Church Bay, is a considerable thickness of yellowish fel- spathic slate and breccia. The dip is north-east, with slight contor- tion on a small scale. Higher beds are seen in quarries at Rhyd Ngharad; the rocks are brecciated in part and very felspathic. Greenish colours predominate. One variety, a pale-green slate, contains cubic pyrites. On about the same horizon, at Ogo Lowry, are green brecciated rocks more highly indurated. The district between Ogo Lowry and Penbryn’r Eglwys is one of the most faulted parts of the island ; the section consists of alterna- tions of Archean and Paleozoic rocks, repeated by faults, and dipping to the north. The wedges of dark shale let in amongst the older group are very clearly seen in the sea-cliffs. The younger shale, of GEOLOGY OF ANGLESEY. 225 course, frequently appears to pass under the felspathic series. Such facts suggest great caution in trusting to dips in areas where the faulting is not clearly seen. The lithology here, as generally in Anglesey, is the true guide. The Precambrian chiefly consists of felspathic breccia. The fragments are usually of similar composition to the matrix; but some are of quartzite. The felspathic fragments weather out sharply where the cliffs are washed by the spray. On the whole this rock strongly suggests the breccia near Nun’s Well, St. David’s. Good exposures are seen at Yynys y fydllyn, Porth yr hwch, and west of Pant yr Kglwys. The thickness, owing to such frequent faulted repetitions, probably is not great. The section at Pen bryn’r Eglwys is of much interest. In this. great headland, which juts out at the north-west corner of Anglesey, are rocks which are described in the section of the Survey Map as ‘‘ oneissic rocks pierced by granite veins ;” and these are represented as passing on the south into “ metamorphic foliated rocks.” The latter are the felspathic shales and breccias already noticed. The description of the former also requires modification. The breccias pass up into the southern slope of the headland, where they are succeeded by a band of quartzite. This is over- lain by a considerable thickness of greenish felspathic beds of ordinary Pebidian type and not much altered. ‘The highest ridge of the promontory consists of this rock. Following these strata across the strike to the north, they are seen gradually to change, the rock putting on a glazed aspect, and mica appearing in small quantities on the lamination-planes. This altered material soon passes into thoroughly foliated gneiss and granitoidite. It was rather startling to find true metamorphic rocks associated with comparatively unaltered felspathic beds of the newer series. The granitoidite is very similar to that of the older gneiss; and my first impulse was to refer it to that group. In sucha shattered district as Anglesey, especially in such an area as its north-west corner, a faulting-up of the older series seemed not unlikely. But, after careful examination, I found it impossible to accept this supposition. The ‘transition between the unaltered rock and the schists is gradual and complete. Or 09 bO am 2,8 “We Ny v2 | _ a, Gneissic series ja | | a, Q.8 Discussion. Dr. Hicks agreed with Dr. Callaway that there are two Pre- cambrian series in Anglesey. He differed from the author, how- ever, in regarding the so-called granitoidite as constituting the lowest and not the highest member of the so-called Gneiss series. He stated that some of the breccias associated with the halleflintas contain pebbles of the granitoid rocks, and are therefore of younger age than the latter. He admitted, however, that some of the points must be regarded as in an unsettled state, owing to the faulted condition of the district. Prof. Ramsay argued against the principle of identifying rocks as of different ages by their mineral characters as studied by the micro- scope. He maintained that the altered rocks of Anglesey are the metamorphosed representatives of the Cambrian, because the un- altered Cambrian are found striking directly towards the altered strata, and both are overlain by the Arenig. Mr. W. W. Smyrtu also argued against the recognition of a number of different formations on mineral evidence alone without any aid from organic remains. He thought the so-called “ gneissose rocks ” differ widely from typical gneisses, and that the granitic series belonged to the class of ill-defined granite rocks known in Corn- wall as “bastard granite,” the plain Saxon of which was perhaps preferable to the cacophony of “ granitoidite.” He thought that the evidence brought forward was insufficient to upset the detailed map of the Survey. Prof. Bonnzy was inclined to agree with Dr. Callaway’s interpre- 238 ON THE GEOLOGY OF ANGLESEY. tation rather than that of Dr. Hicks, as the latter undervalued the amount of metamorphism the Anglesey rocks by the Menai Straits had undergone. He differed from Prof. Ramsay as to the value to be attached to the microscopic study of rocks; he did not regard the Bangor rocks as partially metamorphosed, and he found no trace of the progressive metamorphism insisted on by Prof. Ramsay. He defended the use of the term “ granitoidite,” which, though open to objections, was less misleading than the base-born term proposed by Mr. Smyth. He thought that the microscopic study of the rocks supported Dr. Callaway’s conclusions. Dr. Cattaway agreed with Prof. Ramsay as to the great value of the Survey work. He contended that the principle of the identi- fication of rocks by their mineral characters is a safe one when applied in the same district. The sections described proved that the granitoidite was the summit of the gneissic series. Many of the gneissose rocks were true gneiss of very typical character. The difficulty through faulting was counteracted by comparing areas. ON THE LIMESTONE OF DURNESS AND ASSYNT, 239 16. The Limzstone of Durnuss and Assynr. By C. Cattaway, Esq., M.A., D.Sc, F.G.S. (Read January 5, 1881.) Iv is well known that the late Sir R. I. Murchison based his deter- mination of the Ordovician (Lower Silurian) age of the greater part of the Scottish highlands upon the discovery by Mr. C. Peach of Ordovician fossils in the Durness limestone. The views of Mur- chison were shared by many distinguished geologists, and have been generally accepted as one of the most assured and important conclu- sions of modern geological science. Wishing to ascertain for myself the truth of this opinion, I devoted a portion of last summer to the study of the limestone at Durness and Inchnadamff. I selected these localities because they alone are alleged to have yielded fossils from the limestone, and because Murchison regarded them as of primary importance in the construction of his argument. My researches led me to the conclusion, not only that the sections were broken, and therefore untrustworthy, but that the relations of the several rock- groups were inconsistent with the supposition that the limestone passed below any part of the newer metamorphic series. On my return from Sutherland, I found that many of my observations agreed very closely with those of Prof. Nicol; and I am glad to bear testimony to the accuracy of an author whose work has hardly received due recognition. It is certain that he made out many im- portant points which escaped the observation of his distinguished rival; and I must, in candour, acknowledge that he has anticipated many of the results which I have now to submit to the Society. I can only claim to have ascertained some additional facts, which seem to me to strengthen the case against the received opinion. DuRNEss. a. The Limestone in Relation to the Flagqqy Gnevss.—In his later papers, Murchison admitted that the limestone was separated from the flaggy beds* by a “ great fault,” by which the “ upper quartzite ”’ was thrown down out of sight; but he, notwithstanding, held that the limestone was ‘“ overlain by the upper series of quartzose and gneissic rocks,” apparently on the ground that both dipped in the same direction, to the east. Admitting his facts, his reasoning is obviously fallacious, and might be employed to prove the most op- posite conclusions. But I cannot accept his facts. Neither the limestone nor the flagey group dips to the east. The true relations of the two formations are shown in the annexed plan (fig. 1), which is, in outline, a reduced copy of the new ordnance map. The most important dips are shown by arrows. * These flags are a true schist, though they are far less coarsely crystalline than the Lewisian. 240 C. CALLAWAY ON THE LIMESTONE Fig. 1.—Sketch Map of the Durness Area. (Scale, about ? inch to 1 mile.) Mh. vy NW “ h ‘ B WwW WNW ees CET TTN == lauartaite. ie Koiitis Flagey Gneiss. Limestone. EEE] Blown Sand. f. Faults. The fault f#, runs from sea to sea, east and west, exactly at the base of Far-out Head, which is entirely composed of thin-bedded gneiss and quartz-schist, dipping steadily to the north-east, except at some points a mile north of the fault, where the dip is conse- quently unimportant. West of the ruined church, where the limestone contains numerous fossils, the dip is north-east, so that 1t might seem as if it passed beneath the schist. But here the nearest exposure of the latter is over half a mile to the north-east, the intervening area being occu- pied by blown sand, so that no junction can be seen. Further ex- amination proves that this dip of the limestone is exceptional. Following the strike of the schist to the south-east, we trace it right up to the fault. Both rocks are here clearly exposed, the flags on the shore and in the adjoining field, and the limestone in a low cliff, which is an inland extension to the west of the precipice of Creag Chearbach. The fault runs along the base of the cliff. The flags dip uniformly to the north-east, the nearest section being only twenty yards from the limestone, which, at this point, dips east- | 22 S8050- Bay Gneiss. OF DURNESS AND ASSYNT, Q41 south-east. Several dips taken in this locality were the same, others veered round to the east, but none to the north of east. Following the cliff along to the west, and noting the prevalenco of south-easterly dips, we come, at about two furlongs east of Baile- nacille, to a small arch of limestone, with dips ranging from south- east to north-east. Such local variations are obviously unimpor- tant. In the area south of the fault the dips are almost uniformly to the south-east, as may be seen at many points between the inn and Sangomor. Coming back to the west of the old church at Bailenacille, we find the north-east dip soon changes to east, and then to east-south-east, and on the Kyle of Durness the limestone forms an escarpment over a mile long, striking south-south-west and dipping clearly to the east-south-east. At many other points in the area round Durness the same dip is seen. It is then certain that the mass of limestone lying south of the flags, and in immediate contact with them, dips south of east, northerly dips being quite exceptional, and then occurring only at a distance from the schist. As the latter dips steadily to the north- east, it is difficult to see how it can be conformable to the limestone. But a more comprehensive view of the district presents the re- ceived view in a still more incredible light. East of Sango Bay is the Smoo mass of limestone. Though separated from the Durness area by a faulted strip of gneiss, it preserves the same south-south- west strike. Hast of Sangomor it steadily dips east-south-east, at Smoo it is about horizontal, but west of Sangobeag it turns up to the south-east and a syncline is formed. On the promontory of Leirinmohr some of the dips are nearer south than east; but in no case have I observed in the Smoo mass the north-easterly dip of the altered series. : As Murchison himself admits that higher up the valley the limestone is faulted against the old gneiss on both sides, it is need- less to follow it further. b. Geiss of Sango Bay.—Between the two limestone areas is a band of schist. On the shore it is well exposed, not ‘“ thrust about in dire confusion,” but forming a symmetrical half-dome facing to the north-east. On the east side, within a few yards of the lime- stone, and dipping towards it obliquely (that is, to east-north-east, at 40°), is hornblendic and chloritic gneiss. ‘Towards the west the dip curves gradually round to north-east and north, and the gneiss is then underlain by dark mica-schist, which soon occupies the shore, the dip turning round to north-north-west, and finally to north-west, as if it would pass beneath the limestone which crops up in the shore close at hand, and is seen in the cliffs with its usual low south-easterly dip. There is no doubt that this strip is faulted (f, and f,) against the limestone on both sides. his gneiss is of the same kind as that which underlies the flagey schist on Far-out Head and east of Loch Hrriboll, and it must not be confounded with the Lewisian, which in this district is very massive, coarsely crystalline, and almost vertical inits dip. Accord- 242 C. CALLAWAY ON THE LIMESTONE ing to received views, this newer gneiss must overlie the limestone ; and, if so, the force which contorted the former must also have affected the latter. But though the limestone comes up to the gneiss on both sides, its gentle south-east dip is not changed. It would appear to be more reasonable to infer that the limestone was deposited on the contorted gneiss, and that the latter was subse- quently thrust up through the former between two parallel faults. c. The Quartzite——The bay between the limestone promontory of Leirinmohr on the west and the headland of Lewisian gneiss on the east is occupied by the quartzite, which is also seen sloping down from the fianks of Ben Keannabin to the bay and forms a small outlier on the headland just named. It dips uniformly to the north-east. In the western angle of the bay it is faulted ( f,) against the limestone. Both limestone and quartzite are crushed into thick breccias at the junction, and on the bank immediately above the limestone is seen dipping to the north-west, the strikes of the two formations being, as in the former case, nearly at right angles. To sum up these results, it is clear that the metamorphic rocks, quartzite and schist, have been affected by a force tilting them up to the north-east, while the limestone forms a syncline whose axis strikes to the south-south-west. How the limestone can hold a conformable relation to the altered groups is a problem which the advocates of the received view may fairly be called upon to solve. Though the metamorphic rocks occur in three distinct patches, it is probable that they form a true succession. To the east of Loch Kirriboll, the quartzite is overlain by gneiss of the Sango-Bay type, which is surmounted by the flaggy group on Loch Hope. In Durness the Sango-Bay gneiss must overlie the quartzite, since the latter rests on the Lewisian gneiss; and if the flags of Far-out Head were ~ prolonged on the strike to the south-east, they would overlie the newer gneiss. ASSYNT. The section on which Murchison placed chief reliance is the suc- cession on the south-west slope of Cnoc an drein, above the church. As I entirely differ in my reading, I submit a section of the ground (fig. 2), in which I have carefully excluded hypothesis and have simply inserted the facts observed. Fig. 2.—Section above Inchnadamff Church. S.W. ) NE. L. Limestone. DL. Dark Limestone. F. Felsite. WL. White Limestone. Q. Quartzite. 7. Fault. — a ee OF DURNESS AND ASSYNT. 943 Just above the road we come to a band of dark limestone, which is succeeded by a white zone. Beyond a little hollow, quartzose flags are followed by similar dark and white limestones, which are probably a repetition; then in succession we have felsite, lime- stone, and felsite. Hitherto the dip of the bedded rocks has been to the north-east. Above the felsite is a considerable thickness of quartzite dipping east, so that if it overlies the limestone it must be unconformable. Higher up the quartz-rock dips for some distance to the south-west. Approaching the summit of the hill, felsite appears, and just beyond is massive quartzite dipping steadily to the south-east for a considerable distance. It is obviously unsafe to base a succession upon such a broken section as this. Passing over the intrusions, it is certain that more’ than one fault * occurs, and that the strike of the limestone is at right angles to the main mass of the quartzite. But a still more decisive refutation of Murchison’s views remains. Following the strike to the south-east it is seen gradually to recede from the quartzite ridge, while opposite dips by degrees come in on the northern side of the limestone band, which thus expands in the plateau of Stronchrubie into a broad syncline, the northern side of which dips away from the quartzite, that is, to the south- west. Towards the ridge the limestone dips grow steeper and aft last approach the vertical. Climbing the face of the mountain above, the quartzite is seen to dip to the north of east, so that the dips of the two formations are in opposite directions. As the northern side of the syncline gradually disappears towards Cnoc an drein, it is obvious that it is cut out by a fault, so that the southern side of the basin is brought against the quartzite and appears to dip below it. ‘These facts are illustrated in fig. 3. Fig. 3.—Plan of Limestone and Quartzite at Inchnadamff. wy AN Ca ae * Almost certain at F, and F,. 244 C. CALLAWAY ON THE LIMESTONE There is another difficulty in the reception of the old view. According to Murchison, the limestone is conformably overlain by the great mass of quartz-rock which rises into the lofty peaks of Ben More. If so, then the limestone basin of Stronchrubie must have been covered by an equal thickness of quartzite. It is for the followers of Murchison to show how denudation could have cleared off the entire mass of such an intractable rock as quartzite from the limestone, and yet have made so little impression upon the mountain- ridges which overhang. The quartzite which slopes down from the south-easterm face of Queenaig passes up through a continuous quartzite ridge into Ben More. It is not pretended that the limestone occurs in this line of section; but it is very singular that it should have thinned out just where its presence would have been of decisive value. One important and suggestive fact remains. At both Durness and . Assynt the limestone forms a symmetrical basin, the axis of which does not agree with the normal strike of the rocks which are sup- posed to overlie, but which coincides with the axis of the existing valley in which vt les. In Assynt the axes strike to the south-east, at Durness to the south-south-west. I do not at present press these conclusions beyond the districts observed, but submit them as an instalment towards the solution of a great question. Discussion. The PrestpEnt said he had twice visited the area, and felt diffi- culties in tracing the succession of the rocks, especially as to the asserted recurrence of the quartzites. The fossils proved the lime- stone to be of about Arenig age; certainly the limestones seemed to lie in a synclinal basin on the quartzites. Prof. Jupp said that, after several visits to the district, he felt great difficulty in offering an opinion as to the succession of beds ; he thought the simple sections commonly drawn to illustrate the geology of the country did not hold good. He considered that Prof. Nicol had made good his position as to there being only one quartzite and one limestone; but as to the relation of these to the gneiss, there was great difficulty in arriving at a satisfactory con- clusion. Dr. Hicxs agreed that the limestone series sometimes lay in synclinal folds of the quartzite; but he thought there might be two quartzites, as in a case he had recently noticed in another region along the same line further south. He was thoroughly satisfied that the series was newer than, and did not underlie, the eastern gneiss. Mr. Huptrston observed that Dr. Callaway’s mapping of the Durness area coincided with Sir R. Murchison’s section as to the limestone being troughed, and also as to the reversed position of the two gneisses. Further south the quartzites and limestones ap- peared to be intercalated, as at Loch Erriboll, where Orthoceras had OF DURNESS AND ASSYNT. 245 been obtained from the quartzite, tending to show that the two rocks hang together. The Assynt series, though less distinct than that at Loch Erriboll, was not inconsistent with such an idea. The unequal tension of the two rocks might in part account for local discordance of strike. Dr. Cattaway could not agree with Mr. Hudleston as to his interpretation of the Durness section; such discordant strikes were incompatible with the idea of a conformable succession ; nor did he agree with his view of the Loch-Erriboll rock. He accepted the age of the limestone, Q.J.G.8. No. 146. S 246 R. ETHERIDGE ON A TRIGONIA FROM THE 17. On a New Srrcizs of Triconta from the Purseck Bens of the Vatz of Warpour. By R. Erazripes, Ksq., F.R.S., Pres. G.S. With a Note on the Strata, by the Ruv. W. R. AnpREws, M.A. (Read December 15, 1880). Earty in this year the Rev. W. R. Andrews, of Teffont Rectory, near Salisbury, placed in my hands what at first sight appeared to be a new species of Z’regonia from the Purbeck rocks in the Vale of Wardour. This new species was found by Mr. Andrews in the rail- way-cutting one mile west of Dinton Station in the Vale of Wardour ; and its geological or stratigraphical horizon is the “ Cinderbed ” so well known in the several sections in the Isle of Purbeck and else- where. This bed is here composed of hard grey marl and pale brown soft limestones, variable in their proportions. Ostrea distorta and casts of TZrzgonca abound in this Cinderbed, which here measures about 2 ft. 6 in. in thickness. The section accompanying this paper (fig. 2, p. 252) was constructed by Mr. Andrews, and most correctly represents the succession of the Lower and Middle Pur- becks at the spot in the vale whence the new Trigonza came. It was from this marine “‘Cinderbed ” at Durlston Bay that Hemicdaris purbeckensis, Forbes, was obtained, associated with Ostrea distorta, a Trigonia like 7’. gibbosa, Cardium Gubbsir, Perna sp., and remains of fish. No other locality or section of the Dorsetshire Purbecks has yielded the above. ‘There cannot be any doubt that this 7ri- goma is not T. gibbosa. The Rev. O. Fisher obtained T'rigonia gib- bosa(?) from the middle Cinderbed of Fovant, Wilts; and the Rey. Mr. Andrews has obtained four other specimens of the same species from his immediate area; so that the genus is well represented in the Middle Purbeck of the Vale of Wardour. The Lower Purbeck beds of the vale lie below the datum of the railway and the level at which many of the quarries are worked, the “‘ Juxeat ” beds being about the lowest seen. The beds exposed in the railway-section near Teffont include the base of the Wealden, all the Middle Purbecks of the Vale of Wardour, and the top of the Lower Purbecks. The middle group here is about the same thickness as the beds in the Isle of — Purbeck ; those at Ridgway Hill and Mewps Bay are about 50 ft. ; and at Worborrow Bay the middle series measures 90 ft., and at Durlston Bay 150 ft. There is no Upper Purbeck in the area under observation ; the hard marl with eroded surface is the highest member of the Middle Purbeck, the succeeding 17 ft. being probably Hastings Sand at the base of the Wealden. Description. Section Glabree. TRIGONIA DENSINODA, Ether. (Fig. 1.) Shell greatly elongated or lengthened posteally, the anteal or PURBECK BEDS OF THE VALE OF WARDOUR. 247 anterior side rounded and somewhat steep; the base (inferior or ventral border) flattened; the postero-dorsal (umbonal or superior) -border concave and much elongated. Fig. 1.—TZriyonia densinoda, Ether. Umbones pointed, small, but prominent or elevated and slightly recurved, antero-mesial, or placed within the anterior third of the valves. Area moderately wide, flattened and filled with reticulated costelle, which are minutely and delicately nodulated or tuberculated ; it is bounded by two carine, the outer densely or closely nodulated, in- creasing in width from the umbo to near the extremity of the pos- teal side, then merging into the system of tuberculated transverse costee ; or the extreme posteal border possesses many irregular lines of growth due to the elongation of the tubercles. ‘The inner carina, or that bordering the escutcheon, consists of a series of minute tubercles which are immediately surmounted by the tranverse wavy ridges of the escutcheon. The median sulcus, or mesial furrow, is well defined. The escutcheon is remarkably large, having trans- verse Wavy rugee resembling those occurring in the Neocomian Qua- drate. The shell (valve) possesses regularly and concentrically arranged tuberculated or knotted rows of coste; about 40 range from the umbo to the posteal attenuated border. Those tubercles upon the posterior third of the valve, below the area become elongated and less regularly arranged than those of the umbonal region, where they are closely concentrically arranged. Affinities and Differences. —In outward ornamentation 7’. densinoda s 2 248 EK. ETHERIDGE ON A TRIGONIA FROM THE mihi, much resembles 7’. tenwitexta from the Portland Oolite of Portland, Devizes, Crookwood and Tisbury; but the shell is more depressed and lengthened posteriorly than 7’. tenwtexta, and does not possess the antecarinal sulcus or space of that species, which occurs in all the known Jurassic Glabrae, and which is so essentially characteristic of the Portland group. The ornamentation is that of the Upper-Jurassic Glabree ; but the escutcheon, which is remarkably large and possesses transverse ruge, agrees with that of the Neocomian Quadratz, to which it allies the species. All the Neocomian Quadratze have the escutcheon ornamented, and therefore distinct from that part of the tuberculated Jurassic species in which itis plain or unornamented. Again, the Cretaceous Scabree have also ornamented escutcheons; but the latter are alto- gether different and far removed from the Jurassic Trigome. I therefore regard this new species of Trzgonia as atransition form, tending to connect the Portlandian Glabre, through its ornamenta- tion, with the Neocomian Quadrate, through the ornamented and peculiar or characteristic escutcheon. Its stratigraphical position at the highest portion of the Jurassic system indicates a zoological change in progress as regards the genus Zrigonia, constituting an important feature in the history of the genus, no other example of which, so far as I am aware, is known in rocks representing the Purbeck formation, either British or continental. This new species falls under the group or section Glabre, founded by Agassiz upon very insufficient “data, and described by him as beivg without ornamentation, having no tubercles or costa. The shells of this group are usually ‘“inflated-ovate or ovately-oblong ; and the area is only slightly separated” from the rest of the valve. Mesially, or anterior to the position of the marginal carina, there is a smooth space which commences at the apex (umbo) or near it, and gradually widens downwards to the posteal border. This smooth space is more depressed than any other portion of the valves, and is distinctly impressed by lines of growth. In this group the “ anteal ” portion of the valves possesses costee more or less prominent, usually small, closely arranged, and either plain or tuberculated. The number of species belonging to this sec- tion is seven, six of which are Jurassic, the remaining form being Cretaceous (2. excentrica, Park.). Our species is closely allied to those forms known in the Portland rocks under the name 7’. gibbosa, Sow., 7’. Damonana, de Lor., 7’. Manselli, Lycett, and 7’. tenwitexta, Lycett, and. especially to the last named, from which, however, it differs in form and characters. Nore on the Strata. By the Rev. W. R. Anprews, M.A. Mr. Andrews has drawn up the following description of the Purbeck beds of the Vale of Wardour ; and any of the sections would suffice to read aright the succession and position of the fossil under consideration. PURBECK BEDS OF THE VALE OF WARDOUR. QAO Mr. Andrews says, “the Purbeck strata occupy an area of from five to six miles square at Teffont Ewias, Chicksgrove, Lady Down, &c. in the centre of the Vale of Wardour, and rest conformably upon the Portland group, sometimes, but not always, separated by a thin band of dark clay. These estuarine deposits have been consi- dered to belong to the Jurassic system, in consequence of their pre- senting remains of animal life nearer in affinity with the Jurassic period than with the Wealden above, and also from the fact that the marine Portland beds pass into the freshwater Purbecks without un- conformity, and generally with an ancient land surface at or near the junction ; the same area too (for the Portland beds are always capped by the Purbecks) which formed the bottom of the Portland sea, became when raised, first the support of the ancient forest, and then, when again slightly depressed, the delta of some great river, and this without any unconformity or denudation, which would imply a great lapse of time. “ Plambee ese. ee ceceivwcsee 2d Protozoal .m.h-secmer. teers aah alaral Sele PEO OZOR Aad tassadesoaeat onesies Ti OZOA Oks v.05. .clotos setae SA Aa sce Mean eae le kee PACHIMOZOM SG ssc ose sssscheccess soe Hiehimodermata..........secsss- 9 8 2 PARIMVE NI G reieciS5stsi sic sv eiselddsenieick CS) TE ee es Z PU RUMCEACEA. 9... ccc acseaeasieesleecec LZ) UG) |) ete ea eee I BGVOZOA Mec eccctvactnctesddecet. 3 © ee 2 2 3 7 2 ale btaehiOpOda.<........7-.-r0+- +s: Ne Wsh |) A : Bollietael le A Bamellibranchiata ..........:. A | Or Besee. | lhag 2 | 1 1 2 GaAsKerONOCS .. 05 a.icasde es dear. 4; 4] 3 z meh 2 1 2 2 Peimeecopodari2.. :sstikeeeii:. DAN fo) Z z e 1 i Te ELC LCLOPOGa4....2:... sesc.08eeees 1| 6 ae ilpcee aleges MIC CPUMOPOUA 3.:...0c.-25 +e. 52 Missa line 5) 3 zt A real - A 31 26 Te (PSG 8 ces aed a a G33 ISD) ea ea ee I) a 116 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. LiANDEILO. It may be thought by some superfluous to discuss the history of the Llandeilo rocks proper; but it is part of my purpose to do so, for two reasons :—first, to show their true relations (affinities and differences) to the Arenigs, which they overlie and with which they have long been confounded ; and, secondly, to show the geographical distribution of .the Llandeilo species and their paleontological rela- tions to the Caradoc rocks above and the Arenigs below. The three classical localities for the Llandeilo rocks and fossils are Llandeilo, Builth, and Abereiddy Bay north of St. David’s. Each of these areas possesses a characteristic and definite fauna ; on the whole the species in each locality are much the same; still the facies is peculiar in each. It is also of much interest to clearly understand the true position of the great band of dark roofing-slate of the ‘ Ffestiniog quarries” &c. that underlies the Caradoc or Bala series, occupying the position of the Llandeilo flags. “It is probable that the lowest portion of the Llandeilo group may occur in the Arenig and Ffesti- niog mountains, and also in the Arrans ;” but the evidence is slight and the materials (in the shape of fossils) are scanty. Salter obtained Bellerophon perturbatus in dark slates near Bangor, and also other fossils in the slates, identical with those of the Llandeilo rocks, ‘‘ which overlie the Arenig porphyries.” It is doubtful if the black slates of Anglesey are the equivalents of the dark earthy slates that range east of the Arenig mountain. The Anglesey slates contain Asaphus Powis and Phacops apiculatus, both rare in the Llandeilo group; but the Asaphus occurs both in North and South Wales, Phacops apiculatus only n Anglesey. The Graptolites, through Climacograptus celatus(?) and Didymograptus Murchison, distinctly enable us to refer the beds to the Llandeilo group. The Llandeilo rocks and their fossils in North Wales are confined to few localities, and are but feebly exposed anywhere; and the whole fauna of this formation in that area numbers 47 species. Highteen of these belong to the prolific class Brachiopoda, 15 to the Crustacea. The Llandeilo rocks of South Wales contain 88 species; and about 40 are common to North and South Wales. Itis important to show those species which unite the two areas ; but for my purpose I name only the Crustacea (Trilobita), Hydrozoa, and Brachiopoda. Of the former class 8 species are common, viz. Asaphus tyrannus, A. Powisii, Ogygia Portlocku, Calymene cambrensis, C. Blumenbachii, Trinucleus concentricus, T. favus, and the Ostracode Beyrichia complicata. | | The Hydrozoa are not less important; they are Duplograptus folr- aceus, D.dentatus, Didymograptus Murchison, Leptograptus flaccidus, Dicranograptus ramosus, D. formosus, D. Nicholsom, Climacograptus bicornis, C. celatus, C. confertus, CO. Sharenbergr, and Dicellograptus sextans. The Brachiopoda number still more, as we should expect from their habit and relation to the sea-beds upon which they live. They are all common and well-known species; and 8 of the 13 forms ANNIVERSARY ADDRESS OF THE PRESIDENT. Le? belong to the genus Orthis: they are Crania divaricata, Siphonotreta nucula, Orthis vespertilio, O. turgida, O. Actonie, O. calligramma, O. striatula, O. elegantula, O. insularis, O. bifoveata, Strophomena rhomboidalis, Leptena tenussimistriata, and L. sericea. Cardiola interrupta is the only Lamellibranch out of the known 6 species, and Bellerophon perturbatus and B. bilobatus the only 2 Hete- ropoda out of 7, which, with Favosites fibrosus, help to make up the 41 species connecting the North and South Wales Llandeilos; so that, of the 47 North-Wales species, 40 are common to both areas. No less than 88 species occur in South Wales. The whole Llan- deilo fauna for the British Islands numbers 80 genera and 175 species ; of these ] may mention that Ireland has yielded 51 species, Scotland 66, and Shropshire only 26: these latter are chiefly from the Shelve district, and are mostly Crustacea and Brachiopoda. Our knowledge of the number of species occurring in Scotland is due to the researches of Mr. Lapworth in the Moffat area. In the St.-David’s area and throughout Pembrokeshire, Brecon- shire, and Caermarthenshire, the typical localities of Llandeilo and Builth &c. contain a special fauna. At Abereiddy Bay the black slates and argillaceous limestones are interstratified with felspathic tuff; they rest upon the Arenig group, having a distinct fauna; a thick bed of tuff divides them. The immediate presence and abund- ance of Didymograptus Murchisom and Dicellograptus, Cryptograptus, &¢e. in the slates on the south side of Abereiddy Bay above the fel- spathic tuff, at once clearly marks or determines the base of the Llandeilo beds. Dr. Hicks divides the true Llandeilo series of St. David’s into 3 subgroups*, the lowest containing most Rhabdophora or Graptolites, associated with Trinucleus Ramsayi, Calymene Murchisone, and Theca caereesiensis, which especially characterize this lower division. The middle subgroup consists of dark “calcareous shales, with compact limestone at the upper part;” Asaphus tyrannus, A. pel- tastes, Calymene cambrensis, Ogygqia convexa, and Trinucleus Lloydu are the typical Crustacea, with Lingula granulata, Halysites catenu- latus, and 6 Graptolites. In addition to the fauna, the calcareous nature of the beds lithologically separates this middle subgroup from the, over- and underlying series. Everywhere it is charac- terized by the forms just enumerated ; many of the same occur at Llandewi Velfrey, Lampeter Velfrey, and Musclewick Bay. The Upper Llandeilo is also distinguished by special forms of Trilobita —Barrandia Corda, B. longifrons, Cheirurus Sedqwickit, Ogygia Buchu, Calymene duplicata, Ampyx nudus, and Agnostus M‘Coyit. It will be seen that each division is specialized or can be discri- minated by its Crustacea; the well-known Ogygia Buchu and Cheirurus Sedgwicki, here as elsewhere, are the typical forms in the upper black argillaceous slaty flaggy sandstones. The middle yields Asaphus tyrannus, A. peltastes, and Ogygia convexa, &c. ; and the Lower Calymene Murchisome and Trinucleus Ramsayi; and this grouping holds good wherever the Llandeilo beds occur. Three * Quart. Journ. Geol. Soc. vol. xxxi. pp. 177-180. VOL. XXXVI. 118 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. species only of Trilobita have occurred in the Scotch Llandeilo rocks, viz. Lllenus Bowmannu, Salteria involuta, and S. primeva. Ire- land has 5; but probably many more will occur. I enumerate them so far as we know them—Acidaspis James, Ampyx mammillatus, Ogygia Portlocku, Phacops truncato-caudatus, and Avglina mirabilis. I know no others, although in all the three Llandeilos we have 45 species of Crustacea. Shropshire has only yielded 11 species, and these chiefly from the Llandeilo rocks of Shelve: they are all marked species. In the Builth and Llandeilo areas no species occurs below the tuff-beds of the Lower Llandeilo. Few species of Graptolites are common to the Llandeilo and Arenig rocks of St. Dayid’s, al- though here the succession is so complete, and no less than 18 species there occur in the Arenig and 6 in the Llandeilo. Only 16 species of all groups are common to both formations, although the Arenig fauna numbers 149 and the Llandeilo 175, thus clearly showing the faunal differences. Llandeilo beds are known in Merionethshire, Caernarvonshire, Denbighshire, and Anglesey. Fifteen species have been determined from Merioneth; 13 from Garn, east of Arenig, by far the largest fauna in North Wales. Anglesey from five localities yields 15 species and 18 occurrences, or only 3 species for each locality ; Caernarvonshire, from four localities, 17 species (the chief loca- lity, Teddyn-Dicum, has produced 8 species); and Craig-y-Glyn, near Llanrhyader in Denbighshire, 6 species. No one can doubt, from the above results, that careful research would greatly add to the faunal value of the Llandeilo rocks of North Wales. Prof. Hughes has lately succeeded in ascertaining the presence of Tre- madoc fossils in Anglesey ; and further search at Garn, east of the Arenigs, could not fail to throw much light upon both the Arenig and Llandeilo faunas, and to substantiate still more the relation between North and South Wales. As yet, out of 13 Garn fossils, we know only 1 Trilobite (Ampya mammillatus). In Anglesey, near Llanerch-y-medd, out of 5 species collected 3 are characteristic Trilobites—Phacops apiculatus, Asaphus Powisir, and Calymene cam- brensis. It is the same with Treiorwerth in Anglesey,—clearly showing that a rich harvest of Llandeilo species is yet to be ob- tained in North Wales. The elaborate researches of Mr. Lapworth, in Scotland*, into the history of the Rhabdophora of that and other areas, has greatly modified our views respecting their distribution and range. The oldest fossiliferous strata known in Scotland are the Grap- tolite shales forming the well-known Moffat series, a group of black shales about 600 feet in thickness, and separated by Mr. Lapworth into three distinct paleontological divisions. The lowest (Lower Moffat or Glenkiln Shales) he determines to be of Upper-Llandeilo age; the Middle Moffat or Hartfell Shales above he correlates with the Bala group of North Wales “as an attenuated representative ;” and the third or uppermost (the Upper Moffat or Birkhill Shales) represents the Lower Llandovery. This last determination has an * Quart. Journ. Geol. Soc. vol. xxxiv. pp. 240-346. ANNIVERSARY ADDRESS OF THE PRESIDENT. 119g important bearing upon the distribution of the Graptolithina or Rhabdophora, it being well known that in the Llandovery rocks of Wales no Graptolites occur. The Lower Llandovery of Cardigan and Merioneth, of such great thickness and extent (yet without Grap- tolites), has ‘‘ dwindled down to the thickness of the Birkhill Shales (about 70 feet) in the intervening Lake-district, where it constitutes the Coniston Mudstones—a group of beds almost identical in thick- ness, lithology, and paleontology with the equivalent Scottish deposit of the Birkhill Shales” (Lapworth). The conclusion arrived at by Mr. Lapworth, that the oldest beds of the south of Scotland (the Glenkiln Shales) are the equivalents of the highest Llandeilo rocks, is borne out through his researches amongst the Moffat series in Southern Scotland; and, with him, we must come “to the conclusion that the Lower Silurian rocks of the southern uplands can be arranged in two distinct formations, viz. a lower and very - thin group of fine-grained Graptolite shales, and an upper and com- paratively massive series of arenaceous strata. In their mineralo- gical features and paleontological characteristics the Moffat series differ from any of the typical Silurian rocks of the principality.” Unlike the rocks of Wales, they are almost exclusively Graptolitie in their fossil contents, scarcely any Brachiopoda or Crustacea being known throughout the series. The Llandeilo age of the Lower Moffat shales is further determined through the total absence of those complex Arenig forms of “ Dichograpti, Tetragrapti, and Phyllo- grapti, so characteristic a feature of the Graptolitic fauna of the Skiddaw, Shelve, St. Dayid’s, and Canadian equivalents.” Mr. Lapworth, “ arranging the Moffat strata in the order in which they are displayed, and distinguishing each chief band by its peculiar fossil,” gives the following Table :— (Upper Birkhill | Zone of Rastrites peregrinus. \ ir or Grey Shale | Zone of Monograptus spinigerus. Soa group. Subzone of Diplograptus cometa. | Lower Birkhill oe 4 \ State Lower Birkhill |) Zone of Monograptus gregarius. Llandovery. f or BlackShale + Zone of Diplograptus vesiculosus. | group. Zone of Diplograptus acuminatus. _) pees eel Zone of Dicellograptus anceps. | irae Zone of Barren Mudstones. ; Hartfell 4 stone group. _ { Bal . Sh, i. \ Lower Hartfell ) Zone of Plewrograptus linearis. r i a“es- | or Black Shale | Zone of Dicranograptus Clingani. | { group. Zone of Climacograptus Wilsoni. }} Upper Glenkiln a Fe tiesto | Didymograptus beds. Upper Tteq, | Lower Glenkiln ) p. Llandeilo. Shales. SPAlos Ribbed mudstones. The facies of the Glenkiln shales is distinctively that of the Llan- deilo of South and North Wales, the assemblage being the same: differences occur dependent upon locality and the mineral composition of the beds; but the common Glenkiln forms of Scotland, and those species collected by observers in North Wales (Messrs. Salter, Hop- kinson, Homfray, Gibbs, &c.), serve at once to connect and correlate the faunas of the two areas. 10 species from South Wales and 12 from North Wales are all Glenkiln species and of Llandeilo age; k 2 I20 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. and, adopting the views of Mr. Lapworth, from the absence of Didymograptus Murchison and the genus Phyllograptus in the black shales of the Tremadoc area, and the presence of some succeeding Bala forms, we may infer that these dark Tremadoc shales and the Glenkiln beds are probably of higher-Llandeilo age. This inference is further strengthened by researches made by Linnarsson in the Llandeilo strata “of Central Sweden, where the greater portion of the Llandeilo formation of Britain is represented by the Orthoceras- limestone.” ‘‘ Upon this limestone rests the sheet of dark shales named by Dr. Linnarsson the ‘ Middle Graptolite schists; the lowest part of these schists, according to Dr. Tornquist, affords abundant examples of Phyllograptus and the Murchisoni form of Didymograptus. The highest Swedish beds assigned or referable to the Llandeilo formation afford both the peculiar Scotch or Glenkiln and North-Wales forms, such as Didymograptus swperstes, Lapw., Cenograptus gracilis, Hall, Climacograptus Scharenbergi, Lapw., C. perexcavatus, Lapw. Mr. Lapworth further discusses the relations and affinities of the Llandeilo species of South Scotland (Glenkiln Shales) with the dark shales and flagstones of the Cincinnati group which bound the valley of the Hudson near the. city of Albany and everywhere underlie the Trenton limestone; he shows reason for believing that we may regard them as forming the ‘“ highest division of the Quebec group (Arenig) which emerges unconformably from below the hori- zontal Trenton limestones. ‘The Hudson-River shales may therefore represent the higher Llandeilo beds of Britain.” Lapworth names 14 species collected from Norman’s Kiln in the valley of the Hudson, on the Marsouin river, &c., as being identical with the Glenkiln and Welsh forms. He still further establishes representative affinity between the peculiar Glenkiln species which are absent from the American strata and others which occur in them. Thus :— BrirTisu. AMERICAN, Glossograptus Hincksti is represented by Glossograptus ciliatus, Hmm. Dicranograptus ziczac bs » Dicranograptus furcatus, Hall. Climacograptus Scharenbergi ,, » Climacograptus scalaris, His. Clathrograptus cuneiformis ,, » Clathrograptus Geinitzianus, Hall. ‘‘ All evidence, therefore, tends to prove that the Glenkiln shales belong to the age of the Upper Llandeilo series, and are closely allied to the North-Wales grouping.” From the absence of certain Dzdy- mograptt and Phyllograpti from the Glenkiln division, and their abundance in the lower beds of the Llandeilo of Wales and Sweden, we are obliged to assign the Scotch Glenkiln Rhabdophora “ to the highest division of the Llandeilo, and not far below the base of the Caradoc or Bala’’*. It is of the Hydrozoa, Crustacea, and Brachiopoda, which together make up the 123 out of the 175 known species, that the mass of the fauna of the Llandeilo is composed. The same classes largely transmit their genera and species to the succeeding Caradoc or Bala * For an exhaustive paper upon the Graptolites of the Moffat series, by C. Lapworth, Esq., F.G.S., see Quart. Journ. Geol. Soe. vol. xxxiv. pp. 240-343. The general characters of the Lower Silurian rocks of Scotland and the strata of ~ the Moffat district are also ably discussed. ANNIVERSARY ADDRESS OF THE PRESIDENT. 121 rocks, viz. 16 species of Hydrozoa, 17 Crustacea, and 23 Brachio- peda, or, together, 56 out of the whole 76 that connect the two formations. It is interesting to know also those areas that are spe- cifically most prolific in the above groups, and thus help us to draw conclusions as to the nature of the sea and its bathymetrical condition during the time of the deposition of the Llandeilo and succeeding Caradoc, which are more intimately related than is generally believed. United, the Llandeilo fauna of North and South Wales is constituted by 94 species; or this is the known number for Wales, 88 occurring in South and 47 in North Wales, and 41 being common to both, chiefly among the Hydrozoa and Crustacea. Upon whatever hypothesis we may account for the physical relationship between the Llandeilo rocks of Scotland and Ireland, the connecting species are few (only 27); and no Crustacea are common to these two countries—which, out of a total fauna of 45 species, is remarkable. ‘The Irish species number 6, 3 of which are also South-Wales forms—Ogygia Port- locku, Calymene duplicata, and Aiglina mirabilis. As before stated, Scotland has hitherto yielded only 4 genera and 5 species of Crus- tacea—lilenus Bowmann, Salteria wmvoluta, S. primeva, Peltocaris aptychoides, and Discinocaris Browniana. Praxta.—None. Proroz0a.—Ischadites antiquus, Salt. No other Protozoon has been recognized in the Llandeilo; it is not a common genus in the Lower Silurian rocks. Hyprozoa.—The Llandeilo beds of Scotland appear to be the most prolific in Rhabdophora. Our knowledge of them is due to the researches of Nicholson, Lapworth, and Carruthers in Scotland, all of whom have done able work in elucidating the history and struc- ture of this obscure class. Forty-four species illustrate 18 genera ; but it is chiefly through the larger genera that this Graptolitic fauna is represented. South Wales, as we should expect, is also represen- tative, although poor, as compared with Scotland, having only 18 species, or about half the number. North Wales has 7 species, and. Treland 28. The Welsh and Irish genera have not the same specific value; generally they are much the same, but illustrated by fewer species. This may be due to want of systematic search in Wales, such as the rocks of Scotland have undergone by the above-men- tioned authors. The genera Didymograptus, Diplograptus, Climaco- graptus, and Dicellograptus are those chiefly occurring in South Wales; and Diplograptus, Dicellograptus, and Climacograptus in North Wales, &c. The Llandeilo rocks of Ireland contain species of nearly all the above genera. ‘Thus of the 44 known Llandeilo Rhabdophora we have Species. Ouse WMeS ee. 6.4. Sooke craye Bue 18 INO rb NWiale sues cc Bac a ee o SCO bam sees tie eR Bu Nereida See wi,.. & wa vets 28 85 occurrences ; and 16 species pass to the succeeding Caradoc. 122 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Actrnoz0a.—Halysites catenulatus, Linn., Favosites fibrosus, Goldf., and Monticulipora favulosa, Phill., complete the list of the Llandeilo Actinozoa. No English form is known. North Wales yields Favo- sites fibrosus. All three occur in South Wales, and Fuvosites fibrosus in Ireland. Scotland has no representative. All three species pass to the Caradoc. These are the first corals occurring in the British rocks ; but they expand to 20 genera and 42 species in the succeeding Caradoc. We cannot admit that change of condition only thus favoured the growth and development of the Actinozoa. It is true little limestone occurs in the Llandeilo, and that little in the upper division; but the organic remains in the Upper Llandeilo limestones are chiefly Crustacea. The increase of limestones and calcareous matter in the Caradoc rocks being due to life agency, we may attribute much of it to the secretion of carbonate of lime by the Coelenterata. The only notable or appreciable form is Halysites, which was abundant. Ecutnopermata.—We know of only 4 genera and 4 species of this class, which, like the Actinozoa, may be said to first appear in the Llandeilo. None are known in the Arenig below; but 4 species (1 Menevian and 3 Tremadoc) in fragments have been preserved in older formations. The 4 Llandeilo species are Actinocrinus Wynne, ichinospherites granulatus, Paleasterina Kinahan, and Glypto- crinus basalis. The first 3 are Irish; Glyptocrinus occurs in the South-Wales Llandeilos. Two of them passed to the Caradoc. As with the Actinozoa, this group became greatly developed in the Caradoc, no less than 15 genera and 32 species enriching that for- mation. ‘These, like the corals, must have lived in a moderately deep sea well charged with calcareous matter. ANNELIDA.—It is questionable if more than 6 species (although 9 are placed in the Llandeilo) occur: Paleochorda teres, P. major, and P. minor may be Arenig; they are very sparingly distributed, only 1 species occurring in England, South Wales, and North Wales, and 2 in Scotland. I know of no Irish Llandeilo Annelida. Crustacea.—In this class the Trilobita are of high significance, in both the Llandeilo and Caradoc formations, especially the latter, in which we know of more than 100 British species. The Llandeilo rocks have yielded 18 or 20 genera and only 45 species, or upon the © igs only about 2 species to each genus. Many genera are represented only by 1 species; these are Homalonotus (H. bisul- catus, Salt.), Lichas (L. patriarchus, W.-Kdg.), Stygina (S. Murchi- sonic, Murch.), Oheirurus (C. Sedgwickit, M‘Coy), and Acidaspis (A. Jamesti, Salt.). Much has yet to be done by collectors to increase our knowledge of the specific crustacean fauna of the Llandeilo rocks. South Wales being the classical locality for the Llandeilo group, it is natural that we should expect to find there the largest and most typical fauna, crustacean or otherwise. In one group (Graptolites) we have seen that this is not the case, Scotland being most repre- sentative. Here, however, the Crustacea of South Wales (29 species) are double those of North Wales (15), and ten times more than those of Scotland, which has yielded only 3 species, conditions ANNIVERSARY ADDRESS OF THE PRESIDENT, 123 being more favourable for the support and distribution of certain forms of life in one area than the other. The only 3 Scotch Llan- deilo species known to me are Jilwnus Bowmanii, Salt., Salteria wnvoluta, Thom., and S. primeva, Thom. The 6 Irish species are Acidaspis Jamesii, Ampyx mammillatus, Ogygia Portlocki, Phacops truncato-caudatus, Calymene duplicata, and Aiglina mirabilis. The English forms are chiefly from the Shelve area; and 8 out of the 10 are South- Wales species. The following occupy and characterize definite areas; they are single representatives of the genera named :— ae aoe England.| Scotland.) Ireland. Homalonotus bisulcatus...... * menostus Mi Coyi) s....-1.-..- Pe * Stygina Murchisone ......... : * Cheirurus Sedgwickii......... soo * Acidaspis Jamesii ............ Sas a wh Acts % Lichas patriarchus ............ "stg * 1 4 il Thus 6 Llandeilo genera are illustrated by only 1 species each, England and Scotland having no really marked or characteristic form. Nevertheless there are important Llandeilo species which are entirely restricted to that formation, not ranging either lower or higher. The most important are :— North | South | Wales. Wales, Hie Tana (Secktind | Ireland. Illznus Murchisoni ......... Ogygia angustissima ......... —— Portlockii ............... ——= COINCeENSIS .........0.000. AD WCHM ays slag ost eee aides COMVOMAN io. oe cc cece: Asaphus tyrannus ............ EMGASTES) seo askc.dsectewe- datveostatus.....-.sc-.9a | & = a se Classes. © 5 = a = Ss ae ; 2 ep = o Ay & 5) 2, = =| pics| ° LY ‘S) O wD = 3 = = cS B Es R = ° i) - Zz mM Plante. (PEOLOZOd: Wiaewae eee 1 1 a eerseeceeoree 4 5 1 4 1 3 Z 4 2 Hydrozoa 18 44 a : 18 aS 28 16 . 1 3 3 ACHINOZOR 2. e2s se 3 3 ae : 3 A 3 Echinodermata...... i 4 1 is } cae d 2 j 1 1 i 2 AMMeli Gay seacoast 6 9 2 i 1 3 sta 6 ie) 12 4 9 Crustacea ele Seiwidiolalai ovate 20 45 10 ; 5 29 5 é ry Bryozoa. ng ahi 2 6 8 4 4 Cf Brachiopoda ......... 10 34 r 18 25 9 9 23 Lamellibranchiata. . 5 6 fh A 4 1 3 2 1 2 2 Gasteropoda ......... 6 12 2 1 i 7 2 3 1 1 1 Pteropoda ....d24.065% 1 3 de B. A ies 7 : 1 1 1 2 2 Heteropoda ......... 3 a ; 2 2 3 i 4 Cephalopoda......... 7 ; 1 é 1 Wh 18 27 47 31 31 38 aogooosas 80 1 (Oo 26 47 87 66 51 73 Carapoc or BAta. The rocks overlying the Arenig and Llandeilo groups, and under- lying the May-Hill Sandstone, have received much critical attention from Sedgwick, Murchison, M‘Coy, Salter, and the officers of the Geological Survey. The facies of the fauna is that of the Llandeilo, but greatly developed. The increased fauna of the Caradoc is scarcely recognized until subjected to critical research and analysis. The difference in the fossil contents of the two groups is due chiefly to age and the conditions under which they were deposited. Sedg- wick, in 1853, placed the Bala group in his ‘‘ Upper Cambrian ” division, dividing it into Lower, Middle, and Upper Bala; it in- cludes also the ‘Cambro- Silurian of some authors, a term used or adopted by Prof. Jukes in his ‘ Manual of Geology.’ From the year 1832 to 1853 Sedgwick included the Caradoe in the term “ Lower Bala,” not ieee then the existence of the term Caradoc. Sir H. de la Beche, Prof. J. Phillips, and the Survey, from 1840 to the present date, designate this formation the ‘Caradoc or Bala.” Murchison, from 1831 to 1859, used the term “‘ Caradoc ANNIVERSARY ADDRESS OF THE PRESIDENT. 127 Sandstone.” The term Llandeilo was not admitted by Sedgwick either in the ‘ Synopsis’ or the Woodwardian Catalogue ; therefore great difficulty was and is still felt as to the identification of species long ago collected in certain localities, especially those having reference to the Lower “ Llandeilo of Murchison ” and the Arenig of Sedgwick; the faunas of all three groups require careful revision, in consequence of old or early errors, which of necessity demand it. Sir Roderick Murchison, in 1833, first noticed the ‘ Caradoc Sandstone ”’*; in 1834 the same rocks were described by him under the name of the “ Horderley and May-Hill Sandstone.” In the ‘Silurian System,’ subsequently published, these strata were called ‘Caradoc Sandstone,” the name being based upon their being highly developed in the neighbourhood of Caer Caradoc. In the early history of the Caradoc rocks the fossils of the Pentamerus-beds were included in its lists. The Survey subsequently corrected this, by separating the upper part or highest beds of the Caradoc. In 1852 Prof. Sedgwick and Mr. Salter in part corrected this error, and showed that the fossils of the May-Hill group were very distinct from those of the Caradoc Sandstone. Again, the upper Pentamerus-beds rest unconformably on the true Caradoc Sandstone, and the whole pass under the Wenlock Shale. In Shropshire and at Builth the unconformity is visible; and in the Malvern area, west of the Herefordshire and Worcestershire Beacons, the Upper Llandovery beds lie directly upon the Upper Lingula-flags or “‘ Dictyonema-shales.” Again, the Upper Llandovery beds on the banks of the Onny river lie on the upper part of the Caradoc or Bala beds. West of Wenlock Edge they cover the nearly vertical edge of the Cambrian or Longmynd rocks. “ Probably,” says Prof. Ramsay, “there is no unconformity so complete yet observed in other members of the British Silurian strata.” Besides the researches and large collections of fossils made by the Survey from these rocks, the labours of Sedgwick are preeminently associated with the Caradoc and Bala beds of Wales, through the great memoir on the British Paleozoic Fossils, in which he was so well aided by Prof. M‘Coy, and the “ Woodwardian Catalogue” ft prepared by Salter from the great store of materials in the Woodwardian Museum at Cambridge, and which, many years previously, he had helped Sedgwick to collect, and name, from the classical localities in Wales. To Prof. Ramsay, for his valuable memoir on the Geology of North Wales+, every student is deeply indebted. In this great treatise every detail relative to the physical history and distribu- tion of the Caradoc and Bala rocks and fossils is treated upon ; and to the appendix, originally compiled by Mr. Salter, I have greatly added, especially in that portion treating of the distribution of life, not only for the Caradoc, but through the whole of the for- * Proceedings of the Geological Society, 1833, vol. 1. p. 476. t ‘A Catalogue of the Collection of Cambrian and Silurian Fossils contained in the Geological Museum of the University of Cambridge.’ By J. W. Salter, F.G.S. With a Preface by the Rev. A. Sedgwick, LL.D., F.R.S. ¢{ Mem. Geol. Surv. of Great Britain, vol. iii. Geology of North Wales. 128 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. mations of North and South Wales. The portion of the memoir by Prof. Ramsay devoted to the Caradoc is of necessity large, arising from the extent, magnitude, and importance of the Caradoc rocks as developed in North and South Wales, Westmoreland, Ireland, and Scotland. I therefore, as in the case of the older groups, com- pare or attempt to correlate other portions of Britain with the typical area in Wales, so as to show the present aspect or distribu- tion of the Caradoc fauna, which is so largely developed in the British Islands. To devote much space to the purely geological aspect of the Caradoc would be needless under present circum- stances, as so much has already been done by able observers in the pages of our Journal, as well as in the exhaustive memoir on North Wales by Prof. Ramsay. My tabular results in the appendix to the above memoir, relative to the distribution of the Caradoc fossils both in time and space, will, when published, embody almost all the in- formation known. ‘The result, however, will be referred to here, as being a complete analysis of the 600 species. It is scarcely neces- sary for me to discuss the views of authors as to very minute sub- divisions of the Caradoc group in any given area or under any peculiar condition ; such will always arise under critical examina- tion, extended knowledge, or constant research ; and large as we know the fauna to be, owing to the rocks being so extensively worked in the days of Sedgwick and Murchison, and through the long-con- tinued labours of the Geological Survey, yet many of the zoological groups are still being added to. This is notably the case with the elaborate memoir upon the Girvan fossils by Prof. A. Nicholson and Mr. Etheridge, jun., the first volume of which, containing this addition to our knowledge of the Caradoc fauna of Scotland, is just completed and published. In this work the authors describe 41 genera and 65 species, many of which are new. Perhaps, with the exception of Mr. Lapworth’s paper on the Graptolites of the Moffat series at Girvan and Glenkiln* (mostly Lower-Bala or Llandeilo _ forms) no more important addition to our knowledge of the Caradoc fauna has been made since M‘Coy completed his great work upon the British Paleozoic fossils. The legends attached to the published maps of the Geological Survey, and the explanation of the colours employed to designate the horizons or formations surveyed, show that no attempt was made by the Survey to divide the Caradoc rocks into subgroups, or into Lower, Middle, and Upper Caradoc or Bala; neither does Prof. Ramsay, in that part of his memoir devoted to the Caradoc rocks, attempt any subdivision, but masses the whole group between the top of the Llandeilo and the overlying Llandovery. In the Catalogue of the Cambrian and Silurian fossils in the Woodwardian Museum, Cambridge, Prof. Sedgwick has divided the Bala beds into three subgroups, placing them in his Upper Cambrian stage (the Cambro-Silurian of some authors), the Lower Silurian of Murchison. The grouping adopted in this catalogue is such as to eliminate the true Llandeilo fossils, * Quart. Journ. Geol. Soc. vol. xxxiy. pp. 249-346. ANNIVERSARY ADDRESS OF THE PRESIDENT. 129 and place them in his Lower Bala group. The Bala group, or Upper Cambrian of Sedgwick, consists therefore of ( Lower Bala= Llandeilo flags (Upper Llandeilo of the | Geological Survey, the Arenig being the Lower). Upper | Middle Bala=the Caradoc Sandstone and Bala rocks Cambrian *. § (Geological Survey and Sir R. Murchison). | Upper Bala=the Caradoc Shales, Hirnant Limestone, | and Lower Llandovery rock (Geological Survey). The Upper Llandovery or May-Hill Sandstone unconformably over- lies these. Sedgwick’s Middle Bala, and part of the Upper, are therefore equivalent to the true Caradoc of the Geological Survey; and as such I treat that group here; or the Middle Bala of the Cambridge Catalogue will embrace the whole of the 600 species known as Caradoc forms. I have already discussed the Llandeilo formation proper in the sense now used by the Geological Survey, having relegated those species hitherto called Lower Llandeilo to the Arenig group—the Middle Cambrian of Sedgwicky. The Lower Bala includes certain dark earthy slates and bands of limestone occurring on the east flank of the Arenigs, Mynydd Tarw and Craig-y-glyn above Llanarmon in the Berwyns, the black slates on the flanks of Snowdon, also the arenaceons deposits on the west side of Bala Lake below the Bala Limestone, and that limestone also. The Middle Bala group of the Woodwardian Catalogue embraces the Bala Limestone and its associated sandstones and slates in North and South Wales. In Shropshire it is the Caradoc Sandstone, with its Horderley Limestone. The Coniston and Kildare limestones, and the Craig-Head and Peebles limestones, are all of this aget. Sedegwick’s ‘“‘ Upper Bala” comprehends the Aber-Hirnant beds above the Bala Limestone (with peculiar fossils), the lower part of the Coniston Flags (that conformable to the hmestone), and all beds above the Bala Limestone and beneath the May-Hill Sandstone. Again, the ‘‘ Upper Bala” includes all the beds, whether near Meifod or Welchpool or Llanwyddyn, which lie above the Bala Limestone and under the unconformable cover of the Denbighshire grit and flag. In ascending order the Upper Bala includes (1. The Hirnant Limestone and slate=Coniston Flags, | lower part only (Ashgill, Coldwell, &c.), above the Coniston Limestone. (Sedgwick). | 2. Llandovery beds (Lower Llandovery of Murchison) 2 eh =the Mathyrafal Limestone, near Meifod, of \ Sedgwick. “Tt is the great fossiliferous group of Haverfordwest. The Daiquorhan and Mullock beds in Ayrshire, the fossiliferous rocks of Maume and Cong in Galway belong in part to it’’§. * Vide Woodwardian Catalogue, Cambridge, pp. 25, 26, &e. T Consult the table of equivalents of the strata underlying the May-Hill Sandstone, Woodwardian Catalogue, p. 25. t Vide Woodwardian Catalogue, pp. 26, 39. § Loe. cit. pp. 26, 72. 130 PROCEEDINGS OF THE GEOLOGICAL SOCIETY, It has been stated that the fauna of the Llandeilo rocks num- bers 80 genera and 175 known species. The Caradoc or Bala contains 179 genera and 613 species; or 88 new genera and 375 new species make their appearance either at the close of the Llandeilo or during that period which ushered in or brought about those changes of conditions under which the deposits of the Cara- doc sea were accumulated and its species developed and multiplied, covering as it does so extensive an area in Kurope and Britain, also extending to America, whether homotaxially in that region or con- temporaneously with the deposition in the British area it is difficult to say. That the Trenton and Hudson-River group of North America, and nearly the whole of the Ktage D of Barrande (Bohemia) and Regio D of Angelin (Sweden), are the equivalents of our great Middle Caradoc and Bala, or the true Caradoc, is now not doubted. The fauna of the Bala group, especially in its zoological groupings, differs much from the preceding Llandeilo, although many species (85) are common to the two formations. I purpose treating of the Caradoc rocks and their fossils at some leneth—their stratigraphical place, the geographical distribution of their fossils (especially with reference to the succeeding Lower and Upper Llandovery)—questions which demand critical examination or analysis. The three classes having the greatest number in common are the Hydrozoa (Rhabdophora), the Crustacea (Trilobita), and the Bra- chiopoda. The number of known species in the Llandeilo in each of the above classes is Ey droZz 0a; tert ven.rteee ee 44 Crugtaces ene ee 45 Brachiopeda 1.232% .): 34 In the Caradoc fauna the same three classes number, including those from the Llandeilo, Hydrozoay 2a 38 Crustacea y cee mnne 146 Brachionodan @siosaec. 109 Numerically, the remaining ten classes have little value for compa- rison, although the Actinozoa, Echinodermata, Lamellibranchiata, and Gasteropoda are well illustrated (see Table X.). It is through the above three extensively developed groups, which are mainly moderately deep-sea forms, that the two formations are united. The fact of 73 species out of 175 passing from the Llandeilo to the Caradoc is highly suggestive, and is confirmatory of the views of Sedgwick in uniting the Llandeilo to the Lower Bala; and the distinctiveness of the great Middle Bala is confirmed by the fact that, out of 610 species in that formation, only 102 pass to the Lower Llandovery. Therefore, on paleontological data, we may regard the Llandeilo and Caradoc as being most closely allied. The 102 trans- gressive Caradoc species are distributed through the following classes :— ANNIVERSARY ADDRESS OF THE PRESIDENT. 131 12h: Y.-C) 1 out of 10 known. ) FIVOEOZOG. 2. i nus ly PETG) Colenterata ...... Die eri GOs, | TES Cree Often tN: iRehinedcrmata ucts 721) 496 (O20 ols, Weushacka, uiis.6 NG. De eA Pass to TL, 12 \1(07 20 er eee ee Al Dial 3 = ‘ d ie Pe brachiOpodan sca 93°. 55) LOD) «3, BNET Hamellipranchiata... 3 -,,« 76 «,, Gasteropoda ...... LO eS 08,5 Pieropoda: 4.5.0. - thin ates hy D7 net 9 25 Heteropoda ...... Lp pestn whois Cephalopoda ...... Tie nA] lt. ee OGIO’ wiz: The important zoological groups in the Caradoc are therefore the Hydrozoa, Coelenterata, and Echinodermata—the latter especially through the Cystideans, no less than 8 genera and 23 species illus- trating thisrare group. The Crustacea number 146 species, of which 106 are peculiar; 20 species are from the Llandeilo and 16 pass upwards to the Lower Llandovery, leaving thus for the Caradoc and Bala beds the largest Crustacean fauna known in the Lower Palzozoic rocks. Only 4 species of Bryozoa out of the 21 appear to pass to the Lower Llandovery; and they all commence in the Caradoc: they are Fenestella subantiqua, d’Orb., Ptilodictya dicho- toma, Portl., P. coslellata, M‘Coy, and P. lanceolata, Goldf. We should expect more species in common in this group, from the fact of their being usually a moderately deep-sea family, and less subject to vicissitude or change than littoral or sublittoral species. The - Brachiopoda of the Caradoc and Bala group number 109 species, representing 16 genera; they are more numerous than in other of the Paleozoic rocks. 33 of the 109 pass to the Lower Llandovery ; and we have seen that the connexion with the Llandeilo below is through 26 species, leaving, therefore, 50 as Caradoc proper. As regards geographical distribution, North-Welsh forms greatly pre- dominate, owing probably to the larger amount of research to which the rocks of the four counties of Caernaryon, Denbighshire, Mont- gomeryshire, and Merioneth have been subjected. Ireland also has yielded 60 species, and Scotland 40. Their more gomrighee distribu- tion will be further discussed under that head. The littoral or shallow-sea condition of the Caradoc is further illustrated by the occurrence of a large number of bivalve Mollusca. No less than 76 species are known; and of these only 3 species pass to the Lower Llandovery, viz. Pterinea retroflewa, Wahl., Orthonota suleata, His., and Mytilus mytilimeris, Cony. ; and 2 of these 3 forms pass to the Ludlow and the Upper Llandovery, viz. Pterinca retroflexa and Mytilus mytilimeris ; so that 66 species of Lamellibranchata be- long to the Caradoc exclusively ; and, strange as it may appear, these 3 are the only species known im the whole of the Lower Llandovery. 132 PROCEEDING OF THE GEOLOGICAL SOCIETY. The Bivalvia of the Caradoc exceed in number those of any other known formation below the Carboniferous Limestone. Careful analysis shows that the Arenig rocks contain only 6 species, the Llandeilo 6, the Caradoc 76, Lower Llandovery 3, Upper Llandovery 29, Wen- lock 45, and the Ludlow 71 species; about 13 species range from the Caradoc upwards or through to the Ludlow, 7 to the Upper Llandovery, 3 to the Woolhope, 11 to the Wenlock, and 12 to the Ludlow—this of course being inclusive, as some of the same forms appear more than once in their range. It is not improbable that errors as to species occur, especially when we have to do in many instances with mere casts in these arenaceous deposits. In both the Lower and Upper Llandovery groups the fossils are badly preserved, being casts only. If the Lamellibranchs are largely represented in the Bala group, the Gasteropoda are almost equally so, by 14 genera and 53 species; 2 only (Murchisona simplex, M‘Coy, and Turbo tritorquatus, M‘Coy) are common to the Llandeilo below; 10 pass to the Lower Llandovery ; 9 direct to the Upper Llandovery ; 7 species are common to the Caradoc, Lower Llandovery, and Upper Llan- dovery ; they are Cyclonema crebristria, M‘Coy, Holopella cancellata, Sow., H. tenuicincta, M‘Coy, Murchisonia cancellata, M‘Coy, MM. pulchra, M‘Coy, Trochonema triporcatum, M‘Coy, and Rhaphistoma lenticulare, Sow. Only 3 species pass to the Wenlock rocks: one of these is Holopella cancellata, Sow. ‘The Gasteropod fauna there- fore is as significant as the Lamellibranchs, no fewer than 32 of the 53 species being confined to the Caradoc or Bala group. 16 of the 53 are peculiarly Irish forms, and 5 Scotch, leaving 21 for dis- tribution through the North- and South-Welsh and English beds, of which 9 occur in the Caradoc of Shropshire. I have no determined species from Westmoreland ; North Wales yields 26 species, South Wales only 3, viz. Cyclonema crebristria, Holopella cancellata, and Patella saturni, and these only in Caermarthenshire. This group, like the bivalves, indicates shallow-water conditions. The Pteropoda (4 genera and 12 species) are, with 4 exceptions, confined to the Caradoc. Conularia Sowerbyt passes to the Lower Llandovery, Wenlock, and Ludlow ; ELeculeomphalus scoticus, M‘Coy, is Llandeilo and Upper Llandovery ; 9 of the known species are Irish, and 3 of them Scotch; so that the Lower Llandovery is only directly allied to the Caradoc by one form (C. Sowerby) out of the 12 occurring. This is even more strongly manifested by the associated’ class Heteropoda, of which, as in the Pteropoda, only 1 species in 15 passes to the Lower Llandovery—Bellerophon carinatus, Sow., being, so far as I know, the only form in this group connecting the two formations. Bellerophon bilobatus, Sow., B. perturbatus, Sow., and Maclurea macromphala, M‘Coy, are also Llandeilo. The Upper Llandovery has 3 in common, Woolhope 4, Wenlock 2; and 2 pass to the Ludlow. Specifically, 6 pass up, leaving 8 as be- longing to the horizon. The Pteropoda, being strictly or essen- tially pelagic, give us little clue as to bathymetric conditions at the time of deposition. Neither do the Cephalopoda; but no single class is so preeminently Caradoc: of the 47 known species, only 1 ANNIVERSARY ADDRESS OF THE PRESIDENT. 133 (Orthoceras arcuoliratum, Hall) unites it with the Llandeilo group (and this I believe to be a doubtful species); and only 7 connect it with the Lower Llandovery, and 3 of the 7 do not range higher. The 7 species are Jituites cornu arietis, Sow., Orthoceras ibex, Sow., O. vagans, Salt., O. annulatum, Sow., O. Barrandu, Salt., O. politum, M‘Coy, and O. tenuistriatum, Minst. The 3 species that directly unite the Caradoc and Lower Llandovery are three of the above— O. vagans, Salt., O. annulatum, Sow., and O. Barrandii, Salt. 9 range to the Upper Llandovery, 4 to the Woolhope, 9 to the Wenlock, and 3 to the Ludlow. These are the appearances of species that ascend through and into the higher formations. 30 of the 47 species are confined to the Caradoc, not ranging higher, whereas 15 species pass to the rocks above. Ireland is represented by 24 of the 47 species, 18 of which do not leave the Caradoc horizon, and 17 are strictly confined to the Irish deposits. Of the Scotch Caradoc Cephalopoda we only know of 6 restricted species, viz. Orthoceras audax, Salt., O. arcuoliratum, Hall (also Llandeilo), O. vaginatum, Schloth., O. politum, M‘Coy, O. prumevum, Forbes, and O. tenwistriatum, Munst.; and only 15 species out of the known 47 occur in the Scotch rocks. GEOGRAPHICAL DISTRIBUTION. I have now to treat of the geographical distribution of the extensive series of species occurring in the Caradoc and Bala rocks, the large fauna (614 species) being an additional reason for so doing, more especially considering the widely spread area occupied by the Lower and Middle Bala groups. It would be impossible to give ali the localities whence our knowledge of their distribution has been _derived; but the chief in North Wales number between 30 and 40, and in South Wales about 10. The Shropshire and Westmoreland localities are less numerous, but are prolific in species. The following 10 areas, including about 50 localities in the British Islands, have yielded the 613 species known. Their occurrences or appearances number 1555. - ( Caernarvonshire......... 105 species, distributed in 10 chief localities. North } Denbighshire ............ 131 af ° Ghees - Wales. | Montgomeryshire ...... 1388 4 Be is A 3 Merionethshire ......... 162 Fe i LOweaes $3 South { Pembrokeshire ......... 93 es Ms Rn iss Pe Wales. Neer ccchirs mere 72 fs a pes 6 end leeresie ara Poe aces os: 123 8 * | Westmoreland ...... 123 SUC ENING | Goh SAS COREG REEMA aEEe Bere 302 [iyailninG! CL EAe seer 306 1555 The actual number of species (so far as we know) occurring in North Wales is 270, in South Wales 134, in Shropshire 123, in Westmoreland 123, in Scotland 302, and in Ireland 306. The intimate relation of the Caradoc to the succeeding Lower VOL. XXXVII. 134 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Llandovery rocks obliges me to show their specific affinity, for the purpose of ascertaining the paleontological value of the Lower Llan- dovery (the Upper Bala of Sedgwick). Analysis shows that, of the 613 Caradoc species, 103 pass to the Lower Llandovery ; but, strange as it may appear, 107 are common to the Caradoc and Upper Llan- dovery, although we know of the stratigraphical break that occurs between the two formations. This relation is chiefly through the Actinozoa, Hydrozoa, and Brachiopoda, as we might expect from their bathymetrical position and habitat. Of the 40 known corals, 22 (or more than one half) range to the Llandovery rocks, 20 occurring in the Lower and 17 in the Upper Llandovery; or all but 3 are both Lower and Upper Llandovery ; and 37 species of Brachiopoda, out of the 109 known, pass to the Llandovery, 33 of which are Lower Llandovery. Of the 16 species of Annelida, only 2 pass to the Lower and Upper Llandovery, Tentaculites anglicus and Cornulites serpularis ; they range also to the Ludlow. The class Echinodermata, represented by 32 species, is remark- able for the occurrence of 8 genera of Cystidea and no less than 23 species. They essentially characterize the Caradoc and Bala rocks ; with the exception of one species (Hchinospherites arach- noidea, Forbes), which passes up into the Lower Llandovery, all are peculiar. Itis only in the Caradoc and Wenlock strata that the Cystidea are a well-developed and characteristic group of the Kchi- nodermata; they are replaced in the Carboniferous rocks by the Blastoidea (Pentremates and Codonaster) and do not appear again. The Crustacea are the largest and most important class in the Caradoc. The species number 146, representing 37 genera—Ireland yielding 90 (the largest number of species), Scotland 78, North Wales 50, and South Wales 30; or taking North and South Wales as one area, it has 59 species, 16 genera and 21 species being common to North and South Wales. Shropshire and Westmoreland are nearly equal, the former having a known crustacean fauna of 31 species and the latter 35. The genera most largely represented are Calymene 8 species, Phacops 14, Jilenus 13, Remopleurides 8, Inchas 6, Stawrocephalus 4, Acidaspis 8, Ampyx 6, Asaphus 6, Chewurus 6, and Trinucleus 5. The remaining 26 genera (many of which are Phyllopods) are represented by one or few species. The 37 genera of Crustacea appear or occur 217 times in the geographical areas named in the table of distribution and the horizons they pass to above, and the species appear 433 times. This Table clearly shows us how much has yet to be done before we can obtain reliable zoological data, or be assured that the several classes and genera have been even yet fairly ulustrated, feeling at the same time that, except by exhaustive and careful col- lecting, we never shall arrive at even a fair illustration of the fauna of any given group. For example, the 9 genera of Caradoe Pro- tozoa are represented by only 10 species, or a fraction more than 1 species for each genus. 8 of these genera belong to the Spongida ; and only among these do any 4 of the genera, illustrated by 1 ANNIVERSARY ADDRESS OF THE PRESIDENT. 135 Species each, occur in any one locality, as in Shropshire, Scot- land, and Ireland. North and South Wales have yielded only 2 genera, each with 2 species, in any of the six counties. The Actinozoa even afford us stronger evidence still of the importance of specific evidence. In this class we know through the fullest research that 20 genera and 40 species occur, thus averaging only 2 species for each genus. Many genera (13) have as yet yielded only 1 species; they are Plasmopora, Syringophyllum, Halysites, Stylarea, Omphyma, Thecostegites, Alveohites, Aulacophyllum, Calo- stylis, Cyathophyllum, Pistulipora, Tredadium, and Prosopora. We cannct suppose for one moment that these genera are not repre- sented by more species; it is a matter of research only. The genus Havosites has 8 species, Heliolites 6, Petraia 6, and Monti- culipora 4. The 20 genera make 90 appearances through the 10 areas, and the species 188; 32 of the generic and 68 of the specific appearances occur in the formations above the Caradoc, -and show the relation of one group of strata to another, as deter- mined through paleontological research. The Lower Llandovery, which succeeds or is the natural termination of the Caradoc, con-~ tains 9 genera and 20 species of the whole fauna (7°); and the Upper Llandovery 6 genera and 17 species of the whole. The coral fauna of the Caradoc at once makes its appearance underived, the older Llandeilo possessing only 3 species of Actinozoa, Halysites catenularvus, Favosites fibrosus,and Monticulipora frondosa; whereas, next to the Wenlock, the Caradoc possesses the largest Coclenterate fauna of the Lower Palzozcic rocks. Nowhere, either in Europe or America, does this class appear with so large a generic grouping. 20 of the 40 species pass to. the Lower Llandovery ; most of the same appear in the overlying and unconformable Upper Llandovery. The Crustacea pre-eminently characterize the Caradoc and Bala rocks, and constitute the largest group in the whole of the Palo- zoic series; 37 genera and 146 species have been collected, de- scribed, and registered through the labours of the Geological Survey. of the three countries, England, Ireland, and Scotland, as well as by the researches of Sedgwick, Salter, M‘Coy, Baily, Sharpe, Prof. Hughes, &e. No zoological group is better understood, none more important to the student of stratigraphical geology. The Lower, Middle, and Upper Bala beds could with difficulty be read or under- stood without a minute acquaintance with the Trilobita of this vast middle series of Lower Paleozoic deposits. They are to this group of rocks what the Ammonitide and Kchinide are to the Mesozoic series. The order Trilobita is illustrated by 27 genera, the re- maining 10 are mostly Ostracoda; among them is Z'wrrilepas. Caernarvonshire has yielded 11 genera and 17 species. North Denbighshire i ei ee Bow US. Wales. }) Montgomeryshire id Sie ars) 29s Le | Merionethshire Fe 1G) a ae DO ee South f Pembrokeshire % BG S26 24e oe Wales. { Caermarthenshire Mi DSH § 200 os 12 I 36 PROCEEDINGS OF THE GHOLOGICAL SOCIETY. Rochad Shropshire has yielded 15 genera and 31 species. Oni ‘etaecie ie ist B86 Scotland as ol ie, 78% ee Treland ee Silene 90a The most important genera, or those of chief stratigraphical value and in which the species are numerous, are Acidaspis (8 species), Ampyx (6), Agnostus (5), Asaphus (6), Calymene (8), Cheirurus (6), Homatlonotus (4), Illenus (13), Lachas 6, Phacops (13), and Remo- pleurides (8). The genera essentially characterizing the Caradoc’are Harpes, Salteria, Remopleurides, Cyclopyge, Dionide, Trresias, and Cyphonscus, most of which are only represented by one or few forms. Scotland and Ireland possess the richest assemblage of species ; 26 genera and 63 species occur in the former, and 23 genera and 77 species in the latter area. Only 15 of the 123 species of Trilo- bita pass to the Lower Llandovery; they are Acidaspis Brighti, Calymene Blumenbachu, C. brevicapitata, C. caractacr, C. Allportiana, Cheirurus bemucronatus, OC. clavifrons, Encrinurus punctatus, Cy- phaspis megalops, Cybele verrucosa, Illenus Rosenbergi, I. Bowmanni, I, Thomson, Lichas laxatus, and Phacops Brongnarti. 8 of these same also pass up into the Upper Llandovery, 6 to the Wenlock, and 5 to the Ludlow. The long-ranged species are chiefly those illustrating the largest genera, such as Calymene Blumenbachui, C. Allportiana, Cheirurus bumucronatus, Encrinurus punctatus, Cyphaspis megalops, and Phacops caudatus, all of which species appear in the Ludlow and then cease to exist, the Devo- nian rocks having none in common; yet Bronteus and Harpes are repeated from the Caradoc in the Middle Devonian, both in Britain and on the continent. We must remember, however, that the marine Devonian nowhere visibly overlies the Silurian rocks in Great Britain, and the Old Red Sandstone contains no true marine form anywhere. The order Ostracoda, illustrated by Beyrichia, Leperditia, Cythere, Primitia, and Entomzs, needs little more than notice here; they have received at the hands of Professor Rupert Jones the closest scrutiny both zoologically, paleontologically, and stratigraphically ; few men have so largely added to our knowledge of the orders Ostracoda and Phyllopoda. Bracutoropa,—And next to the Crustacea in force and classifica- tory value we must place this group of Mollusca or Molluscoida. Numerically in the whole of the Cambrian and Silurian rocks the Crustacea include the largest number of species, 550 being known ; whilst of the Brachiopoda we know 456. Individually no class sur- passes the Brachiopoda through all the Palzeozoic and Mesozoic rocks, many genera being, however, far more richly represented than others. In the Caradoc this is notably the case. The genus Orthis has in Britain alone, we know, through the large collections that have been made and the elaborate researches of Davidson, no less than 110 species, and culminates in the Caradoc. In tracing the numerical and stratigraphical value of the genus, we find that in the Cambrian ANNIVERSARY ADDRESS OF THE PRESIDENT. 137 group up to the close of the Tremadoc the genus Orthis possesses (so far as we know) only 4 species, viz. Or this Hicks, O. Carausi, O. lenticularis, and O. Menapie; in the Arenig 10 species, the 3 last-named forms being in common with the horizon below; in the Llandeilo 13 species, 4 of which are Arenig also, viz. Orthis alata, O. calligramma, O. remota, and O. striatula. In the Caradoc and Bala group the species have increased to no less than 41, all having extensive geographical distribution; 7 are peculiarly Irish. In the Lower Llandovery we know of 20 species, 17 of which are Caradoc, the only 3 peculiar Lower Llandovery forms being Orthis Bouchardi, O. reversa, and O. mullochensis. Thus through this one genus alone we see the close alliance of the so- called Lower Llandovery (the Upper Bala of Sedgwick) with the Caradoc. The moderate depth at which they lived is clearly indicated by the coarse and varied arenaceous nature of the deposits, for there is no evidence to show, even by the zoological grouping, deep-sea conditions. No Lingule are known in the Lower Llandovery rocks ; although 6 species occur in the Caradoc and 5 in the Upper Llan- dovery. The fauna generally may be regarded as one accumulating under decreasing depth or slow elevation over given areas. 8 of the 20 Lower Llandovery species pass to the Upper Llandovery or May- Hill beds. To show still further the decline in specific as well as individual members of the genus Orthis, I may mention that in the Upper Llandovery there are 10 species, but not a single form belongs truly to that horizon ; Orthis rustica, the only true Upper Llandovery species, passes to the Wenlock rocks, so that no form of Orthis is special to the Upper Llandovery. The species of Orthis in the Wenlock rocks number 16, of which 8 come from the Llan- dovery and Caradoc and 4 pass to the Lower Ludlow (Orthis biloba, O. crassa, O. elegantula, and O. hybrida), so that Orthis Hdgeliana, Salt., O. Lewrsi, Dav., O. Hughesw, Dav., and O. mullochensis, var., or variety of O. reversa, are new and restricted forms. None pass to the Devonian. I have selected this genus in the Caradoc for com- parison with the other Silurian groups on account of its magnitude and stratigraphical value. Two other genera, Leptena and Stropho- mena, Which first appear in the Llandeilo, also characterize the Cara- doc by greatly increased specific development. These are Leptena with 4 species and Strophomena with 2, both individually numerous in North and South Wales, In the Caradoc Leptena yields 8 species and Strophomena 19, and for the first time we meet with Rhyncho- nella with 8 species ; but no form is known to occur in South Wales, and only 3 have been found in North Wales. Ireland and Scotland yield 7 of the 8. The following Table shows the specific and stra- tigraphical value of these four important genera from the Cambrian to the Upper Silurian :— 138 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. E ie} 7S = A : 7 1 a 5 . bo] ob : 0 5 | Z qd = S) S tite 6 : Be 28 dvley Wiese lice. Ise dec Snel Bes ts |. ae | een en es eo eal Oe cue bese) 2 le See ewes koe | =) = Orthismas ctcuteeelt eee Pe 8) | 14s AY 20s Oe Gs Heptenarc...oc-cececen| Sec sine Pieepie Nicene cf agin, ta Atle ore eke | mm oe Strophomena.........] ... ihe 2) .19: |) EE a Rhynchonella ...... Lh 8.) | BOs eo O; 1) L| 8] 7 | 21°) 76 | 42 1 (34 aoe Those genera having only one or two species are doubtless of equally high stratigraphic significance. Such are Orthisina, Stipho- notreta, Porambonites, Triplesia, Merista, and Meristella. LAMELLIBRANCHIATA.—The known species of Bivalvia in the Caradoc and Bala rocks are more numerous than in any other Lower Paleozoic group. ‘Thus the Tremadoe rocks of St. David’s yield to us the earliest Bivalve fauna, consisting of 5 species; the Arenig has 6, the Llandeilo 6, the Caradoc 76, the Lower Llandovery only 3, the Upper Llandovery 29, the Wenlock 45, and the Ludlow 71. These numbers help us to see, by comparison, the zoological value of the class Conchifera in the several strata. Ctenodonta, Orthonota, Modiolopsis, Pterinea, Ambonychia, and Palearca are the largest genera, or those: haying the most species, by comparison with the older and younger‘fortiiations. The Caradoe Conchifera excel in number all beneath the Carboniferous. The following Table shows this, through the 6 Caradoc genera above named :— B| B . Bil eS Sp mo | os Z 4 = ) 6 i 8 8 ‘ mB | LS a = A S eS a hae = S last = ls a ra an ee 3 Gos ao | q q < ge esecalere hl |) ale |e)?|2\s | See eeeecn ee eens fora) Ss )e <= Ctenodonta ......... a Dei he Tile Eee 6 3 4 Orthonotay eee loe. MIS | ee 5 1 6 3 | 16 Modiolopsis ......... 4) 1| 2:) 16 | 029) SSc em mes IPterinvesc ee ere wae Albers | aes 6 1 6) TE we Ambonychia ......... dae See gee cee SOT eee iPaleearca, | eeeete mene shies Tae ee 2 a STAR MS 438 <=. "| —_ |J "|_| | || |_| | |__| 4 | 57 2 D1 Dae ares Cardiola with 3 species, Mytilus with 4, Pleurorhynchus, Ano- dontopsis, Cucullella, Megalomus, Clidophorus, Arca, and Lyroderma having each 1 species, complete the Conchifera of the Caradoc. Their / ANNIVERSARY ADDRESS OF THE PRESIDENT. 139 geographical distribution is significant, but clearly shows the im- perfection in collecting, and how much has yet to be done to demonstrate their relation to the extensive areas over which they ara distributed :— Caernarvonshire has yielded 6 genera and 16 species. Denbighshire a a (RE Montgomery shire ,, 8 eS 14. —SC««, Merionethshire __,, 6 “ Los Pembrokeshire _,, none none Bo thenskite s 2 i. 2 We Shropshire ut 5 if TOe ees Westmoreland 7 3 i Ouhaktss Scotland a ih R Bye Ireland ‘5 10 55 Side ee Nothing short of strict analysis could impress upon us the fact of such unequal distribution as the above examination shows. No species of bivalve is known in Pembrokeshire, a district rich in other groups. Only 1 genus and species in Scotland (Pleurorhyn- chus dupterus, Salt.), 2 only in Caermarthenshire (Ambonychia triton and Ctenodonta varicosa, Salt.), and 3 in Westmoreland (Ctenodonta anglica, D’Orb., Pterinea tenwstriata, M‘Coy, and Cardiola inter- rupta, Sow.). The physical geography of the area, either through barren or interrupted coast-line, rather than movements of land or depth of sea, would most probably account for this unequal distri- bution and poverty of species in one area, and their comparative abundance or fair representation in another. Looking at the horizons which the above 6 genera illustrate, and through which they pass, it seems hardly explicable that there should be only 3 species of Lamellibranchiata known in the Lower Llandovery, viz. Pterincea retroflexa, Mytilus mytilimeris, and Orthonota sulcata ; and these are Caradoc and Upper Llandovery also, showing us that not a single species belongs or is confined to the Lower Llandovery. This result, compared with the large bivalve fauna in the Caradoc or Lower Bala beds (57 species), and under conformable stratification, is scarcely to be accounted for, except through elevation of the sea- bed and change in bathymetrical conditions sufficient, with all other conditions, to cause the extinction or removal to another area of the Caradoc Lamellibranchs. Similar results are obtained with the Gasteropoda; the 53 Caradoc and Bala forms dwindle down to 16 in the Lower Llandovery. The Llandovery species will be noticed in their place or order subsequently. As before, Ireland yields the largest number of species (37). . GastERopopa.—Like the three preceding classes, the group Gasteropoda is largely represented in the Caradoc strata ; 14 genera and 53 species have been determined. No species has yet occurred in Pembrokeshire, and only 3 in Caermarthenshire (Holopella can- cellata, Sow., Patella saturni, Goldf., and Cyclonema crebristria, Moy). Neither have we evidence of any species from Westmore- I40 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. land, although, considering the somewhat abundant forms in Shrop- shire (9) and Scotland (11), and that the Westmoreland area inter- venes, we might expect some to occur. The genus Huomphalus in Ireland (with 7 species), Holopewa in Merioneth (7), Murchisonia and Rhaphistoma in North Wales generally (12), constitute nearly one half the univalve fauna—T'rochonema, Turbo ?, Trochus?, and Pleu- rotomaria having but few species in each. 8 genera and 19 species actually pass up to the succeeding Llandovery, Wenlock, and Ludlow rocks above. The Lower Llandovery receives 10 species, and the Upper 16; the Woolhope 1 (Turbo tritorquatus, M‘Coy), the Wen- lock 2 (EHuomphalus alatus, His., and . sculptus, Sow.). 3 species range into the Ludlow—Huomphalus alatus, Holopella cancellata, and H. conica. ‘Three species of Gasteropoda are common to the Llan- deilo below—Cyclonema crebristria, Murchisoma simplex, and Turbo tritorquatus. The relationship of the two Llandoveries is marked ; the Lower has 103 species of all classes in common and the Upper 107, Ireland and Scotland possessing the highest numbers, being re- spectively 304 and 206. PrreropopA.—Ten species of this class occur, the largest pelagic Pteropod fauna known in the Paleozoic rocks ; they occur in nearly equal numbers through the 10 areas, Ireland still yielding the highest number, nearly three times as many as Shropshire and Merioneth. No species is known in Pembrokeshire. Hetreropopa.—This order is represented in the Caradoc and Bala rocks by 11 species of Bellerophon, 2 of Kcculiomphalus, and 3 species of Maclurea. In the British area the two last-named genera are restricted in their geographical range to Ireland and Scotland, and stratigraphically to the Arenig or Llandeilo and Caradoc. Maclurea is known only in Scotland; MM. Logani, Salt., M. Peach, Salt., M. macromphala, M‘Coy, and M. magna are said to be (in Scotland) of Llandeilo age. The Durness species, M. Logam and M. Peachia, with their associated forms, I believe to be Arenig. Piloceras in- vaginatum and Orthoceras mendax occur in the same beds. Maclurea abounds in the Chazy Limestone of the United States and Canada —hbeds probably to be correlated with our Lowest Llandeilo, the Upper Cambrian or Lower Bala group of Sedgwick. J. M‘Coyz, M. macromphala, and M. magna are admitted to be with us of Caradoc age, and certainly occur low down in the series. Clonmel, Stinchar river and Bugon &c. in Ayrshire, and the Durness Lime- stones yleld M. Peachu and M. Logan: abundantly. I have preferred to place Kcculiomphalus with the Heteropoda rather than the Pteropoda, although Portlock’s genus may be one of the latter group. CrpHatoropA.—No less than 47 species have been determined from the Caradoc series ; their maximum development is in Ireland, where 6 genera and 24 species occur; no species is known in Shrop- shire. Westmoreland and Scotland register 12 each. The Irish genera are Cyrtoceras, Koleoceras, Lituites, Poterioceras, Trocholites, and Orthoceras. As a generic fauna this is by far the largest in ANNIVERSARY ADDRESS OF THE PRESIDENT. I4I the Paleozoic rocks. Only one species (Orthoceras arcuoliratum, Hall) is common to the Caradoc and Llandeilo; but 15 species range upwards—7 to the Lower Llandovery, 9 to the Upper Llan- dovery, 4 to the Woolhope, 9 to the Wenlock, and 3 to the Ludlow ; or the 15 make their appearance or occur 32 times in their range through the higher divisions of the Silurian rocks ; 32 species are therefore essentially Caradoc or Bala forms. The numerical distri- bution of the Caradoc species through North and South Wales is re- markably uniform. 12 species occur in the four North-Welsh coun- ties, and 8 in the two South-Wales counties. Shropshire possesses no Cephalopod, yet 57 genera and 1283 species range through 15 zoological classes occurring in that county; but even this number represents numerically only a little over 2 species for each genus. Table X. is prepared on geographical data, the universality or extension of the Caradoc rocks and species not admitting of any other mode of expressing their distribution; and to analyze the group under subdivisions of the Caradoc or Bala, as given by various authors, would end in confusion, owing to the application of the terms Lower, Middle, and Upper Caradoc, or Bala, not being the same inall. Text-books have not yet defined the limits of these divisions, either zoologically or geographically ; in the former sense the greatest difficulty would be felt over extended areas, through want of true succession and continuity in strike. A careful study of the Woodwardian Catalogue demonstrates the difficulty of treating the Caradoc or Bala group as a whole in any other way. The column headed ‘ Localities and Numbers” in that Catalogue at once shows that it is only through space distribution that we can realize the value of the subdivision. In the Lower Bala group of the Upper Cambrian, as established by Sedgwick, Mr. Salter has placed 45 genera and 82 species; in the Middle Bala 82 genera and 212 species; and in the Upper Bala group 43 genera and 100 species. That the fossils of the “‘ Lower Bala” of the Woodwardian Cata- logue represent the Llandeilo proper of the Geological Survey there can be no doubt ; and out of the 82 species catalogued, only 17 pass to the Middle Bala group, and 8 to the Upper Bala. The Middle Bala of the Cambridge Catalogue receives the fine assemblage of fossils collected from the Bala limestone, sandstones, and slates of North Wales, the Caradoc sandstones and the Horderly limestones of Shropshire, or the true Caradoc. It includes also the fossils of the Coniston Limestone, the Kildare, Craig-Head, and Peebles Lime- stones. The Hirnant Limestone is not included in Sedgwick’s Middle Bala, but constitutes the base of his Upper Bala group. The 212 species largely illustrate the characteristic fauna of the Caradoc, and the group clearly shows the value of the division (especially with the elimination of the Llandeilo forms) and its separation from the Lower Llandovery (the Upper Bala of Sedgwick), into which only 33 of the 212 species passed. These proportions in the Cambridge collection (as catalogued), showing the community of forms between the 3 divisions, show how well the collection was I42 -PROCEEDINGS OF THE GEOLOGICAL SOCIETY. made, and also the care bestowed upon the stratigraphical grouping of the fossils. The Upper Bala group of Sedgwick includes the Aber-Hirnant Limestone, which occurs above the Bala Limestone, also the lower part of the Coniston Flags, the beds above the Bala Limestone, and below the May-Hill Sandstone; it therefore includes the Lower Llandovery or Mathyraful Limestone near Meifod, and also the rich fossiliferous series at Haverfordwest, the Dalquorhan and Mulloch beds of Scotland, and the Maume and Cong beds of Galway *. The Woodwardian Catalogue enumerates 43 genera and 100 species from this group (the Lower Llandovery of the Geolo- gical Survey), 18 genera and 36 species of which also occur in the Middle Bala below. Lower LLANDOVERY. In my analysis of the Lower Llandovery species from the rocks of that age in North and South Wales, Scotland, and Ireland, I have brought together and tabulated all that is known relative to the range and distribution of life through this Middle Silurian or Upper Bala group of Sedgwick. ‘The table shows the value of this upper member of the Caradoc or Bala beds, and its relation to the transi- tional Upper Llandovery, or the group so called, which stands be- tween and connects the Lower Silurian of the Survey with the well- defined succeeding Wenlock and Ludlow series. Careful analysis of the species in both their stratigraphical and geographical distri- bution shows how small a specifically characterized group the Lower Llandovery appears to be, for only 115 species out of 204 are really Lower Llandovery. The total number of genera and species known for all Britain is 68 genera and 204 species; but 50 genera and 104 of these species transgressed or came from the Caradoc and Bala beds below, thus reducing the actual Llandovery fauna to 18 genera and 100 species. The intimate connexion with the Upper Llan- dovery is chiefly through the Actinozoa (20 species), Brachiopoda (38), Crustacea (10), and Gasteropoda (8). No form of Protozoa, Echinodermata, or Pteropoda is common to the two Llandoyeries in any area in Britain. No species of Echinodermata occurs in North Wales, Ireland, or Westmoreland. No Lower Llandovery Annelid is known in North Wales. Neither do we know of any Lamellibranch, Pteropod, Heteropod, or Cephalopod in the same area. I will now discuss through tabular analysis the distribution of the chief zoological groups in the Lower Llandovery, viz. the Echinodermata, the Crus- tacea, Brachiopoda, and Gasteropoda. The remaining 7 of the 13 groups (there being no plants) are conspicuous through feeble re- presentation, due doubtless to geographical changes towards the close of the Caradoc and Bala period. Hyprozoa.—The careful researches and generalizations of Lin- narsson in Sweden, Lapworth, Nicholson, Carruthers, and Hopkin- * Vide Cat. Coll. Cambrian and Silurian Fossils in Mus. Univ. of Cambridge, pp. 25, 26, 39, & 72, for greater details. ANNIVERSARY ADDRESS OF THE PRESIDENT. 143 A 2 TE Sm aD ee =o ee AD 4H won on a 52 mn wo =e | From Llandeilo. Q bg Q kg wae ese Soe ee a [Sect |) 2 Bee Sy See Ess eh ey 3 = Oa a OO AR Eee os wets 2 Se S Sie eo 6 § Seo OE o Ss 2 eS SN io = si «63S fen en erie a of Ee Go £ : Bee 6S. oO: & Se os fo} : a © - Seale coe J het el laden = ar baa aie ie SS; : g9 3) OLAS of aCe ge ae = . wm 2 6 By Qe NARI = algae ail iees eal la R a a : | 2 SGT SR

bO Cc aw worroroaoantwnare oo Oo Ke | | | i | Genera. Ip | | | : ~ . OF | wp oe ee < ao be we Ue be AR BAR AA HE Caernarvonshire. uo : iA ‘ ; ° G2 [ow an ue wa sa So oe EY we ca Zo : =~ | Denbighshire. a 5 eg [ee an ae oe Ee So Ta Sy we xe Gx : sw | Montgomeryshire. = | ss Q ne =i COS Lec) ARC cole ee : Meri thshir S aS ai) tm 4P oF LG B& Go AA GX OR > erlonetnsnire. ae Lar} OVE & TM Sl oe Sr : 5 Ss qo BE AS Fe Bos bop Pembr k hir g E. EN : 5 a RO wo ~~ : roKkesoaire. = Q S | RO rR oe we nm Bo re TO AA ww on wes | Caermarthenshire.| = ae ! | fh Fy ; S =) ; = we > oH on oa Da oe we CF AA A® SA AA abo Shropshire. = . | F | | 8 Se eo ee ee Oe exes, Sm nono) com he W t r i d B os = pon VS ee ae TS) estmoreland. or | | TS an _ ) Ae a9 : De wr Pegiemo se oF SS ee OH ge OF 2 ae wo aa | Scotland. | | se | oe 2 Of won iS as os aw =i oh 82 oa Pa EN | Treland. | | = og | We He He So we Bo aw 55 ww ww So x» ee | Lower Llandovery. | | o F i S Classes. = 4 Sf ao o Z| 2) S eS e : = = oc : 6 jts 2. s g a fe = = | Oe g 2 3 = 5 a & |) aad = © 8, S 5 8 2 Sai Fy to) a A D D 5 Ee ia Plantae ses seteceiee Fil) LEIROLAOVAOE YS “AU onae mace pane 3 A 3 a 1 5 7 7 6 Ep ||| LelylxoyZory eessabeddser 7 50 A we te & 3 5 ¢ 5 8 7 8 BP || EGBUIVOYACE opodaaoaves 12 26 an A ao oh 2 | Echinodermata...... 2 2 3 E 2 1 9 2 1 1 2) Atameladays tereenee tees 2 3 Be: z iu u 2 10 1D im 4 8 10 7 + Ore@rustaceaiwscsencees 13 25 P ae 1¢ Z o AaB RY OZOBiasceemelorec oe 2 5 a 2 z 4 2 8 7 8 10 3 | Brachiopoda ......... 10 59 Bo ae ne a 10 3 | Lamellibranchiata..) 3 3 a Z a 2 2 a 1 6 6 > | Gasteropoda ......... 3) 15 u : 2 5 g 1 | Bteropodarsscvadaces- 1 1 i 1 1 ° 1 a 1 1 | Heteropoda ......... 1 6 3 x 4 2 2 | Cephalopoda ......... 3 8 2 3 1 1 50 22 52 5 4 ; 68 204 43 120 130 98 a5 nice Uprrer Luanpovery, on May-Hitt Grovp. Perhaps no formation in the Lower Palseozoic series is more diffi- cult to understand, either physically or zoologically, than the “ Upper Llandovery,” or May-Hill group. Its place, stratigraphically, was long ago settled by Sedgwick, who was the first to point out the necessity for separating these beds from the Lower Silurian (his Cambrian), and removed them from the Caradoc in 1853, proposing the name of the ‘‘ May-Hill Sandstone ” for these beds, being above all his Cambrian Rocks; and this is so throughout the northern hemisphere. On the west flank of the Malverns, at Woolhope, May Hill, and ANNIVERSARY ADDRESS OF THE PRESIDENT. I51 Tortworth, in many localities in Shropshire (Norbury, Cherbury, Church Stretton, &c.), and at Llandovery, Presteign, &c. in South Wales, the Upper Llandovery or May-Hill beds are typically exposed. Many names have been proposed for these variable beds, such as Upper Caradoc, Pentamerus-beds, Wenlock Grit, &c.; but, through their chief development at Llandovery and May Hill, in Gloucester- shire, the double name of the ‘‘ Upper Llandovery,” or ‘‘ May-Hill Sandstone,” has been given to them. In North Wales the Upper Llandovery rocks are absent ; m South Wales they first appear in Marloes Bay, appearing here and there across Pembrokeshire. Near Llandeilo, in Caermarthenshire, they lie at the base of the Upper Silurian rocks, being most variable in thickness. They transgress or lie indifferently and unconformably on Lower Llandovery, Caradoe, or Llandeilo beds. In Shropshire they are markedly unconformable to the Caradoc. In the Longmynd area, under the condition of a caleareous conglomerate, they rest upon all the older rocks exposed. That a great physical break takes place between the two Llandoveries there can be no doubt. They rise from under the Wenlock Shale at Noeth Grug, strike to near Llandovery and Pen-y-lan; again they range uninterruptedly from Marloes Bay, in Pembrokeshire, where they are highly fossiliferous (28 species occurring there), through Caermarthenshire to Builth, and on to the Longmynd and typical Silurian country of Shropshire. They everywhere rest unconform- ably upon the older rocks, sometimes lying on the denuded edges of the Lower Llandovery or on the Caradoc Sandstone ; at Builth and the Longmynd on the contorted and denuded Arenig, Llandeilo, and Cambrian strata. The absolute unconformity of the Upper Llandovery beds to the strata below, coupled as it is with changes of species, is doubtless connected with a lapse of unrepresented time. Whether that time be of great duration or not we have no means of judging; but, looking at the intimate connexion between the fauna of the Lower Llandovery and that of the Upper, we are led to suppose that it was not of sufficiently long duration to cause either the extinction or migration of the older fauna or the wmtroduction of a new one (only 4 genera seem to have appeared); for we have seen that the Lower Llandovery transmitted 45 genera and 104 species, out of its fauna of 68 genera and 204 species, to the Upper Llandovery. It is therefore evident that upheaval and denudation must have been of comparatively short duration, and little physical change could have taken place in the area occupied by the Lower Llan- dovery after upheaval ; this the physical geography and paleontology of the two groups help to show. | The preponderance of individuals in species of the genera Pentc- merus and Stricklandinia, especially P. oblonqus, S. lens, and S. lirata, which are the best known and most widely spread, is a marked feature in the Upper Llandovery rocks, and imparts to them peculiar and distinct facies. These species occur, though not so plentifully, in the Lower Llandovery. The Malvern and May-Hill form is S. lens, and it oceurs in profusion. Looking at the grouping of the m 2 152 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Upper Llandovery and the distribution of its fauna, we are surprised to find that, of the 261 species which occur in the formation, only 136 species really constitute the Upper-Llandovery fauna. The fact that 104 of the species are Lower Llandovery goes very far to unite the two so-called formations into one, or a Middle-Silurian group, abolishing the terms Lower and Upper. Granting the high value of the unconformity as of paramount stratigraphical importance, yet the community of fossils and general facies of both horizons lead us to regard the species as belonging to neither; for certainly the Lower Llandovery was an expiring close to the Caradoc and Bala, through elevation of the Caradoc sea-bed, and the Upper through continuity of certain species ; and the introduction of new forms commenced at the base of the Upper Silurian of Sir R. Murchison and the Geological Survey, from which, in the Wenlock rocks, were derived 58 genera and 125 species; for out of the 523 Wenlock, 125 are Upper Llan- dovery, 136 species only being the true Upper Llandovery fauna proper. Pranta#.—One species only, a fucoid (Pucordes gracilis), is known in this group of strata. This is probably only one of the many forms of Annelide or molluscan tracks or burrows, which simulate the habit of some of the marine alge. Pror0z0a.—Clathrodiciyon vesiculosum, Nich., and Cliona ( Vioa) gracilis constitute the only two Protozoa known. The former occurs in the Llandovery rocks of Scotland; the latter appears to be a dichotomizing, burrowing sponge, found in the shelly structure of a Pterincea (P. demissa) from the Upper Llandovery of the Mal- vern Hills. This is probably the oldest known burrowing sponge ; its habit is quite that of the modern genus Cliona. The Welsh Llandovery rocks, although so carefully searched, have yielded no traces of Spongide. We should, however, quite expect to find the group Silicipongie, through the Hexactinellide, repre- sented in these sediments or formations accumulated in moderately deep water, and associated with an extensive Coelenterate and Brachiopodal fauna, such as we have in the Upper Llandovery. The Analytical Table (p. 159) shows that only in two of the ten (Worcestershire and Scotland) are the Protozoa known. iypRozoa.—The genera Monograpius, Oyrtograptus, Duplograptus, and fetiolites alone seem to represent the Rhabdophora in the Upper Llandovery rocks ; but so many species appear to be common to the Lower and Upper Llandovery, that it is with difficulty I am enabled to realize the distinctness of the two faunas. The Irish, Scotch (Girvan, Gala, and Mulloch beds), and the Coniston flags and mud- stones have so many species in common with the Lower Llandovery that only through intimate and practical acquaintance with the two groups of species can they be separated. In the Valentian or Llandovery-Tarannon series of Lapworth, including the Birkhil series and its several zones, so many species of Monograptus, Deplo- graptus, and Cyrtograptus appear to be common, that it is no easy task to determine numerically or statigraphically, especially when geographical distribution is also taken into consideration, their true ANNIVERSARY ADDRESS OF THE PRESIDENT. 153 history. Probably the Upper Llandovery rocks of Ireland yield 6 or 8 species, the Girvan beds about 30, the Gala group the same, the Mulloch beds 25 species, and the Coniston beds 25 species; but many also belong to the Lower Llandovery belowandelsewhere. The species in two of the genera above named, AMonograptus and Cyrtograptus, increase in the Wenlock; in which, if we include the species which occur in the Tarannon Shales lying at the base of the Upper Silurian (Lower Wenlock group), some 9 genera and 20 species may occur. As in the Caradoc, it is extremely difficult to arrive at the number of true species that range through the areas, owing to not having correct knowledge of the species. Actinozoa.—This is, zoologically, an important group in the Upper Llandovery, from the number of genera occurring and the numeri- eally few species, 11 of the 16 genera being represented only by 1 species each. These are Alveolstes, Coenites (probably one genus), Halysites, Labechia, Omphyma, Plasmopora, Ptychophyllum, Lind- stroma, Propora, Pinnacopora, and Streptelasma. Thus only 5 genera constitute a specific fauna, viz. Favosites 4 species, Heliohtes 5, Paleocyclus 2, Petraia 7, and Syringopora 3. Their dispersion or distribution is equally significant, showing that the two chief com- pound genera, Mavosites and Heliolites, are, with one exception (Peiraia), the most widely distributed geographically. Petraca occurs in all the 14 localities* ; and perhaps no single genus has so wide a range or such persistent representative forms, commencing in the Caradoc with 6 species, having the same number in the Lower Llandovery, and 7 species in the Upper Llandovery. It is reduced to 5 in the Wenlock, and 1 (P. bina) in the Lower Ludlow. We know also of 3, if not 4, Petracw in Lower, Middle, and Upper Devonian ; they are, however, distinct species, being P. celtica, Lonsd., P. gigas, M‘Coy, and P. plewradialis, Phil. As regards the number of genera and species numerically of value, or important in certain areas, we find that in Pembrokeshire and Cardiganshire these are but feebly represented. The former county, at Marloes Bay, has given us 3 genera and 4 species, viz. Mavosites asper, Omphyma turbinata, Petraia bina, and Petraia subduplicata; and Cardiganshire only 1, Petraia elongata. Radnorshire has yielded 2 genera and 3 species, Heliolites interstinctus, Petrara elongata, and P. subduplicata. These few and local species are significant and suggestive as to the cause of their restricted numbers. Worcestershire, through the Malvern area, is richest in species, its 7 genera and 18 species being the highest known in the nine areas— Caermarthenshire and Caernarvonshire having 12, Gloucestershire, Scotland, and Ireland each 10 species, and South Wales generally 14. Of the 16 genera and 32 species that range through the Upper Llan- dovery beds, 14 genera and 22 species pass up into the Wenlock. The 3 genera peculiar to this horizon are Propora (P. Edwardsu), Pina- copora (P. Anderson), and Lindstromea (L. levis). The remaining genera and species all pass to the Wenlock series, the Coelenterate fauna of which numbers 76 species. * P, subduplicata has recently been discovered in Scotland. 154 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Ecuinopermata.—lt is clearly shown that this group requires to be more carefully collected and studied. We have 5 genera and 5 species, or only | representative species to each genus. None occur in Shropshire or Scotland. Two species, representing 2 genera, have been met with in Pembrokeshire. The species are Actino- crinus pulcher?, Cyclocystoides Davisu, Salt., Paleaster coronella, Periechocrinus monliferus, and Paleechinus Phillipsie, Forbes— the last being the first known representative of the Kchinoidea in time. The Perischoechinide, to which family this genus belongs, attained its maximum development during the Carboniferous period, both in Britain and America. No fewer than 7 genera have been recognized in America and 4 in Britain. In the test of Paleechinus the interambulacral plates abut against each other, not being bevelled for overlapping articulation, as is the case with Perischodomus, Oligophorus, Melonites, and other genera. The family Archeocidaridee, of which the Carboniferous genus Archwocidaris 1s the type, does not occur below the Carboniferous rocks in any known area. The modern representative, ‘‘ Calveria” (C. hystrix) of Prof. W. Thomson, has been dredged in 445 fathoms water (2670 feet) in lat. 59° 38’ N., long. 7° 46’ W. Calveria differs from all known recent Kchinoidea through the structure of the plates of the test, which overlap, instead of abutting against each other, as in all other genera. Again, the plates composing the interambulacral areas over- lap from the apical disk towards the mouth, and those of the ambu- lacral area from the mouth towards the apical disk, or in the reverse direction. This ancient type appears in the Chalk as Echinothuria (E. floriformis); and now, through the dredgings of the ‘ Porcupine,’ we have proof of ifs continuity from the Cretaceous epoch. AnnrLIpA.—Three genera, with four species, two of which claim attention here, viz. Ventaculites (ornatus, Sow.), and Cornulites (serpu- laria, Schloth.). They have occurred in almost every locality since their first appearance in the Caradoc and Bala beds. Tentaculites ranges through the Caradoc everywhere except in Scotland ; it is abundant in the Lower Llandovery of South Wales; but its maxi- mum of development takes place in the May-Hill Sandstones, where it is a marked feature in the fauna. It is sparingly exhibited in the Wenlock shales and sandstones, Cornulites having replaced it indi- vidually in those beds. TYentaculites anglicus, T. ornatus, and 7’. tenwis are also Ludlow species; and the long-lived 7. anglicus either passes to the Middle Devonian slightly modified, or is replaced ~ by Schlotheim’s species 7’, scalaris. The 3 genera and 4 species (Cornulites serpularius, Tentaculrtes anglicus, T. ornatus, and Spirorbis Lewis) pass to the Wenlock. The Cephalobranchiate or Tubicolar Annulosa (Annelida) play an important part in the life and distribution of the class through time. Their habits insure persistency, their structure security, and their bathymetrical ranges are such as to lessen their chances of modification through relative changes of land and sea. No form of the order Errantia, or Dorsibranchiate Annelida, has been detected in the Llandovery rocks, although, doubtless, they existed, as many ANNIVERSARY ADDRESS OF THE PRESIDENT. 155 . of the so-called fucoidal remains are only tracks of these creatures ; and their presence has been demonstrated by Mr. G. J. Hinde through his discovery of the jaws of many species of the Poly- cheta in the Silurian, Devonian, and Carboniferous formations. In the Wenlock alone he has discovered the remains of 6 genera and 24 species; and in the Ludlow 2 of the same genera and 6 species. No one can doubt that, by careful sifting and preparation, the clays and shales of the still older rocks may yield a rich harvest of the remains of the Annulosa and soft-skinned Echinoderms. This process is now bearing fruit through the energetic measures em- ployed by Mr. G. Maw, in having large quantities of the Wenlock Shales washed, sifted, and examined. Numbers of new forms have been discovered. The same result has been obtained through the determined and persistent search made by Mr. Bennie (the Collector for the Scotch Survey) for remains of the Holothuride in the Car- boniferous shales of Scotland, and also for the Chitonide. ‘The results obtained by this patient worker, and the species collected, are now being described and figured for the Glasgow Geological Society by Mr. Etheridge, Junior. It is now the age of microscopical investigations, and ere long microscopic paleontology will stand on the same footing with petro- logical and crystallographic investigations. Crustacea.—The distribution of some of the species of the Crus- tacea through the Lower Llandovery is more constant and uni- versal than in any other Silurian deposit. Calymene Blumenbachu, Eincrinurus punctatus, Ilenus Thomsom, Proétus Stokesw, and Pha- cops Stokesi occur in almost every locality, and all 5 pass to the Wen- lock. 7 of the known 24 species do not pass to any higher horizon than Upper Llandovery; they are Illenus emuius, I. Bowmanm, I. Mac- callunn, Phacops Weaveri, P. obtusicaudatus, Lichas laxatus, and Tri- nucleus concentricus. .'To show how equally the species are distri- buted, I may state that in Caermarthenshire 7 genera and 12 species occur ; in Pembrokeshire 7 genera and 11 species ; in Worcestershire 7 genera and 13 species ; in Shropshire 7 genera and 15 species; and in Ireland 7 genera and 14 species; and, singularly also, 7 genera and 14 species pass to the Wenlock. As we have before seen, 10 species are Lower Llandovery and Caradoc—thus, as regards the Crustacea, rendering the value of the Upper Llandovery as a distinct paleon- tological group more questionable still. Bracuiopopa.—tThe species in this class are more than double the number of those of any other in the Upper Llandovery rocks. We recognize 65 species and 13 genera. The geographical distribution of Atrypa (5 species), Leptena (4), Orthis (10), Pentamerus (4), fihynchonella (9), Stricklandinia (2), Spirifera (5), and Strophomena (11) is complete through South Wales, Gloucestershire, Worcester- shire, and Shropshire. These 8 genera include the mass of species (50 of the 65) that range through the fourteen localities I have selected, and whence I have drawn the materials for analysis, and I believe I have omitted none. Lingula with 5 species, Crania and Discona each with 1, Chonetes with 3, and Meristella with 6 species 156 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. are sparingly and irregularly distributed ; but we should expect this from the habits and habitats of 3 at least of the genera. No less than 36 species come up from the Lower Llandovery (22 of the same are Caradoc); and 41 of the 65 range into the Wenlock beds. As among the Crustacea, few species belong to the Upper Llandovery proper, showing how slight is the evidence afforded us to separate the Lower and Upper Llandovery. The stratigraphical unconformity is, however, completely borne out by a corresponding paleontological break. This is most decisively seen from the Actinozoa, the Crustacea, the Brachiopoda, and the Lamellibranchiata; indeed Car- diganshire has as yet only yielded | species of Actinozoa, Petraia elongata, M‘Coy. The Upper Llandovery rocks of the nine chief areas given below have yielded the following number of genera and species of Brachiopoda :— Counties. Genera. Species. Remibrokeshines 7.40 ae. 11 19 Caermarthenshire ........ a, 38 Cardiganshirenc. seen oe none. none, IRGOGESOS So esessacece 6 De Gloucestershire .........-. 9 it Wrorcestershires 4 ae att 35 Shmopsiune yeti ae ae 13 13 SCOGlAMGs 2s cc ct ee aiee mee 9 De, Terelamndiy: case ee eo oun 10 19 No less than 11 genera and 41 species pass up to the Wenlock group, thus showing that a greater number of species in this class pass up than belong to any even of the most prolific areas. LAMELLIBRANCHIATA.—Considering the few species known in the Lower Llandovery (3), we have here a marked change in the greatly increased fauna of bivalve Mollusca, which have increased from the 3 genera and 3 species of the Lower Llandovery (Pterincea retro- flexa, Mytilus mytilimeris, and Orthonota sulcata) to 12 genera and 29 species. 2 of the 3 species above named pass to these upper beds, so that the specific fauna is essentially new; in the end, however, it gives to the Wenlock 7 genera and 15 species; or 50 per cent. pass to the Wenlock. The Worcestershire area is most prolific, 12 species occurring. Only 3 species are known in Scotland, Anodontopsis bulla, Pterinea Sowerbyt, and Cardiola striata, and 3 in Ireland, Anodontopsis bulla, M‘Coy, Pterinwa bullata, M‘Coy, and Ctenodonta subcylindrica, M‘Coy. No species is known either in Radnorshire or Caermarthenshire. (Vide Table No. XII.) GastEropops.—Kqually balanced with the above group, the dis- tributional value being through species of Holopella, Acroculia, ha- plistoma, and Turbo. These species, too, have a wide geographical distribution, Worcestershire, as before, possessing the greatest num- ber of species (9), although here only equal to Ireland, which has also 9 species; while Caermarthenshire has 7, the remainder of the 13 genera and 28 species being equally distributed. 6 genera and 8 species pass to the Wenlock; they are Acroculia halts, ANNIVERSARY ADDRESS OF THE PRESIDENT. 157 Euomphalus alatus, E. frenatus, E. sculptus, Holopella obsoleta, Loxonema sinuosum, Murchisonia articulata, and Turbo tritorquatus. No form occurs in the Upper Llandovery of Cardiganshire. Preropops.—Only | species is known in the Upper Llandovery. Conularia cancellata, Sandb., occurs in Gloucestershire and Shrop- shire, and is known in no other formations or localities in Britain. The more abundant C. Sowerbyi is Lower Llandovery, Wenlock, and Ludlow. It is questionable if C. cancellata is really a British species or only a variety of C. Sowerbyt. 7 of the 10 areas have no represesentative of this group. Herzroropa.—Leculiomphalus (Cyrtolites) and Bellerophon, the . former having 2 species (#. levis, Sby., and £. scoticus), the latter 9, range pretty equally through the Upper Llandovery rocks, B. dilatatus, B. trilobatus, and B. carinatus haying the widest or most extended geographical distribution. The rarer forms, Hecul- omphalus scoticus, M‘Coy, Bellerophon subdecussatus, M‘Coy, B. wen- lockensis and B. obtectus, Phill., have as yet only occurred each in one locality ; 5 of the 11 have appeared before in the Caradoc; but only one species, B. carinatus, seems to have been Lower Llandovery. These Pelagic Mollusca, especially the family Atlantide, have no fixed habitat, their distribution being quite independent of the nature of the sea-bottom, so that the Bellerophons of the Silurian, Devonian, and Carboniferous periods occur indifferently in mechanical deposits of the most varied petrological or lithological characters. The great Poreellia (P. Woodwardit, Sby.) of the Middle Devonian, or the two species P. striata and P. Symondsii of Phillips, occur in all sorts of deposits all through the history of the genus, of which 12 or 14 species range from the Devonian to the Trias, Belgium having afforded many forms. Conrad’s genus Cyriolites, probably the Ecculiomphalus of Buckland, is represented in Britain by 3 species —H. Bucklandi, Caradoc only; £. levis, Upper Llandovery, Wen- lock, and Ludlow; and Z. scoticus, Llandeilo, Caradoc, and Upper Llandovery. Of the 11 Upper Llandovery species 6 pass to the Wen- lock, 5 being species of Bellerophon, and 1 Eeculiomphalus (EL. levis). CrpHatopopa.—Of the 6 genera occurring in the Upper Llan- dovery, viz. Actinoceras, Cyrtoceras, Lituites, Phragmoceras, Treto- ceras, and Orthocerus, one (Actinoceras nummularium) appears for the first time, and the rare forms Phragmoceras compressum and Cyrtoceras upproximatum have but local geographical distribution. The last named is from the Upper Llandovery of Malvern only; P. compressum the same, being of doubtful occurrence in the Ludlow rocks. We have seen that the Caradoc rocks of Shropshire have not yielded a single species of Cephalopod; but here, in the Upper Llan- dovery of the same county, we have 3 genera and 9 species; yet 8 genera and 47 species occur in the Caradoc rocks. This fact, coupled with others to be arrived at through the tables of distribution, tends to show the shallowing of the Caradoc sea and slow elevation of the land through the latter part of the Caradoc period, and the time represented by the deposition of both the Lower and Upper 158 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Llandovery series. 4 genera and 9 species pass to the Wenlock. Ireland has only 2, Orthoceras coralliforme, M‘Coy, and O. subgre- garium, M‘Coy. Radnorshire has 1 species, Litwites cornu-arietis, Sow. Taking South Wales generally, only 6 species are known there, and these represent 4 genera—Actinoceras 1 species, Lituites 1 species, Orthoceras 3 species, and T'retoceras 1 species. The Upper Llandovery rocks of Scotland give us 4 species, 3 Orthocerata and 1 Phragmoceras (P. compressum),. We therefore find that, out of the whole fauna of the Upper Llandovery rocks, numbering 240 species, only 91 are confined to them and do not pass up. The following list shows the special Upper Llandovery fauna :— i“ RObOZ08) 44 Weinents ee 2 species. iNctamOzoa,! Joe aa are. Ae 10 “3 Kehinodermata.......... 3 a Crustacea vik ae. ee 8 ee TTY OZ OA cphisual.-p senses net eenae evel igs Brachiopodarn es -.ntiee ee 22a ae Lamellibranchiata ...... 11 Be Gasteropoda)j2 OV isin kek ening Pteropodas. secigisiat wick & il: < Meteropodauc...u ee. aes RO ee Cephalopoda ss. er Dit.) U5 i WENLOCK. I have stated that 58 genera and 125 species are common to the Wenlock rocks and the Upper Llandovery; in other words, these 125 species pass up from the Llandovery to the Wenlock formation. Next to the Caradoc and Bala, the Wenlock group, embracing the Tarannon Shales, Denbighshire Grits, and Woolhope beds, possesses the largest Silurian fauna known; it numbers 168 genera and 530 species. Hight chief groups or classes out of the fourteen furnish the majority of the species, in each case attaining their maximum in the Wenlock Limestone. ‘They are the following :— said now oa yi ats ees 6 genera and 30 species. 2s Achimozoa nts dea a 25 45 ROr es 3. Echinodermata ...... 28 genera and 68 species. 4: Orustacealys)) 28). 210.8 27 3 Vestn Ones oi) brachiopodaniee ase 21 F Oss ees 6. Lamellibranchiata .... 16 He AB yk Ge (Gasveropoda (ia. s en: 9 : PATRIA &. Cephalopoda =) .. 78 5 y DOT es I may add the Bryozoa, for at no period in the history of the 159 THE PRESIDENT. S OF ES SARY ADDR ANNIVER Taare XII.— Upper Llandovery. Geographical Distribution. net = o i w o Pat ot ats) P: i re q a a | 4 ae ass Se Classes. a g a = S g a g Os to oir : \ ng :: A rz 2 a =) 08 ag rd Wenlock. wb) 3 n ia &o Oo a4 o © n m 5 a o ES Ce Wee en coh ge erate eegin Sh calgon er enemas Sg | 3 Ee 3 = ae: = 3 3 be i= 5 es B) a, ) @ @ S A oa! te ae So co) By a) MD or Oo o) ete wn O A) HO RQ Plone atcmamectsrakealt acl 1 oa 6 oe : 1 EOUOZO 8 ax the ates catewete, 2 UE) Vee a vee ves wee ves te 1 “ vee i Heled |) HAV GIOZOA:. vac hes ccavtevcescule — 4 33 *F ... on ie ve tee oe nee vee 2's 30 9 : " 3 5 3 10 6 7 4 7 a0 UNGEINOZOBN. cate: A ctor oul RHO 32 4 12 i 3 14 10 18 9 10 22 Eechinodermata..............., 5 i) 5 vee oC 1 5 I 2 vee 2 /NOIMIOES coemaooddgecorcodecon|=. & 4 3 “ ves 1 3 3 3 3 vee ' 4 8 7 7 6 7 6 7 7 6 a Onustacode. ncn all Pie || art 13 : 8 13 & 13 15 vee 7 14 2 . é N 2 2 BY 1 1 1 < 2 BUYOZOS ayaacnaieesianescsecl 2 7 vee 3 vee ves 4 3 1 2 : 1 3 10 . 7 tC 11 6 9 11 13 PU EMECHIONOCay ins. cme: LG 65 Ae an im As ate 27 35 13 000) 12 44 2 Lamellibranchiata .........) 12 29 7 4 vee see 3 8 2 é ee 1 15 6 : ‘ 4 3 9 3 4 4 3 ; g 1) Grasteropoda...c.c.e.0 LO 28 3 7 3 11 é 9 5 nee 4 8 UREGHOMOG Esta dew. tevteenccw cs 1\ Tit exe fo BEC Guie ade A eee i 2 1 1 1 1 1 1 1 1 2 ? Re terO pode. .ce.cmeacnnl wee Hel IY PE i . 1 3 1 3 7 os 4 6 1 | Cephalopoda ...... cenceualh Ome ie Ong 4 aa 1 é : 3 ve 4 9 A5 38 42 22 54 41 48 41 6 35 59 104 thse eeeeees 91 261 H 658 90 I 40 106 70 VOR ke 25 74 126 | i ae iS a 2) eS Np I Sa Le lA Nhe al ee ee al eee eel ee ARR ge * It is difficult to decide with regard to these species which are Upper and which Lower Llandovery. t L also give these 6 genera and 25 species in the Upper Llandovery as well as the Lower, for I really know not whose view to adopt as to their stratigraphical position, Mr, Avyeline and Prof, Hughes place them in the Tarannon Shales, Mr. Lapworth and Dr. Nicholson in the Lower Llandovery. 160 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Lower Paleozoic rocks were they so largely represented, and all the known forms in the Wenlock group (11 genera and 24 species) occur in the Wenlock Limestone and Shale, and in those of England alone, only 1 species occurring either in Scotland or Ireland. Protozoa.—Chona prisca and Spongarium Hdwardsii are both Denbighshire-grit species, the latter being also Wenlock and Lud- low; Cnemidium tenue occurs in the Wenlock only, and Ischadites Grindrodi in the Woolhope; the other well-known species, J. Keng, Murch., has a wider range both in time and space, and has occurred in the Wenlock rocks beneath the Cretaceous series at Ware, in Herefordshire, in the Silurian cores brought up from a depth of 1000 feet, during trials for the better supply of London with water. 30 Wenlock species were obtained here from a few feet of cores; they are identical with the Wenlock species of Dudley and Wenlock Edge. (Vide page 229, in the Chapter upon the extension of the older rocks under the London area.) The singular and still doubtful genus Stromatopora (NS. striatella) occurs plentifully in the shales and limestones of the true Wenlock beds, and is also found in the Aymestry Limestone. This genus comprises a large number of Silurian and Devonian fossils of every size and form. The affinity of the Stromatoporids is still doubtful; by some authors they are placed with the Hydrozoa, in the subclass ‘“‘ Hydrocorallina.” The doubtful Devonian genus Caunopora suggests much research, both from its abundance and peculiar structure. In Britain Stroma- topora first appears in the Caradoc and Bala group, ranging upwards into the Middle Devonian of Devonshire. Hyprozoa.—The Wenlock rocks (assuming the Tarannon and Denbighshire beds to be at their base) have yielded 6 genera and 30 species. Stratigraphically they occur chiefly in the Tarannon beds, or probably the Denbighshire Grits. So far as | know there are none in the Woolhope beds, and I omit them from the column headed Wenlock, as they occur chiefly at Builth and in North Wales, and in this case would be repeated. The T'arannon or Denbighshire beds have yielded 5 genera and 23 species; North Wales 1 genus (Monograptus) and 8 species; South Wales 3 genera and 15 species ; Westmoreland 3 genera and 16 species; Scotland 3 genera and 5 species. Acrinozoa.—At no period during the progress of Palseozoic time was there so large and rich a Coralliferous fauna as during the Wenlock period; numerically the Middle Devonian about equalled the Wenlock species in number, the genera and species being in both deposits much the same. The Wenlock rocks of Britain yield 25 genera and 76 species, the Devonian 24 genera and 52 species, and the Carboniferous 36 genera and 141 species. I believe every known Wenlock species (there are 76 of them) occurs in the Wenlock Limestone and its subordinate shales. In the lower division, especially the Tarannon Shale and the Denbighshire Grits, the species are few, not more than five being known in the former and four in the latter. ANNIVERSARY ADDRESS OF THE PRESIDENT. 161 The Tarannon species are Havosites asper, I’. fibrosus, F. gothlan- dicus, Petrara bona, and P. subduplicata. The four Denbighshire-grit species are also Favosites aspera and I’. fibrosus, Petraia subduplicata and Syringopora serpens. The upper member of the Lower Wenlock (the Woolhope beds) yields 16 genera and 26 species, most equally distributed generically. The North-Welsh Wenlocks yield 7 genera and 11 species, and the South-Wales beds 10 genera and 18 species ; Scotland 8 genera and 11 species; and Ireland 14 genera and 23 species. We must not forget that 14 genera and 22 species came from the Llandovery rocks, and 18 pass to the Ludlow group, 13 of which were also Llandovery ; thus 41 species are peculiarly Wenlock. Ecutnoprermata.—As compared with the Upper Llandovery, with only 5 genera and 5 species, and the Lower Llandovery only 2 species, the Wenlock sea must have been highly favourable to the development of the Echinodermata ; no greater development of life is known in any formation, for 60 of the 68 known species of Echinodermata were new appearances in the British area. Many are American ; 65 of the 68 are Wenlock Limestone; no species is known in Scotland, and only 1 in Ireland (Acéenocrinus Wynne, Baily); only 1 species occurs in N. Wales in the Tarannon Shale (Actinocrinus pulcher), and only 2 in the Woolhope (Lucalyptocrinus polydactylus, M‘Coy, and Pisocrinus pilula, De Kon.). We should expect that out of so large a fauna many species would pass to the higher divisions of the Upper Silurian, but only 4 genera and 6 species pass to the Lower Ludlow. These are of the Crinoidea, Actinocrinus pulcher, Ichthyocrinus pyriformis, and Ichthyocrinus M‘Coyanus ; of the Cystidea, Pseudocrinites magnificus, P. quadrifas- ciatus; and of the Asteroidea, Rhophalocoma pyrotechnica. Ichthyo- crinus pyriformis is the only species occurring in the Aymestry Limestone ; the Upper Ludlow only holds 1 species in common with the Wenlock, the long-lived Actinocrinus pulcher. No less than 20 new Kchinodermal genera made their appearance in the Wenlock sea, and 17 of them are confined to the Wenlock rocks; the 4 genera that pass to the Ludlow are Ichthyocrinus, Taxocrinus, Pseudocri- nites, and fhophalocoma. ‘This marked, sudden, and important addition to the 3 orders of the Echinodermata can only be explained upon the theory of migration from some prolific area; whence it is most difficult to suggest ; but looking at the great generic and specific development of the Crinoidea, Cystidea, &c. in N. America, and the great resemblance, if not identity, of the faunas in the two areas, would lead one to surmise that it was from the west rather than from Europe that the Wenlock sea derived these HKchino- dermata. Awnetipa.—No other British Palxozoic strata possess so many species of Annelida; they number 35. The Upper Llandovery possesses only 4 species, the Lower Llandovery 3, the Caradoc 16, the Llandeilo 3, and Ludlow 17 species. This exuberance in the Wenlock strata is due to the researches of 162 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. one naturalist *, through whose careful investigations no less than 6 new genera and 24 new species have been added to the Annelidan group of the Wenlock. We know of but one Wenlock species in Scotland or Ireland. All except one species occurs in the true Wenlock Limestone and shales: 4 species pass to the Lower Ludlow (Helminthites, sp., Serpulites dispar, S. longissemus, and Sprrorbis Lewis) and 8 to the Upper (viz. Cornulites serpularius, Serpulites dispar, S. longissimus, Tentaculites anglicus, T’. ornatus, Tr ach yderma squamosa, and 2 species of Arabellies). CrustacEA.—Ten new genera appear here for the first time; it is important to name them—Turri lepas, Aichmina, Cyphaspis, Deiphon, Eurypterus, Hemiaspis, Pterygotus, Thlipsura, Cythere, and Cythe- rellina. 'The entire Crustacean fauna consists of 27 genera and 78 species; the Upper Llandovery transmits to it 7 genera and 14 species: 74 of the 78 species occur in the Wenlock Limestone ; and the 4 wanting species are Homalonotus cylindricus, Ceratiocaris ludensis, Primitia excavata, and P. lenticularis. The Woolhope beds hold a remarkable Crustacean fauna, comprising 13 genera and 24 species : this class and the Brachiopoda are the only two marked zoological groups in this division of the Lower Wenlock series ; 12 of the Trilo- bita are Upper Llandovery, and only 1 is distinctly a Woolhope species, viz. Homalonotus cylindricus, Salt. The remaining 3 species of Crustacea in the Woolhope are Primitia excavata, P. lenticularis, and Cythere Grindrodiana, Salt., all belonging to the Ostracoda ; they are of no value. We thus rather unexpectedly find that the Woolhope formation possesses groups of fossils almost equal to those of the Wenlock Limestone; such as the Actinozoa with 26 species, 18 of which are also Upper Llandovery; the Crustacea with 24 species, 19 of which are Trilobites; the Brachiopoda with 17 genera and 56 species out of the 21 genera and 96 species known in the Wenlock Limestone. The poverty of the Crustacea in the Tarannon and Denbighshire Grits might be expected from the lithological characters of the rocks; only 8 of the 78 species occur in the former, and 5 in the latter horizon. North Wales has only yielded 5 genera and 11 species, and South Wales 9 genera and 14 species ; only 3 species occur in Scotland—Turrilepas Wirightianus, Beyrichia Kleedeni, and Phacops Stokes. 'The Irish species number 8—Calymene Blumenbachit, var. Allportiana, Cheirurus bimucronatus, fnerinurus punctatus, EL. variolaris, Phacops caudatus, P. nudus, P. Stokesti, and Proétus latifrons. The order Merostomata makes its first appearance in the Wenlock through Hurypierus punctatus, flemiaspis horridus, and Pterygotus problematicus, all in the Wenlock- Limestone series. This group of Crustacea has been extensively and critically worked out and illustrated in a masterly manner by Dr. H. Woodward, F.R.S., in the volumes of the Palzonto- graphical Society, the Quarterly Journal of the Geological Society, and the ‘Geological Magazine.’ The suborder Eury pterida i is repre- * G. J. Hinde, Quart. Journ. Geol. Soc. vol. xxxvi. pp. 368-378. ANNIVERSARY ADDRESS OF THE PRESIDENT. 163 sented by 5 genera, 3 of which commence or first appear in this division of the Upper Silurian strata. No less than 12 species of Primitia, besides other Ostracods, have been determined by Prof. Rupert Jones, F.R.S., two of them being Woolhope (Cythere Grin- drodiana and Primitia lenticularis). Of the whole group of Crus- tacea, 13 genera and 18 species pass up to the Lower Ludlow, whose Crustacean fauna numbers 18 genera and 47 species. Bryozoa.—Of the 11 known genera that occur, 6 are new in the Wenlock. No less than 24 species are known in the shales and limestones; none have occurred in the Tarannon Shales, Den- bighshire Grits, or anywhere in North Wales or Scotland. In South Wales 2 species are known, Drastopora consimilis, Lonsd., and Péilo- dictya lanceolata; the same species occur in Ireland, and, with Escha- rina angularis, Lonsd., pass to the Lower Ludlow. Fenestella assimilis, F. subantiqua, and Ptilodictya lanceolata were also Lower Llandovery species. The 6 genera not known in the older rocks, and which first appear in the Wenlock, are Cellepora, Ceriopora, Diastopora, Discopora, Hscharina, and Polypora. Most of the Paleozoic genera belong to the order Gymnolemata, suborder Cyclostomata. We know of no Paleozoic genera extending into the Secondary Period, where this class attains its maximum paleontological development, the Cretaceous system alone having yielded more than 200 species. BracwropopA.—In the Wenlock rocks, like the Caradoc, the Bra- chiopoda greatly predominate over most of the other groups. We have determined 15 genera and 109 species in the Caradoc; and now in the Wenlock are enabled to show that 22 genera and 101 species occur, being an increase of 7 new genera (namely, Athyiis, Cyrtia, Fiichwaldia, Nucleospira, Obolus, Orbiculoidea, and Retzia), the species being fewer so far as we know through the literature of the class. 21 genera and 96 of the 101 species belong to the Upper Wenlock or Wenlock Limestone and shale*; all the known 22 genera are also represented in the same horizon. The Woolhope beds yield 17 genera and 56 species, every Woolhope shell being also Wenlock. Orthis mullochensis, Dav., or its variety O. reversa, may stand alone as a Woolhope species, but every other form is equally Wenlock. Certain horizons in the Wenlock rocks are richer generically and poorer in species than others; this is shown in the Tarannon Shales and Denbighshire Grits, where, in the former, 10 genera occur with only 15 species; it is the same with the Denbighshire Grits, in which we have 14 genera and only 19 species. No genus or species is peculiar to either the Tarannon or Denbighshire beds; all are good Upper Wenlock forms. ‘Those species having the longest range in time in the Wenlock group number about 11; they also have a correspondingly wide range in space; they are :— * The researches of Mr. Davidson, F.R.S., upon the collection made by Mr. G. Maw in these beds has enabled him to add many new species and one or two new genera to the fauna of the Wenlock rocks. 164 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Atrypa reticulata. Rhynchonella nucula. Chonetes striatellus. Spirifera elevata. Crania implicata. exporrecta. Leptzena transversalis. Strophomena pecten. Meristella tumida. rhomboidalis. Orthis elegantula. All those named have appeared in the Llandovery or Caradoc before, and all pass to the Ludlow; but their universality through the Wenlock renders them conspicuous or long-lived, though not perhaps the best witnesses in determining the age of any particular horizon through the Wenlock deposits; those species which are short-ranged, or essentially typical or confined to particular beds, should be named, as contradistinguished from those given above. No species is peculiar to the Tarannon Shale, although 10 genera and 15 species range through this series of beds. No species essen- tially marks, distinguishes, or is confined to the Denbighshire Grits ; yet 14 genera and 19 species also occur in this horizon. Although 17 genera and 56 species are known in the Woolhope beds, yet only 2 species are restricted to them; these are Orthis mullochensis, Dav., and Pentamerus undatus, Sby. This last species occurs in the Woolhope (?) of Ireland. Every Tarannon species except Lingula Symondsw had previously occurred, or appeared in one or other of the two Llandovery horizons or in the Caradoc. All the species in the Denbighshire Grits except three, Discina Morrisi, Meristella tumida, and Rhynchonella navicula, also have appeared in and passed up from the same deposits. Thus we find that the restricted Brachiopodal fauna (so far as we at pre- sent know) in these beds at the base of the Wenlock is extremely scanty. Out of the 101 known Wenlock species, 56 had not occurred in any earlier formation, but first appeared in the Woolhope and Wenlock Limestone; these newly introduced Wenlock forms repre- sent 20 genera. I deem it important to name them and give the numbers of the species in each genus; they are :— VNIgwAAt) Godgsosdacgac0504900000. o.species: §|) Ololtisy so peusasceses eee eeeee 3 species. Athy Dain cccssase care sleeniands-bi- Deities Orbiculoideayeac.-ceeeree is Orantas pecccen tee cosas ccc So One Orthig.:....Jssoshe ee eee Ons frat (ON Acti codaudocanaddeqoqoqouandee A Re Pentamenus ys -cseee ee eeeee Sar Diseina! tkvcerc teers coanees ge aa BRetzial since Waetenceeeeeeeeene Pap iWichiwaldiaeeece eee ree EN 4 Rhynchonellal) -s-ceeseeee Aner, LDS Oh AE), ma pdcagsussuscoso0coo Zieteee Siphonotretayy... cases eerie 1 ye IGHAGWIEY, Soncssgnuebsosreooods000 oa Spirilera. 2.) wcckece eee Subs Meristellapea tee 4 , Sirophomlenals- ceca eeeeeeee Sis ee INucleospiraie. a--ccnsde ser gael ir foes PTriplesia..c..25. ds. saan EO Fes Out of so large a fauna as 101 species, 96 of which occur in one division (the Wenlock Limestone and Shale), we should expect to ANNIVERSARY ADDRESS OF THE PRESIDENT, 16 5 find a large proportion absolutely restricted to that horizon. Of such restricted species in the true Wenlock rocks (limestones and shales) we have 24 or 25 per cent. None of the following species occurs in any other horizon; and they may therefore be accepted as definitely typical Wenlock species— bs S ) Oo eS o = [es a) i) iS 5 HA Fi4liale ia 2 1 3 1 1 1 2 1 3 1 1 1 5) 1 3 3 3 23 : r 0 8 15 16 5 2 2 16 25 7 10 8 5 3 26 76 11 1&8 sxe 11 1 2 28 1 4 * 1 2 65 1 5 00 1 3 13 4 4 sf 1 3 34 4 6 © 6 3 14 2.7 5 9 s 8 5 2:5 74 11 14 3 3 10 2 4 24 eee 2 e 10 14 17 21 9 12 11 15 19 06 96 23 39 20 1 7 3 15 8 11 ? 2 10 5 43 14 26 1 4 4 9 E 7 1 1 5 8 24 8 al te " 1 l 3 1 1 1 ee 1 ee 3 1 i « 1 1 1 2 1 1 Mins 3 3 5 4 3 9 3 3 5 4 3 1 vi 4 26 12 13 3 26 39 67 1G1l R47 68 3 29 54 57 137 473 98 160 16 45 * In the column headed Wenlock beds I omit the Rhabdophora; they appear in the localities, Pass to Lower Aymestry Upper Ludlow. Limestone. Ludlow. 3 2 2 3 2 2 1 2 9 ee 4 16 8 7 4 1 2 6 1 2 3 3 4 4 3 6 13 5 8 18 6 10 2 1 2 1 12 10 iil 32 20 17 13 3 9 21 3 24 6 4 4 9 6 6 2 Be Uh 2 2 2 Pee ne {he tens. 2 TPs 2 3 11 2 3 71 38 51 126 562 817 170 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Luptow. 392 species, representing 137 genera, is, as nearly as can be ascer- tained, the numerical value of the Ludlow fauna. No true Ludlow rocks are certainly known to occur in North Wales, excepting perhaps in a small part of Montgomeryshire, south-east of Newtown. The Ludlow formation is zoologically allied to the Wenlock through 129 species. The chief classes, or those most represented numeri- cally, are the Actinozoa with 12 genera and 17 species, the Echino- dermata with 12 genera and 21 species, the Crustacea with 29 genera and 97 species, the Brachiopoda with 13 genera and 48 species, the Lamellibranchiata with 17 genera and 71 species, the Gasteropoda with 9 genera and 33 species, the Cephalopoda 6 genera and 37 species, and lastly the Vertebrata, which first appear through the Fishes, represented by 10 genera and 14 species, omitting the genus Onchus, the remains of which may be “telsons” or tail-spines of Crus- tacea. The comparatively small fauna of the Aymestry Lime- stone (53 genera and 84 species) is, although local, a remarkable one, and clearly shows the nature of the sea-bed and sedimentary material in relation to the habits of life of the species represented. Mostly it is a recurrent fauna from the Wenloek, 57 species being common to the two, these 57 representing 38 genera, clearly showing the rarity and paucity of species illustrating the genera, and the lenticular or accidental nature of the Aymestry Limestone between the Lower and Upper Ludlow beds. The classes that appear to be most fully represented are the Coelenterata (Actinozoa), the Crusta- cea, Brachicpoda, Gasteropoda, and Cephalopoda. No fish-remains are known, although the two Ludlow series contain them*. The feeble representation of the Ludlow rocks in North Wales, as before stated (only 19 species), by comparison with Westmoreland (127) calls for critical examination into the physical rather than the pale- ontological aspect of the question; and this is still more prominently seen through the still larger fauna in South Wales and Shropshire, the former yielding 163 species, the latter 187; and further through the intermediate areas termed Herefordshire (where 114 species are known) and Worcestershire (121). Could we but see and examine the continuity of the old Wenlock and Ludlow sea-bed and its deposits between Denbighshire, Westmoreland, and the south of Scotland, now under the St.-George’s Channel (or denuded away), we should doubtless find traces of the Lower Wenlock and Ludlow rocks, uniting the two now disunited areas. We cannot doubt that this continuity once existed, though probably the rocks of the two areas were deposited under very different and local conditions. Petrologically they essentially differ ; but paleeontologically or speci- fically in certain groups the agreement is close and decisive, omit- ting the largely represented Annulose or Crustacean orders Meros- * Mr. Salter included the Aymestry Limestone with the Lower Ludlow, as being a calcareous condition of it. Mr. Lightbody was always impressed with the fact that the calcareous nature of the beds above the Aymestry Binatone influenced the fossils to a marked degree. ANNIVERSARY ADDRESS OF THE PRESIDENT. I71I tomata and Phyllopoda in the Ludlow rocks (45 species), which stand alone as a peculiar and local group; their presence there- fore need not enter into the calculation*. The great extension of the Wenlock promontory to Wellington, running parallel to the Dudley beds, and the overlap and cover of the Triassic rocks to the north on to the Mersey, Liverpool Bay, Lancaster and Morecambe Bays, go far to lead us to believe that the missing evidence of the agreement between the Ludlow of the two areas of Wales and Westmoreland, as well as the unsatisfactory correlation of the Den- bighshire Grits and Tarannon Shales with the rocks of the Lake country, can be accounted for ; for it must be admitted that much has yet to be done with these Lower Wenlock rocks of Denbighshire, Westmoreland, &c. The Kendal group (or the Ludlow rocks) and the Ireleth Slates, which are the equivalents of the Wenlock of the south, can certainly be correlated paleontologically, dissimilar as they appear physically. The Scotch uppermost Silurians are exhibited only in three localities; to the south of Kirkcudbright and in the Pentland Hills both the Wenlock and Ludlow groups occur. The American series termed the Lower Helderberg group are the true equivalents of our Lower and Upper Ludlow, and contain a number of species in common. ‘The exact equivalent of the Oriskany Sand- stones may be our lowest Devonian. In North America their place is either at the top of the Upper Ludlow or between that and the lowest Devonian rocks, or ‘‘Corniferous beds” of the American geologists. I now attempt to analyze the extensive fauna of the Ludlow group, accepting the triple division unconditionally ; or, in other words, retaining the Aymestry Limestone as a distinct subfor- mation, not attached either to the Lower or the Upper Ludlow; its accidental position carries with it certain fossils having an important bearing upon physical geology. Prantra.—For the first time above the Wenlock we meet with what may be termed plants proper; yet the two genera Chondrites and Fucowdes are doubtful. 5 genera and 5 species are known; 4 species occur in the Upper Ludlow, and 3 of the same in the passage-beds between the Upper Ludlow and Lower Old Red Sand- stone ; none are known in the Lower Ludlow or the Aymestry beds ; nowhere in the Ludlow of either South or North Wales, Westmore- Jand, or Ireland are plant-remains known. Actinophyllum plicatum occurs in Shropshire and Herefordshire; Chondrites verisiemilis 1s an Upper Ludlow form in Scotland. The seed-like bodies of doubtful affinities (Pachytheca spherica) occur in the tilestones and bone-bed of the uppermost Ludlow of Shropshire, Worcestershire, and Here- fordshire. Actinophyllum and Pachytheca here first occur, and do not pass to the higher formations. The passage-beds contain Acti- nophyllum plicatum, Chondrites verisimilis, Pachytheca spherica, and Pachysporangiwm pilula. 15 species of Crustacea, 19 species of Lamellibranchiata, 6 species of Gasteropoda, 4 species of Hetero- poda, and 6 species of Fish, or nearly 70 species, illustrate 10 classes * Phyllopoda (Ceratiocaris 14 species), Merostomata (Hurypterus 10 species, Hemiaspis 6, Pterygotus 9, Slimonia 3, and Stylonurus 3). 172 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. in this borderland between the Upper Ludlow and the Old Red Sandstone. Prorozoa.—A series of sponges (Calcispongie) numbering 7 genera and 10 species occur in the Ludlow rocks, 3 of which are also Wenlock. 6 genera and 8 of the species belong to the Upper Ludlow; Amphispongia, Favospongia, and Tetragonis are new forms, and appear for the first time, and, with Mavospongia Ruthven, Ischadites Kenigr, and 4 species of Spongariwm, die out in the Upper Ludlow. 3 genera and 5 of the species occur in the Ludlow beds of Westmoreland; they are Favospongia Ruthven, Spongarium equstriatum, S.imterlineatum, S. interruptum, and Tetragonis Danbyt. The conformable Old Red Sandstone in the Ludlow area, or anywhere along the extended strike and junction of the two formations, con- tains nothing in common with the Upper Ludlow save a few fish and crustacean remains, all ceasing with the Downton beds and tilestones. The Eurypterids (Sty Wlowan us megalops and Hurypterus pygmeus) pass to the lower beds of the Old Red Sandstone; but through all the Old Red of South Wales (or the Silurian area) or Gloucestershire or Somersetshire, from its base to its junction with the Lower Carboniferous, nothing has occurred save fragments of Cephalaspis and Holoptychius, no other class in the animal kingdom being represented. Hyprozoa.—Only 1 genus and 8 species of this class seem left to bring to a close the history of the Graptolite group of the Hydrozoa. Monograptus has 8 species, which are all Lower Ludlow and occur in the Ludlow area. In no region of the globe where the Silurian rocks have been deposited have any species of this class trans- eressed or passed into higher formations. About 210 species belong- ing to 30 genera have lived during the Silurian epoch, ranging from the Arenig with 42 species, the Llandeilo 44, Caradoc 38, Lower Llandovery 50, Upper Llandovery 12, Wenlock 23, to the Ludlow with 8, their maximum development being at the commencement of the Lower Silurian of certain systematists. They came into existence with unexampled prodigality and suddenness, but gradu- ally died out as conditions grew less and less favourable for their sustenance and development. America, Bohemia, Scandinavia, Australia, and Britain have all largely possessed the same hydrozoal fauna, illustrating the same or similar deposits. Whether homo- taxially or not, most of the same genera have occurred in rocks of the same age as those of the british Islands, and always under the same physical conditions. The labours and researches of Hall, Barrande, Geinitz, Linnarsson, Nicholson, Lapworth, Hopkinson, Carruthers, and M‘Coy in this field of zoological history will ever reflect honour on their memory. To Mr. Lapworth graptologists owe much; no other author has given such attention to the Rhab- dophora, whether we regard his specific work or that devoted to their history and geological distribution through time, and his phi- losophical views relative to their affinities, structure, and systematic classification. Actinozoa.—Not a single species out of the 17 occurring in the ANNIVERSARY ADDRESS OF THE PRESIDENT. DRS Ludlow beds strictly belongs to them or is confined to the Ludlow ; all have previously occurred in the Wenlock. 17 species are Lower Ludlow; 8 of the same occur in the Aymestry Limestone, and 7 in the Upper Ludlow, chiefiy the Favosites (5 species) ; 13 species occur in Shropshire, 6 in Westmoreland, and 12 in Worcestershire. Monticulipora papillata is the only Scotch species known. 6 species occur in Ireland (Fwosites asper, F. cristatus, F. fibrosus, Alveolites Labechu, Cr yathophyllum truncatum, and Monticulipora papillata) only 1 species is known in North Wales (Heliolites tubwlatus) ; and. 6 genera and 9 species in South Wales. ‘The accompanying Table, showing their geographical distribution, will be clearer than descrip- tion, where the species are so irregularly distributed. oO . 3 eeliccs Ri es see ob ve o © e oS “3 3S Cet] Se a il O + S 5 Ludlow Species. Peal ete Hee eral Et resem ee le eaten = | ts ete ara SS ath By eles as) = (e) S © on) Gs | & S B = S a On| Wasa eees S) Sees) is = So | = aD |) 24 |p| Eo es ca | Alveolites Labechii .................. % % % % * Coomites mbertextus .c.2..<.5 .csecss. * Cyathaxonia siluriensis ............ ea al wae Cyathophyllum truncatum ......... “3 % ae * Favosites GDSTTE “yscqssoandodnosdaroce % % * Benalla % cristatus SE TaN ota eae baa. || sc00 * ae * x Bee =A DNOSUG CS AS ticle osu -wetdc dev doek % x * % % —— gothlandicus ....... pactasaets * x | * | x Horbesi. Fistulipora decipiens. Halysites catenularius............... See ae le % Heliolites interstinctus ............ Sean |e 0) |e ¥ % x fubulatus.....s..:. EC or ne AOA ae asst * * % Monticulipora papillata ...........- Stay eeu kere Ana Deen fe IE gh 9G- hea Ompbyma turbinata ..............- Ree ee Sea aes | aE * Pabraley OYA. ......+csceey ese Repth arr Ser diese malts | Fae Syringopora bifureata.......... oobl|.cok: |leodoin lesion lle we Be MPCUISE waht vssiaeshiewies sacievee * % * 9 US Aa 76 6 1 6 EcurnopERMATA.—This group in the Ludlow comprises 12 genera and 21 species. There are 3 genera and 4 species of Crinoidea (Actinocrinus pulcher, Ichthyocrinus pyriformis, I. M‘Coyanus, and Taxocrinus @Orbignyt). Of the remaining 17 species 4 are Cys- tideans (Hchinocystites pomum, H. uva, Pseudocrinites maguificus, and P. quadrifasaatus), all Lower Ludlow; and 13 Asteroidea. 2 species of Palwaster (P. Ruthvent and P. hirudo) and Pale- asterina primeva, with Protaster Sedgwicku, occur in the Upper Ludlow of Kendal, Westmoreland. 4 Palwocome (P. Colvini, P. cygnipes, P. Marston, P. vermiformis) are all Lower Ludlow from the Leintwardine beds near Ludlow. Palwodiscus ferox, Protaster leptosoma, P. Miltom, and P. Sedgwicku are from the same horizon and locality, as well as Rhophalocoma pyrotechiica—Shropshire thus 174 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. yielding 8 species of Lower Ludlow starfishes and Westmoreland 4 Upper Ludlow. The Lower Ludlow Echinodermata therefore number 11 genera and 19 species, the Upper Ludlow 6 genera and 7 species. None occur in Herefordshire, Scotland, or Ireland. 4 genera and 7 species (3 Crinoidea, 1 Cystidean, and 3 Asteroidea) came from the Wenlock beds. In the Ludlow promontory and all along the line that marks the junction of the Old Red Sandstone with the Upper Ludlow all Silurian life ceased. The slow elevation of that region caused the Silurian Annuloid fauna to dwindle and pass away almost entirely. No vestige of it again appeared in the true Old Red area anywhere; and the want of evidence to show what rocks may be below the Lower Devonian of the Devonshire and Somersetshire areas leaves the question of stratigraphical or zoological continuity one of doubt and in the region of the unknown. Ireland or North Devon may yet solve the problem, for it is here we should expect it. The new Ludlow genera were Hchinocystites, Paleasterina, and Tetragonis, none of which leaves any successors behind, none being known either in the Devonian or Carboniferous rocks. ANNELIDA.—The Annelida are a singular group in the Ludlow rocks; 9 genera and 17 species have been obtained from the several horizons (4), and their geographical distribution is wide also. 15 species occur in the Upper Ludlow; the characteristic Cornulites serpularvus is the most abundant, occurring at all 4 horizons, and in all localities except Shropshire and Ireland. The Lower Ludlow has yielded Cornulites serpularius, Serpulites dispar, S. longissimus, Spirorbis Lewisv, and Trachyderma squamosa. The Aymestry Lime- stone contains 4 species, all of which are also in the Lower Ludlow. 4 of the 17 species are also Wenlock. 3 species of Afnonites and 3 species of Arabellites, belonging to the order Polycheta, are determined from jaw-remains by Mr. G. J. Hinde*; they are Upper Ludlow, and help to swell the Annelide fauna. 3 species of the order Tubicola occur in the Passage-beds. 9 species range through South Wales, North Wales has 4, Westmoreland 6, Shropshire 5, Worcestershire 7, Herefordshire 4, and Scotland 3; but none, so far as I can ascertain, occur in Ireland. Crustacra.—The two great orders of the class Crustacea in the Ludlow rocks are the Merostomata and the Phyllopoda. Of the former we know 32 species—Lurypterus 10, Pterygotus 9, Himan- topterus 1, Slimona 3, Stylonurus 3, and Hemiaspis 6. Of the latter, through Ceratiocaris and Dictyocaris, we have 16 species ; and there are possibly one Amphipod (Necrogammarus Salweyr) and a Peecilopod (Neolimulus falcatus, Woodw.). These 32 species of Merostomata and Peecilopoda swell up the Crustacean fauna at the expense of the Trilobita, which number only 10 genera and 20 species, 2 genera only of the 10 (Homalonotus and Phacops) passing to or occurring in the Devonian rocks. The Ostracoda number 5 genera and only 13 species. The Crustacean remains termed Asta- coderma by Dr. Harley, number 14 so-called species; all are in the * Quart. Journ. Geol. Soc. vol. xxxvi. pp. 368-376. ANNIVERSARY ADDRESS OF THE PRESIDENT. 175 Lower Ludlow. The whole Crustacean fauna, therefore, of the Ludlow rocks is 29 genera and 97 species. The Lower Ludlow contains 18 genera and 47 species, the Aymestry Limestone 9 genera. and 12 species, and the Upper Ludlow 23 genera and 71 species. The Passage-beds into the Old Red Sandstone have yielded 9 genera and 15 species, 4 being Ostracoda, 1 Phyllopod (Ceratiocaris decorus), and 10 Merostomata (Eurypteride). No Trilobite occurs in the Passage-beds. This Passage-bed Crustacean fauna is what we should have anticipated ; the Brachiopoda and Mollusca proper show the same indications of a change from a deeper to a shallow coast- line, and a fauna struggling to maintain life against new and adverse conditions. Shropshire through the Ludlow area possesses 18 genera and 46 species, Worcestershire 17 genera and 25 species, Hereford- shire 14 genera and 26 species, Westmoreland 8 genera and 15 species. Scotland, chiefly through the Phyllopoda and Merosto- mata, has 12 genera and 22 species. Ireland only possesses 2 (Calymene Blumenbachu and Phacops caudatus). This deficiency of the Crustacea in Ireland, as indeed of all the classes except the Brachiopoda, bears out the fact of the smallness of the Ludlow fauna in Ireland. I am obliged to carefully detail the number of genera and species that range through and enter into the distri- bution of the Crustacean fauna on account of its peculiarities. 13 new genera made their appearance during the Ludlow period. 4 Ostracoda, Cypridina, Entonus, Kirkbya, and Moorea ; 4 Merosto- mata (Hurypterida), Himantopterus, Slimona, Stylonurus, and Parka; Xiphosura 1, Neolumulus; Phyllopoda 1, Dictyocaris ; Am- phipoda 1 ?, Necrogammarus ; Crustacean remains, Astacoderma (14 species). 3 genera (Stylonurus, Cypridina, and Parka) occur in the Old Red Sandstone, but not in connected areas. 16 genera and 26 species are common to the Wenlock and whole of the Ludlow, thus reducing the true Ludlow Crustacean fauna to 71 species instead of 97. Bryozoa.—Heteropora crassa, Cerropora sulcata, and Ptilodictya lanceolata are the only 3 Ludlow Bryozoa. The first and last are Wenlock also. They occur in the Lower Ludlow and Aymestry Timestone of Worcestershire and Shropshire. Ceriopora sulcata occurs in the Ludlow of Ireland. No species occurs in the Upper Ludlow or Passage-beds, either in South or North Wales, West- moreland, or Scotland. The nature of the Ludlow shales and mud- stones doubtless was the cause why the Actinozoa, Hydrozoa, Kchi- nodermata (Crinoidea &c.), and Bryozoa, all mostly dwellers in clear water, are so sparingly developed and distributed through the Ludlow rocks, the Aymestry Limestone being of little palaeontological value owing to its lenticular disposition, mode of occurrence, and uncertain continuity. Bracuropopa.—Although by no means a small fauna, the Ludlow Brachiopoda numbering only 48 species and 13 genera are less than half as numerous as those of the underlying Wenlock rocks. 41 of these 48 had hved in the Wenlock seas and passed to the Ludlow ; therefore only 7 species are Ludlow proper; and only 3 species passed to the Devonian—Altrypa reticularis, Strophomena 176 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. rhombordalis, and Lingula cornea. The last named in the Ludlow country passes into the lowest beds of the Old Red Sandstone. The others named are “ Devonian” of the North and South Devonshire areas, and not known in the Old Red proper. In the Passage-beds into the Old Red of Herefordshire, Shropshire, and Worcestershire some 5 or 6 species occur; they are Chonetes striatellus (latus), Dis- cma rugata, Lingula cornea, Lingula nanima, and Strophomena rhomboidalis. They are associated in places with a few Asiphonida of shallow-water habits, Pterinea, Modiolopsis, Orthonota, &e. ‘The Lower Ludlow has yielded 13 genera and 38 species, 33 of which are Wenlock. 12 genera and 25 species occur in the Aymestry Limestone, and 11 genera and 24 svecies in the Upper Ludlow; but no genus is peculiar to either the Aymestry or Upper Ludlow beds, the whole being represented in the Lower zone. ‘The only species in the Ludlow not Wenlock are Lingula cornea, L, lata, Orthis canahculata, O. lunata, Rhynchonella pentlandica, and Strephomena ornatella. In other words, only these 6 species of Brachiopoda are really Ludlow forms. No species is known in North Wales, but 28 have been collected in South Wales. Here- fordshire, Worcestershire, Shropshire, and Scotland have each yielded 11 genera, and respectively 23, 27, 30, and 15 species; Westmore- land has 8 genera and 14 species ; Ireland 9 genera and 18 species. LAMELLIBRANCHIATA.—Both the Lower and Upper Ludlow pos- sess a large Molluscan fauna; the Lower 14 genera and 54 species, and the Upper 15 genera and 56 species; the intermediate Ay- mestry Limestone only 4 genera and 7 species—Pterinea lwans, P. retrofleca, P. Sowerbyi, Cardiola striata, Grammysia cingulata, Orthonota rotundata, and O. semisulcata. The chief genera are Pte- rincea with 12 species, Anodontopsis 7, Orthonota 16, Modiolopsis 5, Ctenodonta 4, Cucullea 4, and Grammysia 4 species. The Passage- beds contain 9 genera and 19 species, yet no single species occurs in the Lower Old Red Sandstone; with the change of lithological characters all life seems to cease also, or no record is left. For the life-history of those rocks, equivalent in time to those of the Old Red Sandstone in the classical Silurian area, we have to consult the De- vonian series of North and South Devon and Cornwall, where a fauna unsurpassed in perfection, if not in magnitude, in all the in- vertebrate classes exists—a group made famous in British geological history through the labours and researches of Sedgwick, Murchison, De La Beche, and Phillips. Westmoreland and Cumberland, through the researches of Sedgwick and M‘Coy, Harkness and Nicholson, in the Kendal area, exhibit a large Ludlow Lamellibranchiate fauna ; 12 genera and 37 species are known. Shropshire has 9 genera and 16 species, Worcestershire 8 genera and 17 species, Herefordshire 6 genera and 12 species, Scotland 6 genera and 13 species, and Ireland only 2 genera and 3 species. North Wales hitherto has yielded only 1 genus and 4 species; 13 genera and 21 species ranged from the Wenlock into the Ludlow, thus altering the numerical value of the Ludlow bivalves proper to 50 species, none of which passed the Passage-beds between the Upper Ludlow and Old Red Sandstone. ANNIVERSARY ADDRESS OF THE PRESIDENT. Maa Extensive patches of Upper Silurian rocks on the eastern side of the Silurian area and border of the Old Red, such as Woolhope, Usk, the long strike of the Malverns on to May Hill, thence across the Severn at Purton Passage to Tortworth, indicate great physical changes in the borderland between the fossiliferous Upper Silurian and the unfossiliferous Old Red; and whatever change brought in the marine Devonian of North Devon, south of the Mendip axis and on the same strike, and all the South-Wales Old Red beneath the South-Wales coal-field to the Bristol Channel, certainly must ac- count for the loss of all the Ludlow species ; that the Devonian area was one of long and continuous depression south of the latitude of the Mendips there cannot be any doubt, and probably the mass of Old Red Sandstone occupying Caermarthenshire, Monmouthshire, Breconshire, Herefordshire, and Worcestershire was being at the same time slowly elevated. GasTEROPODA.—9 genera (six of which had appeared in the Wenlock) and 33 species, 11 of which were also Wenlock, range through the Ludlow rocks; the Lower Ludlow has yielded 8 genera and 15 species, the Aymestry Limestone 5 genera and 11 species, and the Upper Ludlow 9 genera and 21 species. 3 genera and 6 species occur in the Passage-beds into the Old Red Sandstone ; they are Holopella conica, H. gregaria, Holopea obsoleta, Mur- chisonia torquata, Platyschisma helicites, and P. Williamsi. They, however, cease here, none living into the red beds of the Old Red Sandstone. North Wales has yielded only 2 species, Holopella gracilior and H. gregaria ; Scotland only 4 species, Acroculia anti- quata, Huomphalus funatus, Platyschisma semulans, and P. helicites ; Ireland has only registered 1 species, Huomphalus alatus; Here- fordshire 7 genera and 11 species, Worcestershire 6 genera and 9 species, Shropshire 7 genera and 14 species, and Westmoreland 6 genera and 14 species. 6 genera and 13 of the 33 species come up from the Wenlock and range through the Ludlow rocks; they are Acroculia haliotis, Cyclonema coralli, C. octavia, Kuomphalus alatus, E. carinatus, E. funatus, H. rugosus, Holopella gracilior, H. obsoleta, Loxonema elegans, L. sinuosa, Murchisonia Lloydi, and Mu. ar Beclaties. 9 species only are Lower Ludlow. All become extinct in the Tile- stones and Passage-beds. Preroropa.—Conularia subtilis and C. Sowerbyi, with Theca Forbesi, are all the species (3) illustrating this pelagic group. ‘The two last named are Upper Ludlow. Neither the Aymestry Lime- stone, Passage-beds, North Wales, or Worcestershire yield either species, nor ‘do we know of any species in Ireland. C. Sowerbya occurs in Scotland; all 3 species in Westmoreland and 2 (Theca Forbes and Conularia Sowerby yt) in Shropshire. These last are also Wenlock species. Hereroropa.— Bellerophon (6 species) and Heculiomphalus levis constitute the Ludlow stock of Heteropoda. Heculiomphalus levis, Bellerophon expansus and B. dilatatus occur both in the Lower and Upper Ludlow. 3B. Murchison, B. obtectus, and B. trilobatus are Upper Ludlow only ; 4 species occur in the Passage-beds, B. cari- 178 PROCEHDINGS OF THE GEOLOGICAL SOCIETY. natus, B. exepansus, B. Murchison, and B. trilobatus; 6 species occur in the Upper Ludlow of Westmoreland; Shropshire and Worcestershire possess 3 of the same species. None are known in the Ludlow rocks of North Wales, Herefordshire, Scotland, or Ireland. B. carinatus, B. trilobatus, and B. dilatatus are also Wenlock species. CrpHaLopopA.—No less than 24 species of Orthoceratites, 6 species of Phragmoceras, and 3 of Litutes, with other forms, make up the 6 genera and 37 species known in the Ludlow. Ascoceras Barrandii constitutes a new generic type. The Lower Ludlow has yielded 4 genera, Latuites (3 species), Orthoceras (15 species), Phragmo- ceras (6 species), and EHxosyphonites (2 species). The Aymestry Limestone has yielded 3 genera and 6 species— Tntuites giganteus, Orthoceras 4 species, and Phragmoceras ventricosum. The Upper Ludlow 5 genera and 22 species—Ascoceras Barrandi, Litwites giganteus, Orthoceras 17 species, Phragmoceras nautileum, and Tretoceras semipartitum. In the Passage-beds we have noted Orthoceras bullatum and Tre- toceras semipartitum. 16 species of the genus Orthoceras occur in Westmoreland, mostly Upper Ludlow, and no other genus has yet occurred there. Shropshire possesses 4 genera and 22 species, em- bracing most of the species in the Upper and Lower Ludlow divisions ; Herefordshire 5 genera and 21 species, Ascoceras Barrandu, Intwites 3 species, Orthoceras 11, Phragmoceras 3, Exosiphonites 2. Worces- tershire has 10 species. Orthoceras Maclareni and O. subgregarium are the only Scotch forms known ; and O. subgregarium is the only Irish species. 11 species come from the Wenlock rocks to the Lower Ludlow, but 16 Wenlock species range through the Ludlow group ; Ascoceras is the only new genus; and none ranged above wmto erther Old Red Sandstone, Devonian, or Carboniferous rocks. Pisces.—No Vertebrata have yet been discovered in earlier rocks than the Lower Ludlow. Scaphaspis (Pteraspis) ludensis, Salt., is the first fish known, and the only species in the Lower Ludlow; none have yet been detected in the Aymestry Limestone, but every known Ludlow form occurs in the Upper Ludlow, and five of them in the Passage- or junction-beds, through which about 20 species of the Silurian fauna pass to the Lower Old Red Sandstone, 7 of which are Crustacea (Merostomata), 6 Fishes, and 2 Cephalopoda, &e. The Crustacea common to the Ludlow and Old Red, and all be- longing to one order, are Hurypterus abbreviatus, HL. acumimatus, and E. pygmeus, Pterygotus problematicus, Stylonurus Powriet, S. me- galops, and Parka decipiens. The fishes that connect the two for- mations, but only in the Silurian area along the frontier of the Old Red, are Auchenaspis Salteri, Cephalaspis Murchison, C. ornatus, Pteraspis Banksii, Scaphaspis ludensis, and Eukeraspis pustuliferus. All belong to the Ganoidei. DEVONIAN. Prant#.—We have no clue as to the region in which the Devonian Plants first appeared; and yet small as is the flora of the British Devonian as compared with that of North America it 6éeT ~t i) Ne We tt i) n x nO OM NY OM Fe oy YY gay PY FA Do we woo From W enlock. “mojlpvI— ATX aay, PORYP EP OY OR RE By ao S, t+ & & 8 SS Gl SS ile ep Ay eo. SS Spletese | 6D Shy este oO 2 eo =| & = < 1 MEEOne es ce ee UG St Neos ea Sis Sco Siavces 2 & 2 8 S 5 © Vices SS eis: 8 e es 8 8 © s Moms se Seema Shs Bos os Q e : © ; © MG s 3 : 5 5 6 — : Oe Oe ae ke ke © wa ee BP: ye MEE sf eS Pee fe Pk o : RR RES Bo RoR Sees ; a " — ee Se Sil Sannvoataow Sow BHA oq Genera. Se lt =A eS eo Tp CO me b> — . Plea ao SH Hot TH Tw S oxy Species. DD No i} _ Ce te pK bt th et ° =F i bat et one ketiet DODO on® BA oo wo SO ID emo. ao OK AL 0 Lower Ludlow. 2 go S : é : try Lime- ee - OO He ue aN =e bone 5S? Aw wo Oe 5 am ¢ ee = i ) © — ie. _ tS SS om De s) Th = ° Ss wo | fo Sa on we So TO RE i ER ga se se 5 oa aa | Upper Ludlow. (Se Ow ° ° Di De on = 36 | OS BS S08 Coy aor Rowse we i 3 5 we | Passage-beds. 2 i= ° e ° e 25 wun ae eet pone a) as ae : Ere OD HH OD + : : | South Wales. pe] DSH Ef owe am ff we aw nm ee See 2 | North Wales. i or by Lo! [al CO et = . | We st 1: d @ ee C at Bre wo KO aK KO nP PR AD AK : Ge 9 estmoreland. ec | rs fas) ; ES 2 [me ph SH we Ex Seo So we Sy eu eo Ge om ee eo | Shropshire. Ez. oe 2 =) eS Be we 5 wa Zo SY ww SY am on To as ww | Worcestershire. a | om = e ; ae BS | me Ee we fa Do Bowe Se ae se we non | Herefordshire. ° ~ | i uo | Se pH awe Do of: RE we me aT | Scotland. nS | ee He wt Bo me won : Oe Treland. Ne ak Ne OO? DO et (9) z “UVIMOAR(T 0} ssug 180 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. shadowed forth the prolific flora of the Carboniferous epoch, which in the British Islands numbers 82 genera and 330 species; but the Devonian, which preceded it, has, up to the present time, yielded only 12 genera and 18 species. In America, however, about 95 genera and 160 species have been described through the researches of Dr. Dawson, Lesquereux, Vanuxem, Hall, Hartt, &c., especially the first-named author, who has devoted a long life to the eluci- dation of the Devonian and Carboniferous flora of the American region. 7 of the 12 British Devonian genera occur in America, viz. Calanuites, Lepidodendron, Psilophyton, Sagenaria, Spheno- pteris, Stigmaria, and Trichomanites ; but our ill-preserved speci- mens scarcely admit of specific determination, in fact even their affinities are often difficult to make out; and the poverty of our flora is clearly seen from the few species (18) occurring to illustrate the 12 genera. On such slender evidence it is difficult to conceive from what area our Devonian and Old Red Sandstone flora was derived; whether migration took place from Central Europe or America. The facies appears to be American rather than European, this being the case with the Scotch and Irish floras, especially -through Psilophyton, Paleopteris, and Sagenaria. The Lower Devonian (Lower Old Red) has yielded 2 species, Lepidodendron nothum, Unger, and Psilophyton Dechenianum, Carr., with coniferous remains from Caithness and Wick in Scotland. The Middle Old Red contains Caulopteris Peachi, Salt. (=? Psilophyton robustus, Dawson), and the two mentioned as occurring in the Lower beds. The Upper Old Red species (12) approach closely to the Car- boniferous. 7 genera and 12 species range through the Upper divi- sion—Advantites 1 species, Calamites 1, Filicites 1, Sagenaria 5, Sphenopteris 2, Trichomanites 1, and Anorria 1. Thus only 18 species occur in the widely extended and thickly developed deposits of the Old Red of Scotland, the Silurian Old Red, and the rocks of the Irish area. Prorozoa.—Scyphia turbinata, Spherospongia tessellata, and 5 species of Stromatopora all occur in the Middle Devonian; Cawno- ora, Phillips, Sparsispongra, D’Orb., and Coscinopora of Goldfuss are probably synonyms of Stromatopora, The Middle Devonian lime- stones of Torquay and Newton Abbot abound in the ever-varying forms of these incrusting Protozoa. The Devonian forms differ essentially from the Wenlock and Ludlow species, the type in those rocks being S. striatella, whereas the characteristic species in the Middle Devonian are S. concentrica, Goldf., and S. placenta, Lonsd., or the Caunopora of both Phillips and Lonsdale. No Protozoa occur either in the Lower or Upper Devonian, being strictly confined to the massive limestones of South and North Devon. S. placenta, S. ramosa, and S. verticillata are strictly British. Actinozoa (Actinoidea, Dana; Coralliaria, M.-Kdw.).—Perhaps during no period in the physical history of the British Islands have we had such a remarkable assemblage of Actinozoa as that which so essentially and specifically characterizes the Middle Devonian rocks of South and North Devon. Out of the 24 known genera and ANNIVERSARY ADDRESS OF THE PRESIDENT. 181 52 species, no single form passes to the Carboniferous, and none are common to the Silurian rocks in any area; they stand alone, and are sufficient in themselves to maintain the integrity of the Devonian system. This is equally definite and distinct throughout the Kuro- pean or American areas, or wherever the Devonian rocks are deve- loped. 15 genera illustrate the Zoantharia rugosa and 8 the Z. tabu- lata. The genus Acervularia numbers 7 species, Alveolites 4, Cya- thophyllum 12, Favosites 5, Smithia 3, Endophyllum 2, and Petraia 2. The remaining 18 genera are only represented by 1 species each ; 20 species are common to the rocks of North and South Devon, and 25 occur on the continent (Rhenish Prussia, Belgium, and France) ; none pass to the Carboniferous rocks in any region. ‘The Lamelli- branchiata, Gasteropoda, and Cephalopoda are all of equal strati- graphical value. The Middle Devonian group contains all the 24 genera and 48 of the 52 species. The Lower Devonian has hitherto only yielded 4 genera and 7 species; the latter are Alveolites sub- orbicularis, Cyathophyllum helianthoides, Petraia celtica, P. gigas, P. pleuriradialis, P. bina?, and Pleurodictyum problematicum ; and the Upper Devonian also 4 genera and 7 species, viz. Amplexus tortuosus, Cyathophyllum cespitosum, C.ceratites, Pistulipora eribrosa, Michelinia antiqua, Petrara celtica, and P. pleuriradialis. The researches of Dr. Nicholson upon the Devonian Corals of North America have thrown much light upon their history, habits, and mineralization ; and itis to be hoped that his labours will be embodied in a volume upon the Rugose Corals, equal in interest and value to his late contribution on the Tabulata. Doubtless the Paleeozoic Actinozoa of the British rocks now require critical revision, especially the Silurian and Car- boniferous groups. The Upper and Lower Devonian rocks are chiefly composed of slates, with here and there impure limestones. They therefore possess no coral fauna; whereas the highly developed masses of limestone around Torquay and Newton Abbot are simply Devonian coral reefs of great magnitude. In North Devon, between Lfra- combe and Linton, the limestones are lenticular, thin, and im- pure; yet more than one half of the entire known Devonian Acti- nozoa have occurred in them, and we have evidence of beds below at low-water level near Llfracombe yielding even finer specimens than at Torquay. They have to be searched for in North Devon and West Somerset ; patient working over that extensive area be- tween Barnstaple and the Foreland with unbiassed views and a knowledge of the fauna which exists altogether independent either of the Silurian or Carboniferous, would readily convince those who have never examined the county that the rocks of North Devon between Baggy and the Foreland have nothing whatever to do with theCarboniferous system. ‘The fossils alone, setting aside strati- graphical evidence and succession, determine the relation of this area to Belgium, the Rhine, and France; and their continuity under Somersetshire, Wiltshire, Berkshire, and Middlesex on to the continent is no longer matter of doubt or speculation, for the philosophical and far-seeing views and hypotheses of Mr. Godwin-Austen and Prof. VOL. XXXVIL. 0 182 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Prestwich have been fully realized, through the determination of the Devonian and Silurian rocks in Middlesex and Hertfordshire, immediately below the Cretaceous series. EcHINopERMATA.—/ genera of Crinoidea, 2 Asteroidea, and 1 Blastoid are all that are known of this class in the British Devo- nians. 80 genera have been described from the Devonian rocks of Europe and America. By comparison, therefore, the British fauna has scarcely any claim to recognition, and 5 of the 7 Crinoidea are also Carboniferous in Britain. It cannot be said, therefore, that the species in this division of the order are representative in Britain. The 10 genera and 24 British species are of necessity unequally distributed. Actinocrinus tenwmstriatus, Phill., Cyathocrinus megastylus, Phill., and C. pinnatus, Goldf., are the only forms known in the Lower Devonian ; 6 genera and 12 species occur in the Middle Devonian, viz. Actinocrinus 1 species, Cupressocrinus 2, Cyathocrinus 3, Hexa- crinus 3, Platycrinus 2, and Taxocrinus 1 species. The Asteroidea all belong to the upper division, and are mostly from the North-Devon beds, which with them have also yielded 7 Crinoidea. We have therefore 8 genera and 14 species in the Upper Devonian beds of North Devon—Protaster 2 species, Paleaster 2, Helianthaster 1, Pentremites 1, Adelocrinus 1, Actinocrinus 1, Cyatho- crinus 5, and Taxocrinus 1 species. Of the whole fauna (24 species) 3 genera and 6 species pass to the Carboniferous; they are Actino- crinus triacontadactylus, Cyathocrinus ellipticus, C. geometricus, C. pinnatus, C. variabilis, and Pentremites ovalis. Three of these are European Devonian. AnnELIpa.—Serpula advena, Salt., occurs in the Upper Old Red Sandstone of Caldy Island, and Yentaculites annulatus?, Schloth., in the Middle Devonian of North Devon; the last named is abundant (in places) in the limestones near Ilfracombe. CrustaceA.—All the 4 orders, illustrated only by 20 genera and 37 species, are represented in the British Devonians ; 45 genera and 290 species, however, have been described from Bohemia, Germany, Spain, Asia, Africa, America, &c. 20 of the foreign genera and 200 species are Trilobita ; the remaining 90 are Merostomata, Phyllopoda, and Ostracoda. As in the case of the Echinodermata our Crustacean fauna is any‘thing but representative, yet it has a characteristic facies that carries with it the conviction of distinctness. Only 6 genera of Trilobita are known, viz. Bronteus (flabellifer), Cheirurus (articulatus), Harpes (macrocephalus), Homalonotus (elongatus), Pha- cops (Sspeciesincluding the subgenera 7rimerocephalusand Crypheus), and Phillipsia (Brongniarte); the doubtful genera have been relegated to their supposed places. 9 genera and 24 species of all the orders are Lower Devonian (6 genera are Trilobita), 5 genera and 6 species are Middle, and 7 genera and 9 species are Upper Devonian. Eurypterus (6 species), Stylonurus (6 species), and Pterygotus (4 species) are all, with one or two exceptions, Lower Old Red Sand- stone types. None of these are known to occur in the typical Devonian area, being either Scotch or in the Silurian region. The singular genus Prearcturus (P. gigas, Woodw.) of the family Idoteide is ANNIVERSARY ADDRESS OF THE PRESIDENT. 183 from the Old Red of Herefordshire; this and Proricaris M‘Henrici of Baily form the only 2 new British genera introduced into the Devonian fauna. Phillipsia Brongniartii is the only crustacean that occurs in common in the Devonian and Carboniferous rocks ; it is the Asaphus obsoletus and A. granuliferus, Phill.; so that, small as the Devonian Crustacean fauna appears to be, it is nevertheless distinc- tive. A. Romer, Sandberger, Dalman, Brongniart, Richter, Minster, Beyrich, Steining, Roualt, &c. on the continent, with Salter, Phillips, Woodward, Jones, R. Etheridge, jun., and Baily in Britain, have all greatly enriched our knowledge of the Devonian Crustacea. Bryozoa.—The Tubuliporide through Ceriopora, the Retioporide through Henestella, Henntrypa, Polypora, Ptylopora, and Retepora, and the Hscharide through Glauconome, are represented through the Devonian rocks by the above 7 genera with 11 species; all the genera are equally Carboniferous, but only 4 species are common to both formations-—Ceriopora similis, Phill., Fenestella plebeia, M‘Coy, Glauconomne bipinnata, Phill., and Polyporalaxa,Sandb. The Lower Devonian possesses 2 species only, viz. Henestella antigua, also oc- curring in the Middle and Upper, and Retepora repisteria, which is also Middle Devonian. 6 of the 7 genera and 7 species are Middle Devonian—Polypora, through its representative species (P. lawa), being Upper and Carboniferous. 5 genera and 6 species are Upper Devonian ; they are both South- and North-Devon forms—Cerio- pora (Millepora) gracilis, Phill, Fenestella antiqua, Goldf., F. prisca, Goldf., #. plebera, M‘Coy, Glauconome bipinnata, Phill., and Polypora lava. The known Devonian Bryozoan fauna (Kuro- pean, American, and British) consists of 26 genera and 115 species ; of these we have only 7 genera and 11 species. These species range through North and South Devon and South Cornwall; they are rarely well preserved and always difficult to determine. Critical analyses of the species have been undertaken by Messrs. Shrubsole and Vines, who in time will revise the entire group. Bracuropopa.—With the exception of the fishes of the Old Red Sandstone (125 species) this is the largest group in the British Devonian rocks. We should expect this when we know that no less than 61 genera and over 1100 foreign species have passed through the hands of European, American, and British zoologists and palzontologists, and all have been described ; of these 1100 species only 116 are British; and of the 61 known genera we possess 26. Calceola (1), Davidsonia (3), Cyrima (A), Rensseleria (11), Camaro- phoria (1), Stringocephalus (1), and Uncites (1) are the genera new to Britain, none of which made their appearance in our area until the Middie periodof the Devonian deposits, the most marked and prolific of the three horizons. With the exception of Cyrtina, which exhibits _4 species, each of these genera is represented only by 1 species. Hall’s genus Rensseleria yields 10 as the total value of the genus ; but only 1 species occurs in Britain (Rensseleria stringiceps, Rom.) ; 8 of the rest are American, and 1 species is Coblentzian. JI call attention to these 7 genera because they are essentially Devonian ; the remaining 19 appeared in the Silurian rocks, and lived on through the Carboniferous. 02 184 PROCEEDINGS OF THE GEOLOGICAL SOCINTY. The Lower Devonian yields 9 genera and 21 species—Athyris 1, Atrypa 3, Chonetes 3, Leptena 1, Or this 3, Rhynchonella 3, Spirifera 3, Spur iferi ina 1, Sér eptorh ynchus 3. The Middle division contains 23 genera out of the 26, and 80 of the 116 species. The Upper contains 14 genera and 37 species; of these 10 genera and 16 species pass to the Carboniferous. It is essential to the history of the Devonian fauna that I name the genera—Athyris 2 species, Cho- neétes 1, Discina 1, Lingula 1, Productus 1, Rhynchonella 3, Spirifera 3, Streptorhynchus 1, Strophomena 1, Terebratula 2. The large genera are Lthynchonella 16 species, Spirifera 20, Streptorhynchus 6, Orthis 6, Cyrtina 4, Productus 4, and Terebratula 4. Regarding the great discrepancy or smallness of this peculiar fauna as compared with that of the continent and America, we must have regard to the smallness of the area now exposed in England as compared with the original area occupied by the Devonian sea, the accumulations of which are now covered by the Secondary and Tertiary rocks of the west and east of England, the Devonian floor or old surface being hidden east of the Quantocks, North Devon, and Torquay. We have proof of a rich Upper Devonian fauna in the rocks under London and Turnford, at the depth of 1000 feet, and below the Cretaceous series ; between these two places and North Devon we can at pre- sent only surmise the plane they occupy. The volume of the Paleontographical Society’s publications by T. Davidson, Esq., F.R.S., &c., devoted to the British Devonian Brachiopoda, is worthy of the fame of its distinguished author; in it are described 116 species, the arguments for and against their genuineness being impartially and consummately reasoned out, figures of all the species and doubtful forms are given; and when we know the patience and skill required to adjust and discriminate so dismembered and small a group as the British Devonian Brachiopoda out of such a mass of material occurring in the European and Ame- rican fauna (1100 species), we may well be thankful that there are men who have the required leisure and knowledge and who devote their lives to one subject. Associated with Davidson in the Devo- nian fauna, the names of Von Buch, Schlotheim, Schnur, Dalman, Sandberger, Hall, Billings, Dall, Conrad, Romer, Vanuxem, Bar- rande, De Verneuil, Roualt, Phillips, Sowerby, King, and M‘Coy must be prominently noticed. LAMELLIBRANCHIATA.—The mass of the species of the class Conchi- fera occurring in the Upper Devonian, and especially abundant, are Asiphonida, through Avieulopecten, Pterincea, and the Mytilidee. 20 genera and 39 species are known, and 29 species illustrating 11 genera are in the Upper division ; yet of the whole bivalve fauna only 4 genera and 5 species pass to the Carboniferous series in North Devon, viz. Aviculopecten granosus, Sow., A. plicatus, Sow., Pterinea damnonensis, Sow., CuculleaGriffithu, Salt., and Curtonotus unio, Salt. The Lower Devonian is even poorer, only 3 genera and 4 species occurring; these are Aviculopecten polytrichus, Phill., Pierinea anisota, Phill., P. spimosa, Phill., and Otenodonta Kratche, Rom.; the last mentioned is the only bivalve species actually con- ANNIVERSARY ADDRESS OF THE PRESIDENT. 18 5 fined to the Lower Devonian. ‘The Middle Devonian is represented by 13 genera exhibiting 23 species— e & a oS Classes. © S © S53 =) 3 n 2) ° 5 Qj oO H mt foal eS Sen feece me Sees By aae my O D = = =) py 2 2 ffpitlernitea Weeecewereeese-- 12 18 : 3 12 2 WOCOZOR, 2 48 50 80 | 33 40 PD YETSTTLTIES Re ae + its 2 lsc 5 IRRECCEATINUS, cc cteccs cs icc sciessecest ae 4 1 1 LUTTE O) Seinen i OC 2 | nef} Al LEGO. GOk ae eRe Eee eee , 1 ee ees "PATER Rtn aae ene ae 5 5 3 | 6 EGRIGOHOMYS ....,..0.sncereceese 8 2 8 5 J. ETRE et ecm cccoes snes cuecete i x2) aeelitysee 9 PETEPOMICES Sones ec sscsccbcveccotess 3 6 7 be 2 85 74 | 186 67 pa eS Oo The preponderance of species in one genus, as in Aviculopecten, is unexampled in any other British formation. 150 species occur in Britain and Europe; 50 have been described from America, and all differ from our forms. The species occurring in the Belgian Car- boniferous rocks are known to us through the researches of De Koninck. In Ireland, mainly through the researches of M‘Coy and Baily, the Integropallialia appear to have been exhaustively recorded. The genus Avicula, in the Irish Carboniferous rocks, numbers about 23 species, Aviculopecten 80, Pecten 5, Pteronites 5, Posidonomya 8, Pierinea 3, Pinna 5. Scotland: Avicula 10, Avicu- lopecten 50, Pinna 4, Posidonomya 2, Pteronites 6. England: Av- cula 16, Aviculopecten 48, Pinna 5, Posidonomya 8, Pieronites 3, and Pterinea 1. These comparisons are important, as showing geogra- phical distribution as well as age. The collective fauna illustrating this group in all Kurope beyond Belgium does not exceed 50 species, showing either want of research or the small development of the Carboniferous rocks east of Rhenish Prussia. LAMELLIBRANCHIATA.— Dimyaria*.—No less than 43 genera and 245 described species occur in the Carboniferous rocks, the largest number being in the Carboniferous Limestone (80 genera and 182 species); the underlying Lower Limestone and Shales yield 23 genera and 103 species; and the Calciferous beds of the north of Kngland and Scotland 17 genera and 38 species; only 9 species are Yoredale. Many species are necessarily common to the three lower horizons ; but the faunal contents are as stated. The Coal-measure species will be noticed in their place. Thus the united British Lamellibranchiate (or Pelecypod) fauna numerically reaches 54 genera and 424 species, 179 being Mono- myarian Asiphonida (Integropallialia) and 245 Siphonida (Sinu- pallialia). Hitherto I have not mentioned the Coal-measure Dimy- * 18 families, including the Arcidz, Trigonid, and Unionida. 218 PROCEEDINGS OF THE GEOLOGICAL SOOIETY. aria, or those occurring in the Millstone Grit and Lower, Middle, and Upper Coal-measures. With the exception of the Lower Coal- measures, which have yielded 12 genera and 42 species, the three remaining horizons are comparatively poor, as would be anticipated from the estuarine condition or nature of the deposits. Anthracosia, Anthracoptera, Aximus, Ctenodonta, Edmondia, Modiola, Myacites, Myalina, and Schizodus are the genera that characterize the Coal- measures, or beds above the Yoredale series. 6 genera and 9 species are all that are known from the Millstone Grit; the genera are Anthracosia 2 species, Aanus 1, Ctenodonta 3, Edmondia 1, Lunulacardium 1, Myacites 1. The genera in the Lower Coal-measures, 14 with 44 species, are still more estuarine in habit, many allied to the Unionide and Myadz—Anthracomya 5 species, Anthracosia 6, Anthracoptera 2, Axinus 2, Conocardium 1, Ctenodonta 6, Edmondia 3, Leptodomus 1, Modiola 4, Myacites 4, Myalina 6, Pleurophorus 1, Pullastra 1, Schizodus 2. The Middle Coal-measures have, as yet, only yielded 4 genera and 16 species—Anthracomya 6 species, Anthracosia 6, Anthracoptera 2, and Myalhna 2 species. The Upper Coal-measures contain the same 4 genera with 11 species. The above, with the table of distribution, clearly shows the changes from the deeper-sea fauna of the Limestone series to the shallow and estuarine accumu- lations of the Coal-measures, the gradual dying-out of those genera, essentially dwellers in clear and deep water and with associated sedimentary matter, as well as the almost total extinction of the molluscan fauna with the elevation of the sea-bed, a few genera only living on into the Permian sea. The accompanying Table gives the numerical value of the species of 18 of the chief Dimyarian genera out of 43 for England, Scot- land, and Ireland; also Belgium, for comparison, which is the only country in Europe where the Carboniferous rocks (Lower series) are well developed. The researches of Prof. De Koninck, of Liege, have afforded me data for this column. I have before stated the small number of known species in the European area beyond Belgium ; to show, however, the relationship between the American fauna and our own, I have appended a column also, which shows excess in the number of species in most of the genera. Such holds good with the older Paleozoic genera also; but, through the flora and Mollusca of the Coal-measures, America and Scotland are intimately associated, and, in a similar but less degree, Ireland, ANNIVERSARY ADDRESS OF THE PRESIDENT. 219 Table of 18 of the chief Carboniferous Genera, showing their specific or numerical value through the five areas or countries named. | | [mel oe dies : 5 Oo | ElElglelgig. Genera. eh ot So lcen ine lee q s 9 | 'o ee Bla ae |Alsai6 Allorisma (Myacites) ...... 6} 3] 4 28 | 41 ATC APACOINYA.......00ss-000- Salas 2 | 13 |Chiefly Coal-measures (6 species). RIUBEAEORIG) <.650.5 00.0 20200- 10; 9 23 | 8 | 50 | Ditto (6 species). Lacs, 2 ee AT SOn StOn ke ht 44 LAShit =e Rel ae 100 Sta Et Cardiomorpha ............... 9} 6| 5]} 86} 18 | 69 Conocardium .........:..... Gullean Ol (Ss) 39 EGtemodouta, <.:.......<...-.-. 11 | 15 | 15 | 9 | 34 | 84 | 6 Coal-measures. Cypricredia. ....-............. eo OM 1Gy TL os JSC ES Nit 12 eee 10 | 10 | 12} 2 | 25 | 59 | 8 Coal-measures. LG) A 10; 8} 12]... | 5 | 35 |4 Coal-measures. BURY EMBs soa. -cas-..eesees 3; 2] 3] 380] 6} 44 WAM eee. 32-2.--.02.-.00- 6; 9; 1| 3] 18 | 37 |6 Coal-measures. Sanguinolites................ LOv) 20) | 1645327 |-68 SCLNIZ00 si 3| 8] 7]... | 13 | 31 | 2 Coal-measures. BCAGWHEMED -20.2.2000s00.000- ees | Bani ak 6 | 15 i Timrce. ——————— Were era omni orl oul t4 SL. TUT 3A Ease eens LST PRG at ha bern | 106 /102 115 [171 [222 716 Sorznoconce1a.—6 species of Dentaliwm are known ; but none pass above the Carboniferous Limestone. D. priscwm is the only species that occurs in 3 horizons—the Calciferous Sandstones, Lower Lime- stones, and Carboniferous Limestone. JD. scoticum is only found in the Calciferous Sandstones; the remaining 4 belong to the Carboni- ferous Limestone, and do not range higher. GaAsTEROPODA.—223 species and 29 genera, all belonging to the division Holostomata of the order Prosobranchiata, constitute the univalve fauna of the Carboniferous rocks. All the genera (29) and 202 species occur in and range through the Carboniferous Limestone ; or nearly every species occurs in this horizon. 16 of the 29 genera are distributed through the Carboniferous rocks of Europe and America, and are therefore of zoological as well as stratigraphical value. I give the following Table of these genera to accompany that showing the distribution of the Lamellibranchiata (Dimyaria) ; by comparison with the molluscan fauna of Europe and America we may hope to obtain some clue relative to the migration and dis- persion of the Mollusca from some original area. These 16 chief genera are represented in England by 176 species, in Scotland by 90, in Ireland by 111, in Belgium (chiefly through De Koninck) by 176, and in the American Carboniferous rocks by nearly 200 species. This generic relationship with America is important as determining similarity of bathymetrical conditions, temperature, and food, and con- nexion through coast-line or land now lost. 9 genera and 21 species nt 220 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. occur in the Calciferous Sandstone—Huomphalus 4 species, Lacuna 1, Inttorina 2, Loxonema 2, Macrocheilus 4, Murchisonia 3, Natica 1, Naticopsis 1, Pleurotomaria 3. The Lower Limestone Shales contain 50 species, which, with one exception (Turbo appropinquans), are also those of the Carboniferous Limestone above; or 49 of the 50 occur also in the massive lime- stones. The Yoredale beds (the Upper Limestone Shales of some authors) are poor in Gasteropoda; only 7 genera and 10 species appear to be known. Clearly this must be the result of imperfect collecting rather than of almost total absence of species. We know that both the Lower and Upper Limestone Shales are impure argillaceous limestones, and not favourable to the presence of the Mollusca at the time of deposition or to the preservation of organic structure afterwards: but we see here the 7 genera only represented by 10 species; they are :—Huomphalus catillus, Sow.; Loxonema constricta, Sow.; Macrocheilus curvilinea, Phill.; Jf. imbricatus, Sow.; M. rectilinea, Phill.; Murchisonia fusiformis, Phill.; Naticopsis plicistria, Phill.; Plewrotomaria lim- bata, Phill.; P. twmida, Phill.; and Yurritella tenwistria, Phill. The Millstone Grit, as we should expect, contains but a small Gasteropod fauna; such arenaceous deposits suit not the habits of this group of Mollusca; only 3 genera and 3 species are known— Natica variata, Phill.; Pleurotomaria limbata, Phill.; and Murchisonia fusiformis, Phill. The Pelagic Cephalopoda in the same beds, as we © shall see, number no less than 33 species; but the sandy and muddy beds of the shore was not their habitat; they are not such good wit- nesses in the beds in which they are found as the ordinary Lamelli- branchiata and Gasteropoda. The Lower Coal-measures alone yield Gasteropoda; none are known either in the Middle or Upper. 7 genera and 19 species are recorded from this division :—Huomphalus Glover, Brown; Juttorina obscura, Sow.; L. solida, De Kon.; Loxonema galvam, Baily; L. minutissema, Baily; L. Owen, Brown; L. reticu- lata, Brown; Macrocheilus, 4 species ; Natica vetusta, Sow.; Nats- copsis plicistria, Phill.; Plewrotomaria limbata, Phill.; P. usocona, Sow.; Zurritella, 4 species. Middle Coal-measures none. Upper Coal-measures none. ANNIVERSARY ADDRESS OF THE PRESIDENT. 221 Geographical Distribution of the Gasteropoda through 16 characteristic Genera. | | | ne] ne) ra osteo | Ae es ieeeceices | SPE. Nishi 1 oes oc Ree ae | = <3) 2) Ss ea bales 6 “ESET TOL a eae ins ee 6 5 8 13 22, 54 CClifiae, | go ree 3 3 4 15 2, 27 BU OMIDMALUES Jt enscnvesseeeses 20 Wi 20 20 23 | 100 LORS ae a a 2 1 os 2 yes 5 WMERCROCMOIUG! { 1. ..0--cc0--00 lesvses 17 8 12 5 19 61 EOWA A oo ea. = cvnon noses sens 12 Na os 5 2 19 RECIEVE)... soc sorcccceccsseees 15 O 6 19 20 67 MUSE TURN coco es ceacccwswecnose 13 9 9 14 16 61 PE AVICOSIS mod. ais o's ox a'view oneiecnt 2 1 3 a mE 6 [ESE 12) ° 5 Seen eee 7 ae 5 6 ay 18 IPMAMICTORMUS, -coceccsecccecccsecsee 4 ee 3 “ss 2 9 IIABYRCHISWID, 2... 05-2 cecc econo sae 9) 2 4 1 + 16 IEIOMEGLOMIATIG «.-5ccscccneeecnseces 46 24 16 59 65 | 210 MIME TOUS 2s oo coc la occecccccevscaecs 6 1 4 aan 2 13 TNS 19)” GA ae i 6 1 4 fy af, 16 | SOS TEL GSIEE: Gea Seo ere 12 i] 13 12 20 68 176 90 | 111 WG. tb 197-2 F750 Preropopa.—The genus Conularia first appears in the Carboni- ferous rocks in the Lower Limestone Shale, and ranges through all but the Upper Coal-measures. This genus dies out in the Coal- measures. Many of the nodules in the “ Penny-stone” of Coal- brookdale contain fine examples of C. quadrisulcata, Sow., the only determined species known. An undetermined species occurs in the Calciferous Sandstones of Woodhall, Scotland. America yields 17 species of Conularia to our 2; our C. quadrisulcata is not known out of Britain. Herrrovopa.—Bellerophon and Porcellia (the former with 27 species and the latter 4) exhaust the species in the Carboni- ferous rocks. At no period in Paleozoic times did so many species of Bellerophon exist in one horizon. 24 of the 27 species are found in the Carboniferous Limestone; and 6 of these 24 species range higher; they are :—JBellerophon apertus, Sow.; B. decussatus, Flem.; B. Dumonti, D’Orb.; B. hiulcus, Mart.; B. Oldham, Portl. ; and B. Uru, Flem. 4 species (B. costatus, Sow., B. decussatus, Flem., and var. undatus, Kther., and B. Urii, Flem.) are Calciferous or Tuedian species. B. decussatus, B. apertus, B. hiulcus, and B. Urii range up to the Middle Coal-measures. 5 species occur in the Lower Limestone Shale; but none are peculiar to those beds. 17 Species are essentially Carboniferous Limestone. The Yoredale beds have yielded 4 species—B. apertus, B. decussatus, B. hiulcus, and B. Urn, all species having a long range. The Millstone 222 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Grit contains the same forms. The Lower Coal-measures yield the same, with the addition of B. navicula, B. Oldhamu, and B. Dumontic; none occur in the Upper Coal-measures—this order, like all others in the Carboniferous series, dying out at its close, not to appear again. The genera Bucania of Hall and Huphemus of M‘Coy are synony- mous with Bellerophon. Ireland has yielded 21 species of Bellero- phon, Scotland 15, England 17, Belgium 25, and America 39 species ; only 19 species range through Europe, 8 of them being British; and 14 of these 19 are Russian. CrpHatopopa.—No fewer than 169 species are distributed through the British Carboniferous system. 4 genera and 6 species occur in the Calciferous or Tuedian beds, 5 genera and 22 species in the Lower Limestone Shale, 8 genera and 140 species in the Carboniferous Lime- stone ; 5 genera and 33 species are Yoredale ; 3 genera and 30 species are Millstone Grit; 3 genera and 24 species occur in the Lower Coal- measures, 3 genera and 6 species in the Middle Coal-measures, and 1 genus and 2 species in the Upper Coal-measures. These 169 species represent 8 genera—Actinoceras 1 species, Cyrtoceras 3, Discites 17 (subgenus of Nautilus), Goniatites 59, Nautilus 36, Orthoceras 48, Poterroceras 3, and Tirigonoceras 2 species. The accompanying Table of the 8 British genera shows the number of species in each country, including, as in the two prior Tables, Belgium and America for comparison, in which it will be seen that the 169 species, through the 500 occurrences, are thus distributed :— the number of species in the 8 genera in England is 126, in Scot- land 63, in Ireland 107, and in Belgium 90. Comparison carried beyond Europe to America, as before, shows generic affinity through 114 species, hardly a single form being British. All Europe, only yields 95 species through 11 areas; research and the accident of our possessing these rocks highly developed is the reason why our mol- luscan fauna is so extensive. 8 é : : q rd oo ro 3 BP lik Os lly ee ca ees a a) 3) 5 a0 ee WA CEIMOCELAS Ye face sean oe ekeseeenee 1 1 1 1 2 6 Oyrtoceras.:, hiss sc. seeasenlesdwcess 2 3 aie 8 5 18 DIS Cites, Sees. f ee Ne asset eek 10 3 LS idee 1 32 Goniatites 2047 tcacnaneccteeties 59 15 32 25 32 | 163 Nautilus: tin tscccscsecec neuen ea ea: 18 14 21 22 54. 129 Orthoceras /eeerk eee cee Reena eee 33 24 36 29 20 142 IPoteriocerasin. econ ee eee 2 3 2 au Fad a Tri gOnOGeLrasnnc:t pa scmeedennesee 1 a 1 1 on 3 126 63 | 107 90 | 114 | 500 ANNIVERSARY ADDRESS OF THE PRESIDENT. 223 VERTEBRATA. Piscrs.—120 genera and 290 species of fish and 26 genera and 33 species of Amphibia compose numerically the Vertebrate fauna of the Carboniferous rocks. The comparatively rich assemblage of ichthyic remains in the Old Red Sandstone bears no proportion to the great increase and development of this class in the Carboniferous system. Only two orders are represented in the Carboniferous rocks, viz. the Ganoidei and Elasmobranchi (Placoidez, Ag.), the Ganoidei through the suborder Crossopterygide by 4 of the 6 recog- nized families, the Saurodipterini, Glyptodipterini, Ctenodipterini ? (Dipnoi)*, and Celacanthini. The Elasmobranchi mainly corre- spond to the Chondropterygide or the Cartilaginous Fishes of Cuvier (the Holocephali and Plagiostomi of Owen and the Selachia of Miller). The distribution and appearance in time of the two orders appears to have been much the same; the Plagiostomi first appeared in the Ludlow rocks, the Holocephali through the Chimeroids in the Devonian rocks of N. America (genus fhinodus of Newberry). The section Cestraphori of the order Plagiostomi 1s abundantly represented in the Carboniferous rocks from base to summit, spines and teeth of the several genera occurring everywhere, although very locally, the Cienacantht, Gyracanth, Homacantht, Oracanthi, Onchi, and Lepta- canthi being the most important, and occurring as spines (Ichthyo- dorulites). The genera or groups illustrated by palatal and other teeth are Cochliodus, Deltodus, Psammodus, Petalodus, Otodus, Ctenoptychius, Cladodus, Glossodus, Diplodus, Helodus, &e. Many of these generic groups will be greatly reduced under stricter and more complete research. My duty here is not to criticise, but rather to bring together for a special purpose the labours of others as at present received. The lowest recognized rocks of the Carboniferous system, the “‘Calciferous Sandstone” of the Scotch geologists, or Tuedian of Northumberland, confined to the north of England and south-east- ern Scotland, contains 13 genera and 14 species—Cladodus 1 species, Cienacanthus 1, Ctenodus 1, Ctenoptychius 1, Diplodus 1, Hurynotus 1, Megalichthys 1, Nematoptychus 1, Pecilodus 1, Pygopterus 1, Rhadi- nichthys 1, Rhizodus 2,and Wardichthys 1. Thus, with one exception, each genus is represented only by 1 species. Surely much has to be done both with the Ganoidei (8) and Placoidei (5) when better materials come to hand ; for many of the genera must be established upon slender grounds. THe Lower Limestonzs anD SHatEs have yielded 29 genera and 50 species ; and if this lower member be associated with the thick or * T am aware that Dr. Traquair would remove the Glyptodipterini from the Crossopterygious Ganoids and place them sectionally in the order Dipnoi (vide Trans. Roy. Soc. Edinb. vol. xxvii. for the sections and families proposed by Dr. Traquair). 224 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. true Carboniferous Limestone, the two together would represent 80 genera and 170 species, 17 genera and 22 species being peculiar to the Lower Limestone Shales. No one can doubt that these dark, impure, argillaceous limestones were the commencement of the purer calcareous series above; and paleontologically they cannot be sepa- rated. Closer research will probably show that the 17 genera now known to occur only in the Lower Limestones exist in the succeed- ing series, Carponirerous Limestonz.—This finely developed group of rocks in the British Islands, which has received so much attention both under geological and paleontological aspects, is possibly better understood than any other division of the Carboniferous system. The limestone beds known as the Mountain Limestone have yielded no less than 63 genera and 147 species of fishes; but the Yoredale series which succeed them yield none ; and only 3 genera and 3 species are known in the Millstone Grit (Megalichthys Hibberti, Rhizodus Hrbberti, and Acanthodes Bronni). This sudden cessation or non- occurrence can scarcely be accounted for on lithological grounds, or original sedimentary accumulation, such as would be applicable to the habits of certain genera of the Mollusca. Bathymetrical con- ditions may have influenced this paucity in the whole fauna of the Yoredale and Millstone Grit, which is so apparent on inspection of the Table of Distribution. 38 genera of Placoidei are confined to the Carboniferous Limestone; and 12 genera of the same order are from both Lower Limestone Shale and Carboniferous Limestone: on the other hand, it is remarkable that only 6 genera of Ganoidei are strictly or essentially confined to that horizon ; they are Acrolepis, Asterolepis, Coccosteus ?, Cycloptychius, Phyllolepis, and Platycanthus. (I give them as near the truth as possible, as the species of Ganoidei in the Carboniferous system are in as much confusion as the Placoidei*.) As before stated, the Yoredale rocks(Upper Limestone Shale of some areas) have not yielded remains of fishes. We are in doubt as to the real position of the Upper Limestones of Durham and Northum- berland, or what, if any, should be assigned to the Yoredale group. Nevertheless [ am not aware of any Vertebrata occurring in the Yoredale beds. The Millstone Grit, as before stated, has only (so far as I know) yielded 3 genera and 3 species, and these in the form of scales. | Lowrr CoaL-MEAsuRES.— Whatever conditions caused so complete a break in the distribution of the fishes in our area zoologically, stratigraphically, and geographically during the deposition of the rocks mentioned, the return of the same genera and many of the same species in the Lower Coal-measures is equally important. We have evidence of the remains of 52 genera and 112 species, the whole composed nearly equally of the orders Placoidei and Ga- noidei. The numbers may be thus expressed :—In the Lower Coal- * The Ganoid Fishes are under revision by Dr. Traquair, whose accurate knowledge of the order will be brought to bear upon their structure and classi- fication. ANNIVERSARY ADDRESS OF THE PRESIDENT, 225 measures 32 genera and 50 species of Ganoidei occur, in the Middle Coal-measures 5 genera and 6 species, in the Upper Coal-measures 6 genera and 8 species. The Placoidei are represented in the Lower by 22 genera and 40 species, in the Middle by 10 genera and 14 species, and in the Upper by 6 genera and 8 species. Regarding the Coal-measure fishes as a whole, however, they stand in the Table as Lower Coal-measures 52 genera and 112 species, Middle 14 genera and 20 species, and the Upper 12 genera and 14 species. The numerical distribution for Britain and Belgium is— [ST EG oe 136 species. seaulands | oss Socey 5... P52) es Mrclanieey ps 2 ee ad 21) miner ibelstunyeS a eee ou S. ist) Wagers ‘The Carboniferous system is finely developed in Belgium. 37 of the same genera occur in the American rocks, 10 being Ganoidei and 27 Placoidei, the two orders there yielding 240 species. Amputpra.—Of the 4 orders in the class Amphibia, the Labyrin- thodontia alone occur in the Carboniferous rocks. 26 genera and 33 species are known in Britain. Without exception they all occur in one or the other of the divisions of the Coal-measures. Prof. De Koninck, in his great work (‘ Ann. du Mus. Roy. d’Hist. > Nat. de Belg.’ tome ii., Faune du Calc. Carb. de la Belg. pt. i.), has described 29 genera and 44 species, all from the Carboniferous Limestone. The fauna of the Coal-measures in Belgium is small compared with that of Britain. Tue Lowsr-CosL-MEASURE genera are Ainphicelosaurus, Amphi- saurus, Anthracerpeton, Brachyscelis, Batrachiderpeton, Discospon- dylus, Dolichosoma, Erpetocephalus, Ichthyerpeton, Keraterpeton, Lepierpeton, Loxomma, Ophiderpeton, Labyrinthodontosaurus, Lepto- gnathosaurus, Macrosaurus, Megalocephalus, Orthosaurus, Parabatra- chus, Pholidogaster, Streptodontosaurus, and Urocordylus. Tar Mippre-Coat-MpasvRE genera are only 3—Anthracosaurus, Megalerpeton, and . holiderpeton. Loxomma and Pieroplax occur in both Lower and Upper Coal- measures. We are chiefly indebted to Professors Huxley and Miall for elaborate researches into the structure and affinities of the extinct order Labyrinthodontia, a group eminently characteristic of Carbo- niferous time; for, with the exception of RAinosaurus from the Lias and the Jurassic Brachyops, no remains of this order have hitherto been discovered in rocks younger than the Trias. 226 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. TaBLE XVI.— Carboniferous. ——— *UBIMLIOg | 07} 03 ssed an ne oe yey} soroodg | *soansvoul-[v0D | .@ © soddq ae aN =e OO AR v0 NS av on "SQITISVOTU-[VOD) | ~@ a o® AA TO art Oo Av SF ae sa MO eTPPIAL a mc =e *SOUNSVOT-[VOD) | ga Glo an JOMOTT 2% rie SE OA COD Ore may coke Se re nin am mg on an ae e e . e e pr) “ITH OUOISTTNAL | aa SS og yeaa HUES Re Pe 22 me an nt a, ae 3 | oe a 3 Q . ° a “BpIq, 8TBpat0 X ° rn? XT} ae oe : : : os 2 On n> aeoav “eo : : 2 “QUO SOUT Oa 6 0o ot Ho ot @ : ° > ov 8 At of os te ” ae snoIE;IUOGIeO nn a ome SN ot ae rie ae 2 S 7 Nr nr ne ae ae 8 ret) (GU OS asi ae ‘OW SOTBYN 9U0JS | ae me CF pq MO AA GORD : ° 2 90 HO Ut ae wa @2 ae mo ROCnE a -OUI'T -ToMOorT NN aN Ca Cs cb : 2 rN nD “NO a Dale) ; mM e¢ | le “eudIspUv ; ‘ ; : 3 5 ° iS enous ee Cy ik} © 08 miet Ar e e A . ov a2 ae =e =a @ ey 305 ay : a m = c : A e SoS 1m @ me od ben) set @ om 1a spdg BS ARPPSeagrVPRRERSS oO PD OD 1O aR) ma re [ey | mam AN N mw a : PN ~~ = © © co iis BIOUOD | SYS) SSS G8) Ge) COGN ca eee “A aR &|3 e eoreceee eoveees ° ° ° ° ° ° ooreroeeseon . . . ° . . eee oseos ° ° Se eeeeoestoseenses eoeeesegnten ry @eceececoereoree ate inci : ; 3 Bhp o pacar: ear ee Sua : aS Ct Sit iA BS, . CMe mee Orin a) a) o. es 8 (8 he : Mews ve BF ane : : Ba fi: 2 B street em na: > bi oioi hot @ 2.) 2 wa > “— 2: aM io Sia es Oe hea a 1S of. 'y ee : m2 od Geach 2 of ob od 8 te 8 gS Be BS a Sai: 2 8 PP oe RP Be ee = . Sh MOT tire Sole Cee oni re 2 Oats ort wm 8 PM POP Poi b bob Pop) eR 2 Soweto Boab sb but ues See se : SENN = yeni: MME aS ec) ae Ji obo db tpi @ b) e080 Se ps rs 696.89 8,8 S$ °' gg 2 8&8 6 22> ee \3 8 SS 3S S386 5 8 Ss 8 0 Zor ee ~~ (2) og Se 2 aC 2 8. 2) © igi fe ommomes is : om SC) ©) o ey ¢ 8 BHSBsS 2822 BS 8° 822 eee — 4 oH . Ja & mae e5aRea48 8 nommhonm < ne ee ete e Me “UBIUOARCT WOd AN e's m0 et wo ee tu ow HN Me aa ANNIVERSARY ADDRESS OF THE PRESIDENT. 227 Tre EXTENSION oF THE OLDER OR Patmozorc Rocks BELOW THE NEWER oR SECONDARY FORMATIONS. The position and extension of the Paleozoic rocks beneath the newer formations of the British Islands is a problem of deep interest, and is now occupying much attention in consequence of the facts brought to light by the numerous trials lately made either for the supply of water or in search of minerals. It may be said that ever since the remarkable trials for water at Harwich and Kentish Town, and also the still more remarkable generalizations (almost predic- tions) of R. Godwin-Austen, Esq., in 1856*, and Prof. Prestwich in 18727, with reference to the possible extension of the Coal-measures beneath the south-eastern part of England, the minds of pure geolo- gists have been excited by speculative views, and desires to arrive at some knowledge of the extension or distribution of old land or Paleozoic surfaces, so as to restore to the eastward in England the physical geography of those groups of rocks which now constitute so grand a feature along the western side of England, Wales, and Scot- land, but which are lost or covered up beneath the unconformable newer or Secondary rocks. West of long. 1° 30’ the greater part of the exposed rocks are Palxozoic, ranging from the Cambrian to the Coal-measures, their general strike being about N.E. and 8.W. East of this meridian are Secondary and Tertiary rocks of great thickness, which doubtless cover the easterly extension of the Pale- ozoic series towards the European continent. The geographical changes of land and sea must have been numerous from the time of the consolidation of the Cambrian sea-bed to the close of the Carboniferous epoch, the rocks of the latter period being deposited in depressions and valleys of the older, with succession or position due to the removal of the subjacent rocks. With the old and wide extension of these earliest-formed masses we are becoming daily more familiar. Looking at the physical structure of the south-western and north-western parts of the British Islands, and the great mass of the older Paleozoic rocks of North and South Wales, it is evident that from the Cheviots to Cornwall the oldest rocks in Kurope are exposed, their eastern extension being hidden. The Northumber- land and Yorkshire coal-fields down to the latitude of Notting- ham are covered and deeply buried by the Triassic, Jurassic, and Cretaceous rocks. South of Nottingham these old land areas are again exposed; the Charnwood rocks of unknown age, the associated coal-field of Ashby-de-la-Zouch, with the Warwickshire and South- Staffordshire coal-fields stand out like islands in the midst of the great Triassic plain of mid-England; they are the last isolated exposures or remnants of Paleozoic land seen south of the great Penine axis. A line drawn from the Malvern range, due south to the Mendips, and thence to Torquay, will define absolutely the exposed line of demarcation between the Paleozoic and Mesozoic rocks. The mass of North and South Wales stands out in bold relief westward * Quart. Journ. Geol. Soc. vol. xii. pp. 838-4 +t Popular Science Review, vol. xi. p. 241 (1872). 228 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. of the Severn valley. The Old Red mountains and older Silurian rocks which border the northern edge of the South- Welsh coal-field, as well as the Devonian promontory of Cornwall and the mass of North Devon, isolated as they appear to be from the unconformity of the Secondary rocks, are only apparently so through the great over- lap. Could we uncover and expose the old Paleozoic floors or land- surfaces with all their irregularities, doubtless we should find that the eastern face of the Paleozoic plain would stretch away under the north-eastern and south-eastern counties and the German Ocean, the newer rocks filling up the irregularities in the old land-surface—this denuded plain being either produced by the eroding agency of the Secondary seas during the slow depression of the area they then occupied, or previously sculptured and fashioned into hills and valleys prior to the deposition of the Secondary or Mesozoic rocks. ‘The Irish Sea, the English Channel, and the German Ocean are only hollows in the land occupied by the several seas around the British coasts, any important change in which would alter our relative position to the continent either to the north, the east, or the south, and in case of upward movements, would reveal those accumulations which have gone on since the close of the Glacial epoch. Could the valleys of the North Sea or St. George’s Channel, the English Channel, and the German Ocean become again dry land, we should again be restored to and form part of the great European plain or plateau, and those stratigraphical masses that are now abruptly cut off at the coast all round the British Islands would be traced in broken continuity over their once continuous or origi- nally connected area. Further, could we strip off all the Secondary and Tertiary rocks, and reveal or expose the extension of the older or Paleozoic series towards Germany on the east, and France on the south, then the vexed question of the old physical geology and geography (paleography) of Britain and the relation and correlation of our area with that of Europe would be revealed; the once continuous terrestrial surface joining us to Europe, and probably America, on which grew and flourished the flora which furnished the materials of our coal, could be determined; the probable relation of the underlying or partly contemporaneous Devonian to both the Silurian and Carboniferous; the reason for the isolation of the Old Red Sandstone in different geographical areas, marine in one area, freshwater in another—the one with a well-defined base and top, the other having as yet no discovered base, but having a well-defined passage into the Carboniferous; such and a hundred other problems would be solved could this old floor be ours to examine. A rise of a thousand feet would reveal much of all the hidden older land east of long. 2° W., or all eastward of that which extends from the southern termination of the Penine chain and Charnwood Forest ; for we now know that as far south as Northampton, and at less than 1000 feet*in depth, the Carboniferous Limestone occurs. Still further south, and but little deeper (1184 feet), the Coal- measures have been proved at Burford, in each case yielding the characteristic fossils, the limestone at Northampton being crowded ANNIVERSARY ADDRESS OF THE PRESIDENT. 229 with 2 species of corals, Lithostrotion irregulare and Lonsdaleia flo- riformis ; the Coal-measures at Burford with Oyclopteris orbicularis, Neuropteris, and Pecopteris. By degrees we are arriving at sufficient data to enable us to judge somewhat of the physiography of these older accumulations or formations and arriving at their distribution. The sources of the oldest sedimentary strata will probably ever remain a mystery; the materials that supplied these oldest British strata have wholly disappeared ; the extent, distribution, and dimen- sions of the Paleozoic series all afford indications and proofs of the vast regions in the north and west which have been thus denuded or washed away. This brings me to the consideration of the proof of the extension of the Upper Silurian and Devonian rocks eastward of the Malvern chain, the Staffordshire beds (Dudley), and the Bristol Coal-basin, obtained at no less than five places, if not six; the seventh was unfortunately never completed. The oldest rocks yet touched are the Wenlock; they were deter- mined at Ware, in Hertfordshire, at 795 feet, immediately beneath the Gault, during the process of boring for an extra supply of water for the New River Company. These beds dip at an angle of 40° to the S.E.; consequently their strike is from the N.E. to the S.W.; and probably the Devonians rest upon them in succession, as they are known to occur at Turnford, 7 or 8 miles to the south of Ware, and continuously on to London. No rocks, however, except 1 foot of Lower Greensand (‘“ Carr stone”), came in between the Silurian and the Gault—the Devonian, Carboniferous, and all the Lower Secondary rocks being unrepresented. This feature in the paleogeography of the eastern region of England, long ago anticipated in some form by Austen, Prestwich, and Hull, has now been verified, and the age of the rocks determined. The line occupied by the Wenlock rocks may be higher or nearer to the north than was hypothetically believed by the authors above mentioned, but not more so than the course or strike of the Silurians and Devonians would probably take, having regard to the position of the Malvern, Woolhope, May- Hill, and Tortworth Silurians to the west, although the Ware beds can hardly be referred to the type of Silurian rocks that exists at the places above named. It will therefore be asked whether this Wenlock at Ware is of British or continental type; in other words, ean it be correlated with our Welsh or English Upper Silurian, or is it of the Ardennes type? Do they constitute a portion of the Staf- fordshire (Dudley) or Shropshire (Wenlock) Silurians, spreading away eastwards towards Belgium? or are they a prolongation of the Silu- rians of Belgium to the west, or a western extension from the con- tinent? In other words, does this Wenlock at Ware belong to the edge of another basin or coast-line, an extension from Western Europe or Scandinavia, or an easterly expansion of the Upper Silurian of the Silurian area? The facies of the fossils and the characters of the rock in all essentials are decidedly British; yet there is much resemblance to the Scandinavian fauna, a prolongation of the rocks containing which is by no means impossible or impro- VOL. XXXVII. if 230 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. bable. Onthe whole, however, I am inclined to regard the Wenlocks under Ware as part of the continuous and old denuded Upper Silurian surface of those rocks now conspicuously exposed in Herefordshire, Shropshire, and Worcestershire. In direct E. and W. line, and latitu- dinally, the Wenlock rocks of Malvern, Woolhope, and May Hill may (under the view of continuity) claim connexion; but I regard the Ware fossils as having more affinity with the Wenlock-Edge group, or that series which underlies the western side of the great mass of the Old Red Sandstone. The fossils also equally resemble those of the Wren’s Nest at Dudley, where the dull earthy limestones as well as the crys- talline ones (as at Ware) occur. The 33 species of fossils noticed in the cores at Ware are species for species identical with those of the Wenlock Edge or the Wren’s Nest (Dudley). Could we remove the overlying Mesozoic series between Ware and Burford, and again expose the Coal-measures known to occur there at the depth of 1180 feet, then should we better understand the thinning-out of the Triassic and Jurassic series eastwards towards this Silurian ridge or plateau. We can hardly .now doubt the extension of the Burford Coal-measures, in all ‘probability terminating against the Silurian or Devonian series in the: Ware region, at Turnford, and near London. To what distance rocks of these or older date may occur north of Ware further research alone can:decide. We have, therefore, under the so-called London basin an axis of Palzeozoic rocks, two divisions of which are known, the Upper Silurian and Devonian. The Lud- low rocks may be expected to occur under or a little south of Hertford ; for, as we shall see, the Devonians set in between Ware and Turnford, and probably occupy the entire area between that place and London, where it is now well known they occur under Tottenham Court Road. The boring at Ware was carried down 797 feet and into the Wenlock beds to the distance of nearly 50 feet, or, to give particulars :—Gravel 14 feet, ‘Chalk 416 feet, halk-marl 128 feet, Upper Greensand 77 feet, Gault 160 feet, and Lower Greensand (Carr stone) 1 foot; the boring was continued for 50 feet in the Wenlock Limestone, and without the intervention or occurrence of the Devonian. We are thus justified in stating that the old Paleozoic land-surface composed of Devonian and Silurian rocks occupies much of Middlesex and Hertfordshire at the mean depth of 970 feet; they probably extend westward to the exposed Silurian, Devonian, and Carboniferous areas of Cornwall, Devon, and Wales. Harwich, which lies 80 miles to the eastward and 10 miles further north, and probably on the same strike, has revealed Paleozoic rocks at a depth of over 1000 feet ; but the age of these Harwich dark clays or shales was never clearly made out. This question of the depth and geographical extension of these oldest rocks is seldom practically tested ; for no mineral wealth of sufficient value occurs below the Coal-measures (Upper Palseozoic) to induce trial or experiment; and but for the purpose of obtaining water pure and in quantity, it is questionable if this problem would have ANNIVERSARY ADDRESS OF THE PRESIDENT, 231 been solved (at present at least); had the Lower Greensand oc- curred in its normal condition as a water-bearing stratum, or even had the Upper Greensand (above the Gault) yielded water in quan- tity, neither at Messrs. Meux’s, at Turnford, nor at Ware should we have touched the undoubted and unequivocal Devonian and Silurian strata. Few give thought to or are aware of the difference that exists in the thickness of rocks of the same age in different yet not very remote localities. In Britain the Cambrian and Lower Silurian deposits are from 20,000 to 30,000 feet in thickness, whilst in Sweden and Russia their representatives or equivalents in time rarely if ever exceed 1000 feet. This difference is and was probably due to the form and nature of the Pre-Cambrian land on which the newer Cambrians and Silurians were deposited; for there cannot be any doubt that such Pre- Cambrian rocks did and do exist, and were and are widely extended, although concealed over the present known European area, and that their existence was connected with a probable great geographical ex- tension westwards of the British Islands. The plateau governed by the 100-fathom level that surrounds the British Islands is part of this extension, on which all our physical changes have taken place. Probably the crystalline rocks of Scandinavia, parts of North Wales, North-west Ireland, St. David’s, and the Hebrides are exposed areas of this Pre-Cambrian stage of the highest antiquity, and were covered on their submerged and denuded masses by the Longmynd, Harlech, ad St.-David’s rocks, which in their turn were succeeded in some areas by the Lingula-flags, the Tremadoc, and Arenig, life- groups of antiquity so high that we have no formula to express their age, or when life first appeared in the seas of the British Islands. As yet we have no evidence relative to these formations occur- ring eastward of the Penine chain, of which Charnwood, Ashby- de-la-Zouch, and the Warwickshire coal-field are the most southerly exposures ; but we have lately unexpectedly determined the presence of the Carboniferous Limestone below Northampton at the depth of 890 feet*. This has carried still further south the Penaine axis, and would lead us to expect that between Northampton and the exposed Coal-measures of Atherston and Nuneaton an extended coal- field may occur. What relation the Burford coal on the south may have is conjectural only; but I am disposed to regard the whole as one greatly extended coal-tract. We must now regard the Carbo- niferous Limestone of Northampton in longitude 1° 30’ W. as the most easterly known in England; and no Triassic rocks cover this limestone, a few feet only of undeterminable rocks occur between it and the Lower Liast. Whether the New Red hasthinned away entirely here, or it happens to be accidentally absent, is a question; but at Burford on the same strike and relative position there is a consider- able thickness of Triassic rocks, and to the N. and N.W. also every- * Jn the cores brought up I determined the presence of Lithostrotion junceum and Lonsdaleia floriformis in abundance. + The cores brought up are 15 inches in diameter. This great undertaking, as well as those at Turnford and Chatham, were carried on by Messrs. Docwra and Gulland with their new machinery. 232 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. where. The Upper Devonian beds were first determined by myself at Messrs. Meux’s ; they are of the true N. Devon and N. Cornwall Devonian type—dark chocolate-coloured semi-slates containing the characteristic shells Spirifera disjuncta and Rhynchonella cuboides &c. This discovery solved the problem of the existence of Paleeozoic rocks at an accessible depth under London, and of the absence of the Jurassic series. Immediately above the Devonian occurred the Lower Greensand, abnormal in all conditions save the never-failing test of fossil remains, and what few occurred left no doubt as to the age of this at first somewhat doubtful rock. Trigonia aleformis, Cardium Hillanum, Trochocyathus Harveyanus, and Cerithia, &e., added to the general facies, immediately settled the question of age ; neither of the other four borings exhibited any signs of the true Lower Greensand. At Ware and Turnford the Gault rested upon from 8 to 10 inches of theCarr stone, and this was all that represented the Neocomian rocks beyond the 65 feet of chalky oolitic subcal- careous abnormal Lower Greensand; all the borings show that the Neocomian (Lower Greensand) is interrupted on the north by the underground Paleozoic ridge. The geographical extent or surface- area occupied by the Devonian of North Devon between the Pilton beds on the south and the Linton beds on the north (15 miles), where we believe we have the full thickness of the whole series, is as near as possible that known to occur between London and Turn- ford, assuming that the strike of the North-Devon beds is directly towards the London area, and thence on towards Belgium and the Rhine (Eifel and Coblentz). The Devonian cores brought up at both localities (Tottenham Court Road and Turnford) yielded the same characteristic fossils; and the beds dipped at the same angle at both places, 30° $.H.—the plane of the old land-surface being 1148 feet below London, and 980 feet below Turnford, showing a dif- ference in level of 168 feet, due either to dip or denudation. I should believe the difference was due rather to denudation along or over a given plane striking from 8.W. to N.K. Where the junction of the Upper Devonian and Wenlock rocks may take place between Turn- ford and Ware it is difficult to say ; but if we infer that the Ludlow and Lower Devonian beds occur above the Wenlock, dipping south between Ware and Turnford, then there is little room for them, and the Upper Devonian must cease at no great distance north of Turnford. Looking at the thickness of the Lower Devonian in North Devon, and regarding it as maintaining the same in its strike under Wiltshire, Buckinghamshire, Middlesex, and Hertfordshire, there is room in the 8 miles from 8S. to N. at the dip of 30°S. for the presence of the Lower Devonian and Ludlow. We need not assume that the Lud- low beds do cecur, any more than we should expect to find the great sandy and gritty group of the Foreland, whatever it may repre- sent. The old land-surface, therefore, of Devonian and Old Red Sandstone of the southern half of England must have extended from near Yarmouth (lat. N. 52° 50’), descending by a gentle southerly curve to about Leighton Buzzard, thence 1ising to the Wenlock pro- montory, and again sharply deflected S.W. to Milford Haven, and ANNIVERSARY ADDRESS OF THE PRESIDENT. 233 on under the Bristol Channel to Cornwall, all east of the Mendip Hills being now covered by the Secondary and Tertiary strata— thus, I think, showing that a widely spread floor of these Paleozoic rocks occurs under the extensive region of the southern half of England, and which could as easily be shown to occur in the north and in Ireland, but under different conditions. Now that the Ware Upper Silurians have been proved to dip to the south, we may anticipate the occurrence of the more ancient or lower series further north towards Cambridge, Huntingdon, and Peterborough. Whether in their extension northwards they may be of Scandinavian type we can only surmise, unless the Westmore- land and Cumberland Silurians pass under the unconformable and overlapping Carboniferous group of the Penine chain as a con- tinuous floor or old surface to the south. Since the determination of the existence of the Devonian rocks north of the Thames to Turnford, and of the Upper Silurian at Ware, below the Cretaceous rocks at the depth of 800 feet, our views rela- tive to the distribution of the Coal-measures have materially altered. It is clear that all the stratified rocks between the Silurian and Gault and the Devonian and Gault are wanting or not represented ; in other words, the whole of the Carboniferous, Triassic, and Jurassic rocks, and the Purbecks, &c. are missing north of the Thames to lat. 52° 10’. This shows the great unconformable overlap upon the Silurian and Devonian floor to the west. Had the sub-Wealden boring been completed, the problem as to extension would have been solved. Unfortunately for science this undertaking was never completed, and the problem remained unsolved; the ancient land was not discovered. We must rest content with the hypothesis of Mr. Godwin-Austen as to the area where we should expect to find the Coal-measures, probably ranging, as he suggests, under or north of the North Downs. Life must indeed have been most abundant and prolific in the Wenlock sea of the Ware locality; for so rieh in fossils are these Ware beds that no less than 33 species were obtained, all belonging to the Wenlock series, furnishing sufficient data in themselves to establish and determine the age of the rocks in which they occur. The five borings in the London area, or within the London basin, within a radius of 20 miles, Ware being the furthest removed, taking them in the order in which they were sunk, may be thus enume- rated :—No. 1. Kentish Town, 1300 feet deep; the London clay passed through was 350 feet thick, the Reading beds 50 feet, and the Thanet sands 15 feet, Upper Chalk 250 feet, Chalk-marl 30 feet, Upper Greensand 10 feet, Gault 60 feet, and 190 feet of red sandy rock believed to be Old Red Sandstone or a condition of the Devo- nian; the depth passed through was 1300 feet. No. 2. The Crossness new ‘well bore-hole about 1030 feet deep. No London c¢lay proper occurred in this sinking; the alluvial clay and gravel immediately below the ordnance level. was 20 feet thick, and rested upon the Woolwich and Reading and Thanet beds, here about 100 feet thick, the Chalk 620 feet, the Upper Greensand 234 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 33 feet, the Gault 135 feet, succeeded by loose red, coarse, and fine sand, having all the appearance of the New Red Sandstone, and resembling the Kentish-Town Red series. The boring No. 3, the most important in geological results, was that at Messrs. Meux’s, Tottenham Court Road. The London Tertiaries here attained to 156 feet, the Upper Chalk 440 feet, the Lower Chalk and Chalk-marl 215 feet, the Gault and Upper Greensand 190 feet, the Neocomian (of peculiar type) 64 feet ; beneath this occurred the Upper Devonian shales at the depth of 1064 feet, with characteristic fossils, Spirifera disjuncta, Hdmondia, Rhynchonella cuboides, Orthis sp., &e. This was the first indication of the presence of rocks older than the Neocomian east of North Devon and north of the latitude of London, clearly showing the easterly extension of the Paleozoic rocks from the western side of England, and extending towards Holland and Belgium. This boring therefore may be considered classical; it has revealed to us what was surmised intuitively by Mr. Godwin-Austen and Prof. Prest- wich. ‘Turnford and Ware have revealed other facts of nearly equal significance, and at the depth of 940 feet and 800 feet. The 4th boring is that at Turnford, 12 miles north of London, where the London Tertiaries are 100 feet thick, the Chalk 620 feet, the Upper Greensand 15 feet, the Gault 135 feet, the Neocomian, of the Carr-stone type, about 12 inches. We here again prove the position of the dark chocolate-coloured Upper Devonian rocks crowded with characteristic fossils; at the depth of 940 feet I obtained Spirifera disjuncta, Rhynchonella cuboides, Rhodocrinus, Strophomena rhomboidalis, Edmondia, Pterineea, ive TT