Bek Se be Salas Res Pe ass Lohse PW yay fn Pep fede Tee eT Se Fa rata Ae 1 bes titesd au as wi inde ces Bly AeA veh : wu bee Bg ee ii Wb rd eo Lrg aes og Laney tne ae rich hes) TILA Ce ad Pais) eae se) 1 wb lea ye aoe tgs us! ‘ witin Stee Lies ve aves teat Sy RA a Re Pegi nate Eg OR ES NTR CRIN vis Whey she e my WIEN RTA Tete eae cl cay RTE COR DUR 8 sna pete ey INRA ont: yi sarsicst “sats vais ea Vy Suey HS Stig as Sy vineie ING ae * Wy ry i Avieatksn Seasetiedn A, PA leSati AMZN AE YET ’ Wong bie Me Wnty mci egiya BENG fem ae SRL EE Gene Nd LAM StA syn ty 2 C9 Hee I AL A Peat NEWER ET as Heyy Attia puedes et caityt ane Le ne VecieaTnysvoniio © beNity Reid | bo hK ny : ary rUPMsibna yas? peikec Rs Res AS hy ST TENE aT we soe iets Begs Vanes Wen RE Sen Wye 0 rch Jengihagy toa Ay se eles TUG ails 2 fate anata alt ct ie Coney Prati ae eh ; 4 ity eal ti PR Lt AAAAP aie ThE ne nes tee Bi Pay eR Caches i tile Whoa art Scene MEN Se Pr te gee yas eigen be Singh MG aa ‘ oF i pea Eg ay huis sabes SESH Os BST yl eae ig Be ET a 3! i sin 1 dota esse rs Dae Ral eh A meas nen stae ot deta setae raring Bi Vpn sey a: badas Uric eae Boa ipa r iro Cen Shae gwen a ‘ Ceseend nak: a abel Cucr 3 ier busepelbisa cure pea Trani nad te wiphid dor ne SA aewale t Ue ke fate Ob Ue Ete seve vita x aatine ecieseip ci APD nf or Oy vege ip Ayo and stacey! a babAgh prieatne Be Bede SANG BST Ney Nd UH pe as Sa ees Ne Be Vai ba Hae) Hpi att Heese Strat anny! eg se Bin icio here Ue ee aL AEM Pt Dakin anesy er A ns ues F ag carat 1 Tait een yh Goeap Petes Lay) 4 A CTW OCGA EH AINE Ay t.tba * yes\ Se hes aes eae) me hits gba Heike Pn Gh ae) Way eV gE tt WOR or 5 ead FAIS ‘Winn DA, Saeki Ut Reeds Qed wie NUCL eee ea ae Uo ‘i at 4 fe CMe Tro “ a i eye vied te npiy nM ie Aart ; CH bee eH ARE Sg Fie sits 1} ale vA HY NOt ig Tans tin a fie? ei feweene ty aah ovis Sey Ale ‘ ae i i ‘ se ise bea lf a oy . Madi? DE Pea tintian pod smh 28) BY ie Pee Deca Fire Wik fe goes Wee og) panalbet HOS Aee eNO DV ILOE aig canty hae sh Wg katy yt yl filer, ot Satie eG BAF eet eras aN CeveRn Te tertiary fiers ve i 4 » 5 eh ainay a4 as an seia eg gl Lindisene a Seas Hairs est a“ ER ed ei Net te * j ; ican tuaty ane ie ae be fw yugr paagcoe # 2 Aer yphento mre Tie pen nee 5 pga s PG Lites Ma 7 pad eee rn ple, piv eA oe ‘ iy bay sth sean taasess Fe ier rine pele Phy “ats Uden i bossousieunt ees tis insu 999694 Spay denial seueeeut cba Onn at iat Bos Ny eis Figs wap ula wale Pt 1 i) Pena pe Pilea Sead ti es avanie a y; Bi aiyey Pye wa beanies Tad Poe Re , tHE 1 Caiugtes aay ict aR a % 3 is Test hh eh adap Sete Co pda Hs oy Atay (yeaa papain PRM te ae sib higa tara hey [eh Mah 5: spn tipit a ines ven, aH) i Brie ginny a sie i" Lea anil WOU y Ni wees hin lesb 0M Lak fips MMi pares: Vd 9) Seoaie aniston jhe geo yong if pad gee aeys Sidydincpanaie-aeies ips eae Serer eA ean a Bo mp fyi i ih ae ite LT i LIA fit rore i Sr een ales Cy hi as abe pate ri i youae Hi Mb AIM ee aie oy KIS WH * AL “hy y Az = 72 WS GS z dy 7) a y D WE ee n° 4 Yi s > = ise es = . aw z 77) ak a = RARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHALINS oe = S ex ll tut BEN tit = ont fp 2 8X = ¥ Pips 4 iT aN SS ss ts Y PN = E JY LS _— SS fs ie a + ons m WY eg m 2 m Me SS Z ie Z ws SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS n Pet: + WY) : w = a Ni = iy = o Zz 4 VG Y et G B 8 Dp @ 2 = 2,77 /# Z : ; g = SMITHSONIAN INSTITU LIBRARIES SMITHSONIAN Z Pa 2 : : Yip, = 4 m4 ts op fi A e = Ye lie mm ae Sig hh Meat ge he oO oO oO Ze Zz = ARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31¥uv. Sa1uvNaIT LIBRARIES 6 & fr z — SY. fe Oo o ° i 4 Ss) 5. 5 2 CER 5 rates Es = W = - * be a XN = ”) w 7. “” z = ep 2 LINLILSNI NVINOSHLINS S31YVYEIT LIBRARIES SMITHSONIAN Pa ep) = Paget oe) = = = Mare = < z S =a 5 = 5 6 = S a = =) a = = ; > ; a = a oS 7 | SMITHSONIAN INSTITUTION NOILONLILSNI NVINOSHLINS S3IUYV wn > v 3 ep) i aW : w eo : om fe a fe i oe \ Se § [ Bea ae en 1 4% SMITHSONIAN NVINOSHLIWS SMITHSONIAN NVINOSHLIWS Ks io by Whe ie % SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLIWS Saray RIES NOILNLILSNI LIBRARIES ACEP Sa1yVudl Gig INSTITUTION NVINOSHLINS S3tu¥vugii LIBRARIES S31YVUSIT LIBRARIES SMITHSONIAN INSTITUTION NOILOLILSNI S3IuvVusit SMITHSONIAN SRARIES INSTITUTION NOILNLILSNI NVINOSHLINS SSIYY ‘ n rai 27) Pa w = < = Hy s = 4 z =4 tt ye. F ot : : HL? j Z = 2 Fe 2 > = ~ > = > Zz ” neh ee “” Zs _NVINOSHLIWS SStYVYaIT LIBRARIES SMITHSONIAN INSTITU zZ oe Z a z 7) at 7) at no ~ ae 4 a - = laed 3) c ys = wae = = see 4 at fea) — oO. = S i fe) th fe) Zz | Pa wee Zz SMITHSONIAN INSTITUTION NOILALILSNI NV¥INOSHLINS S31yuy me i =z rr : =e Oo “0 O a OX 2) I= 4 oe \ pe ee Ne > rad Wh = =e z = 2 = w mn ” m oD = 22 teem = ” of Tae S vu K XS Ae eH = q hee i SONI AN SINS Pi je = x = < CN iy :: ma =a 7 nf J z 4 z NS SN 1g 4 Pa i@) sf at Wi, + hh oO : am Oo NS Se Y Jif? ISM a NNN Vif 2 yy * 2 = = > ai = Ras s *) z w hy me Y BRARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S31YY: o-= (ep) — ul a: m8 HB Ws ul % ond ee =f AN’ ik ahs \ SN *‘SLAIGONY “INQOO0Y ISAdy, ONO LSouy, HOIMEsaHg , OL OL 1SF OL OL SF Orss= 0 Seed eatin ae Dee miseries " poaleAoded xB, FUOOUT Hoa © | ene eC *-y00}G ‘3099 red @ vIpUy ‘Py “se 6SLF OO ar °°". CIA se Toqmraoag, 4 S1eyUeg oy) 8 soured ‘ UI Po}SeAUL pUNy IY} WO (x¥]-eMODUT Sse) SpUSpPIAI(] 0 OL OL CCC Ca Te Te see et Oe ee le ral LP "'s "25" ETEL 4ST Arvnuee ye sloyUVG oy} 4B souL[e ‘ps F ‘SINGNAV Wp 2 "‘SLAIGO MT] “INQOOOY shady, = ~aNOY AariTay TVOINOTOUY) , Il & Sle Ths Sle PL 0 Pe Re ee ere geet Ses peterodal XBT, stuOoUT 6 ive G CC 0G haces) lod e e HIpleg OLGF UG Gee CIOL STS oun ze (1 ages el OU ee oui UL pazseattt puny oxy To (xv],-etOaUT] SseT) SpUeprAl(T Se al ee eee gs “Tepe Joywon AG|OL FP 6 °'''''*° GTET 4sT Atunuve qv stoyueg og yw soured pe Se Fe ‘SINANAV oD. SG. - ee ‘SLAIN ‘INQOOOW ISAUy, , NO ASHI , G G WSF G6 G PEF PAO Oe ee ee ee “ pataAovar xy, atuoouy Tr 4 a see Peeteee seen s HQOIG rates ; aInjueqa “jue ted ¢ AvmyIey UtayyION WwaIH SOPs Ae ee le ee Ur payseaur puny ay} wo (xe L,-eMMODU] seq) SPUPPTATCL 0 0 46 peasy 4q}e grog cit: GIGL ‘4ST Acunuve ye sdoyueg oy} ye soured ps8 F ‘SINGWAV J ‘Ds F “LNOOOODY LSA "SLAIHOMY - ‘CND ’ é . ° SHORDRE NOUWUVM “H 'S NHTIV “V AUNUH ‘9T6L ‘YI6G hummunp ‘ainsnat], “TTTHNOW ALOACAA ‘90188 0} Toy} puy pur ‘sn 0} poyteserd syuNodDW pUe sYoog oY} IIA quOMIEZeIG StI} poredutoo oavy oA Z O S&F OL 0) fe SPOTS OOK ORLA TL OF Om ans AONE ‘+ paleAodat xey, ouroouy “ Clb OS ee ee ee Pee O01S Timed iad Fg BpeMedn QOOLS UL PoJSOAUL PUN ot{y WO (xv ]-sULOoUT S89) Spuepral(y OY, [9p oe 'SLAIMO MT “LSUNOMG ATLOG 8 0 S&F OloO @ See CEOORNG On Ola Or-CD Aad ct von On, patorovar xy, atmooUy “ AES Re OS aCe) tad £¢ wpeury QOOLF Ul po}SeATI puny oy} WO (xBz-omMooUyT 88901) SpueprIAI, OF, Dees ae ‘SLAIGO AY “LSdnbag NOLSATAN] ‘SGNNA IVIOAGdS JQ) SES; DS (Ne Se eS NS AST Leqmeoa(] WV SLoyuBg oy} Jv ooUBlVE kg es See *SINANAV 8 0 Sos 8 0 && °*°' GIB S81g toqmoseg yu stoyuRg oy} 78 ooUR/vE Aq 1 EES G3 “‘SINAWAV 7 ———— & 61. ts fe @ OP 282 AKA “sTe caqmmeoacy 4B SdoyUBE, oy} 4B oouVleg oo T OL 66 sete eet teeter eer cers eeesetecerercoesers nipmy SE DS oo “SINGNWAV qd & 61 VE 8 CL I Bo00.0-0.0-0-0'0 U2 080-0080" 0) DTD: 0-020 ONOFOs0 poteaooer xey, eumoouy “ F T LG oo1g ‘ques sod ¢ uoneLodi0~ [os "PZ “SL GLOLS Ul paqsoAUL puny oY} WO (xv]-euLCdUT sse[) spueprarq “ eg OL “°''** GIGL 481 Avenue 4% sroyueg ey} 78 eouUltE OL 8a "‘SLALMOO “INNOODY LS], (ONOY NOHNAIg THINVG ; part 1] FINANCIAL REPORT. Statement relating to the Society’s Property : December 31st, 1915. a5 & OW Balance in the Bankers’ hands, December 31st, 1915: » On Current Account ............6:. ce cee cee eee ees WAS 83) 7 », Deposit eA Aan aT sear Sse at Mayne ee RA 500 0 0 Balance in the Clerk’s hands, December 31st, 1915 719 9 Due from Messrs. Longmans & Co., on account of the Quarterly Journal, Vol. LX XI, etc. .... 364 «3 Arrears of Admission-Fees (estimated) ........ 50 8 O Arrears of Annual Contributions .......... Se oo e OnG (Estimated to produce £192 18s. Od.) Funded Property, at cost price :— £2500 India 8 per cent. Stock ............ 2623 19 0 £300 London, Brighton, & South Coast Rail- way 5 per cent. Consolidated Preference StOCK . 6.66 seek eee ee eee e erent teres 502 15 3 £2250 London & North-Western Railway 4 per cent. Preference Stock ............ 2898 10 6 £2800 London & South-Western Railway A per cent. Preference Stock ...........- 3607 7 6 £2072 Midland Railway 23 per cent. Per- petual Preference Stock. .........00000- 1850 19 £267 6s. 7d. Natal 3 per cent. Stock........ 250 0 £2000 Canada 33 per cent. Stock .......... 1982 11 [Norre.—The above amount does not include the value of the Libr and Stock of unsold Publications. The value of the Funded KOO Leet: a 635 18 4 468 12 9 £1104 61 £13716 2 9 ary, Furniture, Property of the Society, at the prices ruling at the close of business on December 3ist, 1915, amounted to £9,837 3s. 2d. | BEDFORD McNEILL, Treasurer. January 29th, 1916. x] PROCEEDINGS OF THE GEOLOGICAL SOcIETY. [ vol. Ixxii, AWARD OF THE WoniAston MEDAL. In handing the Wollaston Medal, awarded to Dr. ALEXANDER Perrovicn Karprysky, to M. Constantin Nazoxorr, Councillor of the Imperial Russian Embassy, the Prestpent addressed him as follows :— Councillor NaBoxorr,— The Council of the Geological Society has this year awarded the Wollaston Medal, its highest distinction, to Dr. Alexander P. Karpinsky, Honorary Director of the Geological Committee of Petrograd, which is responsible for the geological survey of the Russian Empire. Dr. Karpinsky’s activities have extended over a period of more than forty years, and so long ago as 1874 he made one of his most important discoveries, that of a marine formation in the Ural Mountains intermediate between the Carboniferous and the Permian Systems. This Artinskian Stage, as Dr. Karpinsky termed it, has now been traced in Russia almost from the Arctic Ocean to the Caspian Sea, besides being recognized in more remote regions, as in the Salt Range of India. Its interesting fauna has also been the subject of several important monographs, of which one of the most valuable is that on the Ammonoids, contributed by Dr. Karpinsky himself to the Imperial Academy of Sciences of Petrograd in 1889. Dr. Karpinsky has continued to take the deepest interest in the geological problems presented by the Urals, and has treated them with remarkable versatility from every point of view, whether tectonic, petrographical, or paleontological; but as official director of the surveys from 1885 to 1903 he also extended his researches to many other districts, and took a pro- minent part in the preparation of the beautiful geological maps which were issued during his period of active service. The useful Geological Map of Russia in Europe, which he edited in 1893, is especially well known. All Dr. Karpinsky’s work is characterized by the most painstaking thoroughness, of which I need only cite his two exhaustive memoirs on the Carboniferous ichthyolite, Helicoprion, as conspicuous examples. Those who have the privilege of his personal acquaintance recognize in him an un- assuming and enthusiastic student, still absorbed in following and part 1] ANNIVERSARY MEETING—MURCHISON MEDAL. xl aiding the progress of our science, and pre-eminently one whom - the Geological Society delights to honour. The Council will be glad if you will convey this medal to. Dr. Karpinsky as a token of its esteem and admiration, with an expression of its best wishes. Councillor NaBoxorr replied in the following words :— Mr. PREsIDENT,— Please accept my sincere thanks for the honour that you have done me in asking me to come here to-day and to convey to Dr. Karpinsky, with the expression of your good wishes, the Wollaston Medal which the Council of the Geological Society has awarded to him. I feel certain that this great distinction will be deeply appreciated by the recipient of the medal, as well as by the Russian Geological Committee as a high tribute to their Director. My distinguished friend, Dr. H. H. Hayden, Director of the Geological Survey of India, who crossed the Pamirs from India into Russian Turkestan a few months before the war, has often expressed to me the wish and hope that the highly interesting and valuable scientific researches, which have been carried out on both sides of the Pamirs by the British and Russian geologists, may be linked up and conducted on a basis of firmer and more complete unity and coordination. I venture to avail myself of this opportunity of expressing, on behalf of my countrymen, the same wish, and the confident hope that the ties of friendship which now unite Britain and Russia may extend from the fields of battle to the lofty peaks of science and enlightenment. Awarb oF THE Murcuison MEDAL. The PrestpEnt then handed the Murchison Medal, awarded to Dr. Ropert Kripsron, F.R.S., to Dr. F. L. Kircury, for trans- mission to the recipient, and addressed him as follows :— aa Dr. Krrcury,— The Council has awarded to Dr. Robert Kidston the Murchison Medal as a mark of its appreciation of his numerous and valuable contributions to our knowledge of fossil plants, especially those of the Carboniferous Period. For nearly forty years he has devoted xli PROCEEDINGS OF THE GEOLOGICAL socteTy. [ vol. lxxu, himself to an exhaustive and successful study of the external characters of the plant-remains associated with the various coal- seams ; and in this manner he has acquired an unrivalled knowledge of the distribution of the Carboniferous flora, which has proved of fundamental importance both to the geologist and to the practical miner. J] may mention, as examples of this work, his classic memoirs on the fossil plants of the Yorkshire and Staffordshire Coalfields and of Belgian Hainaut. During more recent years he has also extended his researches to various facts of structure and morphology which have a direct bearing on evolutionary problems. His memoir on the fructification of Newropteris heterophylla was the first description of the seed of a Pteridosperm in direct continuity with the frond; while his account of the microsporangia of the Pteridosperms first demonstrated the nature of the male organs in plants of this transitional group. His description of the internal structure of Szgillaria, and his remarkable series of memoirs, with the late Prof. Gwynne-Vaughan, on the evolution of the Osmundacez, must also be specially mentioned. While pursuing his researches he has continually recognized the importance of careful field-work, and has thus made a large and valuable collection of specimens, which has always been placed freely at the disposal of his fellow-paleobotanists. In transmitting this medal, please express our hope that he will treasure it not only as a token of our admiration, but also of our gratitude. Dr. Krrceuiy replied in the following words :— Mr. PrREesipENT,— It gives me great pleasure and satisfaction to be here to-day as Dr. Kidston’s representative, to accept on his behalf this valued award and to convey to you his thanks for the honour conferred on him by the Council of this Society. And I desire to thank you, Sir, in his name, for the kind words with which you have accompanied this presentation. It is gratifying to be the transmitter of the Murchison Medal to one who, a Scotsman himself, has laboured so long and so assiduously in elucidating the stratigraphical bearings of the Carboniferous Flora. Dr. Kidston, I feel sure, would have received this medal with enhanced pleasure, could he have listened to your graceful and appreciative references to his work. He asks me to express to you his great regret that he is unable to be here in person; and part 1] ANNIVERSARY MEELING—LYELL MEDAL. xl I may add that he is detained by responsible public duties, which have the first claim upon his time. | I have received a letter from Dr. Kidston which, with your permission, I will read :— ‘Will you please express to the President my sorrow at not being able to be present to thank the Society personally for the honour that they have done me in presenting me with the Murchison Medal, an honour which, it is needless for me to say, I very much value and appreciate. ‘The award of this Medal brings vividly to my memory that a number of years ago the Society awarded to me the Balance of the Proceeds of the Murchison Geological Fund, and I would like them to know that these proceeds were spent in the purchase of books dealing with Paleozoic Botany. It is only workers situated where not a single book on their special subject of study is obtainable for reference, who can fully appreciate the value of the help that I received from that award, and I hope that the books will eventually be placed where they will be of help to others. ‘I have now only to thank the Council of the Geological Society once more for its kind and encouraging recognition of my work.’ AWARD OF THE LYELL MEDAt. In presenting the Lyell Medal to Dr. Cuartes WinLiAM Anprews, F.R.S., the Prestpent addressed him as follows :— Dr. ANDREWS,— The Council has awarded to you the Lyell Medal as an acknow- Jedgment of the value of your numerous researches in Vertebrate Paleontology. Since your appointment to the Geological Depart- ment of the British Museum in 1892, you have made excellent use of the opportunities for research afforded by your official duties, and have made important contributions to our knowledge of fossil reptiles, birds, and mammals. You were soon attracted by the unique Leeds Collection of Oxfordian marine reptiles, and your studies of this collection eventually culminated in the two handsome volumes of the Deseriptive Catalogue, published by the Trustees of the British Museum (1910-13), which must always remain a standard work of reference on Ichthyopterygia, Sauropterygia, and Crocodilia. Your papers on the South American Stereornithes, on Rails from islands in the Southern Seas, and on Prophaéthon from the London Clay, are equally valuable contributions to our knowledge of extinct birds. Your researches on the fossil mam- mals of Egypt, many of them discovered by yourself, are still more noteworthy ; and your Descriptive Catalogue of the Tertiary xliv PROCEEDINGS OF THE GEOLOGICAL SOCIETY. | vol. xxii, Vertebrata of the Faytim (Egypt), published by the Trustees of the British Museum in 1906, began a new era in the history of mammalian life. Your demonstration of the stages in the evolution of the Proboscidea, and of the relationship between the Proboscidea and the Sirenia; your description and interpretation of the strange Eocene genus Arsinoitherium ; and your recognition of the early differentiation of the Hyracoids in Africa are especially funda- mental contributions to biological and geological science. I would further add that all your writings are characterized by remarkable thoroughness and insight into the meaning of the facts described. As your colleague in the British Museum during the whole period of your service, it gives me great pleasure to hand to you this medal, which the Council of the Geological Society could not have more worthily bestowed. Dr. ANDREWS rephed in the following words :— Mr. Presipent,— I wish to express my most sincere thanks to the Council of the Geological Society for the honour that it has done me in awarding to me the Lyell Medal, and to you, Sir, for the too flattering terms. in which you have made the presentation. I am_ particularly pleased to have received this medal from the hands of one with whom I have been associated for so many years. You will remember that, exactly twenty years ago, you yourself received this award from Dr. Henry Woodward, and that at the same time I received a moiety of the Balance of the Proceeds of the Lyell Geological Fund. If I have been able to accomplish something in Vertebrate Paleontology, it is mainly due to the fortunate environment in which I have found myself. An assistant in the British Museum possesses quite exceptional advantages, having free access to the great libraries and to the ever-increasing collections, and lastly, but by no means least, having many opportunities of making the personal acquaintance of workers interested in his subject. Having enjoyed these privileges, I feel that I have somewhat fallen short of what I ought to have accomplished; but, although it is just now uncertain what the future may have in store for us, I hope that I may still have opportunities of doing further work such as. will justify this award. part 1] ANNIVERSARY MEETING MURCHISON FUND. xlv AWARD FROM THE WoLLAston Donation Funp. The PrestpEnT then handed the Balance of the Proceeds of the Wollaston Donation Fund, awarded to W1ii1am BourkE WRIGHT, B.A., to Mr. G. W. Lampruen, F.R.S., for transmission to the recipient, addressing him as follows :— ! Mr. LamriueH,— The Balance of the Proceeds of the Wollaston Donation Fund is awarded to Mr. William Bourke Wright, in recognition of his contributions to Quaternary Geology. After completing his geological studies under Prof. J. Joly at Dublin University, Mr. Wright joined the Irish branch of the Geological Survey, and came under the influence of yourself when you were engaged in working out the glacial problems of the Dublin district. He took part in the revision of the memoirs and drift-maps of the Dublin, Belfast, and Cork districts, and shared with Mr. H. B. Maufe the discovery of a continuous raised-beach feature older than the Glacial Period. He also observed this pre-Glacial rock-shelf or beach in the West of Scotland, showing that a general subsidence allowed the sea to enter the valleys along the coasts of the British Isles, almost at the present sea-level, before they were occupied by the ic& After some experience both in Scotland and in England, Mr. Wright returned to Ireland, where, as Senior Geologist of the Trish Survey, he has since been successfully engaged on the glacial geology of the Kenmare and Killarney district. Much of his leisure has been devoted to the preparation of an important work on ‘The Quaternary Ice Age,’ in which he has made good use of his observations not only in the British Isles, but also in Scandinavia. Impressed by the value of Mr. Wright’s researches, the Council will be glad if you will transmit this award to him, with its best wishes for the progress of the work which he has so well begun. AWARD FROM THE Murcuison GronocicaL Funp. In handing the Balance of the Proceeds of the Murchison Geological Fund, awarded to Mr. Groraz Wanrer TyRRELL, E.G.S., to Dr. H. Lapworrn, Sec.G.S., for transmission to the recipient, the PRESIDENT addressed him in the following words :— xvi PROCEEDINGS OF THE GEOLOGICAL Society. | vol. lxxu, Dr. Larwortu,— The Balance of the Proceeds of the Murchison Fund has been awarded to Mr. G. W. Tyrrell in recognition of his contributions to the petrology of South-Western Scotland. His keen petrographic insight was first shown in his description of the quartz-dolerite sills of Kilsyth. His results of most general interest to geologists are those connected with the Paleozoic alkaline rocks ; for his investiga- tion of lugarite has added to petrology a peculiar rock-species and important evidence in favour of the differentiation of igneous rocks by the sinking of their heavier constituents. In several papers on the Auchineden Hills he has described their igneous rocks and their glacial and physical features; and, in his recent account of the ravine known as the Whangie, he has advanced conclusive evidence of its formation by earth-movements. As the Senior Assistant in the Geological Department of Glasgow University, and later also as Lecturer on Petrology there, he has done much towards the development of that school of geology. The Council hopes that this award may encourage and assist him in further research. AWARDS FROM THE LyELL Gronogican Funp. The PruestpEnr then presented a moiety of the Balance of the Proceeds of the Lyell Geological Fund to Mr. Marrin A. C. Hinton, addressing him as follows :— Mr. Hinton,— The Council has awarded to you a moiety of the Proceeds of the Lyell Fund in recognition of your researches on the British Pleistocene Mammalia, and as an incentive to further work of the same kind. Under circumstances frequently discouraging, you have for many years devoted yourself especially to the study of the Rodentia and the Insectivora, and have obtained a remarkable knowledge of the skeleton and teeth of certain groups which are most commonly met with among fossils. In this manner you have made discoveries with an important bearing on many problems of Pleistocene geology, which you have never failed to recognize. As one who has followed your work with great interest for several years, I have much pleasure in handing to you this award. part 1] ANNIVERSARY MEETING—LYELL FUND. xlvit The PrestpEnr presented the other moiety of the Balance of the Proceeds of the Lyell Geological Fund to Mr. Anrrep SanTER Kewnarp, F.G.S8., addressing him in the following words :— Mr. Kennarp,— It is particularly appropriate that the second moiety of the Proceeds of the Lyell Fund should be awarded to you, who have worked so long and so successfully with Mr. Hinton at problems of Pleistocene geology in the South of England. In the leisure of a busy life, you also have made yourself thoroughly acquainted with a group of fossils, the non-Marine Mollusca, which are of fundamental importance in classifying and interpreting the various deposits in which they occur. Both alone, and with Mr. B. B. Woodward, you have published many interesting notes and lists of such mollusca from Pleistocene and Holocene deposits in different parts of Britain. The Council desires to acknowledge the value of this work, and I have much pleasure in handing to you é tangible expacssion of its good wishes. xlviii PROCEEDINGS OF THE GEOLOGICAL socinTy. [ vol. ]xxil, THE ANNIVERSARY ADDRESS OF THE PRESIDENT, Arrnur Smita Woopwarp, LL.D., F.R.S. Amone those whom we have lost by death during the past year are two distinguished Foreign Members, Count Solms-Laubach and Dr. ©. R. Zeiller, who devoted themselves to Paleeobotany. For obituary notices of them I am indebted to Prof. A. C. Seward. ‘There are also two Foreign Correspondents, and several Fellows who have done great service to Geology. Hermann Graf zu Sorms-Lavpacnu died on November 24th, 1915, in his seventy-third year. He was elected a Foreign Member ot the Linnean Society in 1887, of the Royal Society in 1902, and of the Geological Society in 1906; in 1911 the Gold Medal of the Linnean Society was awarded to him, and with other foreign visitors he received an Honorary Degree at Cambridge on the occa- sion of the Darwin Celebration in 1909. Solms-Laubach belonged to ‘one of the most ancient German families, who were sovereigns in their own right down to the year 1806’ (‘ Nature,’ January 13th, 1916). He occupied the Chair of Botany at Gottingen, and succeeded the eminent botanist de Bary as Professor in the University of Strassburg. Though well known as the author of many important papers on recent plants, it is the distinguished part that he took in the reformation and revival of paleobotanical research which more particularly appeals to geologists. When he began to write his ‘ Hinleitung in die Paladontologie,’ he found it necessary to study the unrivalled collection of sections in the possession of Prof. W. Crawford Williamson, and for this purpose paid several visits to Manchester, which, as he says in a sym- pathetic biographical sketch of his friend published in ‘ Nature,’ September 5th, 1895, knitted closer the bonds of reverence and friendship between himself and his English colleague. The publication of Solms-Laubach’s book in 1887 led to a greater appreciation of the value of Wilhamson’s work, and his critical treatment of the widely-scattered literature enabled botanists to obtain a general view in truer perspective than had previously been possible of the significance of paleobotanical records. In the preface he wrote :—‘ Botany, which in former times generally treated Paleophytology in a very stepmotherly manner, now part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. xix finds it to be a subject of the highest interest to herself on account of the prominence at present assumed by the point of view of the theory of descent.’ The publication of an English translation in 1891 by the Clarendon Press conferred a boon upon many students to whom Count Solms’s involved and difficult style was a serious obstacle. The book contains a mass of valuable information, based in great measure on an actual examination of the specimens described; it is characterized by vigorous and discriminating criticism of the conclusions of previous writers, and the whole bears striking testimony to the author’s grasp of his subject, his exceptional power of handling details without losing sight of guiding principles, and his wonderful energy. Hor many years Solms-Laubach contributed to the ‘ Botanische Zeitung,’ and later to the ‘ Zeitschrift fiir Botanik,’ eritical summaries of recent work, which have played a con- spicuous part in keeping botanical readers informed of the more important results of paleobotanical enquiry. The great majority of Solms-Laubach’s published papers deal with petrified plants, and it was but rarely that he concerned himself with impressions. He made notable additions to our knowledge of the Paleozoic genera Medullosa, Protopitys, Sphenophyllum, and other types, and supplied fresh data of special interest from an evolutionary point of view. His researches into the structure of Lower Carboniferous and Upper Devonian plants yielded results of the greatest interest; he not only corrected the mistakes of earlier investigators, but presented for the first time an accurate picture, so far as the fragmentary nature of the material permitted, of the morphological characters of some of the oldest known plants. The discovery of several new generalized types gave emphasis to the view that the extinct Devonian genera represent highly-complex terms in a series extending back into ages far beyond those that have left any decipherable records. The account of his investigations on Stig- mariopsis and other plants in the quarries of St. Etienne threw new light on a subject that is still far from exhausted, and in a more recent paper he cleared up certain difficulties connected with the method of growth of the root-encircled stems of Psaronius. In collaboration with Capellini, Solms gave a systematic account of the splendid Cycadean stems from Northern Italy in the Bologna Museum, and, by his discovery of pollen-grains asso- ciated with a female flower, foreshadowed the later discoveries of Wieland, who had at his disposal the much more complete VOL. LXXIT. d 1 PROCEEDINGS OF THE GEOLOGICAL society. [vol. lxxu, American material. His work on the reproductive shoots of the English Cycadean stem Bennettites gibsonianus, origimally described by Carruthers, is especially noteworthy as an example of intensive study combined with rare morphological insight. Papers on the structure of some Conifers from the copper-bearmg Permian beds of Ilmenau and Frankenberg, and on Jurassic Gymnosperms collected in Franz-Josef Land by members of the Jackson-Harmsworth Expedition in 1894-96, are of special value as elucidating the structural features of genera previously known only as casts or impressions. Count Solms added greatly to our knowledge of recent and fossil calcareous Alge, not only in his text-book, but also in his monograph of the Acetabulariez published by the Linnean Society in 1895, and in other papers. His name will long be remembered with gratitude and respect by students of ancient plants; he raised the subject of Palzo- botany to a higher plane, and by his writings, as also in no small degree by his enthusiasm and infectious energy, he was the means of attracting many botanists to a branch of the science which had suffered neglect and had been discredited through its treatment at the hands of authors insufficiently equipped with a knowledge of recent plants. Solms-Laubach paid several visits to this country, where he was always a welcome guest. He was a man of considerable force of character, cultivated, and endowed with a sense of humour; a warm-hearted friend; and an in- vestigator of marked originality, whose work, more especially in the domain of Paleobotany, has had a wider influence than that of many men whose output was larger. PAR Cys CHARLES René ZEILLER died in Paris on November 27th, 1915, after a long and painful illness, at the age of 68. He was born at Nancy, his father being Ingénieur-en-Chef des Ponts et Chaussées, and his mother a great-granddaughter of the Lorraine artist Guibel, sculptor to King Stanislaus of Poland, Duke of Lorraine. He married in 1877 the sister of Léon Ollé-Laprune, the Catholic philosopher and a Member of the Institute, whose religious and philosophical views he shared. Prof. Zeiller was a Member of the Institute, Commander of the Legion of Honour, Inspector-General of Mines, and Professor at the Ecole Nationale Supérieure des Mines ; in 1905 he was elected a Foreign Member of the Linnean Society, and of the Geological Society in 1909. Zeiller rarely travelled in other countries than part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. li his own: he paid his first visit to England in 1909, when he attended the Darwin Celebration at Cambridge; and it is a satisfaction to learn from his daughter that, during his illness, he derived no little pleasure from the recollection of the days spent in this country. His dignified bearing and handsome face made him conspicuous among the distinguished foreigners upon whom the Chancellor conferred Honorary Degrees. Despite ex- acting official duties, Prof. Zeiller found time to make numerous substantial contributions to Paleobotany and, incidentally, to Geology. One of his earliest works, published in 1878 (the year in which he instituted the course of lectures on Paleobotany at the School of Mines), is an account of the plants of the French Coal Measures, the first of a series of admirable volumes on Upper Carboniferous and Permian floras. The two volumes on the botany of the Valenciennes Coalfield which appeared: in 1888 afford a good example of the author’s thoroughness of treatment and lucidity of style; they are not merely important from the point of view of the systematist and stratigraphical geologist, but the enlightened and philosophical treatment of the extinct types in their relation to recent plants gives them a high botanical value. Zeiller’s attitude was thoroughly scientific; he was always ready to consider criticisms and suggestions, and punctilious in his reference to the labours of colleagues: in him Solm-Laubach’s aphorism, ‘Bescheidenheit ist eine Zier,’ was conspicuously demonstrated, The description of the flora of Commentry and of the plants of Autun was shared with Renault. In the volumes on the Brive, Creusot, and Blanzy fossils, Zeiller made many additions of the first importance to our knowledge of several Paleozoic types, notably, in the Brive flora, as regards the main anatomical features of Psaronius. His work on the Coal Measures of Heraklia, in Asia Minor, and his description of Permian plants from Lodéve, contain much of great botanical interest. Without attempting to give a list of the types which he was the first to deseribe, or to enumerate the genera on which his investigations threw new light, one may refer to his memoir on the fructification of Sphenophyllum, published in 1898, as a remarkable example of the valuable results that can be obtained by a patient and skilful examination of unpromising material. Reference must also be made to the discovery of Paleozoic species closely allied to Selaginella, to his more recent researches into the anatomical characters of Lepidostrobus, and to his a2 ln PROCEEDINGS OF THE GEOLOGICAL SOCTETY. | vol. xxii ? discovery of new forms of Fern-like fructifications. Zeiller was the first to recognize the generic identity of the Indian genus Trizygia and the European Sphenophyllum; and his ingenious explanation of the peculiar features of the widespread Indian and Southern-Hemisphere genus Vertebraria is the best that has so far been suggested. One of the outstanding features of Zeiller’s work is the high standard of excellence of his descrip- tions of plants preserved as impressions: the comparatively few papers on petrified plants show that he was also thoroughly competent as an anatomist, but it is the high standard of his descriptive work, the determination to exhaust every method of attack, and his sanity of judgment and breadth of view that give a permanent value to his achievements. Zeiller’s accurate stratigraphical knowledge of the French coalfields enabled him to contribute in no small degree to the better appreciation of the value of plants as indices of geological age. He considerably extended our knowledge of the GJlossopteris flora in Brazil and South Africa, and described some new forms from the Lower Gondwana rocks of India; his paper on ‘ Les Provinces Botaniques a la Fin des Temps Primaires,’ in the ‘ Revue Générale des Sciences’ (1897), is a model of clear exposition and a highly-suggestive presentation of one of the most fascinating problems of geo- graphical distribution and plant-migration. Another contribution of especial interest from the same point of view is his critical examination of the Siberian plants referred by Schmalhausen to a Jurassic age, which led to their recognition as members of a Permian flora closely allied to those of Gondwanaland. Among many publications dealing with Mesozoic floras, the monograph of the Tongking Rheetic plants deserves special mention; it is one of the best works of the kind that we possess, not merely because of its excellent illustrations, but for the wealth of information which it contains and the masterly treatment of the rich material. Zeiller described collections of fossil plants from many parts of the world, from New Caledonia, Madagascar, China, Peru, North Africa, the Balkans, and elsewhere; his ‘Eléments de Paléobotanique,’ published in 1900, though of necessity greatly condensed, is a useful book for beginners as well as for experienced students. For several years Zeiller supplied comprehensive and critical summaries of recent palzobotanical literature to the ‘ Revue Générale de Botanique,’ and the thoroughness with which the laborious task was performed is characteristic of the man; he part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. lui was ever ready to help others, and unsparing of himself in the interests of the subject which he loved. It was a privilege to count Zeiller a friend; he was a man of strong affections, and deeply religious. While we admire his intellectual power and his whole-hearted devotion to the advancement of knowledge, it is his singularly-attractive personality and his lovable character that bulk largest in our recollection of the colleague whom we -have lost. To quote the words used by Zeiller in a biographical sketch of the Marquis of Saporta: he has left for all ‘le souvenir d’un maitre aussi aimé que respecté, en méme temps que dun des esprits les plus éminents dont ait a s’enorgueillir la paléontologie.’ pA Casa Micuren Férix Movrton, who was elected a Foreign Corre- spondent in 1899, was born on May 11th, 1845, and graduated at Brussels in 1867. He was appointed Conservator of the Royal Museum of Natural History at Brussels in 1872, and became Director of the Geological Survey of Belgium in 1897. Between 1875 and 1883 he contributed important papers on the Devonian formations of Belgium to the ‘ Bulletin’ of the Royal Academy of Brussels. He also edited the Memoirs on the Cretaceous and Tertiary formations of Belgium prepared for the Survey by A. Dumont. In 1880-81 he published his useful ‘ Géologie de la Belgique’ in two volumes, and during more recent years he wrote several papers on the Tertiary and Quaternary geology of that country. Epmonp Ricavx, who received an award from the Lyell Fund in 1883, and became a Foreign Correspondent in 1893, was an able amateur geologist residing at Boulogne. He made many important contributions to our knowledge of the geology of the Lower Boulonnais, which were eventually summarized in a memoir published by the Société Académique de Boulogne-sur-Mer in 1892. His acquaintance with the Jurassic rocks and fossils was especially profound, and his amiable services were always at the disposal of those who visited the district for geological work. He died in April 1915. Glacial Geology has lost a pioneer by the death of Prof. JAMES GEIKIE, which occurred at Edinburgh on March Ist, 1915. The younger brother of Sir Archibald Geikie, he was ‘born in liv PROCEEDINGS OF THE GEOLOGICAL soctEty. ([vol. Ixxu, Edinburgh on August 23rd, 1839, and was educated at the High School and University of that city. Having been naturally in- clined to geological studies from’ early youth, he joined the Geological Survey of Scotland in 1861, and became a District Surveyor in 1869. He was employed chiefly in the Lowlands and Southern Uplands of Scotland, often in districts of which the solid geology had already been examined, and his most important duty was to study, map, and describe the superficial deposits or ‘ drift.’ Such formations proved to have a special fascination for him, and he spent several vacations in investigating them in the Highlands, Outer Hebrides, and other areas which were beyond his official sphere. The peat-bogs especially soon attracted his notice, and seemed to him to indicate a succession of climatic changes during the period of their formation, which he described in his first important contribution to Pleistocene geology, ‘On the Buried Forests & Peat-mosses of Scotland, & the Changes of Climate which they indicate’ Trans. Roy. Soc. Edinb. vol. xxiv, 1867. He thus gradually arrived at the conclusion that the Pleistocene glacial period had not been continuous, but had been interrupted by several mild episodes or interglacial periods, and his results were eventually summarized in 1874 in his well-known volume on ‘The Great Ice Age & its Relation to the Antiquity of Man,’ of which new editions appeared in 1877 and 1894. This work was supplemented by another on ‘ Prehistoric Europe’ in 1881. In 1882 Dr. Geikie succeeded his brother as Murchison Professor of Geology in the University of Edinburgh. In that year he contributed an unportant memoir on the geology of the Ferde Islands to the Transactions of the Royal Society of Edinburgh, and expressed his views on ‘The Aims & Method of Geological Inquiry’ in his inaugural lecture at the University. He began educational work in earnest, devoting especial attention to the improvement of geographical teaching; and in 1884 he became one of the founders of the Royal Scottish Geographical Society, which fostered this work. From 1904 until 1910 he was President of that Society, and for many years he was honorary editor of its Magazine. His University students were provided for by his ‘Outlines of Geology,’ which passed through four editions (1886— 1903), and his ‘Structural & Field Geology for Students’ (three editions, 1905-12) ; while the relations of geology and geography were treated in a more general way in ‘ Fragments of Earth- Lore’ (1898), ‘Earth Sculpture, or the Origin of Land-forms’ part 1]. ANNIVERSARY ADDRESS OF THE PRESIDENT. lv (1898), and ‘ Mountains, their Origin, Growth, & Decay’ (1913). In the midst of all these new interests he continued to the end to pursue his enquiries into glacial geology, and an admirable brief summary of his latest conclusions was given in his Munro Lectures in 1913, published in book-form as ‘The Antiquity of Man in Europe’ in the following year. He retired from his Professor- ship in June, 1914. Prof. Geikie was an ideal teacher, both in the class-room and in the field, and gained from his numerous students the same deep esteem as that in which he was held by all who had the privilege of his friendship. He was elected a Fellow of the Geological Society in 1873, but had already contributed his first geological paper on the metamorphic Lower Silurian rocks of Carrick (Ayr- shire) to the Quarterly Journal in 1866. Our Murchison Medal was awarded to him in 1889, and he received the Brisbane Medal of the Royal Society of Edinburgh in the same year. He was elected a Fellow of the Royal Society of London in 1875, and was President of the Royal Society of Edinburgh at the time of his death. A list of his writings and a portrait appear in the ‘ Geo- logical Magazine,’ dec. 5, vol. x (1913) pp. 241-48 & pl. ix. RicuarD LYDEKKER, who was born in 1849 and died. on April 16th, 1915, was a Fellow of the Geological Society from 1883 to 1914, served on the Council, and became a Vice-President in 1894-97. Educated at Trimity College, Cambridge, he graduated in 1871, taking the second place in the first class of the Natural Science Tripos. In 1874 he was appointed to the Geological Survey of India, and proceeded to the exploration of the mountains of Kashmir, on which he wrote an important geological memoir. While there, his opportunities for sport gradually led him to take a special interest in the mammals and birds of India. and on returning to Calcutta he began to study the Indian Tertiary vertebrate fossils in the Survey collection. Planning a series of memoirs on these fossils for the ‘ Paleontologia Indica,’ he soon felt the necessity of ample materials for comparison. He accordingly retired from the Indian service, returned to England in 1882, and had the Calcutta collection sent in instalments to the British Museum, where he completed his work. Between 1885 and 1887 he prepared a Catalogue of the Fossil Mammals in the British Museum, in five parts. This was followed in 1888-90 by a similar Catalogue of the Fossil Reptilia & Amphibia, in four lvi PROCEEDINGS OF THE GEOLOGICAL soctery, [vol. lxxi, parts; and in 1891 by a Catalogue of the Fossil Birds, in one part. The whole series of ten small volumes did useful service to Verte- brate Paleontology at the time, and included the first systematic attempt to correlate European with American vertebrate fossils. In 1893 and again in 1894, lydekker visited the Argentine Republic at the invitation of Dr. Francisco P. Moreno, Director of the La Plata Museum. Here he studied the remarkable collection of mammalian remains from the Tertiary and post-Tertiary forma- tions of the Republic, besides a few dinosaurian bones from a Cretaceous deposit in Chubut. Some _ beautifully - illustrated memoirs in the ‘ Anales’ of the La Plata Museum were the result. By this time Lydekker had begun to realize the important bearing of recent discoveries of fossil mammals on the problems of geo- graphical distribution ; and in 1896 he published the most original and most philosophical of his works, ‘ The Geographical History of Mammals.’ During all these researches he contributed numerous preliminary notes to various serials, including the Quarterly Journal of the Geological Society. In 1889 Lydekker wrote the volume on vertebrates for the third edition of Prof. H. A. Nicholson’s ‘ Manual of Paleontology,’ and in 1891 he joined Prof. (afterwards Sir) William H. Flower in the authorship of a valuable ‘ Introduction to the Study of Mammals.’ From 1893 to 1896 he edited the ‘Royal Natural History,’ to which he himself contributed the sections on vertebrates. From 1896 until his death, much of his time was spent in arranging the public exhibition of Recent Mammals in the British Museum (Natural History), for which he was specially employed by the Trustees. His later studies therefore became more purely zoological, and he published several more or less popular volumes adapted to the needs of sportsmen. He also wrote a ‘Catalogue of the Ungulate Mammals’ and some Guide-books for the British Museum. His industry was phenomenal, but he sometimes bewailed his facility for literary composition which betrayed him into errors that on further reflection became at once apparent to him. In his ancestral home at Harpenden, though oppressed by sad domestic trouble, he continued absorbed in his favourite pursuits until the end, and his last volume for the British Museum was published posthumously. He had a large circle of devoted friends, whose admiration for his character increased the more closely they became associated with him. An excellent portrait of him appears as frontispiece to the fourth volume of the ‘Catalogue of the Ungulate Mammals in the British Museum.’ part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. lv Mr. Lydekker was elected a Fellow of the Royal Society in 1894, and the Geological Society awarded to him the Lyell Medal in 1902. CuarLes CaLtaway, who was a Fellow of the Geological Society from 1875 to 1906, and to whom the Murchison Medal was awarded in 1903, was a pioneer in the study of the British pre-Cambrian rocks, and also made valuable contributions to our knowledge of the Cambrian and Ordovician Systems. He was born at Bristol in 1838, and died at Cheltenham on September 29th, 1915. From 1876 until 1898 he resided at Wellington in Shrop- shire, and began his original researches in the area of the Wrekin. So long ago as 1874 he discovered an Upper Cambrian fauna in the Shineton Shales. A few years later he identified the under- lying greenish sandstone with the Hollybush Sandstone of Malvern. He was thus able to prove that the ancient masses of the Wrekin and the Longmynd represented a pre-Cambrian formation, which the termed Uriconian. Callaway next studied Anglesey, and came to the conclusion that the unfossiliferous metamorphic rocks of that island were also probably of pre-Cambrian age. He then visited the still more difficult region of the North-West High- lands of Scotland, and took an important share in the discussion of the great thrust-planes now recognized there. After so much experience of metamorphic rocks, Callaway began to consider the theory of dynamo-metamorphism, and applied it to the explanation both of the crystalline schists of the Malvern Hills and of certain old rocks in Galway, Donegal, and Wexford. He finally returned to the geology of Anglesey, and re-interpreted it in the ight of new knowledge, attempting to show that large masses of crystalline schists had been produced by intense changes in igneous rocks. This revised interpretation has been largely confirmed by the researches of Mr. Edward Greenly during more recent years, and will be fully dealt with in the forthcoming Geological Survey Memoir on the island. A portrait of Dr. Callaway and a list of his writings are published in: the ‘Geological Magazine,’ dec. 6, vol. 1 (1915) pp. 525-28 & pl. xvin. Arrnur VauGHaN was a brilliant exponent of the modern methods of stratigraphical geology, which he applied with success to the Lower Carboniferous formations. Born in London in 1868, Ivii PROCEEDINGS OF THE GEOLOGICAL socrEery. [vol. Ixxu, he graduated both at Cambridge and at London with high mathe- matical honours, and removed in 1891 to Clifton, where he resided until 1910. In the latter year he was appomted Lecturer in Geology in the University of Oxford, where he died on December 3rd, 1915. Beginning with the study of the Earth’s crust from the point of view of mathematical physics, he gradually became interested in all geological questions; and an intimate association with our late Fellow, Edward Wilson, definitely turned his attention to the stratigraphical distribution of fossils. After various preliminary researches, Vaughan proceeded to examine in detail the sections of Carboniferous Limestone in the Avon Gorge, and the results of his work, published in our Quarterly Journal for 1905, and in the Proceedings of the Bristol Naturalists’ Society for 1906, were not only a fundamental advance in our knowledge of the Carboniferous Limestone itself, but also a stimulating new departure in stratigraphical studies. Further researches in other areas in England, Wales, Iveland, and Belgium, enabled him to show how widely his results could be applied; and he gradually acquired so precise a knowledge of the mutations and variations of the Lower Carboniferous fossils, especially corals and brachiopods, that he was able to attempt correlations which otherwise would have been impossible. His last great work, a comparison of the Belgian with the British Carboniferous, was published im our Quarterly Journal so recently as September last. Vaughan had been in failing health for some years, and much of his investigation was carried on under difficulties, but he ever - retained his cheerful and buoyant spirit and unflagging zeal. He was an inspiring teacher and genial friend, and his untimely loss is mourned by all who knew him. He was elected a Fellow of the Geological Society in 1900, served on the Council from 1912 onwards, received the Wollaston Fund in 1907, and the Lyell Medal in 1910. A list of his writings and a portrait appear in the ‘ Geological Magazine,’ dee. 6, vol. iii (1916) pp. 92-96 & pl. v. OrviLtLE ADALBERT DeErBy, who died suddenly at Rio de Janeiro on November 27th, 1915. had devoted his life to the study of the geology of Brazil, and had done great service to our science. He was born at Kelloggsville (New York) on July 23rd, 1851, and proceeded in 1868 to Cornell University, where he came under the influence of Prof. Charles F. Hartt, who had accompanied the Agassiz Expedition to Brazil in 1865, and part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. lix had published his pioneer volume on the Geology of Brazil in 1870. After graduating in 1873, Dr. Derby became Assistant to Prof. Hartt, who had just been appointed Director of an Imperial Geological Commission of Brazil; and in the following year he went to Rio de Janeiro to act as Assistant in the con- templated survey. On the death of Hartt in 1878 the work of the Commission ceased, and in 1879 Derby was made chief of the geological section of the National Museum of Rio de Janeiro. The most important of his early memoirs on the Region of the Lower Amazons, the Cretaceous Basin of Bahia, and the Diamantiferous Region of the Parana, were published in the ‘Archivos’ of the Rio Museum in 1877 and 1878. In 1886 Derby left Rio and became Director of the Geographical and Geological Commission of the State of Sao Paulo, where he remained until 1905. He then undertook an examination of the diamond-bearing rocks of the State of Bahia, and shortly afterwards returned to Rio as Director of a new Geological and Mineralogical Survey of Brazil, which he carried on energetically until his lamented death. In 1887 and 1891 Derby contributed two papers on the nepheline-rocks of Brazil to the Quarterly Journal of the Geological Society. He was especially interested in petrology and in the occurrence of accessory minerals in rocks, and published many papers in the ‘American Journal of Science’ and in the ‘Journal of Geology,’ which are not only of scientific importance but of economic value. At the same time, he did not neglect any aspect of the science ; and one of his latest papers was devoted to an exhaustive study of the stem of the Permo- Carboniferous plant, Psaronius brasiliensis. Derby was an attractive personality, full of enthusiasm for his science, and always eager to welcome geologists who visited the land of his adoption. He gave special help and encouragement to Prof. J. C. Branner during his numerous researches in Brazil, and he similarly aided Dr. I. C. White in the preparation of his monumental volume on the coal-bearing rocks of Rio Grande do Sul. He was also associated with our Fellow, Mr. Joseph Mawson, when he was resident in Brazil and devoting attention to the geology of the country. In 1896, and again in 1907, I had the pleasure of experiencing his weleome both in Rio de Janeiro and in Sio Paulo, and learned to appreciate the difficulties under which he pursued his work with a strangely unsympathetic Government. He was elected a Fellow of the Geological Society Ix PROCEEDINGS OF THE GEOLOGICAL society. [ vol. lxxu, in 1884, and received an award from the Wollaston Fund in 1892. An excellent portrait of him appropriately appears as the frontispiece of the first edition of Prof. J. C. Branner’s ‘ Geologia Elementayr,’ published at Rio de Janeiro in 1907. Henry Hyarr Howenn was born on July 18th, 1834, and began his lifelong career on the Geological Survey of Great Britain in the autumn of 1850. As Assistant-Geologist, his first task was to accompany Beete Jukes in mapping the South Staffordshire coalfield; and for a few subsequent years he was occupied in the survey of the Midland counties. His most important memoir referred to the Warwickshire coalfield (1859). In 1855 Howell was transferred to Scotland, and devoted himself chiefly to the mapping of the Mid- and East-Lothian and Fifeshire coalfields. In 1857 he was promoted to the rank of Geologist, and he co- operated with (Sir) Archibald Geikie in producing their well-known volume on ‘The Geology of the Neighbourhood of Edinburgh,’ published in 1861. His work in association with Geikie and John Young laid the foundation of our present knowledge of the Carboniferous succession in Southern Scotland. In 1861 Howell returned to England, where he surveyed Jurassic areas in Northamptonshire, Bedfordshire, and Huntingdonshire. He then supervised the staff engaged in the survey of the North-Hastern counties, and was promoted to the rank of District Geologist in 1872. Ten years later he succeeded (Sir) Archibald Geikie as Director for Scotland, and eventually in 1888 became Director for Great Britain. He retired in 1899, after nearly half a century of disinterested public service, highly esteemed by all his colleagues. He had been elected a Fellow of the Geological Society in 1853. An account of the official work of Mr. Howell, with a portrait, appears in the ‘Geological Magazine,’ dec. 4, vol. iv (1899) pp. 483-37 & pl. xxi. He died in June 1915. Wiritam ANDERSON, who was born at Edinburgh in February 1860, and died at Sydney (New South Wales) on May 30th, 1915, was elected a Fellow of the Geological Society in 1899. After studying in the University of Edinburgh, he joined the Geological Survey of New South Wales in 1886, and wrote many valuable reports on the geology and mineral resources of that State, besides undertaking a series of researches on water-supply. In 1893 he retired from the New South Wales Survey, and, after a few years’ part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. Ixi service as Mining Specialist to the Geological Survey of India, he proceeded to South Africa. From 1899 to 1905 Anderson was Government Geologist of Natal, and he published three important reports on his work, with descriptions of the fossils by various specialists. After leaving South Africa, he returned to Scotland, but the climate proved too severe for his failing health, and he finally retired to Sydney. He was an excellent observer, full of enthusiasm for our science, and gained the esteem of all who knew him. South African geology loses one of its most active workers by . the premature death of Herpert Kynaston on June 28th, 1915. He was born on July 19th, 1868, at Durham, and educated at Eton and at King’s College, Cambridge. The Worts Travelling Fund was awarded to him in 1892, and he began his geological career by studying the Cretaceous Gosau Beds of Austria, of which he contributed an account to the Quarterly Journal of the Geological Seciety in 1895. From 1895 to 1902 he was engaged on the Geological Survey of Scotland, and in February 1903 he became Director of the Geological Survey of the Transvaal. Later, on the formation of the Union of South Africa, he assumed the direction of the newly-constituted Geological Survey under the Department of Mines. He carried on his Survey work with enthusiasm and success, and his annual reports were looked forward to no less by the student of pure geology than by the practical miner, whose needs in such a country are naturally paramount. Mr. Kynaston was of a retiring disposition and little known outside his depart- ment, in which he was universally esteemed ; but he took part in the proceedings of the Geological Society of South Africa, and was its President in 1908. He was elected a Fellow of the Geological Society in 1894. The Hon. Rogperr Marsnam (since 1893 Marsyam-Town- SHEND), who was born on November 15th, 1834, and died in April 1915, took a deep interest for many years in the study of stratigraphical geology, and in 1877 presented his valuable collection of British fossils to the British Museum. When in Brazil as Attaché at Rio de Janeiro, 1855-59, he obtained some Cretaceous fishes from Ceara, which he also gave to the British Museum. He was elected a Fellow of the Geological Society in 1859. Ix PROCEEDINGS OF THE GEOLOGICAL society. | vol. Ixxu, Sir SanprorpD Fremine, who was born at Kirkcaldy (Fife) in 1827, and died at Halifax (Nova Scotia) on July 22nd, 1915, became a Fellow of the Society in 1877. During the preliminary surveys and the early years of its construction, he was Engineer- in-Chief of the Canadian Pacific Railway. He was also associated with other railway enterprises in Canada; and his important public work was recognized by the conferment on him of the Knight Commandership of the Order of St. Michael & St. George in 1897. His interests were many and varied, and geological science, both theoretical and practical, was among them. Wiriiam Gryitits Apams, born on February 16th, 1836, was elected a Fellow of the Geological Society in 1865. He was a distinguished mathematician and physicist, and Professor of Natural Philosophy & Astronomy at King’s College, London, from 1865 to 1906. He became a Fellow of the Royal Society in 1872, and an account of his work is published in the Proe. Roy. Soc. ser. A, vol. xei (1915) pp. lxin, lsiv. He died in April 1915. Ricnarp AssHETon was a distinguished vertebrate embryo- logist, interested in all branches of biological and geological science. Born at Downham Hall (Laneashire) on December 23rd, 1563, he died at Grantchester, Cambridge, on October 24th, 1915. He had been elected a Fellow of the Geological Society in 1886, and of the Royal Society in 1914. Herpert Srantey Briony, who was elected a Fellow of this Society in 1911, died on June 6th, 1915, at the early age of 27. He was born in India, and educated at University College, London, graduating as B.Sc. at London University. He joined the Geo- logical Survey of India as Assistant-Superintendent in 1911, and was engaged chiefly in Burma and Kashmir. Forrescur Winiram Mizert, who was born in 1833, and died on February Sth, 1915, became a Fellow of the Geological Society in 1900. He was a well-known authority on the Foraminifera, and contributed an account of those found in the St. Erth Clay to the Palzontographical Society’s Monograph of the Foraminifera of the Crag. part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. lxill AnpDREW Duniop, who graduated as M.D. at Edinburgh Uni- versity in 1863, resided for 47 years in Jersey, and devoted much of his leisure to the study of the geology of that island. He contributed papers to the Quarterly Journal of the Geological Society on the Jersey Brick-clay and ou Jersey Raised Beaches in 1889 and 1893 respectively ; and so recently as November 1914 he communicated to the Society another paper on a raised beach on the southern coast of Jersey. He became a Fellow of the Society in 1874, and died at the age of 73 on December 30th, 1915. GEORGE HENRY HoLiinawortnH was an active member of the Manchester Geological Society, of which he was Treasurer for many years and President in 1903-1904. He was especially interested in the geology of coal-mining, but he also studied other geological questions in Lancashire, and in 1881 he contributed to the Quarterly Journal of the Geological Society a description of a peat-bed interstratified with Boulder-drift at Oldham. He was elected a Fellow in 1879, and died in April 1915. Brensamin Hongarr, who was born at Leeds in 1838, became a Fellow of the Geological Society in 1877. He devoted the leisure of his long life to a study of the geology and natural history of the district round Leeds, and made several contributions to the Transactions of the Leeds Geological Association. A portrait and a list of his writings are published in ‘ The Naturalist’ for April 1915 (pp. 145, 146). WiLLiaAM SIMPSON was an active member of the Yorkshire Geological Society and the Yorkshire Naturalists’ Union. He was especially interested in glacial geology, but also published notes on borings in the Millstone Grit at Halifax, where he resided until 1908. He was elected a Fellow of the Geological Society in 1893, and died at Catteral Hall, near Settle, in March 1915, aged 56. Joun Turner Horprack resided at Norwich, and for a long period took an active part in the public life of the city. He was a Member of the Museum Committee, and a valued supporter of the local scientific societies. In 1899-1900 he was President liv PROCEEDINGS OF THE GEOLOGICAL SOCIETY. |[ vol. Ixxu, of the Norfolk & Norwich Naturalists’ Society, and communi- cated to it in 1906 a paper on excavations in the Castle Mound. He was elected a Fellow of the Geological Society in 1900, and died on November Ist, 1915, aged 67. Henry Roreg, who was born on February 15th, 1839, and died on March 2nd, 1915, became a Fellow of the Society in 1890. He was a distinguished waterworks engineer, and constructed important works for the water-supply of many English towns. In 1905 he visited and reported on the water-supply of British Hast Africa. Ricuarpd Kerr, who was elected a Fellow of the Society in 1882, was engaged for many years in educational work at Folkestone, and subsequently became well known as a popular exponent of science. He died at Mitcham (Surrey) on May 21st, 1915. WitiraMm Hurron WixitaMs, who was born at Clifton, Bristol, in 1875, and became a Fellow of the Society in 1905, was a mining engineer who had done much work in Korea, Southern Manchuria, and India. He helda commission as Captain in the East Surrey Regiment, and was killed in action in France in May 1915. I have also to record the death on December 17th, 1915, of Wiri1am Ruprerr Jones, who was Assistant in the Geological Society’s Library from 1872 until 1918. He was born in 1855, the eldest son of Prof. T. Rupert Jones, F.R.S., and inherited his father’s retentive memory, which enabled him to give most valuable help to the Fellows using the Library. From its beginning until 1912 he prepared the slips for the Society’s annual Record of Geological Literature. part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. Ixv ‘Tue Use or tur Higner VERTEBRATES IN SE RAPHICAL GEOLOGY. JA a Tie study of fossil fishes, to which I referred last year, seems to show that each of the successive dominant groups is sharply distinguished from its immediate predecessor by some fundamental character. marking an advance towards the extreme adaptation for locomotion in water, which was ultimately attained in the Cretaceous Period. It isalso evident that various members of each of these successive groups soon became specialized for every possible mode of lite in the circumstances of the time. Fishes of the same general outward appearance and habit have thus originated repeatedly from progressively higher groups; similar adaptations have recurred with only minor differences ; and nearly the same changes, though perhaps with increasing intensity, have always marked the approach to racial old age. These phenomena are, indeed, so remarkable, that the question arises as to whether animals of apparently the same family, genus, or species may not originate more than once from separate series of ancestors. We may even hesitate further in deciding whether or no the really fundamental advances in life at successive periods have occurred more than once in the faunas of which they are respec- tively characteristic. The study of fishes, however, is scarcely sufficient to solve these problems, because the large majority of the fossils are marine, the animals would spread rapidly and widely, and the limits of the seas in which they lived are never clearly recognizable. The higher vertebrates, which inhabited the land, seem to be a much more hopeful source of necessary facts; for the land has always been subdivided into well-defined ar eas, aint by seas, mountains, and deserts. Animals in these sasrerel areas must often have deve- loped independently for long periods; alterations in the barriers ean be detected by the geologist when he studies the migrations and mingling of faunas ; while, as researches progress, the varying geography of successive periods may be more or less successfully restored. Like the fishes, the terrestrial vertebrates have advanced by successive fundamental steps towards perfection in powers of locomotion, and during this progress have at each stage diverged into various analogous specializations. They have, indeed, advanced one step farther by the final increase in the relative size and Ole DNL e Ixvi PROCEEDINGS OF THE GEOLOGICAL SocrIETY. | vol. lxxii, efficiency of the brain, which culminated in. Man. If, in these circumstances, the same type of animal has originated more than once, it should be possible, in some cases at least, to discover the phenomenon and determine its limitations. Although the enquiries involved are almost entirely biological, it is especially important for the geologist to take note of them, because students of shells from the modern standpoint are unanimous in recognizing what they term homcomorphy. They continually find ammonites, gastropods, and brachiopods, for example, which are essentially identical when full-grown but differ completely in their early stages. When the fossils happen to be perfect, these early stages are, of course, preserved and distin- guishable ; and if they reproduce, even only approximately, the ancestral condition in each case, they certainly suggest that the same type of shell may have arisen from more than one source. If this be so, shells from different horizons and widely-separated localities need very careful scrutiny before they can be used for correlating geological formations ; for, even if nearly similar, they may represent animals that have no really close affinity, and any. diagnostic features that they show may be inconspicuous points which have generally been overlooked as insignificant. For a clearer view of general principles, we therefore turn with ex- pectaney to vertebrate skeletons, which have much more numerous and tangible characters, and approach senility in more varied ways. Even among vertebrates it is by no means easy in every case to interpret the evidence that most concerns the geologist, and I need only refer to the supposed thylacines (Sparassodonts) and the horned tortoises (JMolania) of the Argentine Tertiary, which have often been quoted as specially strong proofs of the former existence of an Antarctic continent uniting the Australian and South American regions. So far as can be determined from their fossil remains, the Sparassodonts of the Santa Cruz Beds of Patagonia differ only in: the slightest particulars from the Pleistocene and existing thylacines of Australia, and they have been placed in the same family by American paleontologists. It is, however, to be noted that the thylacines and Sparassodonts are essentially identical with the primitive Creodonta, which are known to have ranged over all the continental land of the Northern Hemisphere at the beginning of the Tertiary Era. They only differ from these early mammals in certain senile specializations which might be expected in any long-lived group ; part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. Ixvii and, as these specializations are not altogether the same in the two cases, it is probable that they have arisen independently. Again, Miolania is a senile heavily-armoured form of the Pleurodiran Tortoises, which had a universal distribution in the late Secondary and early Tertiary Eras. The arrangement of the bony bosses on the skull is not absolutely the same in the species from Australia and Patagonia, and these indubitable marks of racial old-age may therefore have been acquired by separate groups’ in different regions. In other words, the essential identity of these land-animals does not prove a former direct connexion between Australia and South America; they may be merely survivors of cosmopolitan races at the two extremes of their former range, with certain inevitable marks of senility. In making comparisons, indeed, it is not enough to distinguish the fundamental and merely adaptive characters of animals; it is also essential to note separately those characters which depend on the early, mature, or senile position of the particular animals in the evolving series to which they belong. In the present state of our knowledge there seems to be only one case in which we begin to have materials for forming a judg- ment as to whether fundamental advances—the ‘ expression points’ of Cope—occur more than once. I refer to the acquisition of warm blood and its correlatives or consequences: which started the eareer of the mammals. The only reptiles that ever made a close approach to mammals in their skeleton existed during the Permian and Triassic Periods, and they were precisely intermediate between the Paleozoic Amphibia and the Mesozoic Mammalia. It is, therefore, presumably in Permo-Triassic times that mammals arose. These intermediate or-Theromorph reptiles, however, were spread widely over many lands. Their remains were first found in the Karoo formation of South Africa, were afterwards recognized in India, Scotland, Northern Rusgia, and Central Kurope, and also proved to occur in great numbers in the Permian of North America. Traces of them have also been met with in Southern Brazil. The African, Asiatic,and Huropean remains belong to animals so closely sunilar, that it is probable that they lived on a continuous land- area; but the American forms, although in the beginning not unlike, soon began to evolve into several groups which are almost or completely unknown in the Old World. The American Thero- morphs, therefore, probably flourished in isolation. Large collec- tions from both regions have been studied during recent years by e2 Ixvil PROCEEDINGS OF THE GEOLOGICAL SocIETY. [ vol. Ixxu, Dr. R. Broom, Dr. E. C. Case, and others, and a critical summary of the results was published by Dr. Case last summer.! It now appears that all the specializations in the American groups were in the direction of higher reptiles; while all those in the South African groups made a progressively closer approach to mammals, and as nearly as possible culminated in typically mammalian skeletons. Hence, although we have evidence of two possible sources of mammals, only one appears to have produced them. Secondly, consider a case of less fundamental character, the origin of the Monkeys or lower Anthropoidea. It is generally admitted that they arose from the next lower grade, that of the Lemurs or Lemuroidea, which were almost universal in their distribution over the great continents at the beginning of the Tertiary Era. It is also suspected (though the paleontological evidence is still very scanty) that the two markedly distinct groups of Monkeys, the Platyrhines in America and the Catarhines in the Old World, were derived independently from the Lemurs in these two separate continental areas. If this be so, a zoologist’s ‘sub- order’ has arisen by two parallel, though readily distinguishable, developments from an earlier and lower ‘sub-order.’ Now, the surviving Lemurs of the Asiatic and African regions and Mada- gascar have lived on from early Tertiary times practically un- changed ; but there is one isolated area, the island of Madagascar, where some families during the later Tertiary periods attained remarkable specialization. Unfortunately, we only know the Pleistocene and Holocene members of these families that are discovered in the caverns and swamps of Madagascar; but they represent the culminating genera, and suffice to indicate what happened. In this isolated region the highly-specialized late Tertiary Lemurs curiously resembled in some characters both the Platyrhine and the Catarhine Monkeys. ‘The first portion of skull of Nesopithecus, in fact, appeared to so experienced an observer as Dr. Forsyth Major so monkey-like that he originally placed it among the Monkeys, in a new family intermediate between the South American Cebidee and the Old-World Cercopithecide. I+ is now clear, however, from the researches of Dr. G. Grandidier and Dr. H. F. Standing, that none of these animals really passed beyond the typical Lemur-grade. They eventually grew large, 1 #. C. Case, ‘The Permo-Carboniferous Red Beds of North America & their Vertebrate Fauna’ Publication 207 (1915) of the Carnegie Institution of Washington, p. 121. a part 1} ANNIVERSARY ADDRESS OF THE PRESIDENT. ]xix some of them (Megaladapis) almost rivalling donkeys in size; they also diverged for various modes of life, a few of them even entering the proper sphere of hoofed animals. Removed from the stress of competition with other mammals such as swarmed on the continents, they flourished in luxurious ease ; and we are tempted to speculate whether it was not the strenuous competition in the crowded continental forests of America and Africa (or Asia) that made ‘the difference, and led to the increase of brain-power which is the specially distinctive characteristic of Monkeys. If this be so, the lower race passed into the higher race only where the struggle for existence was keenest. Why the higher race never advanced further in America, while it began at once to give rise to the man-like Apes in the Old World, is still an unsolved problem. The case of the Anthropoidea suggests that a widely-distributed group of early Tertiary mammals experienced in three separate regions an essentially-similar initial impulse to become a higher group, evolved on parallel though not identical lines, and ended at ‘three different stages in its onward progress. It is, therefore, interesting to consider the geological history of some of the families of Ungulata which are already sutticiently well known by a succes- sion of fossils to reveal the main episodes in their career. This history is especially instructive, because all the surviving groups originated in the Northern Hemisphere on two or three large continental areas which were sometimes united, at other times separate; and the fossils are beginning to show how and when the various connexions and isolations occurred. Northern Africa, probably with part of Southern Asia, is one of the areas on which Ungulates made an independent start; Europe, with North-Central Asia, is another ; while North America is the third. The Rhinoceroses are a well-marked family, of which numerous fragmentary fossil remains can easily be recognized and compared. They seem to have arisen at the beginning of the Eocene Period from some small generalized Perissodactyl which had a slender snout and a regular close series of 22 low-crowned teeth in each jaw; and, as ages passed, the nasal bones became enlarged and thickened to support a dermal horn, the front teeth tended to disappear as the lips became more prehensile, while the molars and premolars increased in effectiveness for grinding hard and dry herbage. As in other Perissodactyls, the premolars were at first comparatively simple, but afterwards gradually approached the xx PROCEEDINGS OF THE GEOLOGICAL sociEry. |[vol. lxxu, molars in complexity ; while the successive representatives of the family showed a progressive increase in bodily bulk. ‘These changes took place almost simultaneously both in the Old World and in North America, though in the latter region the whole family died out in early Pliocene times as soon as a rudimentary nasal horn was beginning to appear. After the earliest stage, however, the European series at least can readily be distinguished from the American series by the premolar teeth, and there must have been independent evolution in the two regions. According to Prof. O. Abel, the oldest known recognizable member of the family is the small Prohyracodon from the Middle Eocene of Transylvania ; and both this and ‘the later Mpiaceratherium, from the Lower Oligocene of Northern Italy, agree with the earliest North American Rhinoceroses in having the foremost premolar more nearly like a molar than is the hindmost premolar. They thus show the first step in the complication of the premolars from front to back, which, according to Prof. H. F. Osborn, is the direction of this complication during geological time in all the American Rhinoceroses. On the same account, they differ from all the known Rhinoceroses which followed them in Europe: for in these the complication of the premolars always began with the hindmost, and proceeded forwards. Not only can two distinct progressive groups be thus recognized in two distant regions, but more than one line of development can also be traced within each. The fossils, indeed, seem to justify the conclusion that the Rhinoceroses arose from a common stock, but evolved in several different ways and at varying rates towards the same goal of specialization. It is curious that extreme specialization in the family was reached only in the Old World, and still more curious that the large-horned woolly rhinoceros (2h. antiquitatis), which lived in the Arctic Regions with the mammoth, never accompanied that animal to North America. Equally interesting is the exceptional case of high specialization in the Old-World Pleistocene Hlasmo- therjum, in which the extreme deepening, enlargement, and complication of the molar and premolar teeth seem to have been correlated with the reduced development of the epidermal horn. The large bony boss on the frontals of this animal was probably covered only by a very thin capping of horn. At least, it is difficult to suggest any other interpretation for such a horn, which was found isolated a few years ago by the Prussian Geological part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. Ixx1 Survey, in a superficial deposit in the Luckau district south of Berhn. A detailed study of the tooth-pattern in the successive groups of extinct Horses, seems to reveal the same phenomenon that 1s observable in the Rhinoceroses—the approach to nearly-identical extreme specialization by several separate lineages in approximately the same time. ‘The earliest Eocene forerunners of the Horses are essentially similar in Europe and in North America, and some of the later forms also exhibit many resemblances which may be due to migrations from one region to the other; but the most ex- haustive modern researches leave no doubt that the gradual reduction of the foot to a single toe and the deepening of the teeth for effective grinding of hard food, with various correlated specializations, occurred in several distinct groups of Horses in different localities. The climax was reached during the Pliocene Period, both in Europe and Asia and in North America; while the greatest diversity of form appeared in the latter region and among the immigrants to South America, where all Horses died out before historic times. Even in families of more restricted geographical range, the same unswerving tendency towards a fixed goal is very clearly recognizable. Among the Camels, for instance, which seem to have accomplished the whole of their evolution in part of North America, the characteristic cushioned foot appears to have been produced more than once. All the small early ancestors of the Camels had feet like those of deer or gazelles, with pointed toes ; and Prof. W. B. Scott has observed that even the Lower Miocene genera, which were already differentiated into two groups, still exhibited the same primitive feature. All the later Miocene Camels of both these groups, however, possess the irregularly- nodular ungual phalanges which indicate the presence of the cushions or pads. It is thus evident that two progressive series independently acquired one and the same structure. A study of the extinct representatives of several other familiar Ungulata has led to the discovery of many facts which tend to confirm the results just mentioned. In every case, either the feet become perfectly adapted for rapid locomotion over hard ground, or the teeth acquire more grinding power by deepening, folding, and the frequent infilling of the hollows with cement; or both these progressive adaptations take place at the same time. When, however, this evolution first started in the Eocene Period, the / Ixxii PROCEEDINGS OF THE GEOLOGICAL soctety. | vol. Ixxu, | common ancestors of the groups that were thus destined to flourish were accompanied by other generalized Ungulates which began to acquire running feet without any of the requisite changes in the arrangement of the wrist- and ankle-bones, and often showed a deepening or other complication of the grinding teeth without any infilling of the resultant hollows by cement. These animals with ‘inadaptive modifications’ (as Kovalevsky termed them) were soon handicapped in the race with their better-endowed contemporaries, and all became extinct before they had advanced far. In South America, on the other hand, which was an isolated centre of mammalian development during the greater part of the Tertiary Era, all the early Ungulates began to advance in the ‘inadaptive ’ manner, without any more efficient contemporaries and competitors. Here we are, therefore, able to follow to the end | a course of evolution which was abruptly stopped in the Northern | Hemisphere. It resulted in less variety than is observable among ordinary Ungulates, but some of the later genera are strange mimics of the Rhinoceroses and Horses in outward shape. At least one of the bulky three-toed forms acquired a rhinoceros-like horn: and some of the small one-toed forms exhibited a greater reduction of the lateral toes even than in the horse. The molar teeth were also sometimes invested with a little cement, but they never became such effective grinders as those of the Horses and Elephants. The latest of these South American animals, indeed, soon disappeared when a land-connexion in the Pliocene Period allowed the animals of North America to invade the Southern Continent and compete with them. None of the South American Ungulates developed the sym- metrical pair of toes such as characterizes so large a proportion of the northern forms, but their single enlarged toe, which became the centre of symmetry, was always the third, as in the ordinary tapirs, rhinoceroses, and horses. It is, therefore, interesting to note that among the Australian Marsupials, which have inde- pendently developed a hind foot for rapid locomotion or leaping on hard ground, the largest toe is not the third, but the fourth. So far, unfortunately, we lack nearly all the ancestors of these animals; but available evidence seems to show that they are descended from tree-dwellers in which the first digit had become opposable. When they began to live habitually on the ground and the foot became modified accordingly, this digit dwindled to a projecting rudiment (as preserved in Diprotodon), and the axis of part 1 | ANNIVERSARY ADDRESS OF THE PRESIDENT. Ixxi the foot was displaced outwards from the third to the fourth toe. In other words, an essentially ungulate foot was produced as usual, only disturbed by an initial twist. From these and many similar facts in mammalian evolution, we may therefore conclude that parallel development or ‘homco- morphy’ undoubtedly occurs. Some of the changes are adaptations to the same altering modes of life, perhaps also to the same progressive modifications in the environment; among which may be enumerated the increasing efficiency of the teeth for grinding and of the feet for running. Other changes are the inevitable marks of racial maturity or old age, such as increase of size, the development of excrescences, and the reduction in number of the teeth. While, however, both these series of changes always take place in approximately the same order and may be used to mark the successive periods of Tertiary time by a paleontologist who is accustomed to deal with such evidence, there is still no reason to suppose that identical animals have arisen from more than one source. Distinct families or genera which may be difficult to separate on superficial examination, are readily recognized in most eases when studied in detail with our present knowledge of general principles. It is, indeed, usually possible to distinguish between traits of heredity that are fundamental, and those that are merely adaptive or depend on racial antiquity. The new palzontology is thus as useful in relation to world-problems, as was the old paleontology in determining the relative ages of the rocks in the restricted Huropean area to which it was first applied. The adaptive characters just mentioned change with much rapidity, and allow a tolerably detailed subdivision of the series of strata which contain the fossils exhibiting them. It is therefore interesting to notice that some other characters which can scarcely be correlated with utility or efficiency, persist without change for comparatively long periods and are of little value for stratigraphical geology. Huxley long ago directed attention to the problem of ‘persistent types,’ as he termed them, and most of them are still as inexplicable as they were when first considered. Only in the ease of certain fishes have I suggested that advance was stopped by the appearance of various modifications or specializations in the wrong order. It may be that single peculiar characters persist when they cannot be affected by Natural Selection. One of the most striking examples was described by Egerton in the Society’s Journal for 1871 in a Chimeroid fish from the Lower Lias of Ixxiv PROCEEDINGS OF THE GEOLOGICAL society. {[vol. lxxi, Lyme Regis, which he named Ischyodus orthorhinus (now known to be Prognathodus). The prolongation of the snout in this early Jurassic fish is of precisely the same remarkable shape as that in the existing Chimeroid Callorhynchus. Notwithstanding the many changes which have occurred in Chimeroid fishes since that remote period, including a total replacement of the genera, the pattern of snout in one group at least has persisted. Again, the existing Lamnid sharks are characterized by a curious dwarfing of the third or fourth tooth, or both, on each side of the upper jaw. A study of associated sets of teeth of some of the earliest- known members of the family from the Chalk, proves that this feature was already established even in Cretaceous times: not only in the genera which have survived, but also in the extinct Coraz. Finally, in the Society’s Quarterly Journal for 1910, I pointed out that in the Lower Jurassic carnivorous dinosaurian, Megalosaurus, three or four pairs of the teeth in front of the lower jaw are diminutive, perhaps functionless. Last year, through Mr. W. E. Cutler, the British Museum received from the Upper Cretaceous of Alberta (Canada) the jaws of a gigantic Megalosaurian, evidently of another genus, in which exactly the same reduction of the teeth at the mandibular symphysis oceurs. Through a wide range of time and space, therefore, this small and apparently insignificant feature persisted, and even passed from one genus to another. The strength of heredity is, in fact, one of the most remarkable phenomena with which a paleontologist is repeatedly impressed. Even when one great group gives rise to another, the later type often seems to be handicapped at first by the inheritance of some characters which are no longer congruous. The large majority of the Permian and Triassic Reptiles, for example, although true land-animals, still retained the large head and short neck which were well adapted to the aquatic habits of their amphibian and piscine ancestors. This was first strikingly shown by Seeley’s restoration of the South African Aeirognathus, afterwards by Boulenger’s sketch of the Scottish Velerpeton, and has been emphasized more recently by the numerous restorations of Permian reptiles from North America by Williston, Case, and others. Similarly, when Reptiles passed into Mammals, the relatively-large tail of the gliding or swimming animal was no longer needed; but it persisted in nearly all Mammals at least through the first half of the Eocene Period, and even at the . part 1] ANNIVERSARY ADDRESS OF THE PRESIDENT. xxv present day the abdominal region gradually tapers into the thick root of the large tail in such lowly types as the Marsupial Thylacinus and the Kdentate Orycteropus. Archeopteryx suggests that the persistence of the reptilian tail was also a handicap to the earliest birds. It may be that the same conservative tendency is the origin of various other apparently incompatible structures, for which adaptive explanations have been sought in vain. For instance, it has been surmised that the great bony brow-ridges of apes like the gorilla are needed to resist the strain produced by the working of the powerful jaws; but the same explanation will not apply to the nearly similar forehead of Neanderthal Man, in which the jaws are no heavier than those of many modern men. This must be a case either of direct inheritance or (as at present seems more probable) of curious reversion in a higher group to a tendency characterizing an ancestral lower group. Many similar instances might be cited. The paleontologist is, indeed, now continually tempted to trespass on the domain of the speculative philosopher. He has had to abandon the old vague methods of comparative anatomy by which the solid foundations of his science were first laid. He now sorts out characters into several categories before beginning to compare them, and arrives at his interpretations in accordance with certain general principles which I have tried to illustrate. If, as Sir Jethro Teall has said, ‘the state of advancement of a science must be measured, not by the number of facts collected but by the number of facts co-ordinated,’ Paleeontology is well in the forefront of progress. Its devotees may be wrong in some of their broader speculations, but during recent years they have rarely met with new facts which could not be reconciled with the general scheme of things developing in their minds. The value of its results for the purposes of geology must thus be constantly increasing. Leaving high philosophy, it now only remains for me, in conclusion, to thank the Fellows of the Geological Society for the great honour which they conferred upon me two years ago when they elected me to be their President. I also wish to express my appreciation of the kind help and forbearance of the Officers, Council, Fellows, and Permanent Staff, which have made my term Ixxvi PROCEEDINGS OF THE GEOLOGICAL socrnty. [vol. lxxu, of service to the Society one of the most pleasant and memorable episodes of my life. My own special studies are so completely on the borderland of our science, that I fear I must sometimes have wearied you with matters that seem almost extraneous; but it is one of the greatest charms of geology that it needs co-operation with the whole circle of the sciences to further its aims. You have now turned from paleontology to petrology, and I have much pleasure in resigning the Presidential Chair to an old friend, an accomplished Fellow who has indeed delved deeply into the mineral structure of the Earth. part 11 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. xxvii February 23rd, 1916. Dr. ALFRED Harxker, F.R.S., President, in the Chair. William David Purdy, 7 Christchurch Terrace, Chelsea, S.W., and George James Roberts, Noyna, Avenue Rise, Bushey (Hertfordshire), were elected Fellows of the Society. The List of Donations to the Library was read. The following communication was read :— ‘On the Origin of some River-Gorges in Cornwall and Devon.’ By Henry Dewey, F.G.S. (Communicated by permission of the Director of H.M. Geological Survey.) Lantern-slides were exhibited in illustration of the above paper. A nodule surrounded by a layer of cone-in-cone structure, without any trace of calcite, from Lower Paleozoic rocks near Machynlleth (Montgomeryshire), was exhibited by Dr. A. Smith Woodward, In pldiaera« Vele(Crise March Sth, 1916. Dr. Atrrep Harker, F.R.S., President, in the Chair. Wilfrid Baker, Royal Grammar School, Worcester; and Arthur G. Pomeroy, M.A., B.Eng., 21 Orange Street, Haymarket, S.W., were elected Fellows of the Society. The List of Donations to the Library was read. The PresrpEn’ referred with regret to the death, on March 3rd, of Prof. Jonn Westey Jupp, C.B., LL.D., F.R.S., Past President of the Society. He spoke of the value of Prof. Judd’s contri- butions to geological science, and of his eminence as a teacher of the science, and stated that the Society was well represented at the funeral. ' Dr. AuBREY Srrawan, F.R.S., Director of H.M. Geological Survey, exhibited and described briefly a set of specimens from the Western Front, illustrating the character of the rocks in which trenches, tunnels, etc. are being dug. They included specimens from the Cretaceous and Tertiary formations showing remarkable similarity in characters to the contemporaneous forma- tions in Britain. 4 Ixxvil PROCEEDINGS OF THE GEOLOGICAL socrery. [vol. lxxu, The following communication was read :-— “On some Insects from the British Coal Measures.’ By Herbert Bolton, M.Sc., F.R.S.E., F.G.8., Reader in Paleontology in the University of Bristol. Lantern-slides and specimens were exhibited in illustration of the above paper. March 22nd, 1916. Dr. AurreD Harxker, F.R.S., President, in the Chair. Lieut. Kenneth Neville Moss, B.Sc., Royal Engineers, Fieldgate, Walsall, was elected a Fellow of the Society. The List of Donations to the Library was read. Dr. A. SmirH Woopwarp, F.R.S., V.P.G.S., exhibited speci- mens of the problematical ichthyolite, Celorhynchus, from an Eocene deposit in the Ombialla district (Southern Nigeria), and discussed the nature of this fossil. Microscope-sections of the well-preserved Nigerian specimens confirmed W. C. Williamson’s determination that Calorhynchus is an essentially dermal struc- ture. A similar section of part of the rostrum of the teleostean fish Blochius, from the Upper Eocene of Monte Bolca, near Verona, showed an almost identical structure. The precise nature of this rostrum remained to be determined, but there could be no doubt that the so-called ‘ Celorhynehus’ is the corresponding part, either of Blochius or of an alhed genus. The followimg communication was read :— ‘The Pseudo-Tachylyte of Parijs (Orange Free State) and’ its Relation to “Trap-Shotten Gneiss”” and “ Flinty Crush-Rock.”’ By S. James Shand, D.Se., F.G.S., Professor of Geology in the Victoria College, Stellenbosch (8.A.). April 5th, 1916. Dr. ALFRED Harker, F.R.S., President, in the Chair. The List of Donations to the Library was read. The PRESIDENT announced that the Council had awarded the ' part 1] PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Ixxix Proceeds of the Daniel-Pidgeon Fund for the present year to Joun Kaye Cuariteswortn, M.Sc., Ph.D., F.G.S., who proposes to conduct researches in connexion with the Glaciation of Donegal. The following communication was read :— ‘The Picrite-Teschenite Sill of Lugar (Ayrshire) and its Differ- entiation.” By George Walter Tyrrell, A.R.C.Sc., F.G.S. Lantern-slides were exhibited in illustration of the above paper. A Bronze Bust of Sir Charles Lyell, presented to the Society by Mrs. J. W. Judd, through Prof. W. W. Watts, was also exhibited. May 10th, 1916. Dr. AnFRED Harker, F.R.S., President, in the Chair. Keir Arthur Campbell, B.A., Trinity College, Cambridge ; Ednyfed Wynne Hughes, M.Sec., The Grange, Lesness Park, Belvedere ; and Christopher Luke Waite, Newlands, Pontefract Road, Castleford, were elected Fellows of the Society. The List of Donations to the Library was read. The followmg communications were read :— 1. ‘ Carboniferous Fossils from Siam.’ By F. R. Cowper Reed, MEAS se.D. E.G-S: 2. ‘The Lurgecombe Mill Lamprophyre and its Inclusions.’ By Herbert Gladstone Smith, B.Sc., F.G.S. Carboniferous fossils from Siam were exhibited by F. R. Cowper Reed, M.A., Sc.D., F.G.S., in illustration of his paper. Lantern-slides, rock-specimens, and microscope-sections were exhibited by H. G. Sinith, B.Sc., F.G.S., in illustration of his paper. Sapphire-bearing xenoliths, from a Tertiary sill in the Island of Mull, were exhibited on behalf of H.M. Geological Survey. ~ May 24th, 1916. Dr. AufFrep Harker, F.R.S., President, in the Chair. The List of Donations to the Library was read. Dr. A. Smrru Woopwaxp, F.R.S., V.P.G-S., exhibited De- vonian fish-remains from Australia and the Antarctic Ixxx PROCEEDINGS OF THE GEOLOGICAL SOCIETY. | vol. Ixxu, Regions, and discussed our present knowledge of the Devonian fish-fauna of the Southern Hemisphere. So far as is known, there are no strange elements in this fauna, and the remains Hiccomerad closely resemble those met within the Northern Hemisphere. Even the rocks are very similar to those containing the corresponding fossils in the Northern Hemisphere. There is, as yet. no satisfactory evidence of the basal Devonian fish-fauna such as occurs in the Downtonian of England and Scotland; but both Lower and Upper Devonian forms occur in Victoria and New South Wales (Australia ). A Coccostean related to Phlyctenaspis trom Gippsland (Victoria ), and another related to Macropetalichthys from Goodra Vale (New South Wales), may be regarded as Lower or Middle Devonian ; typical plates of Bothr iolepis from the Harvey Range (New South Wales) indicate an Upper Devonian fauna. The fish-remains obtained by the ‘ Discovery’ Expedition in Granite Harbour (Antarctica) comprise Bothriolepis, another Ostracoderm related to Byssacanthus, Acanthodian scales, Selachian dermal tubercles, a Coceostean, scales of Osteolepidie, and scales of a very small Paloniscid. They must be regarded as Upper Devonian. Dr. Smith Woodward expressed his indebtedness to the Govern- ment Geologist of New South Wales for the loan of the Australian specimens exhibited. Mr. R. Butten Newton exhibited some so-called Orbitoidal Limestones from Dutch New Guinea, the microscopical structures of which were shown by lantern-illustrations. The specimens were collected by Dr. A. F. R. Wollaston during his expedition to that country in 1912-13, on the snow-line of Mount Carstensz at a height of 14,200 feet, this mountain forming the highest elevation mee New Guinea, with an altitude of rear 16,000 feet. The foraminiferal organisms determined in this material included five species of Lepidocyclina (sumatrensis, martine, neodispansa, murrayana, and cf. tnsule-natalis), Amphi- stegina, Carpenteria, Cycloclypeus (ct. orbitoideus), ete.; the marine alga or nullipore, Lithothamnium, was also largely repre- sented. This assemblage compares favourably with that which characterizes rocks of similar age in other Pacific regions, such as Christmas Island (Indian Ocean), Formosa, the Philippines, Borneo, Celebes, Sumatra, Nias, Timor, and Australia, besides indicating a Miocene origin. It was pointed out that the genus Orbitoides of A. VOrbigny had been restricted by Schlumberger (relying on the researches of Giimbel, Verbeek, and others) “to species having rhomboidal equatorial chambers and belonging only to Cretaceous times ; species furnished with rectangular chambers, and recognized as Or thophr agmina of Munier- Chains were limited to the Hlpcene and Oligocene formations ; while Giimbel’s genus Lepidocyclina, with rounded or hexagonal chambers, included species of Miocene and later age. As the result of a study of species from Borneo and the Philippines, Prof. Douvillé had proposed to divide Lepidocyclina part 1] PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Ixxx1 into two sections—Hulepidina and Nephrolepidina: the first including forms of generally large size, recognized as Aquitanian ; the second for those of small dimensions, regarded as Burdigalian— these geological divisions representing the oldest stages of the Miocene System. This distinction, however, was not applicable to the New Guinea limestones nor to corresponding rocks from Christmas Island, as both large and small species occurred in association ; it was, therefore, suggested that the age of the New Guinea material might be referable to the later part of the Aquitanian. Several writers have already written on rather similar limestones from various parts of New Guinea, although we are indebted to Dr. K. Martin for the first announcement in connexion therewith: he reported the discovery in 1881 of Lepidocycline organisms from rocks found in the northern and south-western districts of the country (Geelvink Bay, islands of Kei, Aru, etc.), which he attributed to the older Miocene. ‘The same author also referred to the occurrence of similar organisms in Mount Wil- helmina, obtained by Dr. Lorentz, beneath which the Alveolina Limestone was identified, proving the existence of Hocene rocks. The Cycloclypeus remains, which are of frequent occurrence in the present material, bear a strong resemblance to Prof. Douvillé’s new genus and species from the Miocene of Borneo, known as Spiroclypeus orbitoideus. This genus was stated to have all the characters of Cyeloclypeus, but differs from it in the possession of superficial chamberlets in the shelly layers of the central region. Recent investigations, carried out by the speaker, had proved the presence of this character (hitherto unrecognized) in recent forms of Cycloclypeus from Funafuti; hence the retention of Spiro- clypeus now appeared to be unnecessary. A full report on this New Guinea collection by the speaker had been lately published, entitled: ‘Notes on some Organic Limestones, &c. collected by the Wollaston Expedition in Dutch New Guinea,’—forming No. 20 of a series of reports, and it was to be included in vol. ii of those reports. _ By permission of the Secretary of State for the Colonies, Mr. J. F. N. Green exhibited a manuscript report by Mr. J. H. Haglesome, C.M.G., on the Udi Colliery in Nigeria, containing a number of photographs and plans, and made some observations on the economic importance of this development in Equatorial Africa. There were placed with it for comparison two photo- graphs of the outcrop of manganese-ore in the Gold Coast, also discovered by Mr. A. E. Kitson, F.G.S. VOL. LXXEt. f lxxxil PROCEEDINGS OF THE GEOLOGICAL society. {[vol. lxxu, June 7th, 1916. Dr. AuFRED HarxkeEr, F.R.S., President, in the Chair. The List of Donations to the Library was read. Dr, F. L. Kircuty, M.A., F.G-.8., exhibited a representative set of Mesozoic fossils obtained from deep borings and pit-sinkings in Kent. The specimens were selected from the collections of the Geological Survey, by permission of the Director. The life of the successive zones present in the various sections was illustrated by the arrangement of the specimens in sequence, the series comprising a time-range from the Lower Lias up to the top of the Lower Greensand. The section revealed in the Brabourne Boring is remarkable for the range of formations found there in superposition, and may be regarded as the type-section for the study of the hidden Mesozoic succession in Kent. Specimens were exhibited from that locality and from the shafts at Dover, but these were supplemented by materials obtained more recently from various borings situated east of a line drawn from Folkestone to Canterbury. The speaker described the principal characters of the faunas* as developed in this area, and made incidental references to the nature and distribution of some of the associated rock-types. At some horizons, the molluscan assemblage assumes a particular aspect, by reason of the preponderance of species fitted for life amidst the special conditions of deposition. On the other hand, there are evolutionary phases which recur repeatedly with considerable uni- formity, and seem to arise independently of immediate surrounding conditions. Such are illustrated by the degenerative changes shown to occur in many ammonites, and by some of the forms repeatedly assumed by the members of separate series among Mesozoic oysters. The evidence of ammonites, so important for the purpose of zonal determination, is frequently forthcoming at these localities in Kent, even in borings of narrow diameter; but it is often necessary to rely entirely upon the aid afforded by the more abundant bivalves. Many of these, although belonging to un- described species, are found to have a limited vertical range, and by their distr ibution thr oughout this area, as well as farther afield, prove of much service in these correlation-studies. Specimens of many undescribed species have come to light, as well as others which are known from their occurrence in Continental localities, though not previously recorded in this country. A small series of Jurassic Cephalopoda from Kachpur (Russia), collected by the late G. ¥. Harris, F.G.S., at the time when the International Geological Congress met at Petrograd (1897), was exhibited by James Francis, F.G.S. part 1] PROCEEDINGS OF THE GEOLOGICAL SOCIETY. Ixxxil June 28th, 1916. Dr, ALFRED Harker, F.R.S., President, in the Chair. Herbert Theodore Mayo, B.A., care of Port Master, Rawal Pindi (India) ; and William Guy de Gruchy Warren, Capel Issa, Manordilo (Carmarthenshire), were elected Fellows of the Society. The List of Donations to the Library was read. The following communications were read :— . ‘Ona New Species of Hdestus from the Upper Carboniferous of Yorkshire. By A. Smith Woodward, LL.D., F.R.S., V.P.G-S. With a Geological Appendix by John Pringle, EGS. Paine Tertiary Volcanic Rocks of Mozambique.’ By Arthur Holmes, D.I.C., B.Se., A.R.C.Se., F.G.S. Dr. A. Srranan, F.R.S., exhibited cores from borings in Kent, showing pebbles of coal embedded in Coal- Measure sandstones. With the coal-pebbles occurred a few partly-rounded fragments of chert, and in one of these radiolaria had been identified by Dr. G. J. Hinde. The chert resembled that which had been described from Lower Carboniferous rocks elsewhere. Its occurrence suggested that the sequence of strata had been similar in South Wales and Kent, and, taken in con- nexion with the piping of the limestone-surface at Ebbsfleet and the absence of Millstone Grit in Kent, tended to confirm the view that there is unconformity between the Coal Measures and the Carboniferous Limestone in that county. Mr. F. P. Mennetr exhibited a geological sketch- map of the northern margin of Dartmoor. He said that the central part of Devon was to a great extent a terra incognita ; but, as regarded the fringe of altered Carbon- iferous rocks along the northern border of the Dartmoor granite, he had been led, in the course of observations originally concerned with the petrology alone, to the’ conclusion that it might prove possible to establish a definite order of succession. This was rendered feasible by the occurrence of some well-characterized bands of rock, especially limestones and tuffs, which were exposed in every good river-section. It was true that almost everywhere overfolds, sometimes accompanied by thrusts, were to be de- tected, and tended to make the observer somewhat doubtful of his ground. Nevertheless, it seemed impossible to escape the conclusion that, as one approached the granite from the north, continuously older rocks were met with, and the extremely Ixxxiv PROCEEDINGS OF THE GEOLOGICAL society. [ vol. lxxil, continuous character of some of the beds seemed to show that, despite all minor disturbances, the general sequence could be trusted. The comparison of the different lines of section leading up to Dartmoor showed them to be strikingly similar. The granite was, moreover, intruded all along at precisely the same horizon, and its direct offshoots never reached into the lower of the two important bands of limestone, but were confined to the altered shales at the bottom of the series, which afforded, where fresh, good examples of andalusite-hornfels. The series, which extends from south of Sourton to Drewsteignton, and perhaps right round to Doddiscombeleigh, appears clearly older than the shales which have been so carefully searched for fossils in the Exeter region by Mr. F. J. Collins. These last are considered to be of Pendleside age, and nowhere contain any traces of limestone. The probability is thus indicated that the distinctly calcareous series under consideration may represent part of the Carboniferous Limestone. It may be noted that, although a number of bands of epidiorite representing intrusions of dolerite occur roughly parallel to the strike of the sediments, the contemporaneous rocks are never of such basic character. The main band of tuff stretches from Lake, near Bridestowe, to beyond Sticklepath, and, of the numerous well-preserved rock-fragments that it contains, most are of rhyo- litic or trachytie character, with some which represent aitered andesites. Lantern-slides and specimens of Hdestws were exhibited by A. Smith Woodward, LL.D., F.R.S., V.P.G.S., in illustration of his paper. Lantern-slides, microscope-slides, and rock-specimens were ex- hibited by Arthur Holmes, D.I.C., B.Sc., A.R.C.Se., F.G.S., in illustration of his paper. THE QUARTERLY JOURNAL OF THE GEOLOGICAL SOCIETY OF LONDON, Vou. LXXII FoR 1916. 1. On a New Species of Epvesrus from the Uppnr Carpon- IFEROUS of YorKsuIRE. By Artuur Smira Woopwarp, LL.D., F.R.S., V.P.G.8. With a GrotogicaL APPENDIX by Joun Prinexe, F.G.S. (Read June 28th, 1916.) [Puate I. ] THe remarkable Upper Paleozoic fossil Hdestvs has already been proved to represent a row of symphysial teeth of an Elasmobranch fish, but it has hitherto been found only once in direct association with portions of jaws.! A second specimen, still more instructive, has now been obtained by the Geological Survey from the upper part of the Millstone Grit at Brockholes, near Huddersfield, and Iam indebted to Dr. Aubrey Strahan and Dr. F. L. Kitchin for the opportunity of studying it. The circumstances of its discovery are described in the Appendix by Mr. John Pringle. The new fossil, shown of two-thirds the natural size in Pl. I, fig. 1, displays a single example of Hdestus, with a detached dental crown and another fragment of the same form, near the tapering ends of a symmetrical pair of cartilages (¢) which evidently repre- sent a jaw. Whether they are upper or lower is uncertain, on 1 O. P. Hay, ‘ On an Important Specimen of Hdestus ; with Description of a New Species, Hdestus mirus’ Proc. U.S. Nat. Mus. vol. xlii (1912) pp. 31-38 & pls. i-ii. Q. J. G.S. No. 285. B 2 DR. A. SMITH WOODWARD ON [vol. Ixxu, account of the shortness of the portions preserved; but, as the anterior ends suddenly begin to taper and eventually become very slender, they are probably the ~pterygo-quadrates of the upper jaw. The cartilage is well calcified in very small tesserz, and, as shown both by the portions of jaws themselves and by remains in front of the fossil, the calcification penetrates more deeply than is usual in recent Elasmobranchs. The best-preserved outer surface of the cartilage, on the side of the specimen not shown in the figure, is shghtly marked with scattered fine pittings, such as have already been deseribed in Hdestus mirus.} The row of fused symphysial teeth (s) of the Hdestus type is bilaterally symmetrical, and arched almost in the form of a semi- circle. The eight teeth of which it is composed do not increase much in size backwards, and’ the depth of each crown slightly exceeds twice that of its root. The crown is laterally compressed and triangular in shape, much deeper than wide, with the sharp anterior and posterior edges nearly straight, and the postero-inferior angles much produced into slender, pointed extensions, which clasp the next following tooth about as far as the posterior edge of its crown. The superticial gano-dentine is smooth, but the base of the crown is impressed irregularly with a few large vertical plications or flutings, which are deepest in the anterior half and gradually disappear in the posterior extension. There is also a faint longi- tudinal median ridge on part of this extension. The anterior and posterior edges of the crown are coarsely serrated, the serrations (Pl. I, fig. 7) being about 30-in number on each edge, bluntly rounded (not crenulated), and those in the apical portion inclined upwards. In most of the teeth, however, the serrations are much worn by an apparently single row of opposing teeth. They are especially worn in the foremost tooth, irregularly on its anterior edge, most deeply in the Jower half of its posterior edge. The wear is nearly similar in the second, third, and fourth teeth; while in the fifth and sixth teeth a large worn hollow (w) is conspicuous in the upper half of the posterior edge. The apical portion of the seventh tooth is unfortunately lost, but the eighth tooth 1s complete and unworn, displaying all the serrations as previously described. The separate tooth (¢) behind is crushed at the base, and thus was probably not completely developed ; but the crown exhibits well its smooth face and unworn serrated edges. It may be either a ninth tooth of the series displayed or one of the opposing dentition, which is otherwise represented only by a broken fragment of one tooth in the edge of the shaly matrix of the fossil. The root of the fused symphysial teeth is well seen in front, where it has been extricated from the matrix, but it becomes vague behind where the teeth were in process of formation. In side view (Pl. I, fig. 1) the limits of the successive components are only just distinguishable; but in lower view (Pl. I, fig. 2) 1 OQ. P. Hay, Proc. U.S. Nat. Mus. vol. xlii (1912) p. 32. Ee eS eel cee ee part 1] A NEW SPECIES OF EDESTUS. 3 the divisions between the roots of the three anterior teeth are well marked by lines of calcite. In each tooth the root is much shorter than is usual in Hdestws, scarcely extending backwards beyond the hinder production of the base of the crown. As shown by the foremost tooth (figs. 2 & 3), the anterior margin of the root slopes downwards and backwards, and is compressed to an edge nearly as sharp as that of the crown. Its truncated lower face (fig. 2) is excavated into a triangular hollow, so that the lower face of the arch of clasping teeth is impressed by a wide longi- tudinal groove, of which the shape and depth are well seen in cross- section (fig.4). Sections prove that both root and crown are solid, consisting of the usual vascular dentine of rather open texture. Irregularly scattered over the shale below the arch of Hdestus are the more or less broken remains of comparatively small Orodont teeth (0), of the form commonly described as Campodus or Agassiz- odus. Some of these teeth are much extended laterally, witha low central cusp (Pl. I, fig. 8); while others appear to have had little lateral extension, but a relatively-large and elevated central cusp (figs. 9 &10). The longitudinal ridge on the summit of the crown and all the vertical buttresses on its outer and inner face are sharp and simply serrated; the superficial gano-dentine is otherwise smooth. ‘The serrated buttress of the central cusp is especially conspicuous on its outer face; but both this and the lateral but- tresses, which are widely spaced in from one to four pairs, are much more prominent on the outer than on the inner face. The com- pressed root is of very open texture, and seems to be inclined slightly inwards. The fossil itself affords no definite proof that the Orodont teeth thus described belong to the same jaw as the Mdestus, and their very small size seems at first to make their connexion improbable. Teeth of the same type, however, have already been found in association with Hdestus mirus1; and one jaw of Campodus has been described, in which a single arched series of high-crowned symphysial teeth is relatively enormous.? Moreover, it will be noticed that in the new fossil from Yorkshire there is no difficulty in regarding the teeth of the HMdestus type as an extreme modifi- cation of those of the Campodus type which oceur with them. In the new Hdestus the central cusp of Campodus has become exces- sively enlarged and laterally compressed, while the lateral extensions of the tooth are reduced and attenuated, and their sharp crest is represented merely by a faint longitudinal ridge. The irregular basal plications or flutings in the Hdestus are the last remnants of the anterior (outer) lateral buttresses in Campodus. 'The first approach, indeed, towards this extreme modification occurs in the symphysial teeth of the typical jaw of Campodus (Agassizodus) itself, as shown in the accompanying text-figure (p. 4). Here the 1 ©. P. Hay, Proc. U.S. Nat. Mus. vol. xlii (1912) p. 36. 2 ©. R. Hastman, ‘On the Nature of Hdestus & Related Forms’ Bull, Mus. Comp. Zool. Harvard Coll. vol. xxxix (1902) pp. 55-77 & pls. i-iv. B2 4, DR. A. SMITH WOODWARD ON [ vol. Ixxii, central cusp is much enlarged and elevated, and begins to be some- what laterally compressed, with sharp anterior and posterior (outer and inner) edges; but the lateral extensions of the tooth, although attenuated, still retain the characteristic anterior (outer) buttresses, the apical ridge throughout their length, and the typical root (8). In the new specimen of Hdestus the same enlargement of the central cusp with reduction of the lateral extensions has progressed so far that the central feature predominates immensely, while the diminutive lateral parts curve backwards along the modified root to clasp the next successional tooth (c). In typical Hdestus the tooth is almost entirely a laterally-compressed central cusp, while the clasping root is excessively enlarged (»). Diagrammatic sketches of half of a lateral tooth of Campodus variabilis, wpper view (A), and-symphysial teeth of Campodus variabilis (B), Edestus newton1 (c), avd Kdestus minor (»), left side view, to show progressive modification. A. B. Gc ID. [m = longitudinal ridge. } There seems thus to be no doubt that the small teeth of the Campodus type occurring with the new fossil really belong to it, and a question arises as to how the species represented by this jaw shall be named. Ca ampodus (1844) is an older generic term than Edestus (1856), but it is evident that the teeth to which its definition applies belong to more than one genus: for the sym- physial teeth of Cantos variabilis, as made ‘known by Eastman,! must be regarded as generically dienes from those of the specimen now described, ile the former must have been arranged as a single row only i in one jaw, and as a paired row in the opposing jaw; whereas in the Yorkshire teeth the surfaces of wear show that there must have been an unpaired median row in each jaw. As. the latter arrangement has already been proved by Hay to charac- terise Hdestus, “and as the symphysial teeth in the new fossil merely differ from those of the typical Hdestus in the relativ ely- small size and extension of the root, 1 propose to regard it as 1 Bull. Mus. Comp. Zool. Harvard Coll. vol. xxxix (1902) p. 58. Quart. Journ. Geo. Soc. Vor. LXXIl, Pt. I. Kqsaq'oj\09 ‘asouwag ‘Aaou ‘ds ‘INOLMA4N Siees = de {9p ‘PUeEMPOoN ‘WD part 1] A-NEW SPECIES OF EDESTUS. 5 representing an extremely generalized species of KHdestus. It ditfers from all known forms, not only in the shape and proportions ot the root, but also in the shape of the crown and the significant flutings and faint lateral ridge at its base. It represents, in fact, a very distinct species, which may be appropriately named Hdestus newtont, atter Mr. H. T. Newton, F.R.S., who first recorded the occurrence of Hdestus in British Carboniferous rocks.! EXPLANATION OF .PLATE I. Edestus newtoni, sp. nov.; anterior ends of cartilages of jaw (c), median symphysial dentition (s), a detached symphysial tooth (¢), and small seattered Orodont teeth (0).— Millstone Grit; Brockholes, near Huddersfield, Yorkshire. (Museum of Practical Geology, London.) [ Figs. 1-6 are two-thirds of the natural size, figs. 7-10 are twice the natural size. | Fig. 1. The whole fossil. w=worn surface on edge of tooth. 2. Anterior end of symphysial dentition, lower view. 3. Front view of the same. 4, Vertical transverse section of root of third tooth of the same. Figs. 5 & 6. Transverse sections of second and eighth teeth of the same. Fig. 7. Serratfons of the border of the eighth tooth. 8. Left half of an elongated Orodont tooth (Campodus), front view. 9. Elevated Orodont tooth (Campodus), front view; and (9a) upper view. 10. Elevated Orodont tooth (Campodus), back view. APppENDIX.—NoreE on the WeELL-SECTION at Rock Mints and the Fauna associated with Epresrus. By JoHN PRINGLE, E.G.S. THe well at Rock Mills (Messrs. Joseph Sykes & Co.) is situated in the mill-grounds in the valley of the River Holme at Brock- holes, about 4 miles south of Huddersfield. According to the Geological Survey-map (Quarter-Sheet 88 S.H.), the alluvium in the valley at this pomt rests on shales which underlie the Rough Rock, the uppermost member of the Millstone Grit. The shales appear to vary considerably in thickness in this district. Green ? states that 100 feet of shales are present between the Rough Rock and Grit A about Holmfirth, which lies 2 miles south of Brockholes; while on Holestone Moor, west of Huddersfield, he estimates that they are 200 feet thick between the same horizons. Spencer? has shown that in the district embraced in the next sheet to the north (Quarter-Sheet 88 N.E.) the shales contain a marine fauna; and the discovery of Hdestws should stimulate the efforts of the local geologists to further work on this horizon. 1H. T. Newton, ‘On the Occurrence of Edestus in the Coal-Measures of Britain’ Q. J. G.S. vol. lx (1904) pp. 1-8 & pl. i. 2 A. H. Green & others, ‘The Geology of the Yorkshire Coalfield’ Mem. Geol. Surv. 1878, p. 59. 3 J. Spencer, ‘ Additional Notes on the Millstone Grit of the Parish of Halifax’ Trans. Manchester Geol. Soc. vol. xiii (1873-74) p. 109. or) A NEW SPECIES OF EDESTUS. [vol. lxxii, During the excavation of the well no record of the various strata passed through was made by the sinkers; but it appears, from notes supplied by Mr. J. R. Simpson, of Bale Honley, that the rocks were mainly shales, with some thin beds of sandstone. At the depth of 120 feet the specimen of Hdestus was obtained by Mr. H. H. Freer in a bed of rock—rather friable shale,—which also contained numerous fragments of marine shells. Among the specimens the following have been identified :—Postdoniella levis (Brown), Gastrioceras sp., Glyphioceras reticulatum (Phill.), and Orthoceras cf. aciculare Brown. At the depth of 142 feet was a dark-grey flagey micaceous shale, crowded with shells which appear identical with JJodiola transversa Hind. ‘This shale formed the roof of a 83-inch coal which overlies a hard grey sand- stone. The coal-seam probably corresponds to the thin coal which lies on the top of Grit A in the Huddersfield district. Water was reached at the depth of 163 feet, and the sinking was stopped after being carried down a few feet lower. Our thanks are due to Mr. J. R. Simpson, who brought the discovery of the specimen of Edestus to our notice, and for his kindness in placing his specimens of the associated fauna at the disposal of the Geological Survey. The specimen was presented to the Museum of Practical Geology by Mr. E. Crowther, Managing Director of Messrs. Joseph Sykes & Co. Discussion Dr. A. SprauvAN expressed his appreciation of the skill with which the fossil had been developed at the Natural History Museum. He wished also to take the opportunity of acknow- ledging the obligation under which geologists had been laid by My. EH. Cramtinge in placing this unique specimen in a National Museum, where it will be studied by specialists from all parts of the world. Mr. J. Prinaie remarked that it was seldom that so interest- ing a specimen came before the Society, and he congratulated the Fellows on having had the opportunity of hearing Dr. Smith Wood- ward’s lucid and illuminating account of this Eee known genus. He referred briefly to the eeolocical position of the specimen and to the associated fauna. He had had the opportunity of studying exainples of marine shells obtained from the shales which yielded the Edestus. These comprised species of Glyphiocer as, Gastrio- ceras, Orthoceras, and Posidoniella. Mr. EK. T. Nrewron asked Dr. Smith Woodward whether he’ could explain how it was that, while in Hdestws the base of the tooth seemed to be extended backwards from the crown, in Helicoprion it was directed forwards. Dr. A. Smirn Woopwarp agreed with the last speaker that it was difficult to explain the difference in the direction of extension of the tooth in Hdestus as compared with Helicoprion. The two forms seemed to have had a distinct origin. part1] FOSSILIFEROUS LIMESTONE FROM THE NORTH SEA. 7 2. On a FosstiifeRous Limestone from the NortH Sea.t By Ricnarp Burien Newton, F.G.S. (Read June 23rd, 1915.) [Puate II.] THRouGH the good services of Mr. R. W. Thomson, of the Fishery Office, Aberdeen, the British Museum has been placed in possession of two blocks of limestone, one of considerable size, crowded with the remains of marine shells, which had been obtained from the bed of the North Sea by the steam-trawler, the ‘ Procyon,’ com- manded by Captain Wood, some 80 miles east of Orkney or 100 miles north-east $ north of Buchan Ness. In order to acquire ton ther information as to this unique occur- rence, | communicated with Dr. A. W. Gibb, the Curator of the Geological Museum of Marischal College, Aberdeen, who kindly sent me some interesting particulars on the subject. He was familiar with the rock, having had a sample forwarded to him by Dr. Bowman, the scientific investigator on the North Sea Fisheries’ boat, the ‘Goldseeker, who from considerable experience had gained some knowledge of the physical characters of the floor of the North Sea. Dr. Bowman informed Dr. Gibb ‘that at the spot where it [the limestone] occurs, there is what seems to be a gorge or submerged channel of some kind, with much deeper water than in the adjoining sea, and he thinks that there is a considerable mass of the rock in all probability [in situ ?],as trawlers report that they frequently break their gear upon it.’ Although found beneath the sea, this limestone nowhere displays any unusual abrasion, its aspect being entirely that of a rock which might have been obtained from an ordinary quarry or land exposure. There is nothing in its appearance to suggest transportation by glacial agencies, and it can only be/surmised that deep down in the North Sea a considerable development of the rock may actually occur 7 situ. So far as can be ascertained at present, no similar limestone is known in either England or Scotland. The so-called ‘Crag Strata’ of Aberdeenshire, described by Jamieson as occur- ring beneath the Boulder Clay and consisting of stratified sand and gravel, have yielded marine mollusca of a northern type, mixed with a few Red Crag species, which Searles Wood regarded as of Red Crag age ?; although, according to Clement Reds 3 such beds resemble the ‘Bridlington Crag’ deposits of Yorkshire, and would therefore be of Pleistocene horizon. Mr. Reid further stated that there is no Pliocene in Scotland, the Red Crag shells of the glacial beds of Aberdeenshire having been probably derived from strata of that age lying beneath the North Sea. A very 1 Communicated by permission of the Trustees of the British Museum. 2 Q.J.G.S. vol. xvi (1860) p. 373. 3 ¢The Pliocene Deposits of Britain’ Mem. Geol. Surv. 1890, p. 8 MR. R. B. NEWTON ON A FOSSILIFEROUS _ [vol. Ixxu, similar molluscan fauna to that occurring in the Scottish deposits has been found in the rocks of Iceland which Gwyn Jeffreys con- sidered as Post-Tertiary and Searles Wood as ‘not later than Middle Red Crag.’ ! Reference may also be made to Jamieson’s account of some semi-fossil shells of Arctic character, dredged by Robert Dawson from glacial beds underneath the sea, off the Aberdeenshire coast, from 3 to 8 miles away from land, at a depth varying from 30 to 45 fathoms. Similar shells had even been brought up by fishermen’s lines at a distance of 30 miles from the coast, and none of the species were thought to occur alive in that district.? Besides this, we have Alfred Tylor’s account? of fossil marine Arctic shells dredged off the Shetland Islands by Gwyn Jeffreys in about 90 fathoms of water—the species being stated to occur fossil in Sweden and still living in extreme Arctic seas. These records of fairly late fossiliferous deposits beneath the sea are of great interest, but scarcely assist us in understanding the history of the present rock. This North Sea limestone is of a dark-greyish colour, and is built up almost entirely of shell-remains belonging chiefly to the Pelecypoda. Microscopical tests have proved it to be highly siliceous, thus accounting for its great hardness and tenacity, which have frequently militated against complete development of some of the organisms, especially when the attempt was made to expose in- ternal @haracters: otherwise, speaking generally, the fossils are well preserved, and many of their finer external markings are perfectly displayed. It has been possible to determine 23 different forms of mollusea, ten being Gastropoda, while the remainder belong to the Pelecypoda, a new species being included in that group. If the material had been more dismantled, some additional forms might have been available for study, although in that case the larger block would have diminished in value as a museum specimen. The shells are essentially of southern character, although it is interesting to mention that one of the commonest forms is the large Cyprina islandica, which is so well known as existing im boreal seas and around our British coasts, but has never been recorded from the Mediterranean: its origin, however, is to be found in the Vindobonian stage of the Miocene Period. I have mentioned this species incidentally, because it frequently occurs in the Scottish and Icelandic deposits previously mentioned, which I consider of much later age than the present limestone. Occasional small brown masses of woody structure* occur in Q.J.G.S. vol. xli (1885) p. 96. Ibid. vol. xxi (1865) p. 200. Ibid. vol. xxv (1869) p. 8. The occurrence of drift-wood in the Coralline Crag near Sudbourn Church was mentioned by Clement Reid (- The Pliocene Deposits of Britain’ Mem. Geol. Surv. 1890, p. 28) ; but, so far as can be ascertained, no such material has been systematically described. em Ww Mw part 1] LIMESTONE FROM THE NORTH SEA. 9 contact with the limestone, but of very much softer character, which from a preliminary examination appear to be of a coniferous nature, although requiring further study before a more accurate deter- mination can be attempted. Out of the 23 determined molluses of this limestone 18 (or nearly 80 per cent.) trace their origin from the Vindobonian stage of the Miocene Period, 10 (or about 40 per cent.) may be regarded as extinct, whereas 12 (or 50 per cent.) are found living in recent seas. The extinct species include :— Streptochetus sexcostatus. Ficus | Pyrula] simplex. Ringiculella ventricosa. Yoldia oblongoides. Levicardim fragile. Sinodia tertiaria, sp. nov. Tellina benedent. Panopea menardi. Arcoperna sericea. Nucula levigata. The existing species are found chiefly in British and Mediter- yanean seas, although Spiswla ovalis and Cyprina islandica belong to more boreal waters, and are never found so far south as the Mediterranean. ‘This list embraces :— Nucula nucleus. Dentilucina borealis. Thyasira flexuosa. Cyprina islandica. TIsocardia humana (= cor). Spisula ovalis. Ranella gigantea. Turritella communis. Aporrhais pespelicant. Naticina alder. Actzxon tornatilis. < Scaphander lignarius. The majority of the species recognized are fairly evenly repre- sented in both the Coralline and the Red Crag deposits, 15 occurring in the former and about 12 in the latter. Among them the genus Yoldia, belonging at the present day entirely to Arctic seas, is well distributed through the limestone, specimens having been identified as Wood’s oblongoides, which is known from the Red Crag, Norwich Crag, and later deposits of East Angha, although never yet recorded from the Coralline Crag. Another species, however, of this genus, but not occurring in the present limestone, is Y. semistriata, which is restricted to Coralline Crag beds, and therefore forms an illustration of the Arctic genus Yoldia having been associated with a more southern fauna than prevailed during Red Crag or Jater geological times as represented in England. Further support is given to the southern aspect of the fauna by the occurrence in the limestone of a new Dosinitorm shell, Sznodza tertiaria, which presents affinities that are only to be found in the Indian Ocean and regions of Southern Asia. The comparatively large number of extinet species is greatly in favour of the North- Sea rock being of older horizon than the Red Crag, as among them are Arcoperna sericea, Tellina benedeni and Panopea menardi (= gentilis), which are not known of later age in this country than the Coralline Crag, although Zellina benedeni occurs as well in the Lenham Beds and Panopea menardi in the Boxstones. In con- nexion with the antiquity of this fauna, mention should also be made of the occurrence of Streptochetus sexcostatus and Ficus simplex, 10 MR. R. B. NEWION ON A FOSSILIFEROUS __ [vol. Ixxu, Streptochetus sexcostatus,a Fusiform shell, belongs to the Upper Miocene deposits of Northern Germany and Belgium (Anversian Beds), which are variously regarded as Messinian, Mio-Pliocene, — or the Sarmatian-Pontian Series ; in Holland the species has been recorded as of Vindobonian age. ‘his is its first acknowledgment as a British fossil, although remains of it are found in the Lenham Beds, a fact, however, published since the reading of this paper.! Ficus | Pyrula| simplex occurs only in the Messinian Beds of Northern Germany, never having been previously determined from British rocks. In addition to these species, Gottsche® has recorded from the same horizon and country, Scaphander lignarius, Arcoperna sericea, Levicardium fragile, and Dentilucina borealis, all of which form part of the fauna of the North-Sea rock. On the assumption, therefore, that the limestone belongs to some part of the Crag system, there appears to be ample evidence that it is of older age than the Red Crag, on account both of the total absence of Arctic species, and of the comparatively large number of extinct forms which have been recognized in it. It is more likely, therefore, to be of Coralline Crag age: for, in addition to the fact that the fauna agrees in the main with the molluscan fauna of that period, there are some lithological features which may be worthy of mention. According to the constitution of the Coralline Crag deposits of East Anglia, as explained by Prestwich,? the lower beds of the series contain ‘irregular seams of shelly mestone’; in more recent years Mr. F. W. Harmer has also referred to a similar formation in the same deposits as ‘tabular-layers of limestone, very hard, and difficult to penetrate,’ which were discovered in a boring at. Gedgrave Hall.“ I have not seen specimens of these limestones, although it would be desirable to compare them with the present rock. It does seem possible, however, that this North-Sea material may represent the lower portion of the Coralline Crag, and has, therefore, no connexion with Red Crag deposits, which are through- out of a sandy nature and furnish no evidence of intercalated calcareous beds. Considering the large number of extinct molluscan species fur- nished by the limestone, namely, 10 out of a total of 23 (or about 40 per cent. ), as against Searles Wood’s computation of the Coralline Crag mollusca made in 1874, that out of 391 species 142 were not now living® (thus representing about 36 per cent. of extinct forms), one may infer that, if the North-Sea rock really belongs to 1 R. B. Newton, ‘ On the Conchological Features of the Lenham Sandstones of Kent & their Stratigraphical Importance’ Journal of Conchology, vol. xv (1916) p. 74. 2 Verhandl. Ver. Naturwissensch. Unter. Hamburg, 1876—78, pp. 182-85. 3 Q. J.G.S. vol. xxvii (1871) pp. 123, 125. 4 Ibid. vol. liv (1898) p. 336. 5 See Clement Reid, ‘The Pliocene Deposits of Britain’ Mem. Geol. Surv. 1890, p. 38. part 1| LIMESTONE FROM THE NORTH SEA. eh the Coralline Crag, that formation is probably of earlier age than it is usually supposed to be at the present day; and it is interesting in this connexion to note that, in his earlier researches on the Crag mollusca, Searles Wood was of opinion that the Coralline Crag deposits represented the remains of the Miocene Period in Hngland—a view, however, which he afterwards relinquished ! in favour of regarding them as early Pliocene. Clement Reid placed them in his ‘Older Phlocene” group in association with the St. Erth Beds, the Lenham Beds, and the Boxstones,? whereas Mr. F. W. Harmer regards them as the base of the ‘ Newer Pliocene’ Series and younger than the Lenham and Boxstone Beds, which he restricts to the ‘Older Pliocene’ division as represented in this country.? It should be also stated that long before Reid and Harmer published their views on this subject, Mayer-Eymar had estab- lished the group-name ‘ Messinian’ for the later Miocene deposits of Europe, including in it the Lower Crag because a great number of its species were found in the Helvetian stage of the Miocene.* From my own studies of these mollusca, and especially those of the Lenham sandstones, I believe that no great disparity of age separates the so-called ‘ Older Plocene’ deposits. The faunas may differ somewhat in detail according to the special environments which governed their existence; but they all exhibit a strong southern character, and in other ways appear to resemble more the later Miocene life of Europe than that of the succeeding period. Judging entirely, therefore, from paleontological evidence, and apart from any physical considerations, I am inclined to regard this limestone as of Coralline Crag age. List of Molluscan Species. GASTROPODA. Family Lampusip 2. RaNELLA GrGanrEA Lamarck. (PI. I, fig. 1.) Ranella gigantea Lamarck, ‘ Hist. Nat. Anim. sans Vert.’ 1822, vol. vii, p. 160. Ranella reticularis Hoernes, Abhandl. K.IK. Geol. Reichsanst. 1853, vol. iii, p. 211 & pl. xxi, figs. 1-2. Remarks.——A small example, represented by a cavity in the limestone which, with the aid of a wax squeeze, yields the sculp- ture characters of this species. Distribution.—Vindobonian (Austria); Plaisancian (France and Italy) ; Recent (Mediterranean ). 1 Monogr. Pal. Soc. 1848, p. v (Introduction) ; ibid. 1857, pp. 301, 302. 2 «The Pliocene Deposits of Britain’ Mem. Geol. Surv. 1890, p. 2. 3 * Geology in the Field’ Jubilee Vol. Proc. Geol. Assoc. 1909, p. 90. 4 * Catal. Syst. Foss. Tert. Mus. Zurich’ pt. 2 (1867) p. 13. \ Ses eh LOSE RS *‘(Uesleg 03 1eqTRIGQLD) UBOUBIIOJIpPSP, puR YS 4 x *purlaoy —ULIUBIMOITpPa PT PUB FSI x SCH ESC “OGLE T pure uvauvitezyIpeyy ‘snttes “UVOUIE}IpPaT, PUB YSTLIG 4 See “UBAUBILOJ IPI x M/A) ere) uy) 2 Ale 0 in 99 PB ee | ee | es S| = Sh ai eal ‘SVEQ INDOaYy 8 4 Pi" : ie “SBI OULT[BIOD | Sped WH 4S | | | x é ae else al Se ce | = | a ao | oF lo} s z eae w| 2 @ ee “UBIySal(y | x tas > 2 8 Qs. i=] ss a= ism {oe) pe Bos mie ao Sse) = nai {9°} lex} ae iY) j=] *(UBLUOIIOT, =) UBIUOpery | x x x x x x =| oe Lei & Bele 0g a5] » of] & EB) = =, 2 | = a an (a2) (2, < (q>} ot Ste pe) =} | | ‘VAS HIMON AHL NOW ANOLSHWIT SHOUAMITISSO HY Ha —— NI GNQOd VOSOTIOP-, AHL FO MOVES GNV AW, NI NOWOAATELSt qd ‘uvlueqinbd y | *(Aqtemog *D ep *f) psooniquan pijanorbury (Soq10,7) 20a) U20290 AT *(snauury) iunorjadsad sipy.t.cody ‘ds w1oydouary OSSIY S2wmuU OD Y))A2RVMANT, (yortseg) wards snoyy *(qorsog) sngpysoavas srjayooj.dai4g yoinmury vagunhrh nppownay | “VadOdOULSV) —_—<—<—<—$<$___ ——_—__—_+ —_—__ ‘Seloodg pusw Brlouey “UGIUBLIOFTpPoa TAL jou arial aca [serpy pay Ur eatyvatieq | [Seip pay ur eargearreq | ‘vISW UIeyyNOG HO seas 0} poejoraqysea snes y UBOUBLIOFI pa] PUB YS WRIUBII} -Ipey| JoU—[vo10g pue YSyIAg “UBOTIOULY 910 NT pue “uvouvtnoqrpayy “YS “UBOTIOULY Y910 NT pue “Uvouvireqtpayy “YStyTg “Aup quesoad otf} 48 S¥OQ DIJOTV 0} SSUOTEG snuUer) ‘pueluesry Jou “UvIAvUIpUBOG pus ‘uvauRtojipeyy “Sty URAUBILOIpaT, pues YySsTyTIG “UBIUBLIOII pap, pus sty eenfelaie (Aq.tamog a) $7)900 pynsidgy eget sekvysa(] wpuwueu vadoung ‘diopuegse (A : wy ISA Wapauag vur7197, mers “Aou “ds ‘y2.1029.107 wrypoUng (109 =) (snenuIry) vuvwuny vypuvo0sT (sneuuIy) vorpwnjsr vuudhD * (1ypo01g) aprbouf wnrpimar0daT “ (nsevIMoy) vsonwayl v.vusvhy, 7, |" (sheuUry) s27nav0g vutonzyuaq “(P0OM “A “S) Saprohuojgo vipj0 x Aq19Mog “f vpnbr0H) DynonAT vogunt (sneuUUly) snaponw vynanjT Jrseeee (UtoIg) veorLes vusadoouy Vdqaod ANN TAd *(sneuury) smApuby sapunydvogy (SheUUIyT) s27270U.102 UWODI0F 14 MR. R. B. NEWTON ON A FOSSILIFEROUS _ [vol. Ixxu, Family Fustp#. STREPTOCHETUS SEXcostatTusS (Beyrich). (PI. II, figs. 2 & 3.) Fusus sexcostatus Beyrich, Zeitschr. Deutsch. Geol. Gesellsch. 1856, vol. viii, p. 73 & pl. xxiv, fig. 2 Streptochetus sexcostatus Cossmann, ‘ Essais de Paléoconchologie Comparée’ 1901, pt. iv, p. 31. Remarks.—This species, I believe, is represented in the English Crag deposits by Wood’s Fusus waeli,) a form that I am unable to recognize as the Belgian shell which was originally described by Nyst from the Oligocene (Rupelian) rocks of that country. A perfect comparison, however, is impossible, as Wood’s type- specimens do not appear to be in the British Museum. It is worthy of notice that Wood directed attention to an artistic error in his figure 106, which exhibited denticulations on the inner surface of the labrum, a character which was neither present in the English shell nor in S. sexcostatus from the German Miocene. Distribution.—Vindobonian (Holland) ; Messinian (North Germany); Anversian (Belgium) ; Lenham Beds (Britain). Family Frcrp 2. Ficts sIMpLEx (Beyrich). IL, figs. 4 & 5.) Pyrula simplex Beyrich, Zeitschr. Deutsch. Geol. Gesellsch. 1854, vol. vi, p. 777 & pl. xviii (xv), fig. 3 Ficula simplex Gottsche, a Ver. Naturwissensch. Unter. Hamburg, 1878, vol. 11, p. 182. Remarks.—Bolten’s Ficus of 1798 antedates both Lamarck’s Pyrula and Swainson’s Ficula, all of which are founded on the same Linnean tvpe of Bulla jficus. Distribution.—Messinian (North-Western Germany). Family TURRITELLID &. "TURRITELLA COMMUNIS Risso. Turritella communis Risso, ‘ Histoire Naturelle de Europe Méridionale’ 1826, vol. iv, p. 106 & pl. iv, fig. 37. Turritella terebra J. de C. ‘Sowerby, “Mineral Conchology’ 1827, vol. vi, p. 126 & pl. dlxv, fig. 3, now Linnzeus. Distribution.—Vindobonian (France); Coralline Crag to Post-Phocene (Britain) ; Recent (Mediterranean and British Seas). Family XENOPHORID#. XENOPHORA sp, (PI. IT, fig. 6.) Distribution.—This genus is unknown in the East Anglian Crag deposits, although occurring in the Lenham Beds of Kent. 1 Mon. Pal. Soc., Crag Mollusca, Suppl. 2 (1879) p. 9. & pl. i, figs. 10 a-10c. part 1] LIMESTONE FROM THE NORTH SEA. 15 It belongs more particularly to the Mio-Pliocene strata of Italy, Belgium, Holland, and Austria, besides surviving in recent seas (Mediterranean and Pacific). Family APORRHAID#. APORRHAIS PESPELICANT (Linnezeus). Strombus pespelicant Linneus, ‘Systema Nature’ 10th ed. 1758, p. 742. Distribution.—Vindobonian (Austria, Italy); Plaisancian and Astian (Italy, France); Anversian and Sealdisian (Bel- gium); Lenham Beds (Britain); Coralline Crag to Post-Plocene (Britain); Recent (Mediterranean and British Seas). Family Natterp a. NATICINA ALDERI (EH. Forbes). (Pl. II, figs. 7 & 8.)- 5 WNatica alderi HK. Forbes, ‘ Malacologia Monensis’ 1888, p. 31 & pl. ii, figs. 6-7. Natica (Naticina) alderi Dollfus, C. R. Assoc. Frane. Av. Sci. (Cherbourg, 1905) 1906, p. 368. Distribution.—Vindobonian (Holland) ; Messinian (Ger- many); Redonian (France); Coralline and Norwich Crags to Post-Pliocene ( Britain) ; Recent (Mediterranean and British Seas). Family Rine@rcuLip ». RINGICULELLA VENTRICOSA (J.deC.Sowerby). (PI.IL, figs. 9 & 10.) Auricula ventricosa J. de’ C. Sowerby, ‘Mineral Conchology’ 1824, vol. v, p. 99 & pl. ececlxy, fig. 1 Ringiculella auriculata var. ventricosa, Sacco, ‘ Moll. Terz. Piemonte’ 1892, pt. 12, p. 25; ibid. 1904, pt. 30, p. 110 & pl. xxiv, figs. 25-26. Remarks.—The shorter, broader, and more inflated volutions, as mentioned by 8. V. Wood, are sufficient to separate this species from Brocchi’s &. buccinea, which belongs to the Italian Tertiaries and existing seas. ‘The genus is well known in the Mediterranean. Distribution.——Vindobonian to Astian (Italy); Scaldisian (Belgium); Coralline Crag to Norwich Crag (Britain ). Family ACT HONID A. ACTON TORNATILIS (Linnzus). Bulla tornatilis Linneus, ‘Systema Nature’ 10th ed. 1758, p. 728. Acteéon striatus J. de C. Sowerby, ‘Mineral Conchology’ 1824, vol. v, p. 87 & pl. ccecelx, fig. 2. Acteon tornatilis S. V. Wood, Monogr. Pal. Soc. (Crag Mollusca) 1848, p. 170 & pl. xix, fig. 5 _ Distribution.— Vindobonian (Holland); Plaisancian and Astian (Italy); Anversian and Scaldisian (Belgium); Lenham Beds (Britain) ; Coralline Crag to Post-Pliocene (Britain) ; Recent (Mediterranean and British Seas). 16 MR. R. B. NEWTON ON A FOSSILIFEROUS [Vol. lxxii, Family SCAPHANDRID#. SCAPHANDER LIGNARIUS (Linneeus). Bulla lignaria Linneus, ‘Systema Nature,’ 10th ed. 1758, p. 727. Scaphander lignarius Montfort, * Conchyl. Syst.’ 1810, vol. 11, pp. 334-35. Distribution.—Vindobonian (Austria, Holland, Italy); Redo- nian (France); Plaisancian and Astian (Italy); Bolderian to Scaldisian (Belgium); Lenham Beds (Britain); Coralline and Red Crags (Britain) ; Recent (Mediterranean and British Seas). PELECYPODA. Family Myrrnip. ARCOPERNA SERICEA (Bronn). Modiola sericea Bronn, ‘TItaliens Tertiar-Gebilde’ 1831, p. 112; Philippi, ‘Enum. Moll. Sicilia’ 1836, vol. i, p. 71 & pl. v, fig. 14. Arcoperna sericea Sacco, ‘ Moll. Terz. Piemonte’ 1898, pt. 25, p. 43 & pl. xii, figs. 8-9. Distribution.—Vindobonian (Austria, Italy); Messinian (North Germany); Anversian to Sealdisian (Belgium) ; Coralline Crag (Britain); Astian (Italy). Family NucunipZ. Nvctura nucievs (Linneus). Arca nucleus Linneus, ‘Systema Nature’ 10th ed. 1758, p. 695; ibid. (Gmelin) 1790, 13th ed. vol. i, pt. 6, p. 8314. ~ Distribution.—Vindobonian (Austria); Redonian (France) ; Anversian to Scaldisian (Belgium); Coralline Crag to Post- Phocene (Britain) ; Recent (Mediterranean and British Seas, ete. ). Noucuta tavieata J. Sowerby. (PI. II, fig. 11.) Nucula levigata J. Sowerby, ‘Mineral Conchology’ vol. ii, 1818, p. 207 & pl. excii, figs. 1-2. Remarks.—This large and tumid Nuculoid shell is of frequent occurrence in the idnestiorae. and in general features resembles Nucula cobboldie of J. Sowerby, bub is without the divaricate sculpture of that species. Distribution.—Vindobonian (Holland); Redonian (North- Western France); Anversian to Scaldisian (Belgium) ; Coralline and Red Crags (Britain) ; Astian (Holland). Family NucunanNipH. YoLpIA OBLONGOIDES (S. V. Wood). (PI. II, fig. 12.) Nucula oblongoides 8. V. Wood, Mag. Nat. Hist. (Charlesworth), 1840, n. s. vol. iv, p. 297 & pl. xiv, fig. 4. Leda myalis 8. V. Wood, Monogr. Pal. Soc. (Crag Mollusca) 1851, p. 90 & pl. x, fig. 17, non Couthouy. part 1] LIMESTONE FROM THE NORTH SEA. Le Remarks.—This species is of fairly abundant occurrence, and shows relationship both to Y. hyperborea Torell and tu Y. limatula Say, of modern seas (boreal), from which it chiefly differs in the possession of thicker and more convex valves. Distribution.—Red Crag to Post-Plocene (Britain). Family Locryip a. DENTILUCINA BOREALIS (Linneus). Venus borealis Linneus, ‘Systema Nature’ 12th ed. 1767, vol.i, pt.2, p. 1134. Dentilucina borealis Sacco, ‘Moll. Terz. Piemonte’ 1901, pt. 29, p. 80 & pl. xviul, figs. 23-26. Distribution.—Vindobonian (Holland, Italy, Austria); Re- donian (North-Western France); Messinian (North Germany) ; Anversian, Diestian, and Scaldisian (Belgium); Plaisancian and Astian (Italy, France) ; St. Erth Beds to Post-Pliocene (Britain) ; Recent (Mediterranean and British Seas, etc.). Family THyastRipZ. THYASIRA FLEXUOSA (Montagu). Venus sinuosa Donovan, ‘Nat. Hist. British Shells’ 1801, pl. xli, fig. 2, non Pennant, 1777. Tellina flexuosa Montagu, * Test. Britannica’ 1803, pt. 1, p. 72. Thyasira flecuosa (Leach MS.), Lamarck, ‘ Histoire Naturelle des Animaux sans Vertébres’ 1818, vol. v, p. 492. Distribution.—Redonian (North-Western France); Anversian to Scaldisian (Belgium); Coralline Crag, Chillesford Beds, and Post-Pliocene (Britain); Recent (Mediterranean and British Seas, etc.). Family Carpiip2. L&VICARDIUM FRAGILE (Brocchi). Cardium fragile Brocchi, ‘Conchiologia Fossile Subapennina’ 1814, vol. iu, p. 505 & pl. xiii, fig. 4. Levicardium norvegicum Sacco, ‘ Moll. Terz. Piemonte’ 1899, pt. 27, p. 51 & pl. xi, figs. 41-42. Distribution.—Vindobonian (Austria); Messinian (North- Western Germany); Plaisancian and Astian (Italy). Family Cyprinip#. CyprINA ISLANDICA (Linneus em. Miiller). Venus islandica Linneus, ‘Systema Nature ’ 1767, 12th ed. vol.i, pt. 2, p. 1131 (without reference to a type form); Miiller, ‘ Zool. Danice Prod.’ 1776, p. 246 (no. 2977). Cyprina islandica Nyst, ‘Coquilles & Polypiers Foss. Tert. Belg.’ Mém. Cour. Acad. Roy. Belg. 1843-43, vol. xvii, p. 146, pl. ix, fig. 1 & pl. xi, fig. 1. Q.J.G.S. No. 285. C 18 MR. R. B. NEWTON ON A FOSSILIFEROUS _ [vol. lxxu, Remarks.—The valves of this species are of abundant occur- rence, while some of the adult examples measure from 80 to 85 millimetres in length and height. Distribution.—Vindobonian (Holland); Anversian to Scal- disian (Belgium); Plaisancian and Astian (Italy); Boxstones, Lenham Beds, and Coralline Crag to Post-Pliocene (Britain) ; Recent (British and Boreal seas, non Mediterranean ). IsocaARDIA HUMANA (Linneus). (PI. II, fig. 13.) Cardium humanum Linneus, ‘Systema Nature’ 1758,-10th ed. p. 682. Chama cor Linneus, ‘Systema Nature ’ 1767, 12th ed. vol. i, pt. 2, p. 1187. Tsocardia cor Lamarck, ‘ Hist. Nat. Anim. sans Vert.’ 1819, vol. vi, pt.1, p. 31. Tsocardia humana Dall, Trans. Wagner Free Inst. Sci. Philadelphia (Tertiary Fauna of Florida), 1900, vol. iii, pt. 5, p. 1064, Distribution.—Tortonian (Italy, Holland); Plaisancian and Astian (Italy); Diestian and Scaldisian (Belgium) ; ; Boxstones, Lenham Beds, Coralline Crag, and derivative in Red Crag (Britain); Recent (Mediterranean and British Seas). Family Dosrn1ip 2. SINODIA TERTIARTA, sp. nov. (PI. I, figs. 14-16.) Deseription (Right valve).—Shell solid, ineequilateral, sub- orbicular, height in excess of length, well arched ; posterior region deep, sloping, marginal curvature extensive; dorso-anterior margin oblique, straight, long; ventral margin oradually ascending to unite with the expanded and rounded edge of the lower anterior side ; lunuloid region superficial, large, ovately- elongate, obscurely circumscribed ; hinge area massive, deep, bearing two divergent, thick, robust, cardinal teeth, also a third and smaller cardinal in front, of more or less laminate appearance ; ligamental furrow wide, long, and prominent; pallial sinus triangulate, moderately wide at the base, apex obtuse; internal margin smooth ; surface covered with well-marked, nearly equidistant, elevated growth-lines and numerous finer concentric strize occupying the intervening spaces, the whole crossed by obscure radial striations. Dimensions in millimetres.—Largest example, right valve. Length = 50; height = 53; diameter = 16. Remarks.—The more striking features of this shell include the anterior obliquity, the long and superficial lunuloid area, the nearly equidistant growth-lines, and the extensive curvature of the pos- terior margin. It appears to be allied to Chemnitz’s Venus excisa rather than to the other forms referred to this genus, which are of more distinctly trigonal contour. The shell is fairly common in the North-Sea limestone, although it is difficult to obtain specimens with perfect internal characters; one example, however, does exhibit a partial internal cast of a left valve, showing the presence of a well-marked, obtusely-pointed, and triangulate pallial sinus. It is proposed to name this shell Scnodia tertiaria. part 1] LIMESTONE FROM THE NORTH SEA. — 19 It is interesting to find among these fossils a new Dosiniform shell that can be referred to Jukes-Browne’s S’nodia,! which was established on the type of Dosinia trigona of Reeve, and also includes the Venus eaxcisa of Chemnitz. Sinodia is chiefly distinguishable from the true Dosinia by reason of its oblique and straight dorso-anterior margin, followed by an expanded anterior side. Jt is known only in the living state, being restricted to the waters of the Indian Ocean, particularly the coasts of Southern Asia. Family 'TELLINID ®. TELLINA BENEDENT Nyst & Westendorp. (PI. II, fig. 17.) Tellina zonaria Nyst, “Recherch. Cog. Foss. Anvers’ 1835, p. 4, now Basterot. Tellina benedeni Nyst & Westendorp, Bull. Acad. R. Sci. Bruxelles, 1839, vol. vi, pt. 2, pl. 11, fig. 5 bis & pl. iii, fig. 5, p. 399. =. fallax Beyrich, a manuscript name. Remarks.—This is by far the most abundant molluse in the limestone, and well agrees with the typical form from the Belgian Upper Tertiaries. Examples are chiefly preserved as natural casts which exhibit well-marked muscular and other impressions, as also the extensive pallial sinus with its elevated and obtuse summit. Distribution.—Anversian to Scaldisian (Belgium); Middle Pliocene (Holland); Lenham Beds and Red Crag—derivative (Britain). Family Saxicavip 2. PaNoPHA MENARDI Deshayes. Panopea faujasii Basterot, Mém. Soc. Hist. Nat. Paris, 1825, vol. ii, pt. 1, p. 95, non Ménard de la Groye. Panopea menardii Deshayes, ‘ Dict. Class. Hist. Nat.’ 1828, vol. xiii, p. 22. Panopea gentilis J. de C. Sowerby, ‘Mineral Conchology’ 1840, vol. vii, p- 1 & pl. dex, fig. 1. Distribution.—Vindobonian (France, Poland, Austria, Switzer- land, Italy, Holland) ; Anversian and Scaldisian (Belgium) ; Boxstones, Coralline Crag, and Red Crag—probably derivative ( Britain). Family Macrrip#. _ SPISULA OVALIS (J. Sowerby). Mactra ovalis & dubia J. Sowerby, ‘ Mineral Conchology * 1817, vol. 11, p. 136 & pl. clx, figs. 2-5. Distribution.— Vindobonian (Switzerland); Redonian (North- Western France); Middle Phocene (Holland); Scaldisian (Bel- gium); Red Crag to Post-Phocene (Britain) ; Recent (British Seas, non Mediterranean ). 1 Proe. Malacol. Soc. London, vol x (1912) pp. 100-104. c 2 20 MR. R. B. NEWTON ON A FOSSILIFEROUS [vol. lxxui, PostscRIPt. [Since the reading of this paper two memoirs have been published on some Miocene limestones from the North-Sea basin (Denmark and Germany) containing mollusean remains, several species of which are identical with those found in the North-Sea Limestone. In a Middle Miocene boulder from Esbjerg (Denmark), EH. M. Norregaard! has reported the occurrence, among other shell- remains, of Arcoperna sericea, Tellina benedeni, Naticina alderi, Streptochetus sexcostatus, and Acte@on tornatilis; while Karl Gripp 2 enumerates the same species as occurring in some older Miocene deposits of Denmark and Germany, with the following additional forms :—TIsocardia humana (=cor), Thyasira flexuosa, Dentilucina borealis, Levicardium fragile, Cyprina cf. islandica, Tellina benedeni (=fallax), Panopea menardi, Ficus simplex, and Ringiculella ventricosa, all of which are found in the North- Sea Limestone. Again, attention may be directed to a recently published account of the Lenham Sandstone Mollusca,? which have been referred to a late Miocene age, the following species contained therein indicating affinities with the fauna of the North-Sea Lime- stone :—Aporrhais pespelicani, Streptochetus sexcostatus, Acteon tornatilis, Scaphander lignarius, Yoldia oblongoides, Isocardia humana, Dentilucina borealis, Tellina benedeni, and Panopea menardt, R. B. N. March 27th, 1917. } EXPLANATION OF PLATE II. [All the figures are of the natural size, except where otherwise stated. | Fig. 1. Ranella gigantea Lamarck. (See p. 11.) Prepared from a wax model taken from a limestone cavity. . Streptochetus sexcostatus (Beyrich). (See p. 14.) . The same, showing magnified sculpture. . Ficus simplex (Beyrich). (See p. 14.) . The same, showing details of sculpture, magnified. . Xenophora sp. (See p. 14.) Fragmentary specimen exhibitirg adherent cavities on the whorl. . Naticina alderi (KH. Forbes). (See p. 15.) . The same. . Ringiculella ventricosa (J. de C. Sowerby). (See p. 15.) Front. view, X 3. 10. The same, showing sculpture lines, magnified. 11. Nucula levigata J. Sowerby. (See p. 16.) 12. Yoldia oblongoides (S. V. Wood). (See p. 16.) 13. Isocardia humana (Linneus). (See p. 18.) BD oe 0 LO 65 OO ~T 1 « Mellem-Mioczene Blokke fra Esbjerg’ Danmarks Geol. Underség. ser. 4, vol. i, No. 5 (1916) pls. i-iii, pp. 58. , 2 14 inches, and one of chert about an inch anda half long. (6) Uphill. The Uphill ‘trap’ is a much-weathered, highly-amyegdaloidal rock with a reddish colour owing to the abundant iron-oxide. The amygdules may consist of calcite, or of calcite and chlorite. The specific gravity of the freshest and least vesicular example obtain- able (50) is 2°60, and its silica percentage 43-17. In thin sections, € part 1] IGNEOUS ROCKS OF THE BRISTOL DISTRICT. 35 four of which were examined, the bulk of the rock is seen to be composed of felspar-laths, which tend to be larger than in the Goblin-Combe rocks and in those of Milton Hill and Spring Cove. The felspars, though much weathered, are seen to extinguish practically straight, and their refractive index is higher than that of Canada balsam. Hence it is probable that they are oligoclase. Needles of peroxidized magnetite, similar to those occurring in some of the Goblin-Combe rocks, are sometimes present. Pseudo- morphs after olivine in an iron-stained carbonate are plentiful. Partly owing to the extensive weathering, and partly to the ferru- eious staining, no unaltered augite was seen; but, in one section, small serpentinized augite-grains filling the interstices between the felspars are easily recognizable. (ce) Limeridge Wood, Tickenham. Hand-specimens from this locality are fairly fresh-looking dark- brown rocks, generally crowded with calcite- or chlorite-filled amygdules, though sometimes non-vesicular. Hight slices were examined. In all, the main part of the rock consists of felspar- Jaths generally about 0°5 mm. long, and, as a rule, too much weathered for the extinction-angle to be ascertainable. When, however, the rock is a little less weathered, the extinction-angle is found to be practically straight. Associated with the felspar is interstitial augite much weathered and obscured by iron-oxide. In nearly every section regularly-idiomorphic pseudomorphs after olivine are abundant: in some cases they are in a carbonate, in others in serpentine. The specific gravity of three specimens ranged from 2°63 to 2°68. ‘The silica-percentage of one rock (76) is 43°92, the soda is 0-44, and the potash 4°03. The high percentage of potash and the straight extinction suggest that the bulk of the felspar is ortho- elase. Amygdules formed of calcite, or partly of chlorite, partly of finely-spherulitic chaleedony, are common. The rock is closely related to that of Uphull. (d) Spring Cove and Milton Hill. Spring Cove.—The Spring-Cove lava is a very fine-grained red rock in which practically nothing of the nature of the constituent minerals can be ascertained without microscopical examination. The amygdules and red variolitic patches, which (as already mentioned)! are a characteristic feature, are very irregular in their distribution. Each variolitic patch is a group of imperfect varioles. ‘The macroscopic characters of the rock, and, in particular, the occurrence of imperfect pillows 2 2 and of a eoles. suggest that this rock may be related to the spilites. The specific gravity is also low for basalts. Specimens taken from about the 1 Q.J.G.S. vol. Ix (1904) p. 152. 2 See ‘Summary of Progress for 1893’ Mem. Geol. Surv. 1899, p. 105. D2 36 PROF. S. H. REYNOLDS ON THE [vol. Ixxii, middle of the mass had specific gravities of 2°72 and 2°74, while: that of a specimen from the eastern end was 2°67. The chemical analyses quoted on p. 388 (though the percentage of magnesia and soda is somewhat low for a basalt, while that of the potash is high) do not bear out the idea of a spilitic relationship, the soda especially being far too scanty. The occurrence of well-terminated pseudo-. morphs in carbonate after olivine in certain parts of the Spring- Cove lava is also inconsistent with a spilitic relationship. ‘The felspars are, in every case, too much weathered for the extinction- angle to be ascertained with accuracy, but they appear to extinguish straight. In none of the nine slices examined could any recog- nizable pyroxene be detected. Much brown glassy material is. present in the ground-mass. Milton Hill.—The westernmost exposure is that mentioned by Dr. Strahan! as occurring 160 to 190 yards north of Milton- Hill Lodge at the eastern margin of Weston Woods. Here a compact, black, only slightly vesicular rock (97) occurs in the path along the edge of Furze Close; but the exposure is so bad that it may easily be overlooked. The specific gravity (2°78) is. higher than that of any other rock of Milton Hill. The percent- age of silica and alkalies (see Table I, p. 38) is that of a normal basalt; and the agreement between this rock and those in Goblin Combe is close, both chemically and mineralogically. It has a rather fine-grained ground-mass, composed of eone laths with an ecinetion: angle of 10° to 15°, magnetite and granulitic augite, with phenocrysts of fresh augite and of idiomorphic olivine replaced by green serpentine. The chemical characters and extinction-angle suggest that the felspar is probably andesine. The rock differs considerably from any other seen on Milton Hill. The next exposure les north-north-east of Milton-Hill Lodge at the point 98 in the sketch-map (fig. 6, p. 80), and has already been mentioned as having been made in 1911 in the course of erection of a cistern. No trace of this exposure is now to be seen. The rock here is brown, considerably weathered, and rather strongly amyegdaloidal. In a hand-specimen it bears a general resemblance to that from Limeridge Wood, Tickenham, “tong also resembles as regards specific eraviby (2: 66 to 2 68) and silica-percentage (42: 21). Of alkalies Mr. Radley found 2°89 per cent. of potash and 0:68 of soda. The resemblance to the Limeridge- Wood rock is further brought out in a thin section, particularly as regards the relatively- ‘large size of the felspar-laths in comparison with those in the Spring- Cove rock and that of the cross-roads, Milton Hill. The felspars are too much weathered for anything to be made out with regard to the extinction-angle. Olivine in idiomorphic crystals, mainly replaced by a very pale serpentine, is abundant. 1 “Geology of the South Wales Coalfield: Part I1I—The Country around. Cardiff’ 2nd ed. Mem. Geol. Sury. 1912, p. 29. part 1] IGNEOUS ROCKS OF THE BRISTOL DISTRICT. 3o7 The next point where the lava is exposed is at the cross-roads at the top of Milton Hill. The highest bed seen here, when the exposure was in a better condition for observation than it is now, was a rock which in a hand-specimen was not obviously amygda- loidal, although it shows many ill-defined patches of a green alteration-product. Numerous pseudomorphs in carbonate after olivine occur with iron-oxide aggregated along their margins and eracks. Underlying this is a coarsely amygdaloidal rock with calcite-filled vesicles sometimes an inch and a half long. There are other vesicles filled wholly or in part with a chloritic mineral, while other amygdules consist of calcite enclosing a nucleus of quartz. It is difficult to ascertain whether certain spots are amygdules or imperfect varioles. The bulk of the rock consists of felspar-laths of somewhat variable size, which extinguish straight or nearly so, and have a lower refractive index than the Canada balsam in which they are mounted. These characters indicate orthoclase, the presence of which is further suggested by the high percentage of potash—namely, 5°83, found by Mr. Radley (Mr. Sturgess found 5:34). Much iron-oxide is present. The phenocrysts include large (often idiomorphic) olivines, represented by pseudomorphs in carbonate associated with much iron-oxide, and a few felspars which have very low extinction-angles. Some of the large felspars have isotropic patches which Dr. H. H. Thomas suggests may be due to analcitization.! There are also isotropic patches in the ground-mass which are probably analcite. The specific gravity of the rock (94) is 2°57 to 2°58. A close resemblance exists to the Spring-Cove rock, particularly in the high percentage of potash. A trace of lithia was also observed. The easternmost exposure on Milton Hill is that found by Dr. A. Strahan near Florence Cottage, 700 yards east-north-east of the cross-roads. This exposure occurs in the bed of a small pond about 150 yards west of the Cottage, and in wet weather nothing ean be seen. The rock is very highly vesicular, and so much weathered that no sections were cut and the specific gravity was not ascertained. As has been already pointed out, it closely resembles the rock (102) from the little pond on Milton Hill, the rock in each case probably coming from quite the bottom of the flow. (e) Woodspring. The Woodspring trap in a hand-specimen is a much-weathered dark-greenish rock with numerous amygdules, which in the bulk of the rock are small, rounded, and filled with green material. In places very large calcite-amygdules occur. In thin sections the main part of the rock is seen to consist of small felspar-laths in a greatly weathered state; sometimes, however, it is possible to satisfy one’s self that the extinction is practically straight. Much +2 Analecitization in albitized basic felspars is described by E. B. Bailey & W. Grabham from Arthur’s Seat, Edinburgh, Geol. Mag. dec. 5, vol. vi (1909) p. 254. 38 PROF. S. H. REYNOLDS ON THE [ vol. lxxu, finely-divided magnetite is distributed through the sections. Very numerous olivine-crystals, sometimes represented by pseudomorphs. in carbonate, sometimes by pseudomorphs in serpentine, occur, and. in many cases show crystal outlines. No pyroxene, however, can be identified. LTV. Tapes or CoEmMIcAL ANALYSES. 1. I. II. Tet aac Vi eave | NS(Os ete ee lee 0.69 48°80 50°75 5544 | ~51-76 | WNiOps) teee ele 2.80 as 146 O16 <5 04a ZAllo Osi aaatae ee ee 11°54 17°65 12°15 18°60 | 12°36 Hes Oaieieccs ceo om eae 5°83 4-71 9°26 2709 | 438 | | Re Oipreiait ne hie eae 3°29 2°72 Only =| 4;48 | 460 | MnO es 0°36 25 0-43 ese Ol CO). 6:36 7°65 Blt” | 16 eens end Mic Oar ee Nene ec se ee 4:29 3°30 6°22 ATI) | 95 Tt OR eee a ge 501 4°93 7 | 663 | 3°83 iNaO te el aie 0-72 1:10 055 | 1°79 1:99 | H20 at 105°C. . _ 165 0°69 == { eal { me HO above 105° 0. 3°61 2°77 Se Ree ee ISRO reasteme inte ae 0°35 == = | O56 | HOO ate see atc e ite: 477 5°93 3°76 = pl15 OR ee ore ae eaaanabACe — = oss | Mota seeeee eee 100°27 100°25 98:07 10095 | 100:32 | — =e ES | Specifie Gravities...... 2°59 2°617 27 I. Variolitic olivine-basalt, Spring Cove, Weston-super-Mare (E. G. Radley). Il. Do. do. do. do. (J. H. Sturgess). IIL. Rock from Knowle Quarry near Spencecombe, Devon (957). (J. Grant Wilson.) IV. Ciminite, Fontana di Fiesole, Monte Cimini, near Viterbo, Italy. (H.S. Washington.) VY. Absarokite, Raven Creek, Yellowstone Park, U.S.A. (Ul. G. Eakins.) (9) hele Specific | g: | | Gravities. | SiO,. K30. Na,O. 33. Olivine-basalt, Goblin Combe, east exposure.) 2°84-2°86 | 49°12 | 0°41 | 2°37 | 81. Olivine-b asalt, Goblin Combe, west exposure.| 2°80-2°81 | 48:49 | 046 | 2°39 97. Olivine- basalt, Furze Close, "Milton Hall, 2-78 49°58 O73 | 196 6 Weston-super-Mare ..... - °° | O-75*} 1:81* 98. Olivine-basalt, north-east of Milton- Hill ce ; : Lodge, W eston- -super-Mare ....... CODEN ES |) BAI 2) UES 76. Olivine - orthoclase - basalt, Limeridge ? -99%| 4:02 | 0 Wood, Tickenham .. a 5 BIST) ESP EU NO 50. Olivine- basalt, Uphill cutting ~Weston- super-Mare ..... 23. Olivine-basalt, Middle “Hope, ~Weston- 45°69 | super-Mare ...... 7 Te eng 4 260) = 14317 | a ae 17. Variolitic basalt, Spring Cove, “ Weston- | 49°69 | 5:01 | 0°72 super-Mare ........ | 48°80*| 4°93* | 1:10* 94, Olivine - orthoclase - basalt, cross - - roads, 57-958 146-16 Milton Hill, Weston-super-Mare......... SSO aaIs | part 1] IGNEOUS ROCKS OF THE BRISTOL DISTRICT. 39 The estimates marked with an asterisk in the foregoing table (2) were determined in the Chemical Laboratory of the University of Bristol by Mr. J. H. Sturgess, the remainder by Mr. E. G. Radley. Iam much indebted to my colleague, Lieut. D. H. Innes, B. A., and to my pupil, 2nd Lieut. F. . Wallis, for determining the specific gravities. VY. CoMPARTSON with Rocks oF OfTHER AREAS. Chemical and mineralogical examination show the presence in certain of these rocks of acid felspar: orthoclase (Limeridge Wood Tickenham, and Milton Hill) and probably oligoclase (Uphill) ; while, in regard to their general aspect and composition, and particularly in respect of the low silica- -percentage and abundance of olivine, the rocks are basalts. Rocks showing a similar association of olivine and orthoclase have been described from several other localities. The following are the most important :— 1. Some of the Permian lavas of the Hxeter district! are well known to contain orthoclase in association with olivine; and, though in the field, owing to their generally pale and reddish tints, the difference between these rocks and those of the Bristol area is considerable, the resemblance both in microscopical structure and in chemical composition is quite close. This is particularly the case with the lavas described by Sir Jethro Teall as the Knowle- Hill (Crediton) type. These rocks are, he remarks,? ‘basic so far as silica percentage is concerned, but (which) differ from ordinary basic rocks in containing a large amount of potash.’ An analysis of the rock from Knowle Quarry, Spencecombe, is quoted above (p. 38), and it is evident how close is the correspondence between this rock and that from Spring Cove in the percentages of silica, alumina, lime, and alkalies. A section of the rock from Spence- combe Farmhouse Quarry (HK 3219 in the Geological Survey collection) resembles very closely some of the Milton-Hiull rocks. Sir Jethro Teall clearly regarded the orthoclase in these rocks as original, but remarks? on the reduction of the lime through the leaching- out of the carbonate. 2. Dr. Flett suggested comparison with certain of the rocks of Pelvoux described by Prof. Termier,* particularly with certain ‘diabases’ which, while originally rich in lime, now contain much potash. The changes in these rocks that are entirely attributed to the metasomatic action of surface- water, are twofold: (a) the elimination of the bulk of the lime, and to a less extent of the alumina and magnesia; (6) the addition of potash, which is believed to be derived from the neighbouring mass of the Pelvoux 1 See ‘Geology of the Country around Exeter’ (sheet 325), Mem. Geol. Sury. 1902, pp. 55-85. 2 Ibid. p. 83. 3 Ibid. p. 79. 4 Bull. Soe. Géol. France, ser. 3, vol. xxvi (1898) pp. 165-92. 40 PROF. 8. H. REYNOLDS ON THE [vol. lxxu, granite and gneiss. A series of analyses is quoted of the actual rocks, and these are paralleled in each case with a statement giving the supposed composition of the rock when unaltered. The contrast between the environment of the rocks in the two localities is so great as to show that the richness in potash cannot be due to similar causes. Thus the Bristol rocks are everywhere associated with thick masses of Carboniferous Limestone, and, if there had been any transference of material, might have been expected to have had the percentage of lime increased. No granite or other rock rich in potash is known to occur nearer than Malvern on the one hand, and Dartmoor and Lundy on the other. Hence, any elimination of lime and addition of potash, such as_ Prof. Termier describes, can scarcely be contemplated. 3. The Milton-Hill rocks seem to beara considerable resemblance to the ciminites of Dr. H. 8. Washington,! both rocks containing orthoclase, pyroxene, and olivine, and being devoid of hornblende and biotite. The resemblance is brought out by the analyses quoted on p. 38, particularly as regards the excess of potash over soda, but the percentage of silica in a ciminite is distinctly higher than in the Milton-Hill rocks, and Dr. Washington estimates that as much s 22°8 per cent. of anorthite is present. 4. The Bristol rocks resemble in some respects members of the somewhat variable series from the Yellowstone Park, described by Prof. J. P. Iddings as absarokites.2 The two groups of rocks agree in the occurrence of abundant olivine and augite, in asso- ciation with orthoclase and a variable amount of more basic felspar. The silica-percentage in each case is low, that of the alkalies moderately high with potash higher than soda. Some of the absarokites contain leucite, others quartz. The rock from Raven Creek, of which the analysis is quoted (p..38), seems to come nearest the Bristol rocks; it is fine-grained, and consists of augite, serpentinized olivine, magnetite, and orthoclase, with no leucite and only a very little lime-soda felspar. Despite the unusual association of minerals in some of the rocks which are the subject of this paper, it does not seem possible to avoid the conclusion that the acid felspars are original, and hence that these rocks are comparable to certain of the Exeter traps, to the ciminites of Dr. Washington, and to certain of the rocks” described by Prof. Iddings. It will probably be best to term them orthoclase-basalts. 1 Journ. of Geology, vol. v (1897) p. 349. 2 Ibid. vol. iii (1895) p. 938. part 1] IGNEOUS ROCKS OF THE BRISTOL DISTRICY. eee VI. SumMARY AND CONCLUSIONS. The igneous rocks associated with the Carboniferous Limestone ot the Bristol district may be arranged in three groups :— (a) Normal olivine-basalts (Goblin Combe, eastern and western exposures, and 97 from Furze Close, Milton Hill). All these have a very low percentage of potash. (6) Olivine-bearing rocks, with a rather high percentage of potash and a low percentage of soda (Limeridge Wood, Tickenham, and 98 from north-east of Milton-Hill Lodge ; probably the rocks from Uphill and from Middle Hope or Woodspring belong here). (c) Olivine-orthoclase basalts, with sometimes an imperfectly variolitic structure, a very high percentage of potash, and a low percentage of soda (Spring Cove and cross - roads, Milton Hill). I am greatly indebted to Dr. J. 8. Flett and Dr. H. H. Thomas for help and advice; also to the latter for the loan of sections of rocks from the Exeter district. Sincere thanks are also tendered to the following, for permission accorded to dig trial-holes :— Mr. J. Gibson, J.P. (Goblin Combe); Mr. H. B. Napier, of the Ashton-Court Estate (Limeridge Wood, Tickenham); and the Great Western Railway Company (Uphill). DISCUSSION. Dr. A. SrraHAN commented on the difficulty which had been experienced in tracing these igneous rocks, owing to lack of ex- posures. Much of the outcrop in Goblin Combe, as shown on the geological map, was founded on the occurrence of igneous débris in the soil, and in part only on actual exposures. He was not satisfied that its continuity had been disproved. That the lavas in the various occurrences in Somerset were discontinuous was obvious, but that other evidences of igneous action were confined to such limited areas as the Author claimed, seemed improbable. Geologists were greatly indebted to the Author for the additions that he had made to their knowledge of Somerset geology by opening trial- holes in doubtful ground. Prof. T. FRANKLIN SrBLy enquired whether the Authoyr’s further investigations on Milton Hill had enabled him to utilize the volcanic horizon of Spring Cove for mapping the dip-faults, later in date than the great longitudinal overthrust fault which tra- versed the Weston-Worle ridge of Carboniferous Limestone. The faulted structure of this ridge was a typical illustration of the displacements which had affected the Carboniferous rocks of neighbouring areas. Dr. A. P. Youne said that the Author’s discovery of pillow- lavas and associated rocks would furnish a key to many phenomena observed in this area. The speaker had collected from the shore 42 IGNEOUS ROCKS OF THE BRISTOL DistRIcT. [vol. lxxn, about Woodspring Cove some rock-specimens for which he was in fact indebted to the Author, as they had been obtained during the visit of the Geologists’ Association to Bristol in 1907. The slides. from these specimens showed forms the organic origin of which could scarcely be doubted, recalling radiolaria and other organisms, but too imperfect for determination. ‘These bodies were embedded in the ground-mass of the lava, in such a> way as to suggest that they had been taken up by a subaqueous lava-flow from sediments still in the state of mud. It was believed that the exceptional microscopic occurrences, and the general relations of lava to sediments, could be best explained on the hypothesis of a subaqueous eruption over a floor of imperfectly-consolidated sediment. The Avrnor, in reply to Dr. Strahan, said that the information leading him to map the igneous rocks at Goblin Combe as two separate crescentic masses could have been obtained only by digging many trial-holes. To Prof. Sibly he said that at Worle Hill his attention had been confined to the i igneous rocks, and that he had not attempted to study the general “structure of the hill ; he had, however, little doubt that the failure to find any trace of igneous material east of Florence Cottage and between Spring Cove and Furze Close indicated that the volcanic rocks of the Milton-Hill neighbourhood were cut off at both ends of their outerop by cross-faults. part 1] INSECTS FROM THE BRITISH COAL MEASURES. 43, 4. On some Insects from the Bririsu Coan Measures. By Herbert Bouton, M.Sce., F.R.S.E., F.G.S., Reader in Paleontology in the University of Bristol, and Director of the Bristol Museum. (Read March Sth, 1916.) [Puatres III & IV. | THE present paper consists of a critical study of some hitherto undescribed wings of fossil insects from the British Coal Measures, and of insect-wings previously noted by other authors. J am indebted to Mr. William Eltringham, of Newcastle-on-Tyne, and to the authorities of the Geological Survey Museum, the National Museum of Wales, and the Liverpool, Manchester, nil Newcastle Museums for their courtesy in placing the specimens at my service for study. /HD@OPHASMA ANGLICA Seudder. (PI. III, fig. 1.) 1885. Aidcophasma anglica Scudder, Geol. Mag. dec. 3, vol. 11, p. 265. 1885. Adcophasma anglica Scudder ; Scudder, ‘in Zittel’s ‘Handbuch der Palaontologie’ vol. ii, p. 758, fic. 941. 1906. Adceophasma anglica Scudder; Handlirsch, ‘Die Fossilen Insekten’ p. 125 & pl. xiii, fig. 4. This specimen was obtained by Major Chambers, and presented to the Derby & Mayer Museums, Liverpool, in 1858. It is con- tained in an ironstone nodule, much similar to those of Ravenhead (Lancashire). Its horizon is not known, but it is undoubtedly from the Middle Coal Measures. It was partly described and named, but not figured, by S. H. Scudder in 1885; and a figure of the wing appeared in the same year in Zittel’s ‘ Handbuch der Pala&ontologie.’ This figure was probably due to Scudder, who wrote the entomo- logical section of the work for that edition of Zittel’s manual. The wing lies in a fine-grained ironstone nodule, one half showing a left wing, and the other half of the nodule contaiming the im- pression. Since the specimen came into my hands for examination, I have been able to uncover the apex of the wing and a small portion of the base. A little of the base of the wing is missing, and also the middle portion of the inner margin. The total length of the wing as now seen is 87 mim., and the greatest breadth (across the middle of the wing) 40 mm. When complete, the wing was probably 100 mm. long. The costal margin is gently convex. The sub-costa isa broad flat vein which gradually diminishes in width as it passes outwards to the wing-apex, which it just fails to reach. The radius is an even broader vein than the sub-costa, also flat, and passes outwards, parallel to the sub-costa, to the outer angle of the tip of the wing. It remains undivided throughout its whole length. 44. MR. H. BOLTON ON SOME INSECTS (vol. Ixxii, The median divides low down near the base into two equal branches, the outer of which gives off four successive twigs. ‘The first of these twigs (regarded by Handlirsch as the radial sector) arises before the branch reaches the middle of the wing, and remains undivided throughout the whole of its course to the inner angle of the tip of the wing. The second twig divides twice, the outer and inner twigs of the second bifurcation dividing again, so that the ultimate twigs reaching the margin are six in number; the remaining branches, which are three in number, remain simple and undivided. The anterior branch of the median occupies the greater part of the wing-tip, and ends upon it in seven twigs. The inner branch of the median is simpler in character than the outer, and does not divide until well past the middle of the wing. Fig. 1.—Restoration of wing of Mdcophasma anglica Scudder, showing the general character of the intercalary venation. (Natural size.) SOO pre ee a UCI eT It then gives off four twigs, which pass outwards to the junction of the inner wing-margin with the apex. The first of the twigs forks before reaching the margin, and consequently the inner branch of the median has six ultimate twigs. The cubitus passes out from the base of the wing parallel to the main stem of the median and to its inner branch for the greater part of its length. It sends off at a wide angle, at the middle of the wing, a strongly-eurved branch, which bends first inwards and then outwards towards the apex, breaking up into five twigs before reaching the inner margin. The second of these twigs forks. A second branch comes off a little farther out, and also at a wide angle. This remains undivided along its whole length. The main stem of the median runs out to the margin, forming a small fork immediately in front of it. The cubitus ends in the inner wing- margin in six twigs. The next two veins were probably united a little way out from the base, and their direction is such that the single stem from which they arose may have been given off at the base of the cubitus. If this were the case, the two veins would form an inner part 1] FROM THE BRITISH COAL MEASURES. | 45, division of the cubitus. JI am of opinion that this supposition is the correct one. , The outer of the two veins is undivided, and reaches the inner margin beyond the middle of the wing. 'The second or innermost vein runs fairly parallel to the first along its whole length, giving off, as it does so, four inwardly-directed twigs. The first and fourth veins fork before reaching the margin. The whole vein ends on the margin in six twigs. The venation of the radial and median areas consists of straight or slightly curved cross-veins placed at fairly regular distances. The very wide cubital and cubito-anal areas are crossed by irre- gular wavy veins, which branch, and at times join into a loose network. The anal area is occupied by simple cross-veins. Affinities.—At the time when Scudder described the wing, he was of opinion that it had relationships with Meganeura ( Dict ‘Yyo- neura) monyt, and that it was a member of the group Proto- phasmidz. Handlirsch has, however,! removed it to the group of Palzeodictyoptera incertze sedis. Scudder was undoubtedly mistaken in referring the wing to the Protophasmidee, as a glance at the figure of Protophasma dumasii Brongniart? will at once show. Handlirsch apparently deduced the Palzodictyopteroid affinities only from a study of Scudder’s drawing of the wing. I have not been able to adopt Handlirsch’s view of the structure. of the wing, but have arrived at slightly ditferent conclusions. Handlirsch regards what I have described as an anterior branch of the median vein, as the radial sector. If this view be correct, the radial sector and the median vein are completely fused for a distance of 10 to 15 mm. from their point of origin. The basal portion of the radius is widely divergent from the base of the median—more so, in fact, than at any other part of the whole course of the radius and supposed radial sector. They could, therefore, have come into union only at the actual point of origin of the wing. This may have been the case; but, in my opinion, the radius is wholly simple and undivided, and no radial sector is present. The median vein is free and strongly developed, a remark which applies also to the cubitus. The eubitus may, and probably did, send off the long simple vein lying inwards to the main stem; and the simple vein may also have given origin to the next vein, which divides into three before reaching the wing-margin. The interpretation that I make of the wing is, then, that the sub-costa and radius are simple and undivided. The median and cubitus are large, much divided, and take up the greater part of the wing-area; while the anal veins are few, and but one shows evidence of forking. If my view be correct, the wing is undoubtedly a very primitive type of the Proto-orthoptera, still retaining evidence, in the costa, sub-costa, and radius, of its Paleodictyopteroid origin. 1 “Die Fossilen Insekten ’ 1906, p, 125. 2 Ibid. pl. xvi, fies. 1 & 2. 46 MR. HW. BOLTON ON SOME INSECTS [ vol. Ixxu, Type-specimen in the Derby & Mayer Museums, Liverpool. Locality.—Unknown. Donor: Major Chambers, 1858. Horizon.—Unknown. Middle Coal Measures, South Lanea- shire. (Dictyoneuron) HicerNstt (Handlirsch). (PI. III, fig. 2.) 1871. ‘ Neuropterous Insect-Wing’ Higgins, Presidential Address, Proc. Liver- pool Naturalists’ Field Club, 1870-71, p. 18, fig. 15. 1906. (Pale@odictyopteron) higginsii Haudlirsch, ‘Die Fossilen Insekten ’ p. 125 & pl. xii, fig. 6. This wing was described as follows by the late Rev. H. H. Higgins :— ‘A second and smaller specimen [of insect-wing] was obtained by myself and referred to the genus Corydalis. Mr. J. P. G. Smith compared it with Fulgora. A slight sketch of it was seen by F, Smith, of the British Museum, whom it reminded of Gryllotalpa. Mr. Benjamin Cooke, of Manchester, after a careful examination, says: “I believe the fossil represents the basal portion, about one-third only, of the fore wing of a Chrysopa, Golden-eye, or Lace-wing fly, or rather Nothochrysa, separated from Chrysopa by Mr. Mclachlan on account of the manner in which the third cubital cell is divided.” This cell is remarkably well shown in the fossil, and though I could only judge from memory, I believe it is sufficient to settle its relationship.’ The specimen was afterwards named, but not described, by Handlirsch as ‘(Pal@odictyopteron) higginsii. _Handlirsch’s figure is a copy of that given by Higgins. He did not see the specimen. The impression is that of the basal portion of a left wing, the portion preserved having a length of 82 millimetres. The oveatest width is 22 millimetres, across the distal portion of the anal area. The ‘cubital cell’ mentioned by Mr. Benjamin Cooke is an elongated, somewhat oval area lying in the middle of the base of the wing. It has been formed by a twig of a cubital vein coming off forwards and outwards to join on to the main vein lying in front of it. This is a feature of rare occurrence, and will be considered more fully later. The costal border, which is preserved for a length of 29 mm., is strongly curved proximally, and is gently arcuate as it passes outwards towards the middJe of the wing. The sub-costa isa feeble vein, which rises towards the anterior margin of the wing and passes outwards almost in a straight line. The area enclosed between the costal margin and the sub- aos is very wide proximally, and rapidly narrows down to an acuminate point where the sub- costa reaches the frontal margin of the wing. The complete wing must have been at least three times as long as the portion preser ved, and the sub-costa would therefore join the costal margin about the middle, or a little beyond the middle, of the wing. The radius runs almost parallel to the sub-costa along its whole length, giving off the radial sector a little beyond the middle of the portion of wing preserved. The radial sector comes off from the radius at an “acute angle, and passes outwards parallel part 1] FROM THE BRITISH COAL MEASURES. A7 to it in the direction of the wing-tip, the upper portion of which it doubtless reached. The median and cubital veins present certain unusual problems. In all Paleodictyoptercids, the median and cubital veins each arise and continue outwards for some distance as a single stem. The median shows least tendency to branch near its point of origin, while frequently it continues as a single unbranched vein for the greater part of its length. The cubitus in many Car- boniferous insects often sends off a long anterior branch only a little way out from the base, or early divides by forking. In the present instance, three veins occupy the position of the normal median and cubitus. Of these, two either arise from a common root or so close together as to be indistinguishable, while the third has an independent origin from the other two. Further, the third of these three veins sends off, soon after its origin, a stout outwardly- and forwardly-directed twig, which joins “the second of the two anterior veins and merges with it. If we assume that the two veins, arising from’ a common point, represent the median vein, which forks at the base of the wing, the third of the three veins will be necessarily the cubitus, sending a twig forward to join the posterior branch of the median. In this case the ‘cell’ enclosed by the hinder median, and its forwardly-directed commissure, will be a ‘median-cubital’ cell, and not a ‘ cubital cell’ as assumed by Mr. Benjamin Cooke. On the other hand, as the median vein in Paleodictyopteroids shows little tendency to branch, while the cubitus shows a distinct tendency to branch, we may consider the median vein in the wing- fragment to be a simple vein, undivided throughout that portion of its length which is preserved, and arising in close union with an anterior branch of the cubitus. The cubitus then becomes a double-stemmed structure, the two elements of which unite by a powerful cross-vein in such a fashion as to enclose a basal ‘ cubital’ cell. Notwithstanding the common point of origin of the first two veins under CORSO the latter view seems the more logical one. A third view is that the median divides at the base, and that the anterior branch of the cubitus goes off forwards to join the inner branch of the median, the latter beyond the junction being therefore a medio-cubital structure. If these veins be considered from the last-mentioned point of view, the median is a simple undivided vein in the wing-fragment. It arises close to the radius and the anterior branch of the cubitus, the three veins being crowded together for a short distance, beyond which the median bends backwards from the radius, and passes outwards in the direction of the tip of the wing, in an almost straight line. It is separated from the undivided portion of the radius and from the radial sector by an area the width of which is twice that which separates the radius and sub-costa, or that which separates the main stem of the radial sector from the radius. The anterior branch of the cubitus les close to the median 48° MR. H. BOLTON ON SOME INSECTS | vol. Ixxii, for a short distance, and then bends sharply inwards from the median, passing back in the direction of the middle of the inner margin of the wing. A rapidly-widening triangular area is thus left between the anterior cubital vem and the median. The inner cubital vein pursues an oblique course from its point of origin inwards to the wing-margin, and is widely spaced out from the anterior branch. The forwardly-direeted twig, which has been already mentioned, forms a sigmoidal curve reaching the anterior branch of the cubital vein about the middle of its present length, probably at the end of the first third of the total length. Six anal veins are present, all of which pass obliquely downwards to the inner margin. ‘he second and third anal veins fork before reaching the wing-margin. The intercalary venation consists of relatively-large reticulations in the middle areas of the wing, that is, between the median and the cubital veins; while, in the area between the radial sector and the median, the veins cross in curved lines with no very evident reticulation. ‘The intercostal area bears a few faint cross-veins passing obliquely outwards from the sub-costa to the wing-margin. Affinities.—With so small a wing-fragment, it is not possible to form an accurate idea of the whole structure. It is evident that the radius and the radial sector ran out upon the anterior half of the wing-tip, if they did not occupy the whole of it. The median probably curved inwards, reaching the inner margin near the apex of the wing; while the branches of the cubital and the anal veins occupied most, or all, of the inner wing-margin. The character of the broad intercostal area and the course of the sub-costa, are very similar to what is found in Polycreagra elegans Handlirsch ; but the main features of the wing are of a simpler type, and the intercalary venation is wholly different. The reticulate intercalary venation and the character of the principal veins, which do not show much trace of division in the inner third of the wing preserved, are undoubtedly Dictyoneuroid. Instead of creating a new genus upon this wing-fragment, it seems preferable to follow the example of Handlirsch in his ‘Revision of American Paleozoic Insects’ and style the specimen ‘ (Dictyoneuron),’ retaining the specific name of higginsi. Type-specimen in the Derby & Mayer Museums, Liverpool. Locality.—Ravenhead railway-cutting, near St. Helens. Horizon.—Middle Coal Measures. PALMOMANTIS MACROPTERA, gen. et sp.nov. (PI. III, figs.3 & 4.) The late Rev. H. H. Higgins, in his Presidential Address to the Liverpool Field Naturalists’ Club, 1871 (p. 18), thus records the discovery of this specimen :— ‘Perhaps the finest and most remarkable fossil found in the Rayenhead cutting was obtained by a son of Mr. J. P. G. Smith, of Liverpool, who, with a young friend, Mr. Clementson, from Rugby, visited the cutting several times. It was the wing of a large insect, beautifully preserved in a nodule of ironstone.’ part 1] FROM THE BRITISH COAL MEASURES. 49 This nodule was deposited in the Derby & Mayer Museums at Liverpool, and has been placed in my hands by the courtesy of the Curator, Dr. J. A. Clubb. The larger half of the nodule has lost a portion of one end, which contained the tip of the right wing. The proximal third of the left wing remains with its dorsal surface uppermost, and its ventral surface closely applied to that of the right wing. It is quite evident that the whole of the two wings were contained in the nodule; but, when the latter was being split open, a thin film of ironstone carried away the middle and distal portions of the left wing. The impression of the dorsal surface of the greater part of the left wing remains upon one half of the nodule, and this, combined with the portion of wing remaining on the other half of the nodule, enables one to determine with accuracy almost the whole of the wing structure. ~ One unusual feature in the preservation of the wings 1s that they lie with their ventral surfaces apposed. To bring the wings into this position, one must have become bent under the body, instead of falling sideways or above the back of the insect. The body of the econ would thus, if the wings still remained attached, come to lie between the two. No trace of the body can be seen; but the left wing has a deep inward flexure, such as it would naturally acquire if the body had been carried round with the nght wing and pushed up into the anal area of the left wing. Had the insect’s body been canried round in the way suggested, the right wing would not coincide in position with the left, but be thrust farther out. This latter is actually the case, the outward displacement of the right wing as compared with that of the left being at least 20 mm. The two wing-fragments are quite sufficient to show that, while the complete structures were comparatively short in length, they were yet remarkably wide from front to back. In this respect they are unlike the wings of any known Coal-Measure insect, and would seem to indicate a heavy-bodied insect, with slow but powerful fight. The frontal third of each wing is supported by strong and fairly rigid veins, which become more slender as they pass outwards. The hinder margins of the wings are more mem- branous, and were evidently lacking in a rigidity equal to that of the frontal third. Left wing.—A little more than the proximal third of this wing is present, and in good preservation. It, is 42 mm. long, and about the same in width across the anal region. ‘The impression on the other half of the nodule shows all but the tip of the wing. The main trunks of the costa, sub-costa, radius, and median are stout structures, standing out in good relief from the wing-surface. The cubital and anal veins are but half the thickness of the former, and he sunk in shallow grooves. The costal vein forms the anterior margin of the wing, which is shghtly convex forward, and slopes gently backwards as it passes outwards to the apex of the wing. The sub-costa is separated from the costa by an interval of 4 mm., and passes almost in a straight Q. J. G.S. No. 285. E 50 MR. H. BOLTON ON SOME INSECTS | vol. lxxu, line outwards to the anterior edge of the wing-tip. The sub-costa gives off a numerous series of small, irregular, and occasionally forked branches, which pass forward to the anterior margin. The radius arises in close apposition to the sub-costa, and gradu- ally diverges from it as it passes outwards. It throws off a lateral twig on the posterior side before reaching the tip of the wing. ‘The branchings of the radius and the single main stem of the costa occupy the anterior third of the wing-tip. The main stem of the radial sector probably arises from the radius close to the base of the latter, but this cannot be determined, owing to the absence of the inner anterior angle of the wing. The two veins are close to each other proximally, the radial sector keeping parallel with the radius for a short distance and then curving sharply backwards at the end of the first third of its length. At this point it is widely forked, a forward-directed branch passing outwards parallel to the radial sector, and probably forking before it reaches the Fig. 2.—Restoration of the left wing of Paleomantis macroptera, gen. et sp. nov. (Natural size.) margin of the wing-tip. The inner branch passes obliquely backwards to the inner angle of the wing-tip, giving off three forward-directed twigs, the first two forking before reaching the apex of the wing. The median vein isa powerful structure for the first half of the length, beyond which it becomes attenuated. A little beyond its point of origin it breaks up into two main branches, the outer one passing in a broad curve along the median diagonal of the wing, and forking before it reaches the inner margin. The second branch is not so powerfully developed as the first, and follows a course subparallel to the latter, giving off two simple branches on its inner side before reaching the middle of its length, and forming a small fork near the wing-margin. The median and its branches occupy the outer half of the inner wing-margin. The cubitus is a comparatively weak vein, consisting of two main branches, the outer forking once only, and the inner forking part 1] FROM THE BRITISH COAL MEASURES. 51 in the middle of its length, each twig forking again just before reaching the margin. Seven anal veins are determinable, all simple and unbranched, except the third, which forks twice. The anal veins are weak, radiate fan-wise from the base of the wing, and, owing to the great width of the latter, pass almost directly backwards. The intercalary venation consists of an open meshwork of irregular polygonal cells, feebly defined towards the free margin of the wing. Around the point of attachment of the wing are a few intercalary veins, which run obliquely inwards towards the wing-base. Right wing.—The right wing, of which less than the proximal half is shown, lies with its under side uppermost under the fragment of the left. The portion exposed is too small to allow of a detailed description of the full character of the veins, but sufficient is seen of the wing to determine that the median vein differs from its Fig. 3.—Restoration of the right wing of Paleomantis macroptera, gen. et sp. nov. (Natural size.) fellow in ‘the left wing. But one twig comes off backwards from the second branch, forking at the middle of its length. The outer branch of the cubitus forks twice before reaching the wing-margin, that of the left wing forking once. The greatest depth of the wing is along a line drawn from the costal margin to the middle of the cubital area on the inner margin. The diameter is here 40 mm. ‘The absence of the wing-tip is unfortunate, as it renders the complete outline of the wing a matter of uncertainty. The shape of the ironstone nodule would seem to indicate that very little of the wing is missing, for nodules of this character are usually very uniform in size and outline. The missing portion can hardly have exceeded 20 to 25 mm. in length. The indications of the terminal forking in the radius and radial sector are usually in close agreement with those of the median, and the latter indicate that only a small portion of the wing-tip is missing. Bearing in mind the probable maximum length of E2 52 MR. H. BOLTON ON SOME INSECTS [vol. xxi, the nodule, and the fact already mentioned respecting the terminal veins, we may fairly assume that the wings were at least twice as long as broad. and had a bluntly rounded tip. Their great breadth is an unusual feature in a Paleodictyopteroid wing, and more typical of those of Blattoids. The general character and course of the chief veins are most clearly Paleodictyopteroid. The intercalary venation is much lke that of Hypermegethes schucherti Handlirsch, and forms an open irregular meshwork. Affinities.—Higeins does not seem to have hazarded any view as to the character and relationship of this wing, although, as already seen, he attempted to determine the relationships of the smaller wing-fragment discovered by himself in the same railway- cutting. The wing-structure is so clearly. Palzeodictyopteroid, that its affinities must be sought for in that somewhat generalized group. In the general course and character of the main vems, the specimen shows a close approximation to Lithomantis carbonarius H. Woodward,! but differs from all Lithomantide in the definite reticulation of the intercalary veins. The latter character brings it into agreement with the Dictyoneuridz, more especially with Titanodictya gucunda (Scudder) Handlirsch. The last-mentioned author has already shown ? that the wing-structures of the Dictyo- neuride and Lithomantide are so closely alike that the two families might be combined, were it not that in some of the Lithomantide the: form of the body differs strikingly from that of Dictyoneurids. The Ravenhead specimen possesses characters peculiar to the two families. The Dictyoneurid wing is usually quite three times as long as broad, while in the Lithomantids the hinder wings are much broader in proportion to their length. The almost straight posteriorly-directed anal veins of the specimen and the great depth of the wing are alike Lithomantid in character; the wings are markedly aaverca oneannonns in their inner half, and I feel, therefore, justified in assuming that they are the mner of two pairs almost certainly possessed by the insect. The true relationship seems to lie between the Dictyoneurid genus Trtanodictya and the Lithomantid genus Lithomantis. The specimen cannot properly be included in either, and a new genus is, therefore, necessary. To it I give the name of Pale@o- mantis.® Its characters may be described as follows :—Wings short, and very broad in proportion to their length. The radius, radial sector, and median are powerful veins with few divisions; the cubitus is a weak ven. Anal veins directed almost straight imwards. Intercalary venation forming an irregular 1 Q. J. G. S. vol. xxxii (1876) p. 60 & pl. ix, fig. 1. 2 «Revision of American Paleozoic Insects’ Proc. U.S. Nat. Mus. vol. xxix (1906) p. 675. 3 wzaXatos = ancient. part 1] FROM THE BRITISH COAL MEASURES. 53 meshwork of polygonal cells. The genus is probably one which has been developed from ancestors belonging to the family Dictyoneuride in the direction of Lithomantis. To the species I give the name macroptera,! and define it as follows :—Wings twice as long as broad. Apex bluntly rounded. Divisions of radial sector occupying almost the whole tip of the wing. Cubital and anal veins weak. Type-specimen in the Derby & Mayer Museums, Liverpool. Locality.—Ravenhead railway-cutting, near St. Helens. Horizon.—Middle Coal Measures. SPILAPTERA SUTCLIFFEI, sp. nov. (PI. IV, fig. 1.) Among the various organisms obtained from nodules from a stratum a little above the Arley Mine in the Middle Coal Measures at Sparth Bottoms, Rochdale (Lancashire), is a small, irregular, Fig. 4.—Restoration of the left wing of Spilaptera sutcliffei, sp.nov. (Natural size.) micaceous sandy nodule containing the basal portion of an insect- wing. Owing to the coarse granular structure of the nodule, the finer details of the wing-structure are not so well preserved as is usually the case. The chief veins of the wing are robust struc- tures, and these are, fortunately, well marked and clear. The portion of wing preserved is 35 millimetres in length along the costal margin. Its greatest width is 27 mm.; but, as the inner margin is not present, the total width of the base of the wing must have exceeded 30 mm. ‘The wing is a left wing. A little more than a third of it is preserved, and the total length of the whole wing (when complete) must have been at least 9U min. The insect must, therefore, have had a spread of wing of nearly 200 mm., or about 8 inches. The costal margin is very feebly convex forward, forming, indeed, almost a straight line. The sub-costa runs somewhat parallel to the costal margin, and is sunk in a groove. It gradu- ally approaches the costal margin distally, but to so slight an extent as to render it fairly certain that the sub-costa reached ‘ 1 nakpds = long; trepdv = a wing. 54: MR. H. BOLTON ON SOME INSECTS [vol. lxxii, almost, or quite, the apex of the wing. The sub-costa gives off a series of straight veins, which pass obliquely outwards to the costal margin. How numerous these veins were it 1s not possible to say, as the matrix obscures much of the finer detail. The radius is a strong vein, standing out above the general wing-surface, and keeping : a fairly parallel course to the sub-costa and the wing-margin. It gives off the main stem of the radial sector at an acute angle, at about 22 mm. from the base of the wing. The base only of the radial sector is shown. ‘This arises from the radius at a point about 9 mm. in front of the broken edge of the wing. The median vein is the most important vein of the wing. It arises close to the radius, and continues parallel to it for a short distance, and then forks into two equal branches, both of which become widely separated from the radius, and one from the other. The anterior branch gives off an outer twig opposite the point of origin of the radial sector, which it approaches very closely. The inner branch of the median passes in a straight line obliquely out- wards and downwards in a direction which would bring its ultimate twigs out upon the middle of the inner wing-margin. It gives off, near the base, a posteriorly-directed vem which curves sharply away from the main branch and then passes outwards and back- wards parallel to it, the sharp outward curve of the vein causing the formation of a very wide interval between the two. The cubitus evidently divides quite close to the base of the wing. The outer branch passes backwards and outwards in a gentle double curve, giving off two simple twigs from its inner side. The inner of the two main branches of the cubitus passes obliquely backwards, divided into two twigs near the base of the wing, and the outer of these divides again before reaching the wing-margin. The anal veins were few in nuwber, remains of four only being left. These all pass obliquely backwards, and the outermost forks before reaching the middle of its length. The remaining three are evidently undivided. Very few traces of intercalary venation are distinguishable. Allusion has already been made to a few oblique veins, which cross the intercostal area. The only other intercalary veins are a few almost straight and well-spaced cross-veins which bridge over the wide interval between the innermost twig of the median and its fellow. Affinities.—The mode of division of the radius and the median, the wide area between the inner divisions of the latter and the cubitus, the manifest importance of the median, and the few widely-spaced cross-veins, are all typical Spilapterid characters. The wing cannot be confused with Stenodictya, in which the median is much less developed, and the intercalary venation reticulate. From Becquerelia the wing is distinguished by the non-union part 1] FROM THE BRITISH COAL MEASURES. 5D of the anal veins. I have no hesitation, therefore, in assigning the species to the genus Spz/aptera, and would add the specific name of swtcliffec, in memory of the late W. H. Sutcliffe, who helped so materially to work out the fauna of Sparth Bottoms. This genus contains three known species. From Sp. packardi Brongniart the specimen now described is easily distinguished by the sub-costa being parallel to the costal margin; by the equal separation of the sub-costa, radius, and median in the basal third of the wing; and by the division of the median into two branches quite close to the base. Sp. libelluloides Brongniart is only known from the distal half of a wing, in which the cubitus is a smaller vein dividing in a feeble fork some distance out from the base; whereas in Sp. sutcliffec the cubitus is much stronger, and divides into two, or possibly three, main branches close to its point of origin. Sp. venusta Brongniart is a species established upon little more than the distal half of a left wing. The sub-costa is a much shorter vein than in Sp. swtcliffec, and the radial sector comes off from the radius nearer the middle of the wing. In Sp. sutcliffec the outward branch of the median forks almost in line with the point of origin of the radius—a feature not seen in Sp. venusta. Diagnosis of species:—Sub-costa subparallel to costal margin, median vein dividing near the base into two branches, of which the outer forks just beyond the point of origin of the radial sector. Cubitus a large and much-divided vein. Anal veins few in number. Ty pe-specimen in the Manchester Museum, Reg. No. L. 8197. Locality.—Sparth Bottoms, Rochdale. Horizon.—Middle Coal Measures. HyYPERMEGETHES NORTHUMBRIA, gen. emend. et sp.nov. (PI. IV, figs. 2 & 3.) This specimen consists of a portion of the basal half of a left wing, contained in two fragments of an ironstone nodule, of which one bears the impression andthe other the wing-fragment. The two do not wholly coincide, and consequently the outline drawing of the wing-fragment is built up from the two halves of the nodule. The inner margin, anal area, and base of the wing have been lost, so that a little less than two-thirds of the outer portion of the basal half of the wing is present. The wing was of great leneth, the portion remaining being 63 mm. long, with a depth of 31 mm. at its widest part. The whole wing would have a length of about 126 mm., or 5 inches, and the insect a span of wing of close upon 11 inches. The wing shows the basal portions of the costa, sub-costa, radius and radial sector, median and cubitus; possibly there is a trace of a portion of an anal vein. [My views upon the structure and relationship of this wing have undergone a change since reading the paper before the Society on March 8th. At that time all my efforts to determine the character of the intercalary venation had ended in failure, while 56 MR. H. BOLTON ON SOME INSECTS [vol. lxxu, I had been unable to determine the exact condition and course of the veins in the radial sector and median areas. Dr. F. A. Bather afterwards drew my attention to two photographs made of the specimen at the British Museum (Natural History) in June 1914, in which the interealary venation was clearly shown. ‘The detailed structure of the wing was brought out by the simple expedient of photographing the specimen while it was immersed in water. From a study of the British Museum photographs, kindly lent for the purpose, many of the doubtful points of the wing have been cleared up, and its nearness to Hy yper megethes schucherti Hand- lirsch made more certain. The wing is now redescribed in the light of the knowledge thus obtained.—H. B., March 31st, 1916. | The costa is moderately convex from ibs base to a distance of about 380 mm., beyond which it becomes perfectly straight. Separated from ‘he costa by a very wide area basally is the sub- costa, an extremely feeble and hardly distinguishable vein. It passes straight out to meet the costal margin some distance beyond the middle of the wing. Fig. 5.—Restoration of the left wing of Hypermegethes northumbrie, sp. nov. (Natural size.) The radius arises close to the sub-costa, and remains parallel to it along its whole length. It gives off two branches posteriorly, the more proximal branch passing obliquely inwards towards the inner portion of the wing-apex; while the second or distal branch arises from the radius a little farther out, and keeps parallel to it. I had formerly considered the proximal branch of the radius as the radial sector, and the distal one as a simple branch of the radius only. If the wing be closely related to that of Hypermegethes schucherti, which now seems most probable (as will be seen later) my former view will not hold. It would seem that the proximal branching vein must be regarded as the main stem of the median vein, which has entered into union with the radius, and that the distal branch is the radial sector. Regarding the proximal branch of the radius as the median vein, I find that it diverges widely from the radius. It gives off a forward twig, which runs parallel to the radial sector, and then continues inwards until it cuts across the next vein, the two veins part 1] FROM THE BRITISH COAL MEASURES. 57 being perfectly united at the point of intersection. A comparison of this assumed median vein with that of AH. schucherti is in- structive. In the latter species the branching of the main stem of the median vein arises nearer the base of the wing than does the branching-off of the radial sector. It is, therefore, somewhat in the position of the starting-point of the median vein as a free structure in the present wing. The median vein of AH. schucherti has, however, no union with the radius at all. In that species the main stem of the median runs out towards the wing-apex parallel to the inner branch of the radius, giving off a backward branch which passes obliquely towards, and unites with, the anterior branch of the cubitus. Inthe middle of its length it gives origin to a twig running parallel to the main stem, and midway between that vein and the cubitus. The course of the median vein in the specimen here described is exactly similar to that of the branch of the median vein in H. schucherti, except that the inner branch not only unites with the anterior branch of the cubitus, but continues on the inner side of the latter vein beyond the point of union. The conditions may be summarized by saying that in H. schucherte the median vein is free along its whole length, and gives off an inner branch which forks into two twigs, of which the inner unites with the cubitus; while in the specimen here described the median is united with the radius for some little distance, becoming free before the radial sector is reached and then forking, the inner twig uniting with and crossing the cubitus. The next free vein to the radius is the cubitus. It has a short stout basal stem, which forks into two equal and somewhat widely separated branches, the outer, as I have already noted, uniting with the inner branch of the median. The inner branch shows the beginning of a bifurcation on the broken edge of the wing. ' Nearer the inner margin of the wing are traces of two other veins: the first has its basal portion missing, and follows a course parallel to the inner branch of the cubitus. It is strongly forked. The remaining vein is represented by three detached fragments. If Handlirsch’s interpretation of the wing of H. schu- cherti is followed, we should regard both these veins as anal. Jam, however, of opinion that, while the innermost fragmentary vein may be an anal one, the forked vein, by its position, its forked character, and its stoutness, must be regarded as a portion of the eubitus. I am likewise of opinion that the first, and possibly the second, of the veins marked as anal in Handlirsch’s figure of H. schucherti ought also to be classed as cubital. Both in H. schucherti and in the specimen here described the anal area would have an enormous development and occupy most of the wing-margin if the veins alluded to were wholly anal in character. I feel sure that, if the vein nearest the cubital had been better preserved, it would be found in actual union with the latter at the base of the wing. The interealary venation, the knowledge of which I owe to Dr. F. A. Bather, is peculiar, and almost identical with that of 58 _ MR. H. BOLTON ON SOME INSECTS [ vol. Ixxii, Hypermegethes. The intercostal area is filled by an irregular meshwork of fine veins, with a tendency on the outer and inner sides to a transverse arrangement. Between the median and the cubital the intercalary veins are short, straight, and transverse, and feeble traces of similar veins can be seen in the median area. The cubital area is filled in with a meshed venation, larger and more regular than that of the intercostal area, and this seems to. continue into the anal area. Affinities.—The Palzodictyopteroid characters of the veins. of this wing are clear and unmistakable. The general course and branching Gflthe main veins are undoubtedly like those of Hyper- megethes, and now that the interealary venation is known, this. relationship is put beyond doubt. The points of relationship: between Hypermegethes schucherti and this wing may be enu- merated as follows :— Intercostal area wide and filled in by an irregular meshwork of subsidiary veins. In each case the radial sector comes off farthest from the base of the wing, the forking of the median next, and the forking of the cubitus nearest. the wing-base. The radial, median, and part of the cubital areas are occupied by relatively few transverse veins. The inner half of the wing in each case is filled in by a network of more regular and larger-meshed veins than is the intercostal area. In both cases the median has an inner branch, which enters. into union with the anterior branch of the cubitus. The radius, median, and cubital elements are in both cases closely compacted together, with narrow areas, while in the rest of the wing the veins are well spaced. This assemblage of characters common to H. schucherti and the present specimen quite justifies the inclusion of the latter in the genus Hypermegethes; but, in order to do this, it is. necessary to modify somewhat the diagnostic descriptions of the family Hypermegethide and the genus Hypermegethes. I would, therefore, define the family Hypermegethide as. follows :— Costa marginal, costal area broad, radius simple, radial sector present, median probably dividing at the base into two or more main branches, the first: of which may become united to the radius ; cubitus forked near the base, with its branches widely spaced. Anal veins few, and anal area not exceeding a third of the inner margin.! . Genus HypERMEGETHES. Costal border feebly convex, sub-costa and radius close together along the greater part of their length. Outer median vein fused with tadius or separate. _ Cubitus consisting of two, possibly more, parallel and simply forked branches, dividing near the point of origin. Anal veins few. The specimen here described differs from HA. schucherti in the 1 The general assemblage of characters found in Hypermegethide is, I believe, Sinlelhlie suggestive of the Protodonata, but no definite conclusions. can be formuiated ur til the whole wing is known. part 1] FROM THE BRITISH COAL MEASURES. 59) close approximation of the sub-costa to the radius along its whole length, in the outer median vein being fused for a portion of its length with the radius, and in distally uniting with and crossing the outer branch of the cubitus. These diagnostic features are sufficiently important to justify the creation of a new species, to which I give the name of Hypermegethes northumbria. Ty pe-specimen in the possession of Mr. William Eltringham. Loeality.—Pheenix Brick-works, Crawerook (Durham). Horizon.—Shale above Crow Coal. PsEUDOFOUQUEA CAMBRENSIS (Allen). (PI. IV, figs. 4 & 5.) 1901. guaiee cambrensis H. A. Allen, Geol. Mag. dec. 4, vol. viii, p. 65, text- 1906. ene cambrensis (Allen) Handlirsch, ‘ Die Fossilen Insekten ” p. 125 & pl. xiui, fig. 5. This is a broken left fore-wing, of which the two parts are preserved upon the surface of two small fragments of black shale. One fragment, bearing the Fig. 6.—Restoration of the left basal part of the wing, is in wing of Pseudofouquea cam- the Museum of Pactical Geo- brensis (Alle). logy; the other, containing the outer 28 mm. of the wing, is in the Welsh National Museum, Cardiff. Allen referred the wing to Charles Brongniart’s genus Fouquea, comparing it with F. lacroizi Brongniart, but drawing attention to the marked difference of the cubitus in the two species and to the greater number of twigs into which the main veins divided in P. cambrensis. Handlirsch repeats Allen’s figure, and places the species in a new genus, Pseudofouquea. The total length of the wing remaining is 32 mm., so that 9 mm. have been lost from the tip of the wing in the Geological Survey specimen. The total length is given by Allen as 41 mm. The greatest breadth is 15 mm. In the costal area are very faint traces of a feeble venation of cross-veins, which bend obliquely outwards. In the area between the sub-costa and the radius the veins in the proximal part are transverse, becoming irregular in direction farther out. The rest of the intercalary venation is now so feebly marked as to be indistinguishable, except between the base of the anal vein and the main stem of the cubitus, where it is irregularly reticulate. This is not shown in Allen’s figure. The costa is fairly convex, most conspicuously so in the proximal half. The sub-costa is somewhat widely separated from the costal proximally, and gradually approaches it along its outward course. As the costal border dips backwards about the middle of its length, | Natural size. ] 60 MR. H. BOLTON ON SOME INSECTS (vol. lxxii, the inner half of the intercostal area is double the width of the outer half. The radius isa simple vein, giving off the radial sector at the base, and then passing out towards the tip of the wing parallel to the sub-costa, from which it remains equidistant along the whole to} of its course. The radial sector diverges widely from the radius, and now shows but one posteriorly-directed branch, which forks just as the broken edge of the wing is reached. So widely does the radial sector diverge from the radius, that the area included between them is equal in diameter, immediately beyond the forking of the radial sector, to the whole area lying between the radius and the costal margin at its widest part. Allen shows that two additional undivided twigs were sent off posteriorly from the radial sector, beyond which is now the broken edge of the wing The radius and the radial sector occupied almost the whole tip of the wing. The median vein forks ata quarter of the length of the wing from the base, the outward branch again forking before the middle of the wing is reached. The two twigs So) produced pass backwards and outwards to the inner wing-margin, keeping parallel one to the other and to the innermost twig of the radial sector. The iner branch of the median diverges almost in a straight line from the outer branch, so that a wide area lies between them up to the point at which the inner branch forks. The forking of the inner branch occurs farther out than in the outer branch, and forks a second time before the inner wing-margin 1s reached. The cubitus isa remarkable vein. For nearly half its length it passes in a broad curve outwards and backwards towards the inner wing-margin, without forking. Beginning at the middle of its length, the cubitus gives off alternate twigs upon its outward and inward sides, those on the inner side being much feebler than those on the outer. The two outward twigs are strongly developed, and sweep in a wide curve outwards and down to the wing-margin. The twigs from the inner side of the cubitus are ane in sammnalben, and pass directly inwards to the wing-margin, with the exception of the last, which seems a feeble continuation of the main stem, and curves outwards like the two outer twigs ; the course of the inner four twigs is therefore much shorter than that of the outer twigs, while the area that they occupy on the wing-margin is also much less. This is a feature entirely unlike anything seen in Fouguea. Four anal veins are distinguishable. The inner two arise from a common base; the remaining two are distinct. This, again, is unlike what is seen in Fouquea, where the anal veins branch off regularly from a single stem. The area lying between the first anal vein and the main stem of the cubitus is very wide, much wider, indeed, than any other area in the wing. Quart. Journ. Geor. Soc. Vor. LXX!I, Pc.Ill. JW. Tutcher, Photo Bemrose, Colic Derby COAL- MEASURE INSECTS. (All the figures are of the natural size) Quart. Journ. Geot. Soc. Vor. LXXil, Pr. IV. PE eA CR BO > Soe 2 ace ; J. W. Tutche (cp Photo Bemrose, Collo Derby. COAL- MEASURE INSECTS. (Allthe figures are of the natural size.) part 1] FROM THE BRITISH COAL MEASURES. 61 Affinities.—The sub-costal vein reaches the costal margin at about 12 mm. from the apex of the wing—a feature noted in Fouquea. The radial sector originates nearer the base of the wing than is shown in Handlirsch’s figure. The actual junction is not shown in Allen’s figure. It must have been near the base of the wing, much nearer the base than is seen in Fowquea. The characters of the cubital and anal veins definitely remove the species from Houquea, as 1 have already noted; indeed, the cha- racter of the cubitus, with its strong anteriorly-directed twigs and its feeble inner branches, is wholly unlike that of any other insect, and would alone suffice to justify new generic rank. So far I am in agreement with Handlirsch; but I regard the enlarged areas between the inner divisions of the radial sector and the cubitus, and between the cubitus and the anal veins, as more suggestive of the Proto-Orthoptera, notably Thoronysis ingbertensis Ammon. More than this cannot be said, and Pseudofouquea cambrensis must be regarded provisionally as. Palseodictyopteroid, with a possibility of Proto-Orthopteroid affinities. Holotype specimen: outer half of a wing in the Welsh National Museum, No. 13,120. Middle portion of the same wing in the Museum of Practical Geology, Jermyn Street, No. 7272. Locality.—ULlanbradach Colliery, near Cardiff. Horizon.—Top of the Four-foot Seam, Lower Coal Measures. EXPLANATION OF PLATES III & IV. | All the figures are of the natural size. The photographs reproduced in these plates were prepared by the aid of a grant from the Royal Society. | Pate II. Fig. 1. Left wing of Hdewophasma anglica Scudder. (See p. 43.) 2. Basal portion of a left wing of (Dictyoneuron) higginsii (Handlirsch). (See p. 46.) _ 3. Left half of an ironstone nodule, showing a left wing-fragment of Palzomantis macroptera, gen. et sp. noy., lymg upon the right wing, the latter with its underside uppermost. (See p. 48.) 4, Impression, in the right half of an ironstone nodule, of the greater part of the upper surface of the left wing of Paleomantis macroptera, gen. et sp. nov. PLate LV. Fig. 1. Basal portion of a left wing of Spilaptera sutcliffei, sp.nov. (See p. 53.) 2. Left half of a nodule, showing portion of a left wing of Hy perne- gethes northumbriz, sp. noy. (See p. 55.) . Right half of a nodule, showing portion of a left wing of Hyperme- gethes northumbriz, sp. nov. 4, Portion of a left wing of Psewdofouqwea cambrensis (Allen). (See p. 59.) . Tip of a left wing of the same. (Js) Or 62 INSECTS FROM THE BRITISH COAL MEASURES. [vol. Ixxu, DISCUSSION. Mr. G. W. Youne welcomed the paper as another contribution by the Author on a branch of paleontology but little studied in this country. This was the more to be regretted, as much had been written upon it on the Continent and in America; and he could not help thinking that British entomologists were neglecting opportunities, since we had much excellent material available for study. The species dealt with in the paper showed, as usual, that the Paleozoic insects were so generalized that it was difficult to place them in any of the usualiy-recognized orders; and they embraced not only the ancestral forms of Orthoptera, Neuroptera, and Hemiptera, but many other types that had not survived beyond Carboniferous times. Mr. J. H. Durrant also spoke. part 1] RIVER-GORGES IN CORNWALL AND DEVON. 63 5. On the OrtetIn of some RivER-GoRGES 7m CORNWALL and Devon. By Henry Dewey, F.G.S. (Read February 23rd, 1916.) [Puares V-—VII.] THE river-gorges here described are situated for the greater part in North Cornwall, between Boscastle and Tregardock, though a few occur in other parts of Cornwall and in Devon. Some of the characteristics of the coastal gorges of Devon have already been described by Dr. H. A. N. Arber,! but there appear to be additional features in those of North Cornwall, which are possibly due to their mode of origin. These gorge-like valleys are more or less restricted to areas of peculiar topography, and their origin is connected with the history of that topography; while they form a sharp contrast with the vailey-systems where such a topography does not exist. Thus, the valley-systems of Cornwall are divisible into two groups, each being marked by features peculiar to it; although in some cases the river-valley exemplifies the two types of scenery in different parts of itself. The difference is in part due to their respective stages of develop- ment, but mainly marks a renewal or rejuvenescence of the rivers’ activity. he fall of these rivers near Tintagel averages 1 foot in every 13 feet. The gorges are incised in an upland plain, which truncates all hills at a uniform height of about 430 feet above the level of the sea, while the other series of valleys lies im an area diversified by many irregular hills and buried and submerged estuaries. THe Urpnanp Prat. This feature has become well-known in Cornwall through Clement Reid’s? descriptions.. It is everywhere bounded landwards by a prominent ridge, marking a range of degraded cliffs, which also form the boundary of the hills rising above it to a height of 1000 feet above sea-level. This planation has affected alike rocks of different degrees of hardness, reducing them all to a level surface, and cutting a notch in the higher ground beyond them. When viewed in profile, and especially from places a few miles away, this feature is very conspicuous, as when the Tintagel country is seen from Pentire Head, or the Land’s End from Carn Brea. From Boscastle the plain is continuous for several miles towards the south, where, at Tregardock, the sea has encroached on the 1 80 diameters. (See p. 81.) DIscusston. The Presrmpent (Dr. A. Harker) complimented the Author on his paper, and commented upon the various points of interest. Prof. C. G. Cutis, after congratulating the Author, referred 1 «Hxperimentelle Untersuchungen iiber die Bildung der Minerale im Magma’ Tschermak’s Mitt. vol. xviii (1898) pp. 1-90 & 105-240; also T. A. Jaggar, Journ. Geol. vol. vii (1899) pp. 300-13 ; see also Sir Jethro Teall on ‘The Natural History of Cordierite & its Associates’ Proc. Geol. Assoc. vol. xvi (1899-1900) pp. 61-74. Quart. Journ. Geor. Soc. Vor. LXXII, PrVill. Fig 2. OrtHocLase, Quartz, AND RHOMBOHEDRAL CARBONATE SURROUNDED BY BIOTITE. x34. af 4 Cie H.G.S., Photemicre. Bemrose, Golo. Derby. v We eae : a Bey a3 aleve Rh eae eee Vn PALS aa na fi ‘ pnt Ls Quart. Journ. Geov. Soc. Vor. LXXII, Pr. IX. Fic. |. INCLUSION IN THE LuRGeEcomBE Mitt LAMPROPHYRE. x 25. Fic.2. TuBuULAR CONNEXIONS BETWEEN STAUROLITE AND BIOTITE. x so. H.G.S., Photomicro-_ Bemrose, Colla. Derby. i ata oneal Dt he feo ne hie Ghee Avg part 2] . LURGECOMBE MILL LAMPROPHYRE. 83 to the peculiar chemical composition of the inclusions as indicated by their minerals. They were poor in silica, but rich in alumina and iron, and possibly magnesia. In view of this last constituent, the fact that cordierite had not been observed was noteworthy. Possibly some of the patches now consisting of micaceous aggregates were pinite after that mineral. Suppesing the inclusions to have been shale in the first place— which seemed most likely from the field-evidence—considerable interchange between magma and inclusion must have taken place. Inclusions of Skiddaw Slate in the Threlkeld microgranite offered a marked contrast to this, having been altered, by the addition of a little boron, merely to tourmaline and secondary quartz, a mineralization involving practically no change in bulk-composition. It would seem that an acid magina, already rich in silica and alumina, had little solvent or metasomatie effect upon argillaceous inclusions, while a more basic magma, deficient in silica but saturated with alumina, had both a solvent and a metasomatic action upon such fragments, the silica being leached out, and its place taken by iron, and to a smaller extent by magnesia, to form such minerals as magnetite and staurolite, while the alumina remained behind in the erystallized condition as corundum. Dr. J. W. Evans thought that it would not be right to ignore the possibility that the xenoliths might represent iron-ore from a mineral vein traversed by the intrusion, and thus the corundum might be the result of the removal of silica from the magma in combination with oxide of iron. A still more speculative hypothesis was that they might originate from inclusions of sedimentary rocks, containing free hydrates of iron and alumina, and allied in their nature and mode of formation to laterites. The AurHor agreed with the President that the included fragments were probably derived from some underlying horizon, but considered it unnecessary to assume that this was very different in lithology from that of the shales now exposed in contact with the intrusion. i In reply to Prof. Cullis, he considered that the micaceous ageregates of some of the inclusions might represent altered cordierite, but was not prepared to go further. He thought also that it was extremely improbable that analysis of the exposed country rock would give any useful information as to the nature of the material caught up by the magma. Although this material was probably argillaceous, its analysis might differ considerably from that of the exposed shales. Replying to Dr. Evans, he said that the extinctions of the plagioclase “cores indicated a composition approximating to that of labradorite. He agreed that the percentage of potash was unexpectedly low—much too low to justify the name of minette. An attempt to incorporate fragments of a laterite would account for some of the inclusions, but in an argillaceous country, probably quite capable of supplying the necessary material, such a mode of origin was only remotely possible. SA or MR. G. W. TYRRELL ON THE [vol. lxxu, 7. The Prcrtrn-TescHEeNttEe Sitn of Luear (AYRSHIRE). By GroRGE Water Tyrreci, A.R.C.Sc., ¥.G.S., Lecturer in Mineralogy and Petrology in the University of Glasgow. (Read April 5th, 1916.} [PLATES X & XI.) CONTENTS. Page diy AbMMAROYS LCI HKOIG SoG onbuccnnancidonseocens cuocoeoa see deotoo aco and 84 Mle sHaeldelationse ker eco omae mrt Glen cee as see ene 85 (1) The Glenmuir Section. (2) The Bellow Section. (38) Other Hxposures. (4) Summary. REE Petrocrap lay smi eects ee wiichoimatn sake card teecemah ae 95 (1) .The Contact-Rocks. (2) Teschenite. (3) Theralite. (4) Lugarite. (5) Picrite and Peridotite. TEVie SRC tO Om yar sea oe eer RON Te IOC OR Deco AT a RN aS 114 (1) Mineralogical Variations. (2) Chemical Variations. (3) Average Magma of the Lugar Sill. (4) Composition, Identity, and Banding of the Contact-Rocks. (5) Asymmetry. (6) Density-Stratification. (7) Variations in Texture. (8) Segregation- Veins. (9) Mode of Intrusion and Differentiation. (10) Sinking of Crystals in the Central Ultrabasic Stratum. (11) Comparison of the Lugar Sill with the other Picrite-Teschenite Sills of Scotland. I. InrRoDUcTION.. THE association of teschenite and ultrabasic rocks (picrite and peridotite) in a single rock-body has now been established for several occurrences in the lowlands of Scotland. The Barnton occurrence, near Edinburgh, has been described by Sir Archibald Geile! by Mr. J. Henderson & Mr. J. G. Goodchild,? and by. Mr. H. W. Monckton. At Blackburn, near Bathgate, occurs a picrite, which has been described by the first-named writer as a lava,* but has recently been shown to be intrusive ang associated with teschenite by the officers of the Geological Survey of Scotland.® 1 «Ancient Voleanoes of Great Britain’ vol. i (1897) pp. 449-50, * Trans. Geol. Soc. Edin. vol. vi (1893) pp. 297-300 & 301-302. 3 Q. J. G. S. vol. 1 (1894) p. 39.. + * Ancient Volcanoes of Great Britain’ vol. i (1897) p. 419. ° ‘Summary of Progress for 1904’ Mem. Geol. Surv. 1905,"pp, 118-19. part 2] © PICRITE-TESCHENITE SILL OF LUGAR. 85 The famous picrite of Inchcolm, in the Firth of Forth, is well known from the descriptions of several observers. Dr. R. Campbell and Mr A. Stenhouse, however, in a recent detailed investigation of the island, have shown that at both the upper and lower contacts the picrite passes into teschenite.! On the west coast, Dr. J. D. Falconer has described a picrite-teschenite sill at Ard- rossan, intrusive into the Carboniferous Limestone Series: here again the ultrabasic rock occupies the central part of the mass.? Recently the late R. Boyle drew attention to still another occur- rence at Lugar, near Old Cumnock (Ayrshire). He described the passage of dolerite and basalt [teschenite] through doleritic picrite [theralite| to ‘segregated’ masses of peridotite or picrite. The ultrabasic rock occurs in the central parts of the mass, and passes eradually to less basic varieties towards both upper and lower contacts.2 A picrite associated with teschenite has been dis- covered by Mr. EH. M. Anderson at the Inner Nebbock, Saltcoats (Ayrshire) .* In the course of an investigation of the Permo-Carboniferous alkalic rocks of the West of Scotland® I made a detailed exami- nation of the Lugar sill, and found in it an extraordinary complex of various rocks belonging to the analcite series. This included normal and melanocratic teschenites; a facies with abundant nepheline and ferromagnesian minerals—essentially a melanocratic theralite; and a curious rock composed mainly of analcite and nepheline, with subordinate plagioclase, titanaugite, and barkevikite in very perfect crystals. This unique rock, which it 1s proposed to call lugarite, has now been found in several localities in the West of Scotland. Extremely fresh hornblende-picrites and peridotites, however, form the major part of the intrusion. __ The present paper embodies a complete description of this sill, and attempts an explanation of the processes whereby the different facies have been developed. A comparison with the other occur- rences is also instituted. Five chemical analyses have been made in the course of the investigation, and in connexion with these and for the work in general I have to acknowledge the aid of a grant from the Government Grant Committee of the Royal Society. For the analyses I am indebted to Dla skill of Dr. Alexander Scott, of Glasgow University. Il. Frevp RELATIONS. The intrusion which is the subject of this paper occurs near the village of Lugar in Central Ayrshire, near the eastern border of the county. It is intruded as a sill into a crumbling white and yellow 1 «The Geology of Inchcolm’ Trans. Geol. Soc. Edin. vol. ix, pt. 2 (1908) p. 121-34. 3 ‘The Geology of Ardrossan’ Trans. Roy. Soc. Edin. vol. xlv, pt.1 (1907) pp. 601-10. 3 Trans. Geol. Soc. Glasgow, vol. xiii (1908) pp. 202-23. 4 «Summary of Progress for 1911’ Mem. Geol. Surv. 1912, p. 52. > Geol. Mag. dec. 5, vol. ix (1912) pp. 69-80, 120-31. 86 - MR. G. W. TYRRELL ON THE [vol. lxxii, sandstone belonging to the ‘ Millstone Grit.’ When the area is re-mapped it is probable that these strata will be incorporated in the Coal Measures. The intrusive nature of the igneous rock is proved by its increasing fineness of grain towards both upper and lower contacts, and by the marginal fringes of hardened sandstone above and below the sill. Definitely transgressive contacts are not seen, although in one place a thin band of hardened sandstone is encountered about 12 feet from the upper margin of the sill. So far as can be ascertained, the intrusion keeps approximately to the same horizon. The whole series dips at about 10° north-west- wards, and, from the width of the outcrop, the thickness of the Fig. 1.— Geological map of the Lugar district, on the scale of 1 inch to the mile, or 1: 63,360. ORY “hd Gein © ane Ri B vig. 4 or wee Log n [T.P.=Teschenite-picrite ; E.D.=Hssexite-dolerite sill. Broken lines represent faults. ] igneous rock is estimated at 140 feet. Apart from small irregular- ities the outcrop forms a crescent-shaped strip about 3 miles long and a fitth of a mile wide at its widest part. It extends in a general north-easterly direction from Lugar to a mile and a half beyond the village of Cronberry. The Bellow Water with its con- tinuation, the Lugar, cuts through both extremities of the crescentic outcrop, and gives sections at Logan Bridge on the south, and at a spot a mile north-east of Cronberry on the north. The river also cuts through a slight bulge of the outcrop at its confluence with the Glenmuir Water. The latter stream cuts a deep trench in a somewhat different direction through the intrusion at this place. The outcrop is faulted continually towards the south-east part 2]. PICRITE-TESCHENITE SILL OF LUGAR. 87 by a series of six west-north-west and east-south-east faults (see fig. 1, p.86).. At the north-eastern extremity of the outcrop the sill is cut by a west-south-west and east-north-east fault which severs it from another mass of igneous rock, extending half a mile back towards the west. This mass, howeren is on a higher horizon, and has been mapped as separating the ‘ Millstone Grit’ from the overlying Coal Measures.! If it be taken as part of the Lugar sill, the total length of outcrop becomes 33 miles. The finest sections are found just above the confluence of the Bellow and Glenmuir Waters to form the Lugar Water. Both streams have eroded deep rocky gorges through the sill, the one in a north-easterly, the other in a south-easterly direction. When the water is low, practically every foot of the thickness can be examined either in cliff-section, or in horizontal water-polished areas of rock. In these circumstances the study of the sill can be conducted with facilities unattainable in any of the other occurrences; and the conclusions as to the origin of the different facies arrived at in this case may be considered sufficiently well founded to apply to the other occurrences, in which, although the exposures are not so good, practically the same structure and disposition can be made out. (1) The Glenmuir Section. The Glenmuir Water, cutting through the sill in a general north-westerly and south-easterly direction at right angles to the strike, gives the most complete and typical section. The upper contact is seen at the weir, just at the confluence with the Bellow Water. Hardened whitish sandstone occurs overlying a dense basaltic facies, in a steep rocky bank on the south side of the river. ‘The chilling influence of the contact, as shown by fineness of grain, extends down about 12 feet, and has doubtless been strengthened by the inclusion in the sill at this depth of a thin band of sandstone now metamorphosed to a hard white quartzite. The contact-rock is distinctly banded in layers, often confused, wavy, and bifureating, which differ slightly in colour and texture. The thickness of the bands varies from several inches to very fine linear streaks but faintly indicated by a slight difference of colour. Besides the normal, greyish-black, aphanitic contact-rock, the chief varieties included in the banded material consist of fine- grained, pinkish, and greenish teschenitic rocks, and a very dense, dead-black, glossy, basaltic material, although extremely slight differences of colour and texture serve to bring out the banded structure. These varieties show no sharp contacts with each other, the transition from one to the other taking place quickly but quite gradually. These bands seem to be true schlieren, due to the flow of a slightly heterogeneous magma. In general the streaks are drawn out in bands parallel to the upper margin of the sill. In addition to the banding, the contact-facies is traversed by 1 See the Sheets of the l-inch Map of the Geological Survey of Scotland, Nos. 14 (1868) & 15 (1870). 88 . MR. G. W. TYRRELL ON THE [vol. Ixxii, numerous veins of coarse-grained teschenite with abundant anal- cite, similar to the rocks described below. The upper teschenite.—From a maximum depth of 12 feet or so in the mass, the granularity of the rock increases very rapidly towards the interior. The contact-facies passes into a coarse-grained mottled rock, the prevailing tint of which is pink or dark green, according as the felsic or mafic minerals dominate the colour. In thin section the rock is seen to be a typical teschenite, composed of essential plagioclase and analcite, titanaugite, and ilmenite, with accessory orthoclase, olivine, barkevikite, and biotite. The shescbintie is evenly granular, although the feispars sometimes tend to take on a lathy or columnar form. In another variety the augite 1s conspicuous as long thin black prisms, ranging up to 2 inches in length. A third variety shows very abundant analcite in large pink masses, which are frequently spherical. The rock seems to become richer in analcite as a greater depth in the sill is attained. ‘The thickness of the teschenite cannot be measured in this section, as the contact with the underlying facies is not well seen; but it is usually between 15 and 20 feet. There are, however, considerable variations in the thickness of this band. ‘Along the western bank of the Glenmuir Water, above the lugarite locality (see p, 91), indurated sandstones and contact- basalts occur less than 20 feet above the picrite. Hence the theralite and teschenite intervening between the picrite and the upper contact must be attenuated, or one of them must be absent in this part of the section. The theralite band.—Underlying the teschenite is a ‘fine- grained, almost aphanitic, grey rock, which, under the microscope, is seen to have the composition of a theralite. ‘The mafic minerals, olivine and titanaugite, are dominant over the plagioclase and nepheline. Barkevikite, biotite, and iron-ores are ¥ather abundant accessories, and there is often a small amount of analcite. This rock forms a band perhaps 10 feet thick; but in the Glenmuir gorge its outcrop mainly occurs high up in a vertical cliff, although it may also be examined in small exposures on the eastern side of the stream, and in the river-bed when the water is low. In the latter it may be easily distinguished by its smooth, polished, water- worn surface, in contradistinction to the coarse-grained, pitted surfaces of the overlying teschenite and the underlying picrite. Further details of the theralite band are reserved for the de- scription of the Bellow section, where it is much better exposed. The picrite and peridotite.—By a gradual diminution in the amount of felspar and nepheline, the theralite passes into coarse-grained ultrabasie rocks, composed mainly of olivine and titanaugite, frequently with abundant barkevikite, some iron-ore, biotite, plagioclase, and analcite. These rocks constitute the major portion of the sill, and occupy the interior of the mass. The upper part 2] PICRITE-TESCHENITE SILL OF LUGAR. 89 part of the ultrabasic stratum is rich in titanaugite and barkevikite, as well as olivine, and usually contains some plagioclase and anal- cite. This is a true picrite in the original sense of Tschermak, who applied the term to a melanocratic facies of the Moravian teschenite. In the remainder, however, plagioclase and analcite completely disappear, and olivine becomes the dominant mineral. This rock is a hornblende-peridotite, in which the alkalic tendency is still recognizable by the unusual amount of alkalies contained in the bisilicates. A tongue-like extension of the ultrabasic outcrop occupies the bed of the Glenmuir Water for some distance north- west of the railway-viaduct, the overlymg facies forming the cliffs at this point. Good waterworn surfaces can be examined here when the river is low. The rock is remarkably fresh, and’ shows sharp and irregular variations in granularity. East of the railway-bridge, the ultrabasic stratum forms nearly the whole of a great cliff rising vertically to a height of 90 feet from the north side of the stream. Here, where it is exposed to the weather and is not being continually scoured by the water, 1t 1s much decom- posed, and weathers to a loose crumbling mass, with spheroidal lumps of harder and fresher rock in places. The north-westerly inclination of the intrusion is well shown in the face of the cliff by the inclination of the joint-planes, but principally by the contrast between the jointing of the ultrabasic rock and the over- lying theralitic facies. The former is traversed by a few large irregular joints; but the latter by many, both vertical and hori-. zontal, causing a rude columnar structure. While the transition ~ between the two is not sharp, it is sufficiently well marked to show the general dip of the intrusion towards the north-west. The ultrabasic rock is traversed by a few thin veins of a fine-grained felspathic rock (teschenite-aplite). . : The lower teschenite.—A blank of about 30 feet in the section separates the last exposure of the ultrabasic stratum from the underlying facies, which consists of teschenite made up, for the greater part, of the variety containing numerous blade-like or columnar augites. ‘This is exposed in a thick bar across the stream near the second right-angle turn above the railway-viaduct: its thickness is estimated at about 20 feet. It is important to notice that no theralite intervenes here between the picrite and the teschenite. There is little reason to suppose that it is hidden by the blank in the section, for the theralite is the most resistant rock in the complex to ordinary atmospheric weathering. The lower contact.—The granularity of the lower teschenite declines rapidly, until it passes into the ordinary contact-basalt. The latter shows banding precisely similar to that of the upper contact, and is also traversed by veins of coarse pink teschenite. The two most prominent varieties involved in the banding, which is frequently much confused and contorted, are a- fine-grained pink teschenitie rock and a dense black or grey basalt, the two rocks Fig. 2.—Section across the Glenmuir Water, near the lower contact of the sill, showing the 90 MR. G, W. TYRRELL ON THE [vol. lxxu, being intimately welded together. Immediately beneath the contact the Carboniferous sediments crop out, dipping westwards ab LO’. This gradual transition from the lower teschenite to a dense basaltic contact-facies can be followed on the eastern side of the stream. On the west- ern side there are much more compli- cated relations. Here a portion in which the dense basaltic schhe- ren are dominant over the coarser teschenitic layers may be distin- guished from one in which the reverse relation holds. The former abuts against the latter horizentally, and overlies and under- lies it. There is also a more or less gradual transition between the two portions. These relations are shown diagrammatically in fig. 2. The coarse- grained teschenitic rock forms conspicu- ous, waterworn knobs in the bed of the stream. As the stream iscrossed to the eastern side the fine-grained basaltic schlieren dis- appear, and the tes- chenite passes down normally, as above described, into its con- tact-facies. This can be traced to a small bar of sandstone seen at low water in the bed of the stream. The relations of the various facies at the lower contact are suggestive of considerable movement in the magma before crystallization had taken place to any extent, for the minerals show little or no alignment in the direction of the banding. S.E. Linch = 8 feet.) (Scale : Glenmuir Water banding of the coarser and finer schlieren. N.W. part 2] PICRITE-TESCHENITE SILL OF LUGAR. 91 The lugarite.-—On the west bank of the river, under the railway-bridge and a little to the north side of it, the cliff exhibits a sharp horizontal junction between decomposed picrite and a peculiar rock overlying it. The latter shows a greyish or greenish aphanitic ground-mass, crowded with both stout and slender black prisms. Sometimes small rectangular whitish felspars are seen also. In thin section the ground-mass may be identified as analcite, crowded with needles of apatite, and containing some nepheline. In it are scattered numerous perfectly euhedral prisms of barke- vikite, with subordinate titanaugite and plagioclase, also extremely well formed. This rock forms a stratiform mass about 4 feet thick overlain by the theralitic facies. The junction is quite sharp, as the two very distinct rocks can be got within an inch of each other; but the line of demarcation is, in general, not marked by any definite feature on the face of the cliff. The two rocks are intimately welded without any appearance of passage. The long prisms of barkevikite, however, project from the lugarite into the theralite. At places the junction is marked by a thin, fine, horizontal joint-plane, or by a thin layer of slightly decomposed rock. The contact with the picrite is doubtless of the same nature, but the latter is so decomposed that the junction is quite obscured, Some veins of lugarite penetrating the picrite supply further evidence on this point. ‘These veins or small dykes range up to 4 inches in width, and are largely composed of analcite and barkevikite prisms, which are frequently arranged in stellate groups. The contact of these veins with the picrite is an intimate welding, and the barkevikite crystals project from the sides of the veins into the enclosing rock. The lugarite is also to be found in the Bellow section in the same relative position, but rather poorly exposed. The Glenmuir exposure is only accessible when the stream is low; but south of the viaduct occur numerous fallen blocks from an inaccessible exposure in the face of the cliff. In the other direction the dip of the intrusion carries the lugarite down to the water’s edge, where it is lost. a All the available evidence goes to show that this remarkable rock is intrusive in the picrite. The veins of similar material traversing the picrite transgress sharply the different varieties of the ultra- basic rock. There can be no doubt that these veins proceed from the main mass of lugarite, although the junctions have not actually been demonstrated. While the rock is clearly intrusive, it is but slightly posterior to the main phenomenon of intrusion, for the intimate welding and absence of chilled rock at the margins indi- cates that the ultrabasic rock was still very hot at the time of intrusion. (2) The Bellow Section. Although not so complete, this section supplies many details which are missing or obscure in the Glenmuir section. Just 92 MR. G. W. TYRRELL ON THE [ vol. lxxii, above the confluence with the Glenmuir Water, the upper contact of the sill with a soft, yellow, crumbling sandstone is well seen. The contact-facies is a brownish aphanitic rock, with some sporadic flakes of biotite. It is obviously much more decomposed than the corresponding rock in the Glenmuir section. ‘The flow-banding also is not at all prominent. The contact-facies passes down quickly first into a fine-grained pink and green mottled teschenite, and then into a coarser rock with a conspicuous development of large columnar augites. Although this is the dominant facies there are a few bands of fine- grained material, which are doubtless due to the flow of a slightly heterogeneous magma. Still lower down the rock becomes distinctly more felsic, and shows abundant pink and white spots of analcite. In this variety the augite-crystals are not columnar. This analertic facies is in sharp contact with the dark, fine- grained, theralitic stratum beneath. The stream has cleared this contact rather thoroughly, owing to the differential resistanee of the rocks to erosion. It slides down a polished waterworn surtace of theralite, which is bordered on one side by a low cliff of the coarser” and softer teschenite. Hence the contact can be particularly well examined. The theralite is here seen to be shot through or per- meated with irregular masses, patches, nests, strimgs, and anasto- mosing veins of a pinkish, coarse-grained, felsic rock, resembling the overlying teschenite. These patches and veins do not appear to be intrusive. ‘They have no sharply-defined contacts with the theralitic facies, but the minerals interlock at the margins, welding the two rocks into an intimate union. There is no sign of chilling at the margin of the patches, the normal granularity being main- tained up to their edges. There can hardly be any doubt that the theralitic rock and these felsic veins and patches crystallized practically at the same time. The theralitic rock immediately beneath the band of analcitic teschenite is full of these veins and patches, which are also found at lower horizons (although in greatly diminished quantity), until they finally disappear long before the picrite is reached. ‘The overlying analcitic teschenite is not a continuous band, but is developed irregularly, sometimes dying out altogether, in such wise that the augitic facies comes into contact with the theralite. By the gradual disappearance of the felsic minerals, the theralite passes into picrite, although at one or two places lugarite appears to separate the two rocks, Owing to the conformation of the Bellow gorge, the ultrabasic outcrop has a tongue-shaped lobe directed westwards, which oceupies the bed of the stream, while the over- lying facies form the sides of the valley. The ultrabasic rocks have precisely the same characters as in the Glenmuir section.. At and beyond Bellow Bridge the outcrop widens, but the overlying thera- litic and teschenitie facies may still be traced in the steep wooded slopes west of the bridge. At the first right-angle bend of the stream above Bellow Bridge the picrite is intersected by many small crush-planes filled with ‘beefy’ calcite. At the second bend part 2] PICRITE-TESCHENITE SILL OF LUGAR. 93 a large fault-plane is well exposed in the bed of the stream. This, with its fellows, has determined the course of the stream for some distance. It is a fissure about a foot wide, filled with calcite and a yellow flinty material. The rock on both sides of the major fault belongs to the ultrabasic stratum. It is much crushed, splintered, and traversed by numerous thin, anastomosing veins of calcite. Sedimentary rock, however, is exposed on the inner side of the bend, and consists of black shale upturned steeply at the fault. High up in the bank on the west side of the stream, highly analcitic teschenite is exposed, passing quickly into a dense black contact-facies containing sparse flakes of biotite. At the outer edge of the first right-angle bend above Bellow Bridge occurs a black-and-white mottled rock belonging to the lower band of teschenite. At the fault the continuous section of igneous rock ends, and the stream euts into the sedimentaries. Hence the lower contact is not here visible. (3) Other Exposures. Between the Bellow and Glenmuir Waters, near their confluence, rises a knoll of high ground. The Ayr-Muirkirk railway crosses the Glenmuir Water here, and is carried through the knoll in a shallow cutting. As one ascends from either ‘stream, exposures of the theralitic facies and the teschenite are encountered, capped by a small outlier of the ‘Millstone Grit’ (see map & section, fig. 3, p. 94). The railway-cutting shows nothing but decomposed erumbling picrite. At the extreme north-eastern end of the sill, a mile and a half north-east of Cronberry, a small exposure is seen in the Bellow Water, here known as the Gass Water. At the south-western end of the section a fine-grained marginal facies of pink teschenite (doubtless the top of the sill) is observ ed, followed by grey theralite and decomposed pierite towards the north-east. The lower teschenite is not seen; but, so far as it goes, the sequence here is identical with that in the typical exposures. Another section is seen in the Lugar Water at Logan Bridge, at the extreme south-western end of the outcrop. Here the intrusion is quite thin, measuring not more than 15 or 20 feet in thickness, and is wedging out westwards among the sandstones. Both contacts, bordered by hardened white sand- stones, are seen. The marginal facies is a decomposed brownish aphanitic rock showing a few flakes of biotite. The interior consists of decomposed teschenite, but there is no trace of the other facies. A mass of theralite probably connected with the Lugar sill, but with a distinct dyke habit, crosses the Lugar Water in a noah to south direction, about 250 yards west of Logan Bridge. The contacts are not seen; but, on the northern bank of the river, the: rock forms a small knoll with vertical sides, and has the aspect of a. dyke. In appearance and microscopic structure it is identical with. the dommant phase of the theralite stratum in the Lugar sill. Q.J.G.S. No. 286. I 94 MR. G. W. TYRRELL ON THE [vol. lxxu, Fig. 3.—Geological map and section of a part of the Lugar Sill. N.W. So SEeso Glenmuir Water 1 1 1 5 1 a 1 Oa S 1 [P = Picrite and peridotite ; T = Teschenite ; Th. = Theralite. | (4) Summary. It is now possmble to attempt a general view of the Lugar sill, and of the structure and disposition of the various facies of which it is composed. This mass, 140 feet thick, was intruded into cold rocks, as testified by the chilled contacts at both upper and lower margins. The upper and lower contact-basalts are identical in composition. The chilling influence of the contact is manifest in fineness of grain to a maximum thickness of 10 feet, after which the rock passes rapidly into a coarse teschenite at both margins. part. 2] PICRITE-TESCHENITE SILL OF LUGAR. 95 A movement of the magma has given rise to distinct schlieren, distinguishable by slight differences of colour and texture, at both contacts. The same explanation can hardly be applied, however, to the remarkable differences obtaining in the interior of the sill. Including the contact-basalts and the normal teschenites continuous with them, the rock of the interior is divided into at least three different bands by some process of differentiation or by successive intrusion. First there is a band of ultrabasic rock—picrites and peridotites of coarse texture and remarkable freshness—occupying the major part of the interior, and indeed of the whole mass, and resting on the lower teschenite. The picrite forms the upper part of the ultrabasic stratum and the peridotite the lower. Above the picrite comes a band about 10 to 15 feet thick, of a fine-grained, basic, nepheline-rock belonging to the theralite family, which may be considered as continuous with the picrite and passing into it somewhat rapidly. Between the theralite and the normal teschenite overlying it generally intervenes a thin and variable layer of highly analeitic rock, patches, streaks, and veins of which, or a rock allied thereto, permeate the theralite in its immediate vicinity. Fig. 4 (p. 96) embodies a vertical section of the sill showing the approxi- mate thickness of the various facies, and the map (fig. 3, p. 94) illustrates the surface-distribution of the facies in a limited area at the confluence of the Bellow and Glenmuir Waters. _ The differentiation will be dealt with in greater detail, after the ' discussion of the microscopical and chemical evidence. Ij]. Perrograpuy. (1) The Contact-Rocks. (Fl. X, fig. 1; Pl. XI, fig. 4.) The rocks of the upper and lower contacts are identical in all respects, and will therefore be described together in this section. They consist, for the greater part, of a hard, dense, basaltic rock, usually black or dark grey. In several specimens bands or schlieren differmg in many subtle gradations of colour and texture are seen. Thin veins of coarse flesh-coloured teschenite, indifferently traversing all the schlieren, are rather numerous. Microscopically the rock is holocrystalline, on the whole very -fine-grained, and shows numerous sharp variations in texture and composition. The minerals observed are plagioclase, augite, biotite, analcite, magnetite, and occasionally olivine. The chemical analysis (p. 104) shows that a considerable amount of orthoclase must be present. A thin section of a rock obtained at the lower contact in Glenmuir Water gives a good general idea of the aspect of both contact-facies, and will accordingly be described in detail. There are several distinct types of rock in the slide (Pl. XI, fig. 4). (a) Starting from one side of the section there is first a schlieren composed of an even-grained aggregate of plagioclase-laths, subhedral prismoids of . purplish augite, numerous small ragged plates of reddish biotite enclosing all 12 4.—Vertical section of the Lugar Sill, on the scale of 22°5 feet to the inch. One Se NAS if Was Re - () ~osraN 2" 40, \7Fo/2 9° 2 On BN 888 oN ea. CPA ooo 2-64 40520, 9.5 F. oe = 2-84 ware & Sz - °° 2-92 Beck 3-01 SEP ER SS ELLE) BAGLINAN SSN Re OU DLV) eee ; Sandstone ~— Teschenite-basalt. (Upper contact-facies.) Upper teschenite, with analeite— rich facies at the base. Sharp contact. Theralite : patches of analcitice material at the top. Picrite : increasing density of dots: represents downward concentra— tion of olivine. Lugarite. _Peridotite. Assumed sharp contact: not seem in the field. Lower teschenite. Teschenite-basalt. (Lower contact-facies.) [Figures on the left-hand side represent the specific gravities of the rocks: collected approximately at the points indicated. The curve shows the variation of specific gravity with depth in the sill: the zgzags at the top- are due to the alternation of schlieren of different densities. line curve shows the effect of averaging these variations. The broken- The division. of the sill into three sharply-defined parts is well marked. | part 2] THE PICRITE-TESCHENITE SILL OF LUGAR. 97 other constituents but analcite, and minute uniformly-scattered cubes of magnetite, all embedded in a limpid, colourless, isotropic base which has the cleavage and low refraction of analcite. The latter is occasionally segregated into small areas comparatively free from the other constituents. A few small patches of serpentinized material occur, which may represent olivine. The extinction of the felspar is between 20° and 30°, but the fine grain forbids exact measurement; it is probably to be referred to acid labradorite. The following is a rough estimate of the mineral percentage : — plagioclase (Ab, An,) 45, augite 40, biotite 5, magnetite 5, analcite 5. (b) Adjacent to the above, with a sharp boundary between the two, is a band composed of the same minerals, but distinctly finer in grain. The proportion of augite has increased at the expense of plagioclase and analcite. (c) The next band is still denser. The minerals are the same as above, but are so crowded together as practically to exclude the isotropic base. Magnetite is more abundant, and is sprinkled very uniformly over the field. (d) A coarse vein of teschenitic material separates (c) from (d), which is the densest of all the bands occurring in the slides. It consists of a crowded mass of minute microlites of augite and plagioclase, with very numerous, small, irregular, poikilitic plates of biotite, all dusted over with magnetite. There are sparse microphenocrysts of plagioclase and purple augite. The felspar has dwindled, and the rock approximates in composition to the monchiguites. The colourless isotropic base is still to be recognized with a powerful objective, and can be definitely identified as analcite, where, in places, it is comparatively free from augite and magnetite. Apart from the coarse teschenitic veins the four schlieren described above are the principal textural varieties to be found in the slide. But within each of them occur slighter variations in texture and composition, and no fewer than eight distinct varieties of rock occur within the limits of a thin section half an inch long. These rocks are basaltic in aspect, and in accordance with their occurrence as the contact-facies of a teschenite, they may be called teschenite-basalts. The term analcite-basalt is unsuitable, asit has been used for definite lava-form rocks. Moreover, this rock 1s practically devoid of olivine, which is abundant in the great majority of analcite-basalts. Some of the schlieren approximate to monchiquites, and others to biotite-basalts. The contact-facies gradually merges into teschenite by increasing granularity and enlarged proportion of analcite. The following will serve to give an idea of the intermediate fine-grained teschenitic rock. he thin section is from a specimen obtained immediately above the lower contact-facies in the Glenmuir Water. Micro- scopically the rock is somewhat similar to the coarser schlieren of the contact-facies, but is not banded and is not so rich in ferro- magnesian minerals. It consists of numerous, small, euhedral to subhedral prisms of pale augite, with stout plagioclase-laths, some serpentinized olivine, and leucoxenic ilmenite, in a ground-mass of dusty analcite. Associated with the ilmenite are numerous minute scraps of biotite. A few turbid areas have the shape, low double refraction, and general aspect of nepheline as it is more recogniz- ably developed in rocks to be described hereafter. Apatite needles are abundant in the areas of analcite. The rock may be described as a fine-grained teschenite; or if it be desired to emphasize its origin as a contact-facies it may be called teschenite-dolerite. 98 MR. G. W. TYRRELL ON THE [ vol. lxxii, (2) Weschenite. “CEE XS figs 2.) The differences between the teschenites proper of the upper and those of the lower margins respectively are so slight, that the rocks may be described most conveniently in the same section. The teschenites are rather variable, both in macroscopic appearance and in thin section. Variation occurs in granularity and fabric, but the mineral composition is approximately constant, or, at least, variations in the minerals present or their relative abundance do not carry any of the rocks outside the teschenite group. The dominant type, perhaps, is one which has an effect of very coarse grain, due to the abundance of large, irregular, black prisms or blades of augite ranging up to an inch in length, embedded in an even- erained, apparently-uniform ground-mass consisting of pinkish felspar and analcite. It is noteworthy that the teschenites of the upper margin carry a pink analcite, whereas at the lower margin the analeite is white. The former, therefore, are mottled in black, green (olivine, serpentinous and chloritic alteration-products), and pink; the latter in black, green, and white. The dominant type with columnar augites (Galston type)! is interbanded in the upper teschenite with a much finer-grained rock devoid of the pseudo- porphyritic augite. Towards the junction with the underlying theralite the pink analcite becomes increasingly abundant, and forms the dominant macroscopic element in a thin irregular band immediately overlying the theralite. In the lower band of teschenite a fine-grained type occurs, carrying numerous, small, acicular prisms of augite in a white ground-mass of felspar and analeite, with much hornblende and biotite, and a little nepheline (Catheart type). In thin section the teschenites onde essentially of labradorite, titanaugite, analcite, olivine, and iron-ores, named in order of abundance. As accessories occur biotite, orthoclase, apatite, nepheline, and barkevikitic amphibole. The last-named invariably occurs as a peripheral alteration-product of the titanaugite. The secondary products are serpentine, chlorite, and leucoxene. The texture is medium- to coarse-grained, and the fabric an inter- locking mesh of labradorite and titanaugite in sub-ophitic relations, the large polygonal or irregular interspaces being filled with analcite. The plagioclase forms broad laths, thoroughly euhedral, and highly zonal. It is a medium to acid labradorite (Ab, An,—Ab “An.). The crystals are frequently somewhat albitized, and, adjacent to- large analcite areas, have been irregularly corroded and replaced by analeite. The cleavage-cracks and the interior of a crystal are occasionally occupied by serpentinous material, which has mpi’ from adjacent decomposing olivine. The pyroxene is a feebly-coloured titanaugite, which, in -the dominant type, occurs as large, blade-like, or columnar crystals. These, however, although elongated in one direction, are very 1G. W. Tyrrell, Geol. Mag. dec. 5, vol. ix (1912) p. 74. part 2] PICRITE-TESCHENITE SILL OF LUGAR. 99 irregular in shape, owing to their indentation by the terminations of felspar-laths. In some of the rocks, notably the finer-grained varieties, the titanaugite is perfectly euhedral, and is moulded by felspar. The colour is a pale purplish brown, and is very variable and patchy. The most common appearance is of a darker tint towards the margins, but bands alternatively of lighter and darker colour may occur. This zoning coincides with a similar zoning observable between crossed nicols, and frequently also with an hour- glass structure. It is, therefore, connected with slight chemical and optical variations in the crystals. There is occasionally a distinct pleochroism from purplish brown to a pale sepia. A notable feature is that practically all the crystals are hollow. In the prismatic sections elongated cavities occur along the centre- lines, and are filled with calcite, serpentinous alteration-products, and occasionally even analcite. The cavities are more or less equi- dimensional in the basal sections. They are frequently lined with a highly pleochroic biotite, which extends in ragged indefinite patches throughout the interior of the crystal, and is evidently an alteration-product. The latter frequently has an astonishing pleochroism in shades of brilliant blue, red, and peach-bloom tint, but it is sometimes scarcely more than a discoloration of the augite, so indefinite are its boundaries. The more individualized mica has a more normal pleochroism, but its darker shades have a peculiar ‘beetroot’ tint. Serpentinous material also occurs within the augite, but its relations show that it has migrated from adjacent olivine, entering the crystal through the cleavage and other eracks. The augite often thus presents a honeyeombed appearance in the interior of the crystals, the exterior being almost invariably sound, and free from inclusions and honeycombing. A red hornblende belonging to barkevikite occurs as an altera- tion-product upon the margins of the augite-crystals, especially in the fine-grained rocks of the lower contact, where the augite is euhedral. The boundary between the two minerals is always exceedingly indefinite. A rock from the lower band of teschenite above Bellow Bridge is so rich in hornblende that it deserves the designation hornblende-teschenite. This rock is fine- grained and contains nepheline, thereby approximating to the Catheart type.! . A thin band of wxgirine-augite or egirine frequently occurs on the margin of an augite-crystal, especially where it is adjacent to an area of analcite. Small crystals of «xgirine are occasionally enclosed in the analcite. Analcite fills up large and small polygonal spaces between the felspars and augite; but, where corrosion and analcitization of the felspar has occurred, irregular or rounded spaces are formed. It is mostly fresh, showing the cubic cleavage, and occasionally some anomalous birefringence. The commonest alteration is to an ex- tremely-fine, irresolvable, brown dust; but, with a further degree 1G. W. Tyrrell, Geol. Mag. dec 5, vol. ix (1912) p. 74. 100 . MR. G. W. TYRRELL ON THE [vol. lxxu, of weathering, calcite begins to appear. Beautiful rosettes of a serpentinous mineral are frequently developed in, and apparently at the expense of, the analcite. These enclose flakes of biotite and also the other minerals usually found within the analcite areas. Oceasionally the analcite is completely replaced by the fibrous serpentine. ‘These rosettes have been mentioned by Mr. E. B. Bailey in describing the Glasgow teschenites, and he indicates the need for some special explanation of this phenomenon. He suggests that it may possibly originate from ‘juvenile’ reactions.! The analeite corrodes and replaces the felspars enclosing the cavities in which it has crystallized. The attack has usually spread from the cleavage and other cracks, and the process can be followed from mere incipient analcitization, resulting in a widening of the cleavage-fissures, to complete replacement of the crystal. The latter, however, frequently retains its form ; equally often the felspar forms shapeless, irregular masses entirely enclosed in analeite. The final appearance is of a patch of clear or dusty analeite lying in the midst of a large area of partly or wholly analeitized felspars, mixed with serpentinous, chloritic, and other alteration-products.? These reactions clearly belong to a juvenile stage in the history of the magma, and may:be referred to that late period when the rock was ‘stewing in a hot alkaline solution which ultimately crystallized as analcite. Further evidence as to the original nature of the analcite is afforded by the numerous inclusions of biotite, pyroxenes, apatite, and felspars that it con- tains ; by its association with soda-orthoclase; and by its evident reaction on the adjacent felspars and augite, resulting in the latter ease in the formation of a thin layer of a green soda-pyroxene. Pyroxenes which have been entirely enclosed in the analcite have suffered this change to a much greater extent, leading in some cases to complete replacement. It is remarkable that, while a susceptible mineral such as analeite often remains entirely or comparatively fresh, the other constituents of the teschenites:have undergone so much alteration. This leads to the conclusion that the alteration is not so much due to ordinary weathering as to the presence, during the crystallization of the rock, of a hot, intensely-active, alkaline, and water-rich mother- liquor, which crystallized as analcite after effecting much corrosion among the earlier-formed constituents. Because of the corrosion and replacement effected by the analcite among the earlier constituents of this and other vocks, there is a disposition to regard itas ‘secondary’ in a certain sense of that term But it is as definitely a primary consolidation-product of the teschenite magma, inasmuch as its crystallization took place before the cessation of cooling of the rock, as is the allotriomorphie quartz of a granite. Both represent a final magmatic residuum, which 1 «The Geology of the Glasgow District’ Mem. Geol. Surv. Scotland, 1911, p. 182. 2 H. Rosenbusch, ‘Elemente der Gesteinslehre ’ 1910, p. 300. a part 2] PICRIVE-TESCHENITE SILL OF LUGAR. 105 I. T'eschenite-basalt, upper contact of Lugar sill, Glenmuir Water, Lugar. Chemical analysis by Dr. A. Scott. II. Banded teschenite, upper teschenite layer of Lugar sill, Bellow Water, Lugar. Chemical analysis by Dr. A. Scott. IIL. Banded teschenite, same as II, calculatcd from Rosiwal analysis No. i of Yable I. ; IV. Teschenite, rich in analcite, upper teschenite layer, Bellow Water, Lugar. Calculated from Rosiwal analysis No. II of Table I. V. Teschenite, trom lower teschenite layer, Glenmuir Water, Lugar. Calcu-. lated from Rosiwal analysis No. I11 of Table I. VI. Teschenite, from lower teschenite layer, Glenmuir Water, Lugar. Calcu- lated from Rosiwal analysis No. 1V of Table I. VII. Melanocratic teschenite, trom upper teschenite layer, Bellow Water, Lugar.. Calculated from Rosiwal analysis No. V of Table I. : VIII. Teschenite, Mons Hill, Dalmeny (Midlothian). Analysis by E. G. Radley, ‘Geology of the Neighbourhood ot Edinburgh’ Mem. Geol. Surv. Scotland, 1910, p. 299. The teschenitic character of these analyses is at once evident. For a silica percentage averaging 45, and high ferrous iron and lime, the alkalies run to about 7 per cent. The magnesia is com- paratively low, indicating the poverty of the rocks in olivine. Combined water, due to the analcite present, is of course high. The analyses obtained by calculation from the Rosiwal analyses. are remarkably in accord with the chemical analyses. The greatest discrepancies are in magnesia, potash, and water. The latter may be explained by the fact that no account of the alteration of the rocks was taken in the calculation. ‘The excess of magnesia may perhaps be explained by the complete alteration of the olivine to serpentine in the analysed rocks. The serpentine was analysed as. serpentine, but calculated as olivine. The deficiency in potash is due, as explained above, to the difficulty of detecting and measuring orthoclase in these rocks. The analysis of the Mons Hill teschenite shows the essential identity of the Midlothian occurrences with those of Western Scotland. The richness of some of the Scottish teschenites in potash points to the fact that some of the analcite-rocks are probably to be referred to the monzonite series. With increasing abundance of orthoclase teschenites pass into analcite-monzonites, analogous to. nepheline- and leucite-monzonite (sommaite). The abundance of lime in the teschenite-basalt (1) as compared with the normal teschenite, is due partly to calcite, and partly to a slight enrichment of the contact-rock in augite. The abundance of olivine in the melanocratic type (VII) is shown by a great. excess of ferrous iron and magnesia in the calculated analysis. (3) Theralite. (Pl. X, figs. 3 & 4.) Three phases of the theralite may be distinguished. A black or dark-grey, compact, doleritic rock forms the main mass of the stratum. The lower part, however, carries abundant hornblende, and is slightly coarser in grain; while, near the junction with teschenite, the rock is veined and shot with patches of a medium- erained, ight-grey, analcitic variety. Microscopically, the theralites consist of essential plagioclase, 106 MR. G. W. TYRRELL ON THE [vol. lxxil, titanaugite, olivine, nepheline, and ilmenite, with accessory horn- blende, biotite, analcite, orthoclase, and apatite. ‘The texture is very fine-grained, and hence the rock might be termed nephe- line-dolerite or doleritic theralite. In the principal variety the fabric shows innumerable small, prismoid grains of augite em- bedded poikilitically in a ground-mass of felspar and nepheline, the whole forming a base in which pseudoporphyritic olivine, augite, and ilmenite is set. The hornblende-bearing variety has the same fabric, only a new pseudoporphyritic element, hornblende, being added. The light-grey veins, however, contain no granular augite, and their abundant analcite gives them a fabric similar to that of the teschenites. In the principal variety the felspar forms broad, anhedral plates, and seems to have been the last constituent to crystallize, even the nepheline being euhedral towards it. It is extremely zonal, and is crowded with grains of augite. Its composition is consequently difficult of determination, but the evidence points to the usual acid labradorite (Ab,An,), with marginal transitions to oligoclase. The plagioclase is liable to decomposition, especially in the central portions of the crystals, and passes into an irregular felted mass of scaly mica. The augite occurs in two forms: as innumerable prismoid grains embedded in felspar and nepheline; and as larger, pseudo- porphyritic, ewhedral crystals of a much darker tint than the grains. The colour is a pale purple or lilac, generally darker towards the margins of the crystals. The mineral is often strongly pleochroic, from purplish to pale yellow. It alters, Just as in the teschenites, with the production of biotite. ; Olivine occurs in large pseudoporphyritic crystals, euhedral originally, but now the angles are largely rounded off, and therefore nearly spherical grains are not uncommon. It is very ‘fresh with, at most, a slight peripheral serpentinous alteration. All the fissures are much blackened by separated magnetite. Nepheline forms small subhedral to anhedral masses, and appears to be idiomorphic towards the felspar. It is usually altered to a turbid, streaky, micaceous aggregate, the scales of which are arranged parallel to the crystallographic axis of the mineral. It is occasionally quite fresh, and its identification can be made abso- lutely certain by the usual tests. Tlmenite occurs in skeletal masses associated with biotite, but not so abundantly as in the teschenites. Biotite also occurs as small independent flakes embedded in felspar or nepheline, or as an alteration-product of augite. Analcite occurs very sparingly in the principal variety of theralite, but is much more abundant in the veins towards the top of the stratum. Apatite is very abundant, enclosed in all constituents. A red soda-hornblende, belonging to the barkevikite group, becomes an important constituent in the lower part of the theralite stratum. It usually forms extremely irregular plates, embayed by part 2] PICRITE-TESCHENITE SILL OF LUGAR. 107 the earlier constituents, and occasionally enclosing them. It is pleochroic, with colour-extremes of pale straw-yellow and clear red-brown. The quantitative mineral composition of the main varieties of theralite is set out in Table III, cols.1 & u (p. 109). It will be seen that the mass of the rock is decidedly mafic, much more so than the rocks which have been described as theralite. In addition to the mafic composition, it is characterized by a poikilitic fabric. The light-grey veins, however, are much coarser in grain, are devoid of granular augite and the poikilitic fabric, and show approximate equality between the felsic and the mafic constituents. Contain- ing, as it does, some orthoclase and analcite, this variety closely resembles the true theralites, which are, in general, mafelsic in composition. A mineralogical feature that deserves special mention is the beautiful lilac colour of the augite. (4) Lugarite. (PI. XI, fig. 1.). This rock is found intercalated near the transition between the theralite and the underlying picrite. It has a maximum thickness of 4 feet, and is intimately welded to both the contiguous rocks. Tt also occurs as irregular, anastomosing veins ramifying through the picrite, and varying in thickness from 1 to 5 inches. In hand-specimens it presents a striking and beautiful appear- ance; so much so that it is a matter for comment that the rock has apparently never been noticed before. Boulders of it occur in the bed of the Lugar for a mile or two below the outcrop. It is phanerocrystalline and apparently coarse-grained, consisting of an abundant, continuous, greyish-green ground-mass (analcite, nephe- line, and alteration-products) crowded with shining black prisms of barkevikite ranging up to 3 inches in length. The rock of the main exposure also shows equally abundant, more or less equi- dimensional, black crystals of titanaugite. Occasionally, whitish rectangular felspars may be recognized. Weathered blocks are still more striking in appearance, as the ground-mass becomes white, contrasting effectively with the black prisms embedded in it. Microscopically, lugarite is a very beautiful rock, owing to the brilliant colours of its mafic constituents and their perfect euhedrism. The rock consists essentially of an abundant cloudy greyish base, partly isotropic, and partly cryptocrystalline because of an extremely-fine dusty alteration-product, clearing occasionally to areas of identifiable analcite. This base is crowded with per- fectly euhedral crystals of deep purple titanaugite, red barkevikite ranging up to 8 inches in length, ragged masses of ilmenite passing over to leucoxene, corroded felspars,. and innumerable prisms and needles of apatite. Titanaugite forms well-shaped crystals ranging up to a quarter of an inch in diameter. Its pleochroism is very intense, the scheme being as follows :— DLE Mn enenen clear pale brownish yellow. BM Sabha Yaissthe deep maroon or reddish purple. Yeats acAC He brownish violet, 108 MR. G. W. TYRRELL ON THE (vol. lxxii, The hour-glass and kindred zonal structures are prominent, and the exterior zone is always the more deeply coloured and_pleo- chroic. A green coloration frequently appears on the extreme margin. Occasionally, an augite encloses the termination of a felspar-crystal; and in rare cases it is interdigitated with barke- vikite, the junction between the two minerals being indefinite, but the exterior margins euhedral. A few crystals are honeycombed by irregular cavities filled with colourless isotropic material, pre- sumably analcite. The titanaugite of this rock is identical in its optical characters with that of the ijolite-dolerite (nepheline- dolerite) of the Lobauer Berg, Saxony. The barkevikite forms pertect crystals, and is of a deep reddish- brown colour. The pleochroism is intense, with the following scheme :— XG nents pale yellow. Das BRE uara eat deep chestnut-red. LN Rie ae deep brown-red with a tinge of violet. The maximum extinction-angle in prismatic sections is about 11°. The simple twin parallel to 010 is prominent. The mineral is, therefore, referable to the barkevikite group.! It occasionally moulds the pyroxene, and is then evidently an alteration-product ; but the great majority of the crystals are entirely independent of pyroxene. On the other hand, the smaller crystals are frequently enclosed in the felspar. The felspar, where uncorroded, forms well-shaped rectangular laths, and is an extremely zonal plagioclase. Where the extine- tions can be measured they indicate a labradorite of composition Ab,An,. On the margins, however, nearly straight extinctions are ‘obtained, indicating a transition to oligoclase. “The felspar is nearly always corroded, and all stages in its replacement by dusty analcite and isotropie alteration-products can be followed. The replacement begins along the cleavage and other cracks, and advances until large irregular areas in ‘the interior of the crystal are replaced; while, at the same time, the alteration proceeds from the exterior of the crystal in such a way that the felspar remnants axe often crescentic in shape, and appear as if large pieces had been scooped or gouged from their sides. The final stage is a complete replacement of the crystal; or, at most, small oucuenae fragments of felspar are left. Frequently, however, the crystal form is preserved, and is differentiated from the rest of the ground- mass by a slightly paler tint. Olivine occurs very sparsely as rounded inclusions in the horn- blende or augite; and biotite in small flakes, or as an alteration- product of augite. Both minerals may be completely lacking in a slide. Ilmenite occurs in rather well-shaped hexagonal crystals, and is in process of alteration to leucoxene, giving rise to ‘a marked striation (in black on grey) of three sets of lines in directions intersecting at 120°. It appears to have crystallized in this rock before the Sanibleade or the pyroxene. mee 1 A. Scott, ‘ Barkevikite from Lugar’ Min, Mag. vol. xvit (1914) pp. 138-42. part 2] PICRITE-TESCHENITE SILL OF LUGAR. 109 Apatite in needles and small crystals is extremely abundant, and occurs embedded in all constituents save ilmenite. The longer needles and prisms show the usual cross-fracture, and are frequently bifid at the terminations. It is difficult to study the ground-mass, because of its turbidity : it is composed mostly of alteration-products, a fine brown dust, brightly-polarizing zeolites, and ‘ghosts’ of analcitized felspars; but occasionally it clears to an area of recognizable analcite. Under a high-power objective the fine brown dust resolves itself into clear highly-refracting granules, embedded in a colourless isotropic substance which has a refractive index distinctly lower than that of Canada balsam, and, on the thin edges of the section, a cubie cleavage. It is fairly clear, therefore, that the bulk of the ground- mass is a cloudy analcite. Faint, streaky, paler patches, with an occasional approach to hexagonal or rectangular outlines, are probably to be referred to nepheline. ‘This mineral occurs much more recognizably in a lugarite-like rock in association with the bekinkinite of Barshaw, near Paisley. Two varieties of lugarite are distinguished. In one, titanaugite and barkevikite are developed in about equal proportions (‘Table III, col. iii). he distribution of these minerals is, however, very patchy. Some slides contain titanaugite or barkevikite only ; others contain both. The veins that penetrate the picrite have barkevikite only (Table III, col. iv), the prisms of which are frequently arranged in a rude stellar fashion. Quantitative Mineral Composition of Theralite and Lugarite. These were estimated by the Rosiwal method, and gave fairly concordant results, which are recorded in Table III, below. TasueE III. I 18f III. IV | Plagioclase2 2. e.0202.....c-.s-eo|| 233 16°4 105 14°6 YNTAESY Os ee) Galan RCE Pe ms : : , Nepheline 0. SCS CUA ore. I 166 haps } 490 Titanaugite Seceeee Looe ay eeneee ho SOs 35°9 21:7 ye Barkevikite ...........0.00.:.c0000. As 12:2 172 =| 29°5 Olivine Ee 18°6 87 Be matt Biotite 36 67 a Ae Almeniteser, deen aheyccd ec tney Aed 25 50 ie Apatite ccenntenen ce caer tell 10 31 42 | I. Theralite, Bellow! Water, Lugar. Il. Hornblende-theralite, Bellow Water, Lugar. IJ. Lugarite, main mass, Glenmuir Water, Lugar. IV. Lugarite, veins in picrite, Glenmuir Water, “Lugar. 14. Ww. Tyrrell, Geol. Mag. dec. 6, vol. ii (1915) p. 308. 2 Ab, An, in theralites; Ab,An, in lugarites. Q.J.G.S, No. 286. K 110 MR. @. W. TYRRELL ON THE [vol. Ixxu, A little orthoclase has certainly been overlooked in the lugarites and probably in the theralites. Also, in the lugarites the analcite total contains some nepheline. From the chemical analyses it appears that apatite has been considerably over-estimated in the lugarites. Chemical Composition of Theralite and Lugarite. No chemical analysis was made of the theralite proper. The rock analysed prov ed to be nearer to picrite than to theralite, and its composition is set out in Table VI, col.i(p. 114). The lugarite was analysed by Dr. Scott: its chemical composition is unique, reflecting the unique mineral composition of the rock. The che- mical compositions, as calculated from the mineral compositions recorded in Table III, are collected here for purposes of comparison, and serve, as do the others already tabulated, to demonstrate the efficiency of the Rosiwal method in suitable cases as an auxiliary -to actual chemical analysis. Tasre lV. I. JUG yeh IV. Wercouh AVAL WAG Ke WADU Ey OSG. SiG eee aes 42°47 | 42°75 44-42 | 46°29) 45°56 | 47°35 | 43°70 | 52°73 | 56°75 iO oe Rene 2°59 2°68 1°63 2737 2°58 1°46 0°89 ae 0°30 AlbO3 ..2..2...|) 14520) | laal9 13133 1747 | lav2 | 18°02 19°77 | 20°05 | 20°69 Hes Os eee 3°25 3750 | 9:14 2°24, 4-04 3:07 3°35 343 3°52 Me Op iat 8°73 8:44 6:35 | 7:07 6°75 5°65 3 AT 0°99 0°59 TWENO)” og 55 500 ie 0°05 028 | 007 O11 tr. a tr. Mic Omen ke 12°92 9°47 5: 7d. 210 | 3:02 0°82 3°94 O17 Oul CaO 8:26 8°83 10°60 5°82 8:10 7-82 | 10°30 3°35 | 0°37 Nias Olesen 5°33 6°50 5°60 8°69 8:82 9°76 9°78 7-94 | 11°45 oe 0°92 1:37 181 1:47 0:04, 0°06 2°87 477 2°90 G1 Oe eae) WAI ; 6:12) : As 4°85 3:18 On O10] o21 | 1-75 { Pan | 383 | 409] 0-89 ee ae PSO nike seep 118 091 0°35 0°70 1°61 1:78 1°34 tr. ae IBBXOX(SHKO)) gal) coe ae is 0°09 ae a bey 0711 | none COs aeee rae a .. | none ue si avs 0:93 Be eee 0:06 | 0:04 epenidel vOe AO Ren), eae .. [C128 ue ae 0718 she hae ae ae As tr. Totals ...|100°01 | 99°94 | 100°90 | 100°40 | 100°11 | 100715 | 100°30 | 100°01 |100°18 I. Theralite, Bellow Water, Lugar. Calculated from Rosiwal analysis, Table III, No. I (p. 109). II. Hornblende-theralite, Bellow Water, Lugar. Calculated from Rosiwal analysis, Table III, No. I. III. Theralite, Flurhubl, Duppau (Bohemia). F. Bauer, Tscherm. Min. u. Petr. Mitth. vol. xxii (1908) p. 251 IV. Lugarite, main mass, Glenmuir Waiter, Lugar. Chemical analysis by Dr. A. Scott. V. Lugarite, main mass, Glenmuir Water, Lugar. Calculated from Rosiwal analysis, Table III, No. III VI. Luegarite, veins in picrite, Glenmuir Water, Lugar. Calculated from Rosiwal analysis, Table III, No. IV. VII. Tolite, livaara, Kunosamo (Finland), Anal. N. Sahlbom. Quoted from J. P. Iddings, ‘ Igneous Rocks’ vol. ii (1918) p. 306. VIII. Heronite, Heron Bay, Lake Superior. Anal. H. Ww. Charlton. ecieman’ Journ. Geol. Chicago, vol. vii (1899) p. 435. IX. Analcite-tinguaite, Pickard’s Poimt, Essex County (Mass.). Anal. H. 8S. Washington. Amer. Journ. Sci. ser. 4, vol. vi (1898) p. 182. ‘ part 2] PICRITE-TESCHENITE SILL OF LUGAR. 111 The correctness of the reference of the first two rocks in Table IV to the theralites, and the general accuracy of the Rosiwal analyses, is shown by the accordance of the calculated analyses with the analysis of Rosenbusch’s type theralite from Duppau (Bohemia). The excess of magnesia and the slight deficiency in lime of the Lugar theralites are due to their more highly mafic character; but the alkalies, alumina, and silica correspond remark- ably well with those of the Duppau rock. Turning to the lugarite, the calculated Rosiwal analyses corre- spond well with the chemical analyses, except in lime, potash, and phosphorus pentoxide. The excess of lime and phosphorus pent- oxide is partly due to the over-estimation of apatite, and partly to the overlooking of orthoclase and its measurement as labradorite. The deficieney of about 1 per cent. in potash is, of course, due directly to the latter cause. The lugarite analyses are characterized by a very large content of alkalies along with a comparatively- low silica percentage, and by their persodic character. They fall into the hitherto unoccupied and unnamed persodic subrang of lujavrase (11.7.1.5) of the American Quantitative Classification.1 Systematically, these rocks may be described as ijolites in which the place of nepheline is partly taken by analcite, and in which barkevikite occurs as well as, and sometimes to the exclusion of, augite. One of the analyses of ijolite from Livaara (Finland), corresponds fairly well with that of lugarite (Table IV, No. VII). It is, however, richer in lime and alkalies than the Lugar rock, and contains much less combined water, since nepheline and not analeite is the principal felsic mineral. The only other rocks hitherto described with an analeite-content of the same order as that of lugarite are the heronite2 of Heron Bay, Lake Superior, with 47 per cent. of analeite; and the anal- cite-tinguaite of Pickard’s Point, Essex County (Mass.) with 37-4 per cent. of analcite. The chemical analyses of these rocks are set out for comparison in Table IV. Heronite differs from lugarite in containing a large amount (28-2 per cent.) of orthoclase, which is reflected in the large potash-content of the analysis, although the soda-content compares well with that of lugarite. The analcite-tinguaite of Pickard’s Point contains eegirine and anortho- clase phenocrysts in a ground-mass composed of nepheline and analeite, is richer in alkalies than lugarite, and is devoid of lime- soda felspar. ° (5) Picrite and Peridotite. (Pl. XI, figs. 2 & 3.) The ultrabasic rock which makes up more than half of the Lugar sill is characteristically felspar-free, and is a typical hornblende- peridotite. Some varieties, however, occurring towards the top of ' G. W. Tyrrell, Geol. Mag. dec. 6, vol. ii (1915) p. 361. 2 This rock is now regarded as a decomposed tinguaite; see A. E. Barlow, “Nepheline & Alkali-Syenites of Port Coldwell [Ont.]’ Guide-book No. 8 Geol. Surv. Canada, 1913, foselize Kk 2 112 MR. G, W. TYRRELL ON THE [vol. lxxu, the mass, contain some felspar and analcite, and are richer in the metasilicates. These are true picrites in the original sense of Tschermak. In hand specimens the peridotite is a fresh, medium to coarse-grained, blackish-green, heavy rock, in which olivine may occasionally be distinguished, and especially hornblende in large, lustre-mottled plates. The rock is variable in granularity, and may be devoid of poikilitic hornblende. The picrites are distinctly finer-grained, and show white or pink specks of felspar and analcite. Microscopically, the peridotite consists of olivine, titanaugite, hornblende, biotite, and iron-ores in the proportions shown in Table V, col. iii (p. 113). The olivine, which may form 60 to 70 per cent. of the rock, occurs in more or less rounded, subhedral grains, ranging up to half an inch in diameter. It is occasionally quite fresh, but is usually in all stages of alteration to blue, green, yellow, and colourless varieties of serpentine. The coloured ser- pentines are almost entirely devoid of separated magnetite; but the unaltered olivine and the colourless variety of serpentine contain irregular streaks of magnetite, indicating that the colour of the serpentine depends on whether the iron-oxide is thrown out of combination or not during the process of alteration. The next most abundant constituent is augite, which occurs in polysomatic clusters of small euhedral grains wedged in between the olivines, and where the latter mineral is altered, embedded in serpentine. The eranular habit of the augite is distinctive of this type of peridotite, which is accordingly distinguished as the Lugar type. The earliest described picrite or peridotite of the teschenite series, that of Incheolm, contains titanaugite in large plates which mould and poikilitically enclose olivine. ie A red-brown hornblende, belonging, like the amphiboles of the preceding rocks, to the barkevikite group, occurs in large, irregular, poikilitie plates, enclosing small olivines, and groups of granular augite-crystals. It usually forms about 10 per cent. of the rock. The remaining constituents, biotite and iron-ores, form only about 4 per cent. of the rock. The two minerals are, as usual, closely associated. The rock is almost devoid of apatite: this mineral appears to be associated with analcite, and increases in abundance along with that mineral. The picrites which occur towards the top of the ultrabasic stratum are divisible into two varieties. One, almost devoid of recognizable felspar, contains much analcite, and very abundant augite. It approximates to the lugarite type, and represents the modification of the peridotite at the contact with lugarite. In thin section this rock consists of numerous, euhedral, purple augites, embedded in a turbid cryptocrystalline or isotropic ground- mass containing zeolites, occasional clear analcite, and ‘ chosts’ of felspars, idlantaicall with the ground-mass of lugarite. Bakelite oceurs in ragged plates exhibiting the poikilitic habit and inter- grown with ilmenite. Olivine is comparatively sparse, and is completely serpentinized. part 2] PICRITE-TESCHENITE SILL OF LUGAR. 113 The other variety is an extremely fresh rock, with abundant olivine and some fresh plagioclase. It occurs between the perido- tite and theralite in the great cliff of the Glenmuir Water. It differs from the rock described above in the much greater abundance of olivine, in less analcite and turbid decomposition-products, and in the presence of plates of fresh felspar enveloping olivine and augite, and showing mus usual expansion-fissures. Quantitative Mineral Composition of the Picrites and Peridotites. Table V records the results of the Rosiwal measurement of three of the ultrabasic rocks of the Lugar sill. TABLE V. | I it JOU | Sai Boe oR A rN cara eae esa MUP Cu a I a | Plagioclase ee Asien ORE | Analcite .. ail ES NT Sal oh i Bre is Titanaugite Ue he mn UE euA oneassi su ts ey OS OME allie. O sl 20°5 JB HM ehuaul tourna vie Marni Se ee dea aa eel and Ute 86 10°; | Olivanie ee ice a eon aN ol vienna secileLo 49°] Gas2ieal Biotite , eG a ey tea SNL ou 0-4 2°0 Titaniferous iron-o EO KOM ase Mate eae wee Ae rll 1-0 2:3 dE AV ORR Eesy) Dea ee HCC ALONE ane UrScEee ck iP) 03 | I. Augite-picrite, upper part of ultrabasic stratum, Glenmuir Water. II. Olivine-picrite, same position and locality. III. Hornblende-peridotite, main mass of ultrabasic stratum, Glenmuir Water. It will be seen that the picrites are domafic, and the peridotite permafic. The analcite totals contain a little unidentifiable turbid matter. The diminution in the amount of apatite is a curious feature of the ultrabasic rocks. In the Lugar series, and in the rocks of the Ayrshire petrographical province generally, apatite seems to vary in abundance along with the felsic minerals, especi- ally analeite. A similar but less marked decline in the amount of titaniferous iron-ore also occurs. The order in which the analyses are given is that of increasing depth in the ultramafic stratum. The rapid increase in the amount of olivine illustrates the packing of the olivine-crystals in the lower part of the stratum by settling under the influence of gravity. Along with this there is a decrease in the amount of augite; but ee eres a mineral of later con- solidation in these rocks, remains practically constant. Chemical Composition of the Picrites and Peridotites. Chemical analyses of the augitic type of picrite and of peridotite were made for me by Dr. A. Scott ; and these are supplemented by analyses calculated from the quantitative mineral compositions recorded in Table V, above. 114 MR. G. W. TYRRELL ON THE [ vol. Ixxii,. TAasuy Vi. heed oe | JOU. IV. Aaa hs AVAIL | VII. SiOz ..0. 0) 44d Ana) 49°62) |) 40°85) 1589793" | 4 42°06 | 40°32 TO} els fo78 388 | 158 2:12 1°74 1:93 | 2:66 JNISOR ssondons 759 843 641 3°75 2°75 | 1218 | - 9:46 HesOs ae | CPB 5122 | 2:06 3°53 949 | w67 | 475 WEO) ccosrasoioelt GPRM po Tea) ea 9°86 |* 13:30 | 7:89 | 748 | GOO) aan sede dale OAS) O05. 0103 0:20 Ou || 0:25 | MEO ihe 193) E72) 25 sil) 25:69) 89-88) Ara wie eso CAO) ee aes oH TON) |) HSI 4°64) AVG | 11:29 | 10:55 Nas Olen 4:27 51am | 34 SAA 7B vot Ons. TO Meee dee lg 0:03 0:05 ORO) 1 OPI | IO) NO EO edad 0:73 ei us 5:28 ; 0°57 HO ee 0-48 i Peco nycey in Wer a Bales ho ces, | P30; . Pee O54 1:25 9:49 0:25 0:28 0:34 | 0°68 Bao, SrO ......| p-n.d. MAbs ater 0:06 Witla cre a BieA aT NTH Ch asc ¥ ee a 0:04 0°01 ae Laois ARAB eat COM eae agate San He tr. RMU ICS se Ae ES iclepeeeeane 2 2H se he A O29) ee ORS Totals ...| 100°75 | 10013 | 100°04 | 100°50 | 100:03 | 100:05 | 100:09 \ Pe] ESS (are SE I. Picrite, transitional to theralite, Bellow Water, Lugar. Chemical analysis by Dr. A. Scott. Il. Augite-picrite, Glenmuir Water, Lugar. Calculated from Rosiwai analysis, Table V, No. I (p. 1138). Ill. Olivine-picrite, Glenmuir Water, Lugar. Calculated from Rosiwal analysis, Table V, No. II. ; IV. Hornblende-peridotite, Glenmuir Water, Lugar. Chemical analysis by Dr. A. Scott. V. Hornblende-peridotite, Glenmuir Water, Lugar. Calculated from Rosiwal analysis, T'able V, No. III. = VI. Limburgite, Hahn, Habichtswald (Hesse-Nassan). Anal. Jannasch. Quoted from J. P. Tddines, ¢ Igneous Rocks’ vol. 11 (1918) p. 334. VII. Nepheline-basalt, Uvalde County (Texas). Ana]. W. F. Hillebrand. W. Cross, Bull. U.S. Geol. Sury. No. 168 (1900) p. 62. These are typical ultrabasic rocks in their large content of magnesia, lime, and ferrous iron, combined with low silica; but they are characterized by comparatively high alkalies, as compared with other rocks of the same category. This results from the persistence of analcite into the ultrabasic end of the series, and from the alkali-content of the pyroxenes and amphiboles. In this respect it is difficult to find phaneric rocks to match with them. Some nepheline-basalts and limburgites approach closely in chemical composition (see Table VI, cols. vi and vil). IV. Perroroey.! The differentiation of the Lugar sill may be explained in two ways, according to whether it is considered as the product of a single act of intrusion or of more than one. Both modes of ex- planation involve perplexing features. Postulating the former, the 1 [Since reading this paper I have, with the permission of the Council of the Geological Society, considerably revised the theoretical discussion of the Lugar sill. I have done this in deference to weighty opinions expressed in the part 2] PICRITH-TESCHENITE SILL OF LUGAR. 115 present heterogeneity of the sill suggests a very complex process of differentiation. Moreover, if it were intruded as a homogeneous body of magma the chemical composition of the contact-rocks should be similar to the bulk-composition arrived at by averaging the analyses of the different parts after weighting them according to their volumes. As will be seen later, tie is by no means the ease; and, consequently, if the sill is Abe result of a single act of intrusion, the magma must have been heterogeneous prior to in- trusion. The question of its differentiation is then shifted back to an ante-intrusion stage, and its discussion becomes correspond- ingly difficult, These complications are avoided, to some extent, if the mass be regarded as a composite sill resulting mainly from two acts of intrusion—the first introducing the teschenite at both contacts, the second bringing in the ultrabasic rock of the interior. All phases within the sill, however, are intimately welded together, showing that the later intrusion must have quickly followed the earlier. Moreover, the mineral and chemical composition of the rocks show that the successive intrusions have very close genetic relations, and — have probably arisen. by the differentiation of a single body of magma. The problem of the mode of differentiation then becomes the same as that arising from the first hypothesis. It the sill be thus composite, it shows a surprising lack of xeno- ' erysts and xenoliths, or of veins and dykes, along the main interior contacts. Other features, too, are difficult of explanation on this hypothesis. The subsidiary differentiation observed within the central ultrabasic stratum seems most easily explained by the hypothesis of sinking of early heavy crystals under the influence of gravity, aided perhaps by a concomitant rise of lighter con- stituents. The special features of the Lugar sill will now be treated in detail, especially with regard to their bearmg on the hypotheses outlined above. (1) Mineralogical Variations. EHstimations of the modes of twelve types of rock occurring in the Lugar sill are scattered through the foregoing petrographical discussion upon the paper, and communicated privately ; and also on account of the views on differentiation in general expressed by N. L. Bowen in a recent important paper (‘The Later Stages in the Evolution of the Igneous Rocks’ Journ. Geol. Chicago, vol. xxiii, 1915, Suppl. pp. 1-91). Liquation theories of differentiation stand in need of drastic revision after the evidence brought. forward in this paper, broad-based upon the exact experimental work carried on for many years at the Geophysical Laboratory of Washington, that liquation has never yet been observed in the melts experimented with. Neither has it been observed in lavas, which are Nature’s ‘ quenching experiments.’ Whether we can argue from experiments upon the comparatively minute laboratory scale to age-long magmatic processes under natural conditions, is at least open to doubt; and, until that doubt is resolved, it is permissible to use liquation hypotheses, but to assign to other hypotheses more weight according to their correspondence with known facts and conditions. | mineralogical variation concurrent with 116 MR. G. W. TYRRELL ON THE [vol. lxxii, descriptions (see Tables I, III, and V). Referring to these it may be seen that lime-soda felspar attains its maximum develop- ment, 34:5 per cent., in the teschenites. It dwindles steadily through the. theralites, lugarites, and picrites, and is absent from LOWER CONTACT TESCHENITE (LOWER) PERIDOTITE f WwW = Zz Fi ter 5 Ne ar ‘~ H > ib) SN ae ements sols oy Anya tere eee eT =~ eS fe) = aS) x S oS tS ou > Ve Ss Sy{e Ss 2 = $s 8 S = ~*~ s ~~ . SS ju = E S a S : : ~ ‘'~ Q Ww o ft ate ws ; ne we oo OF Zo © Oe Ww ay ae a2 ic oS | oa UPPER " CONTACT 100 70 60 50 30%) 20 0. the peridotite. Orthoclase occurs most abundantly in the tesch- enites, but is present in small amount (although unrecorded) in all the other rocks, save picrite and peridotite. Itisa mineral hard to detect and measure when associated with abundant lime-soda felspar, but its presence is demonstrated by the potash of the chemical analyses. Analcite is the characteristic mineral of the part 2] PICRITE-TESCHENITE SILL OF LUGAR. dale suite, and is present in all the rocks except the peridotite. It occurs in small amount in the theralites, although unrecorded in the Tables. Where present in notable quantity its amount varies from 30 to 40 per cent. (allowing for the presence of nepheline) in the lugarites, to an average of 15 per cent. in the teschenites. Nepheline occurs in measurable amount only in the theralites (up to 16-6 per cent.), but is almost certainly present in small quantity in the teschenites and lugarites. Titanaugite is the most abundant and constant mafic mineral of the suite. It averages about 30 per cent. in the teschenites, increases to 36 per cent. in the thera- lites, and attains its maximum development, 55°6 per cent., in an augitic variety of picrite. Red soda-hornblende (barkevikite) is absent from the teschenites, save in a domafic schlieren (Table I, No. 5, p. 103), but is present in amounts ranging from 8°6 to 29°5 per cent. in all the other rocks, the latter amount occurring in veins of lugarite devoid of titanaugite. Olivine averages about 10 per cent. in the teschenites and theralites, is absent from the lugarites, and attains its maximum development in the peridotite, where it may form 70 per cent. of the rock. Biotite averages about 3 per cent. in most of the types, and is most abundant (6°7 per cent.) in the hornblendic variety of the theralite. Ivon-ore is most abundant in the teschenites, averaging about 7 per cent., but dwindles steadily downwards through the sill, although it is never entirely absent. Apatite is a comparatively abundant and constant constituent. It averages about 1 per cent. throughout the sill, is absent from the peridotite, and is most abundant in the lugarites, thus illustrating its association with analcite. The mineralogical variation with depth in the sill may be shown by means of a diagram (fig. 5, p. 116). For this purpose it has been found advisable to average the types (see Table VII, below), thus smoothing over minor mineralogical irregularities. Taste VII. Modes of Rocks in the Lugar Sill. TEAS GSS oe ie el ania OG RN Aan at Plagioclase (Ab,An; to] 23:2 | 27-9 | 31°7 | 286 | 19°9 12°6| 2:9 AbgAns). pategclass ieee Seale OR) 56 14 4:7 3 oo MA CIbeH seer eee allie lO Onn eye NGS le : Nephtelinelk eye aaa sae fs Me ah 146 ‘ 457 111 ae Nitanaueite ...0.........- 28°71 | 27-2 | 33:6 | 30°6 | 36:0 10°38) 41-4 | 205 Barkevikite ............... aot ins “a a 61 23°4, 10°6 | 10°0 Olivine (serpentine) ...| 10°6 75 8:7 89 | 13°6 aon || OPE || Ga BIO bIKC Ree eeeee eee ial 15 2:5 25 5:2 tne 2°0 Tron-ores ....../.......... PPTL 9°3 68 WH BRS} 378 3:0 23 Anatitenm, ster rwc cele aes 11 0-6 O09 | 13 Be7 1) OPI aa Felsic constituents ..... 49°5 | 534 | 478 | 49°6 | 345 58°3 140 (0) Mafic constituents ...... 50°5 | 466 | 522 | 50° | 655 | 41:7 86:0 |100 1. Upper teschenite (Table I, 1); 2. Analcitic teschenite (Table I, 2); 3. Lower teschenite (Table I, 3, 4); 4. All teschenite, except melanocratic variety of Table I, 5 (Table I, 1, 2, 3, 4); 5. All theralite (Table III, 1,2); 6. All lugarite (Table III, 3, 4); 7. All picrite (Table V, 1, 2); 8. Peridotite (Table V, 3). 118 | MR. G. W. TYRRELL ON THE [vol. Ixxi, The diagram (fig. 5, p. 116) shows the relation of the mineral constituents to depth within the sill. The depths are plotted as the abscissee and the proportions of the various incr constituents as the ordinates, utilizing the averages Nos. 1, 2, 5, 7, 8, 3, Table VII, in the order given. The “preak between the theralite layer and es teschenite, coors different types occur on opposite sides of a sharp boundary (between 2 and 5), and the similar break between the peridotite and the lower teschenite (between 8 and 3), are clearly shown hy the marked changes in the directions. of the lines at the points indicated. The sill can thus be divided into three sharply-bounded portions, in each of whicha more or less continuous variation may be traced. By far the largest part is that which constitutes the central body consisting of theralite at the top, passing downwards into picrite, and ultimately into peridotite. In this portion plagioclase, analcite, and nepheline show a more or less continuous decrease in a downward direction within the stratum; while olivine rapidly increases in amount, forming about 65 per cent. of the peridotite, and augite and barkevikite find maxima in the picrite and theralite respectively. (2) Chemical Variations. The chemical variations within the Lugar sill naturally reflect and follow the lines of the mineralogical variations. Again, as in the preceding section, the variation in ‘basicity,’ or rather ‘ mafi- city’ (if one may coin such a term), is the significant variation. In this case, however, it cannot be shown by using the silica percentages as 1s customary in chemical diagrams, since the silica percentage varies but slightly throughout the series. The percent- ages of the ferromagnesian oxides show a much more sympathetic Tasie VIII. The ee Maal 1) Ad rae ks 6 | | SiO; 45°09 | 42°61 | 4640 | 42-77 | 4014 | 43-16 | Ti02 2°95 | 2°63 DOtl Ds 7 Ss aiemalee 2°66 Al,03 1549 ) 1470 | i714 | 748 |. 325 | 10:95 | Fes03 3°72 Sidi or leie ey eon ie 377 FeO 7:78 8°59 6:49 | 10:79 | .11:83 8:97 | MeO 668 | 1120 | 1:98 | 1649 | 99:99 | 13:84 | CaO 863 | 8:54 725 | 862 4-40 7°62 NasO SAGAN merle Mme is ALA 2-43 4°61 K30 145 114. 052 | 051). 048 1:07 H30 Dal7) Ve Ome ig AAG Ne OAo le aua ate) 2°69 LAOS ies Ane ec Ae aea home a ed On (alli Mea e(OBS, hla ACSI CORSE TP «ORE 0°65 IRES up Ure sh 1 SEeeaas 10 Oma cre R als O-2N ck avn Motelsie ster! 99°84 | 99°96 10023 | 100°32_| 10022 | _ 99°99 { | 1. Average of teschenites. 2. Average of theralites. 3. Average of lugarites. 4. Average of picrites. ‘5. Average of peridotites. 6.. Average rock of the Lugar sill. part 2] PICRITE-TESCHENITE SILL OF LUGAR. 119 relation to the ‘maficity,’ and may accordingly be used as abscissze in a diagram, against which the other oxides may be plotted as ordinates (see fig. 6). Just as in dealing with the mineralogical variations, it has been found necessary and desirable to average LUGARITE PICRITE THERALITE PERIDOTITE TESCHENITE PERCENTAGES OF OXIDES So in the Lugar sill. Fig. 6.—Diagram illustrating variations in chemical composition PERCENTAGES OF MgO*+FeO the chemical composition of the various rock-types for use in the diagram. ‘The averages thus arrived at are given in Table VIII. In the calculation of these averages most of the chemical com- positions calculated from the graphic analyses have been used, omitting those which duplicate the actual chemical analyses made 120 MR. G. W. TYRRELL ON THE [vol. Ixxn, by Dr. Scott. Thus No. 3 in Table II (p. 104), No. 5, Table IV (p. 110), and No, 5, Table VI (p. 114), have not been used on this score. Furthermore, No. 7, Table II, has been omitted, because it represents merely a small and unimportant schlieren, which would have an effect on the average of the teschenites incommen- surate with its size and importance. The sum of the ferromagnesian oxides being utilized as abscissz, the diagram shows that the order of increasing richness in these oxides is lugarite, teschenite, theralite, picrite, and peridotite. ‘The most striking feature of the curves (fig. 6) is the rapid and regular fall in alumina from lugarite to peridotite, and the slight fall in silica. Soda also falls; but potash, along with phosphorus pent- oxide and titanium dioxide (ignoring small irregularities), remains approximately ata constant level. Lime rises to a maximum in the middle of the series, and drops in the lugarite and peridotite at the extremes. The variation of the curves in this diagram may obviously be correlated with the mineralogical variations which have already been described. (3) Average Magma of the Lugar Sill. In the question of the differentiation of the Lugar sill it is of importance to know the chemical composition of the magma before it split up into its present heterogeneous parts. Assuming that the variation has resulted from the differentiation of an originally homogeneous magma, the composition may be calculated by weighting the analyses of the various components according to their bulk, adding, and then dividing by the number of units taken. The average composition of the principal types is given in Table VIII. The proportions have been taken as follows :— Peridotite (Table VIII, 5) ............... + lekienangs) (Atolls) WARD, 4h) eos cuceg aso ocevouben sre Theralite (Table VIII, 2) .................. 2 Teschenite (Table VIII, 1) ............... Contact-teschenites (Table VIII, 1) ... Le oe Since the analyses of teschenite and of the contact-rocks are for all practical purposes identical, seven parts of Anal. 1, Table VIII, have been taken for the last two items. Lugarite is omitted, as it forms an insignificant proportion of the mass of the sill. The result of the calculation is given in Table VILL, 6. These figures possess significance only if the hypothesis of an intrusion of homogeneous magma, followed by differentiation in place, be accepted. On the hypothesis of successive intrusion it is impossible to assume that the relative volumes of the facies have any necessary connexion with the relative volumes developed within the differentiation chamber. If the former hypothesis be assumed correct the figures show that the original magma must have had a composition intermediate between that of theralite and that of picrite. It had a certain correspondence with that of part 2] PICRITE-TESCHENITE SILL OF LUGAR. 121 theralite, but was richer in magnesia, and poorer in lime, soda, and alumina, indicating a greater richness in olivine-molecules and poverty in those of felspars, as compared with theralite; in short, the magma must have been transitional towards the picrite. Con- sequently, we arrive at the important conclusion that the original magma had a composition different from that of the present contact-rock, and hence, that the sill was heterogeneous when emplaced, and has been shifted since diiferentiation. (4) Composition, Identity, and Banding of the Contact-Rocks. Petrographical examination shows that the upper and lower contacts of the Lugar sill are identical, even down to the smaller microscopical features. Both consist of black, basalt-like rocks, curiously streaked and drawn out im slightly-varying layers, and injected by veins of coarse pink teschenite. They both pass eradually ito coarser and more normal teschenites towards the interior of the sill. The upper teschenite becomes more analcitic as it is traced downwards ; the lower teschenite remains compara- tively uniform as it is followed upwards. These facts show that the teschenite must have been injected into cold rocks, which exercised the chilling effect proper to such contact. The banding is evidence of considerable movement of the viscous magma during or after emplacement. That the movement and banding took place prior to crystallization is shown by the fact that the textures of the various types are perfectly granular. There is no sign whatever of the parallel arrangement of columnar or tabular minerals such as felspar or augite. As has been shown in the previous section, the ehemical compo- sition of the contact-rock is dissimilar to the average rock of the sill, and consequently the facies cannot be considered as due to the differentiation of an original homogeneous teschenite magma. The fact has no significance if the hypothesis of successive intru- sion be accepted. If, however, the sill be considered as due to a single act of intrusion, its heterogeneity must have arisen prior to emplacement. It would have originally had the composition of an ultrabasic theralite or picrite, and contact-rocks of like composition would have been formed. 'To account for the present disposition of the facies it would be necessary to assume that the intrusion ‘has been moved on from its first position, where it was differentiated, into the position that it now occupies, leaving its picritic contact- rocks behind adhering to the old contacts. This view has some support in the marked flow-banding of the contact-rocks; but it is difficult to understand why this movement did not more seriously disturb the stratified arrangement of the various layers within the sill. This assumption of movement of the sill as a whole, subse- quent to differentiation, is not necessary if the alternative hypothesis of successive intrusion be accepted. 122 MR, G. W. TYRRELL ON THE [vol. lxxui, (5) Asymmetry of the Sill. When the arrangement and petrographic nature of the various layers composing the sill is examined in detail, a decided asymmetry becomes apparent. The bands, while generally strictly parallel to the contacts, are not repeated in the same order at the top and at the bottom of the sill. The upper layer of teschenite, becoming richer in analcite downwards, comes to an abrupt end at a sharp junction with fine-grained theralite. The lower layer of teschenite likewise passes very rapidly into the base of the peridotite stratum, although the actual junction is everywhere obscured. The thera- lite band consequently does not appear in the lower half of the sill (fig. 4, p. 96). The asymmetry is much accentuated if the densities and sizes of - the various bands are taken into consideration. ‘The heaviest and largest layer, the peridotite, is not arranged centrally, but is so situated that its centre-line is well below the geometrical centre- line of the sill. On its upper margin itis flanked by a big mass of the somewhat less heavy picrite and theralite, and on its lower margin by a considerably lighter and smalier band of teschenite. Thus the centre of mass of the whole sill must be considerably below its geometrical centre-line. This, of itself, suggests that gravity must have been the controlling factor in the arrangement, at least in the central ultrabasic stratum. ‘The densities of the layers, bearing out the above facts, are shown in the vertical section (fig. 4, p. 96). When plotted, they form a curve which bulges in an asymmetric manner below the centre-plane of the sill. a (6) Density-Stratification in the Sill. Apart from the outer sheath of teschenite and its contact-facies, the remainder of the sill, constituting the central ultrabasic stratum, has itself suffered a subsidiary gravity-stratification. This is indicated by the distribution of olivine, which has collected in the lower layers of the mass. The upper portion of the stratum consists of theralite with 14 per cent. of olivine. This passes gradually downwards into picrite with 30 per cent., and finally into peridotite with 65 per cent. of olivine. The inference is that olivine, the earliest constituent to crystallize, has sunk under the influence of gravity to the lower levels of the stratum. This is in - accordance with the experiments of Dr. N. L. Bowen,! who found that olivine-crystals, forming in an artificial melt approximating in chemical composition to a basic igneous rock, segregated in a dense layer at the base of the crucible in which the melt was contained. 1 «Crystallization-Differentiation in Silicate Liquids’ Amer. Journ. Sci. ser. 4, vol. xxxix (1915) pp. 175-91. part 2] PICRITE-TESCHENITE SILL OF LUGAR. 123 (7) Variations in Texture. Rapid variation in texture is common in alkali-rich rocks, and is highly characteristic of the Lugar sill. These variations are doubt- less connected with local variations in the physical and chemical characters of the magma, especially in regard to the gas-content. A textural feature is the comparatively fine grain of the theralite stratum, which is poor in analcite. At the other extreme are coarse-grained teschenites, especially those rich in analcite. The controlling factor is probably the comparative abundance of fluxes, especially Sadie but the fine grain of the theralite may also be partly due to chilling against one teschenite, as well as to com- parative poverty of aqueous residuum. ‘'The peridotite is medium- to coarse-grained, but this is assignable to the slow rate of cooling im the centre of the sill. The dense teschenitic contact-facies are clearly due to rapid chillmg consequent upon intrusion into cold rocks. The general granularity and lack of parallel orientation among the crystals has already been commented upon. (8) Segregation-Veins. Many veins of a coarse pink teschenite cut the dense contact- rocks, and show that a richly-analcitic magmatic residuum remained in a liquid condition long after the consolidation of the contact- rocks. ‘Thin veins consisting largely of analcite and felspar also eut the peridotite. Both ty pes of vein doubtless represent a slight pegmatitic phase in the development of the intrusion. These veins are coarse-graimed, and show no signs of chilling at their margins. They were doubtless ‘injected while the rocks were still hot. There are also a few veins of a black fine-grained rock resembling the teschenite-basalts of the contacts. These are probably to be interpreted as injections of teschenitic material at a later stage, when the contact-rocks were colder and able to exercise a chilling effect. (9) Mode of Intrusion and Differentiation. Whatever hypothesis of intrusion be adopted, the main differ- entiation of the Lugar sill must be referred to the stage of its history prior to its emplacement. It is not possible, therefore, to discuss it as fully as might have been done if it had occurred 7m situ at a post-intrusion stage, with all the details clearly displayed. The discussion of the petrography makes it clear that all the different facies are genetically related, and it is highly probable that they have arisen by the differentiation of a single body of magma. The differentiation appears first to have produced bodies of teschenite and an ultrabasic rock, which may be designated as picrite.! The mineralogical and chemical relations of these rocks 1 The term picrite was first employed by T'schermak in the sense of a melanocratic derivative of teschenite. 124 MR. G. W. TYRRELL ON THE [vol, lxxn, support the view that the gravitational sinking, either of immiscible fractions, or of crystals, was the chief factor concerned in their differentiation. In any case the details of the process are not open to direct inspection, as the magma has been moved since differentiation, and we are, therefore, limited to indirect inference as to its nature. But the gravitational factor receives further support from the phenomena directly observable in the central ultrabasic stratum of the sill, where a gravitational differentiation occurred subsequent to intrusion. ‘This may be regarded as an indication of the process which took place, one stage previously, in the larger chamber whence the Lugar sill was immediately derived. During the progress of this work I favoured the view that the units of differentiation were immiscible fractions of water- rich teschenite, and comparatively anhydrous picrite, respectively. The sharp interior contact between the upper teschenite and the underlying theralite was regarded as analogous to the sharp contact-plane that is developed between immiscible fractions, such as aniline and water, or various mixtures of metals, which have arranged themselves in order of density. It also became necessary to assume that the sill was the product of a single act of intrusion, and that it was heterogeneous at the time of intrusion. In order to account for the dissimilarity between the teschenitic contact-facies and the bulk-composition of the sill as a whole, the differentiated sill was believed to have received an onward impulse, which pushed it forward into cold rocks, leaving behind its original picritic or theralitic contacts, and establishing new contacts with the succeed- ing layers of teschenite. The recent work of Dr. Bowen has discredited liquation theories of differentiation, and has emphasized the importance of crystalliza- tion-differentiation, where the crystals are continually removed, either by sinking or zoning, from contact with the liquid in which they were formed.! This process receives convincing support from the results of experimental work upon silicate magmas, which has now been carried on for many years in the Geophysical Laboratory at Washington. Jam now inclined to ascribe the main differen- tiation at Lugar to the sinking of heavy crystals in a teschenitic magma. This process is dealt with in greater detail in the next section. The hypothesis of successive intrusion removes many of the difficulties, and makes unnecessary many of the assumptions, mentioned above. On this hypothesis the teschenite was intruded first into cold rocks, forming fine-grained basaltic facies at both margins. While it was still cooling, but probably solid, a thick mass of picrite magma was intruded along its centre-plane. During its crystallization the ultrabasic layer became stratified according to density mainly by the sinking of olivine-crystals, as 1 ‘The Later Stages in the Evolution of the Igneous Rocks’ Journ; Geol. Chicago (1915) Suppl. pp. 1-91. part 2] PICRITE-TESCHENITE SILL OF LUGAR. 125 described in the next section. ‘This, in its turn, was intruded ata later stage, probably while still partly liquid, by a small mass of lugarite, which spread out as a thin sheet at a horizon about a third of the depth of the ultrabasic stratum from its upper surface. If the emplacement of the Lugar sill took place in this way, it is difficult to understand why there are no xenocrysts or xenoliths, intrusive veins, or signs of disturbance, along the main interior contacts. The interposition of the ultrabasic magma must have taken place very quietly and gradually, welding itself intimately to the teschenite without wedging off fragments from the contacts. That the teschenite was fractured and comparatively cool before the intrusion of the picrite is shown by the presence of the black veins of basaltic teschenite, which represent teschenitic magma chilled by intrusion into cold, or comparatively cold, rocks. (10) Sinking of Crystals in the Central Ultrabasic Stratum. Dr. R. A. Daly and others have shown that the earlier and generally heavier minerals must tend to sink in the ordinary silicate magma. Actual cases of phenocrysts that have sunk (or risen) in lavas have been described by authorities of no less weight than Scrope, Darwin, King, E. 8. Dana, and Iddings.! In the great quartz-dolerite sill of the Palisades (New Jersey) J. V. Lewis has shown that a concentration of olivine has taken place near the lower contact, giving rise to a stratum of olivine-dolerite.? This is interpreted as being due to the sinking of early-formed olivine erystals. Dr. Bowen has shown that olivine- and pyroxene- erystals collect towards the bases of crucibles containing a suitable silica melt, but that olivine sinks more readily.2 That the process cannot be more often demonstrated in natural occurrences is probably due to the fact that in most small magmas the onset of an inhibitive viscosity, or of crystallization, is too rapid for the gravitational action to take place ; and in large magmas the final products of the process are not often exposed by erosion. The sinking of olivine-crystals is believed to have taken place in the central ultrabasic stratum of the Lugar sill, where a downward succession from theralite, through picrite, to peridotite, may be demonstrated, a succession characterized by a gradually increasing proportion of olivine. Dr. Daly has calculated that a holocrystalline ‘ basalt’ of sp. gr. 3°10 at ordinary temperatures would have a specific gravity of only 2°83 when molten at 1200° C. The figure 3:10 may be taken as a fair average for the specific gravity of picrite, and we may therefore conclude that picrite molten at 1200° C. would have a * See R. A. Daly, ‘ Origin of Augite-Andesite’ Journ. Geol. Chicago, vol. xvi (1908) p. 411. 2 Ann. Rep. Geol. Surv. New Jersey 1907, pp. 125, 129-33. 3 Amer. Journ. Sci. ser. 4, vol. xxxix (1915) pp. 175-91. Q. J. G. 8. No. 286. L 126 MR. G. W. TYRRELL ON THE [vol, lxxu, specific gravity somewhere near 2°85. Similarly, Dr. Daly has calculated that olivine, which has a specific gravity of 3-40 at ordinary temperatures, would only have a specific gravity of approximately 3°30 when crystallized at 1100°C.! The contrast between the specific gravity of crystallized olivine at 1100° C., and that of molten picrite at 1200° C., is sufficiently great to warrant the conclusion that the olivine-crystals, when formed, would sink to lower levels within the magma, and would continue to do so at decreasing speed until the increase of viscosity con- sequent upon cooling inhibited further movement. It is conceivable that, under favourable conditions, an almost monomineralic layer might be formed in a magma by the sinking of crystals. The favouring conditions would be early crystalliza- tion of a comparatively heavy mineral in a highly fiuid magma, unimpeded by the presence of other minerals. The rate of crystal- lization should be rapid, the sizes of the crystals should be large (as the rate of sinking is proportional to their bulk), and the mineral should appear suddenly in large quantity. Olivine and augite frequently satisfy these conditions, and on sinking they would form, respectively, layers of dunite and pyroxenite. Iron-ores less frequently satisfy the conditions requisite for the formation of monomineralic layers. In most magmas their crystals are small, and the development of the minerals is feeble, though they may be the first to crystallize. Nevertheless, in a few cases, the formation of an iron-ore rock as a result of gravity-settling has been recorded. The sinking of crystals heavier than the surrounding liquid would probably continue to a diminishing extent, through practic- ally the whole period of crystallization. In this case Dr. Bowen believes that the crystals would tend to sink as a swarm, rather than as individuals, with little tendency to relative movement between the different kinds.2 The swarm, however, would be dominated by the mineral crystallizing earliest, by the heaviest or ° largest mineral, or by the mineral cr ystallizing in the greatest bulk with the greatest speed, according to circumstances. The develop- ment of any one mineral within the swarm would be controlled, in general, by a combination of these conditions. Before the density-stratification of the central part of the Lugar sill can be accepted as due to the sinking of early-formed olivine- crystals, it is necessary to explain why augite and iron-ores are not segregated to the same extent. In the first place, the olivine had a start of the other minerals in crystallization. It 1s a rapidly erystallizing mineral, and it was forming in large quantity from a richly-magnesian magma. Hence it would form the major con- stituent of the sinking swarm. Augite, iron-ores, and labradorite began to crystallize somewhat later, andl ‘the two first-named would probably participate in the sinking movement. They would find the magma appreciably more viscous, and the field already largely 1 Journ. Geol. Chicago, vol. xvi (1908) pp. 404-406. 2 Lower: Quartxte Portas! Conglomerate Islay Limestone & Malt of Ca Piv es not to true scale) GEOLOGICAL MAP OF ISLAY, JURA, ete. Quart. Journ, Geol. Soc, Val. LAX, PLAT. (based, with mmor modifications,upon Sheets 19, 20, 27, 28, 35, & 36 of the l-inch Geological Survey Map of Scotland), by E.B.Bailey, w.c.,.p.a..F.G.S. Islay ( East of Loch Skerrols Thrust), Jura, Scarba, Luing, etc. Hasdale Slates (black) Ardmore & Laphroat. Quartzites. Port Elen PhyUites (grey) Scarba Conglomeratic Group. Jura Slates (black & grey) Scarba, Transition Group Lipidiorite Sills Mair Quartxtie Dolomitic Group of N. Islay Upper Conglomerate of N. Islay Lower Fine -Grained Quartzite of Northern klay. Portaskaig Conglomerate Islay Timestone & Garvellach Dolomite LDimestone Mull of Ox Phyllites ( grey ) Maol an Phithich, Qauartxtte \ Dip, amount av degrees & Undiating Dip X Vertical. Strata MORO. GRAHAM LTS, LITHRE, LONDON Se EASDALE) i jorules Shuna iL Ardrishaig Phy lites (grey & green) Degnish Limestone slay (between, [, Skerrols Uhrust & TL. Gruinart Fuult) Jowmore Sandstone 2 /Bhinns of Islay & Colonsay i Lower Torridon yistamy Greiss with Baste Dykes Lewistar, Ardmore Sections along lines A,B, 20, Dubh-Fheith | Raised Beaches of L. Gruimart Hellow [ + [~3] /Tayvalfch { 7 Lavas of Old Red Sandstone Age & C on Map (not to true scale) part 2] THE ISLAY ANTICLINE. 159 IV. BreiiogRarnHy. . J. MaccunLocH.—‘ The Western Islands of Scotland’ vols. ii & iii, 1819. . R. J. Murcutson & A. Grixrz. On the Altered Rocks of the Western Islands of Scotland & the North-Western & Central Highlands’ Q. J. G. S. vol. xvii (1861) p. 171. . J. THomson.— On the Geology of the Island of Islay’ Trans. Geol. Soe. Glasgow, vol. v (1873-76) 1875, p. 200 (a short account was communicated to the British Association in 1871). . One-Inch Geological Survey Map of Scotland, Sheet 20,! 1896. . One-Inch Geological Survey Map of Scotland, Sheet 19, 1898. . One-Inch Geological Survey Map of Scotland, Sheet 27, 1900. . ‘The Geology of Islay (Explanation of Sheets 19 & 27 & part of 20)’ Mem. Geol. Surv. Scotland, 1907. 8. W. B. WricgHt.— The Two Harth-Movements of Colonsay’ Q. J. G.S. vol. Ixiv (1908) p. 297. 9. One-Inch Geological Survey Map of Scotland, Sheet 36, 1909. 10. ‘The Geology of the Seaboard of Mid Argyll (Explanation of Sheet 36) ’- Mem. Geol. Surv. Scotland, 1909. 11. One-Inch Geological Survey Map of Scotland, Sheet 28, 1911. 12. ‘The Geology of Knapdale, Jura, & North Kintyre (Explanation of Sheet 28 with parts of 27 & 29)’ Mem. Geol. Surv.-Scotland, 1911. 13. One-Inch Geological Survey Map of Scotland, Sheet 35, 1911. 14. ‘The Geology of Colonsay & Oronsay, with Part of the Ross of Mull (Explanation of Sheet 35 with part of 27)’ Mem. Geol. Surv. Scotland, 1901. The authorship of the above-mentioned Geological Survey publications— so far as it concerns the subject of the present paper—is, in the main, as follows :—4, 5, and 6: S. B. Wilkinson. 7: B. N. Peach & S. B. Wilkinson. 9: S. B. Wilkinson (Jura) & B. N. Peach (northern end of Jura, Scarba, etc.). 10: B. N. Peach. 11: 8. B. Wilkinson. 12: H. B. Bailey. 13 and 14: W. B. Wright & HE. B. Bailey. (Se) ho Re ATG Or & EXPLANATION OF PLATE XII. Geological map of Islay, Jura, ete. on the scale of 4 miles to the inch, or 1:253,440 ; with sections across those islands (not to true scale). [| for ‘Beinn Bahn’ read ‘ Beinn Bhan’; for ‘ Coillabas’ read ‘ Coillabus’ ; for ‘Dubh-Fheth’ read ‘ Dubh-fheith.’] 5 Discussion. Mr. G. Barrow drew attention to the wide range of the Author’s observations ; with regard to the Moine Thrust, its position on the published maps rendered its occurrence close to Islay impossible. He agreed with the Author that there was a great break between the main part of Islay and the western and almost detached part. But, although there might be a fault as well, the changes seen in the condition of the strata as regards crystallization were due mainly to the thrust that passed along the Great Glen of the Caledonian Canal and continued through the Islay area. As in the case of the Moine Thrust, little-altered or non-crystalline rocks on the south-east side of the Thrust were driven over or on to rocks that are highly crystalline. Referring to his paper published by the Geolo- gists’ Association, and accompanied by a map in which the Highland 1 The corners of these Sheets are shown on Pl, XII, 160 THE ISLAY ANTICLINE. [vol. lxxii, rocks are shown to be divisible into zones of increasing and de- creasing thermal alteration, the speaker pointed out that the thrust drives the outer zone, in which there is no crystallization, on to a ‘zone of high alteration; the absence of crystallization in a con- siderable part of the Highland rocks of Islay is thus no proof at all of newer age—it is natural to the thermal zone, to which they belong. The limestone is so little altered, that its original oolitic str eae is locally preserved despite its great age. , He agreed with the Author that the rocks in Western Islay were oe Lewisian age ; but, in addition, they were part of the Highland rocks, or “oha series of sediments of which the Moine Gneisses formed a part, and had the more common strike of the Highland folding, north-east and south-west. The speaker complained of the introduction by the Author of new names for rocks which had long-established names. Thus, in the south-east of the island the “Port Ellen Phyllite’ has long been known to. be the ‘Ardrishaig Phyllite’ or ‘ Canlochan Schist.’ This band of shale is known to continue across Seotland, and emerges on the coast to the south-east of the Portsoy Quartzite. It is more of a true shale-in the latter area, and contains thinner .if not fewer sandstone-bands than in Islay. Sunilarly, the quartzite ” is the well-known Highland Quartzite, which again crosses the “country to Portsoy ; it is also finer and thinner at Portsoy, thicker and coarser in Islay. Again, the limestone is the Blair Athol Limestone, which also -oeeurs on both coasts; the course of it and the Quartzite are fairly well shown on the naoee recent edition of Sir Archibald Geikie’s -10-mile to the inch geological map of Scotland. The limestone in . Islay also is affected by changing conditions ; while the associated = dark schist is, for the greater part, more sandy and lighter-coloured _than in areas to the north-east. With regard to the stratigraphy of the area, it is known that the side of the Quartzite on which the limestone occurs is the opposite to that on which the Ardrishaig Phyllite (Port Ellen) occurs. This was proved many years ago, although the evidence has never been published. Inspection of the map already mentioned will show that the Blair Athol Limestone crops out, more or less con- tinuously, for many miles on the north-west side of the first outcrop of the Quartzite. On the south-east side of it another persistent outcrop occurs, of the limestone known -as the Loch Tay or Pitlochry Limestone. For many years these were thought to be the same limestone ; but, after a short period of mapping , the speaker found that ‘ants earl not be the case, and, after a “considerable ‘argument,’ Sir Archibald Geikie was satisfied that the new view was correct, and that they were stratigraphically on opposite sides of the Quartzite. There was no question of their not being different limestones—that was settled ; it then became incredible that there could be these two limestones continuing for miles, each confined . to one side only of the Quartzite, without their being strati- _ graphically on opposite. sides of that rock. These results were embodied in Sir Archibald’s Presidential Address to the Geological part 2] : THE ISLAY ANTICLINE. Society in 1891. No one has since questioned the fact being separate limestones, and it is doubtful whether thor a single modern official Survey publication in which the Loe! Limestone is not taken to be below the Quartzite. Had facts been published at the’ time, the reniarkable discussion the Quariaite being at the oe of the Series could not have Thus on the eastern side of Islay the top and the mad Quartzite occur: the latter (as usual) is the coarser, the to finer, often very fine and often remarkably white ; ‘this fine mai occurs on the side next fhe Blair Athol Limestone, and there be in the Highlands altogether many hundred outcrops 0: white margin, though, except by the speaker, they are rarely ever mentioned. The usual erosion occurs below the ‘limes! on and, being in a folded area, it causes the outcrop of the bed approach and recede from the margin of the Quartzite in the usual — characteristic manner seen’on most modern maps. When the limestone is followed round, it is seen that the Quartzite is all one — quartzite; and this: is confirmed by the’ frequent occurrence the typical white margi in the area some distance west of Loch Finlaggan.’ It is now ‘seen that the Boulder Bed (Portaskaig Conglomerate) of this area has-exactly the same relations with thie limestone and quartzite ‘as at Schiehallion, ‘placing the ‘identity: of the various beds beyond serious*doubt; the limestone is the Blair Athol Limestone. (The speaker here placed on the table the Geological Survey l-inch maps, Sheets 27 & 55, so that the Fellows might satisfy themselves as to this identity.) In view of what has been stated, the theory advanced by the ‘Author that the ‘ground’ between the two outcrops of quartzite is an anticline seems impossible: if, as shown, the limestone is above the fine margin of the Quartzite, it must be a tilted syncline, the structure being identical with that at Portsoy in the East of Scotland. The white margin is well seen on the ground, where the so-called ‘Fucoid Beds’ are shown to terminate on the map; the speaker found no trace of Fucoid Beds at the places indicated. The statement that the Author disagreed with the view that the Portaskaig Conglomerate was not the Scarba Conglomerate is not well put: the ground was mapped by Mr. Wilkinson, who held no ‘such view, but: his opinion was ignored; further, Mr. Wilkinson _had no doubt that the’ two conglomerates are on opposite sides of the Quartzite, but he Jamie: whether the side on which vine s limestone occurs was the top. Turning now to the sandstone south-west of the so-called ‘ Tuids Skerrols Thrust,’ the speaker thought these crucially important ; _ he doubted the existence of the thrust at all, and believed that these _ sandstones are brought on by pitching of the folded Quartzite, and that small patches cross the so-called thrust and are infolded with the margin of the Quartzite. These beds, as also their relation to _ the limestone and Quartzite, are better seen in the Bowmore area. One now sees that these sandstones have the persistent flaggy habit and the curious persistent dip in one direction, due to singularly Se * NS 16: THE ISLAY ANTICLINE. [vol. Ixxil, perfect isoclinal folding, that characterize the Moine Gneisses ; frow.a study of the maps, the speaker felt that there was a pos- sibility of the original rocks, from which the more massive types of #elspathic Moine Gneiss were formed, being found in this part of Islay, and this was the cause of his visit to the area. He had now little doubt that these sandstones are the rocks in question. They are practically free from any trace of crystallization, present all the characters that the unaltered representatives of the Moine (meiss should possess, and the Bowmore area shows that they are on the fine margin of the Quartzite and, so far as the folding will allow us to see, they are between the limestone and the Quartzite: that is, they are in the position assigned to them in the a S paper published by the Society. Far from being of any great thickness, they were probably quite a thin group originally. Islay is thus one of the most encouraging areas in the Highlands for further examination, and is worth - mapping with the minutest detail. Dr. B. N. Peacu, in the following written communication to the Secretary, stated that, while making criticisms, he would wish at the same time to state that he considered the view put forth by the Author—that the dolomitic ‘ Fucoid’-like beds are not the highest rocks in the Islay succession, as held by the Geological Survey, but are in turn overlain by the main quartzite, the Jura Slates, the Scarba Conglomerate, and Port Ellen Phyllites, in upward sequence —to be worthy of the closest consi nniicr. although he felt that the Author had not brought forward sufficient evidence to establish thoroughly his contention. If this could be done, many difficulties left unsolved by the Geological Survey would be cleared up :— ; 2 ‘The important feature in the paper is the evidence which the Author adduces to prove a further extension of the anticlinal arrangement of the strata in Islay, east of the Loch Skerrols Thrust. This structure, in a modi- fied form, was previously recognized by the Geological Survey. In the Memoir on ‘The Geology of Islay, it was shown that the Islay limestones, black slates, and phyllites form the core of a compound anticline extending from the Mull of Oa north-eastwards to Kiels, near Portaskaig. Jt was also pointed out that, between Loch Finlaggan and Port a’ Chotain in Northern Islay, the strata, including five subzones, ranging from the Portaskaig Con- glomerate up to the Fucoidal Beds with massive bands of dolomite, are arranged in a compound arch. “In his paper the Author contends that the dolomitic group, which resembles the Fucoid Beds of Cambrian age in Sutherland and Ross, is not the highest member of the sequence in Northern Islay, butis overlain by an upper quartzite on the eastern and western limbs of the compound arch. On the eastern limb this upper quartzite runs from Rudh a’ Mhail to Bonahaven, on the western limb from Rudh a’ Bholsa to Loch Skerrols in the south. He also holds that there is an ascending sequence from his upper or main quartzite of Islay, Jura, and Searba, to the black slates of Searba, Lunga, and Hasdale. ‘IT submit the following criticisms on these opinions :— “1. In proof of the superposition of the upper or main quartzite on the east limb of the compound arch in Northern Islay, the Author states that, west of Rudh a’ Mhail, the dolomitic group appears as an inlier with gentle anticlinal dips passing below the upper quartzite of Rudh a’ Mhail and Port a’ Chotain. In my opinion, the evidence shows that the dolomitie group does not there form an inlier, and that the Port a’ Chotain quartzite underlies the dolomitic group. The horizon of the latter quartzite is proved by the occurrence in it part 2] THE ISLAY ANTICLINE. on its west side, of the characteristic band of conglomerat granitoid pebbles near the top of that zone. The fault separa a Chotain quartzite from the dolomitic beds on the west does 10% ¢ reasoning. If there is an upward succession from the dolomitic gro the main quartzite from this supposed inlier, it can only be found cn i fs margin, where the dolomitie group with a south-eastward dip is expose beach followed by the quartzites of Rudh a’ Mhail. But here the do beds and the quartzites are traversed by shear-lines, showing that the o relation has been disturbed by movement. “The evidence at the western margin of the dolomitie group ef an passage into the quartzites of Rudh a’ Bholsa on the western limb of the the section is interrupted by a fault forming a wide shatter-belt. Hast fault, pebbly quartzites occur for a few yards in contact with the dol» group; but, in my opinion, these pebbly quartzites may represent the boulde bed near the top of the lower quartzite. i “As regards the Author’s statement that the Islay Anticline is appa from the dips in his main quartzite-belts, too much reliance ought not tu placed on this line of evidence in so highly folded a region. In his own ma: and sections he is obliged to invoke inversions of the strata north of Loch ' Skerrols, where his so-called ‘main (upper) quartzites’ dip south-eastwards as_ if passing, In normal sequence, below the dolomitic group ! ‘2. According to the Author’s hypothesis, his main quartzite is overlain by the Jura black slates, and these in turn by the Scarba Conglomerate Group in Islay, Jura, Scarba, and Lunga. If there is a succession upwards from the main quartzite in Islay and Jura, it does not extend into Scarba. For, in Scarba and Lunga, the Scarba Conglomerate contains large masses of the black slates, limestones, and pebbly grits lying to the east of it. If the black slates of Scarba and Lunga belong to the Luing and Hasdale black-slate group, there can be no upward succession from the Scarba quartzite into the Hasdale black slates. *3. The Author states that black slates do not oceur in the island of Shuna, between the Craignish phyllites and the Shuna limestone. Black slates occupying this position were officially mapped by him at the southern end of the island, though now repudiated by him. In my opinion, black slates occupying this position occur not only at the southern end of the island, but in the middle and at the northern end, where they were formerly wrought. ‘4, In the quartzite-belt north of Loch Skerrols, and in the belt between Port nan Gallan and the Sound of Islay, the lower and upper (or main) quartzites are represented in his map and sections as coming together, without the interealation of the dolomitic group. No explanation of this feature is given by the Author. “5. As regards the Maol an Fhithich Quartzite, which is supposed by the Author to underlie the Mull of Oa Phyllites, I think that his conclusion is erroneous: for the junction, where exposed at the northern end, is certainly a line of fault. In my opinion, it is merely a faulted portion of the same quartzite as that which cverlies the Loch Skerrols Thrust, north of Bridgend. ‘6. The Author refers to the Bowmore Sandstones south-east of Loch in Daal as a fine-grained series, but in that very area there are marked pebbly bands to which reference is made in the Islay Memoir (p. 27). ‘7. The occurrence of Bowmore grits and sandstones west of Loch in Daal has been questioned by the Author. These grits are to be found on the shore near Gortan schoolhouse, 1 mile north. of Bruichladdich, where they show _ the same stage of deformation as the other rocks of the Rhinns, facts which have an important bearing on the position of his supposed Loch Gruinait Fault. ‘8. Regarding the Author’s suggestion that the Loch Skerrols Thrust may represent the Moine Thrust of the North-West Highlands, I wish to call attention to the fact that the rocks above and below this line of disruption in Islay are in the same low grade of metamorphism—a feature quite unknown in any region where the Moine Thrust has been mapped. 164 THE ISLAY ANTICLINE, [vol. Ixxi. ‘Dr, Clough’s view that the Lewisian gneiss and unaltered Torridonian rocks of Iona are separated from ‘the highly crystalline. Moime Schists of the, Ross of Mull by the Moine Thrust is quite in accordance with the phenomena connected with this line of movement in the north. ‘Talso agree with ‘Dr. Clough’s’ suggestion that the position of the Great Glen Fault lies between the Ross of Mull and Colonsay, on the ground that the strata on the downthrow side of the fault are in a low grade of meta- morphism, compared, with the Moine Schists of the Ross of Mull and Morvern, which existed as crystalline schists before the intrusion of the igneous rocks by which they are now traversed.’ The contributions by Mr. eee and Dr. Peach were for warded tio the AurHor, who sent the following reply : « *T believe that there is good foundation for some of Mr. Barrow’s views regarding the equivalence of rock-groups found in Islay and Perthshire respectively. In fact, a comparison was early initiated by Macculloch. On the other hand, I think that Mr. Barrow employs a very dangerous method when he confidently interprets the.Islay structure and succession in the light of his Perthshire experience. He comes to the conclusion, for instance, that Central Islay is of synclinal structure, although, in my opinion, it is easy to demonstrate the reverse in the field. He also regards the Islay Limestone as of later date than the Quartzite, although this, too, is contiery, to the local evidence. ' *T do not think that the complexity of ie Eiehiand soblew has been sufficiently appreciated in the past. Before a satisfactory general solution is bttained, I am conyinced that we shall have to:acenmulate careful field- Shservations for several years to come. »-*The points raised by Dr. Peach are best coneiderea servatum :— 1. I consider the sections along the northern coast of Islay to be quite con- vineing in regard to structure and sticcession; and I helieve that the descrip- tions given in my paper will be found to be accurate. In answer to the criticism that I have to invoke inversions north of Loch Skerrols, I should like to.point out that-this is not a part of the district where one would attempt to decide the anticlinal or synclinal structure of the Islay Foldas a whole. It us on a flank, whereas:the northern coast exhibits the axial region. .. 2. Dr. Peach’s second criticism is based upon the claim that the source of certain boulders contained.in the Scarba Conglomerate in Scarba and Lunga ‘ean be identified on the localvevidence. As a first suggestion, I think Dr. Peach’s interpretation of Scarba excellent; but, in the light of the fuller ‘evidence afforded by Jura and Islay, I am convinced that it now needs | ‘modification. 3. I mapped black slates as occurring in Shuna when I. accompanied ‘Dr. Peach to that island within a few months of my. joining the Geological Survey. Dr. Peach at the time told me that a particular belt consisted of “bleached black slate.’ Since then I have revisited Shuna, and compared ‘the rocks in question with such small areas of bleached black slate as are sometimes found. immediately adjacent to outliers of Old Red Sandstone farther north, and am convinced that a mistake was made. i 4, The reason why a detailed explanation is. not given, is. that it is impos- sible to decide between the various alternatives that.are available. 5. The evidence upon which my conclusion. is based is given in my paper, - - 6. Allalong the Bowmore shore the sandstone is thoreughly fine-grained, as stated in my paper. 7. It is not denied in my paper that the Bowmore Sandstone may have been rightly identified on the shore near Gortan schoolhouse. The discussion of the evidence for the Loch Gruinart Fault explicitly includes this possibility. 8. The absence of highly crystalline schists immediately above the Loch Skerrols Thrust is emphasized in my paper, and is taken into account in con- sidering the possible equivalence of the Loch Skerrols and Moine Thrusts.’ [September 10th, 1917. | ADMISSION AND PRIVILEGES OF FELLOWS OF THE GHOLOGICAL SOCIETY OF LONDON. Every Candidate for admission as a Fellow must be proposed by three or more Fellows, who must sign a Certificate in his favour. The Proposer whose name stands first upon the Certificate must have a personal knowledge of the Candidate. Fellows on election pay an Admission Fee of Six Guineas. The Annual Contribu- tion paid by Fellows is Two Guineas, due on the Ist of January in every year, and payable in advance; but Fellows elected in November or December pay no Contribu- tion for the current year. The Annual Contribution may, at any time, be compounded for by a payment of Thirty-Five Pounds. The Fellows are entitled to receive gratuitously all the volumes or parts of volumes of the Quarterly Journal of the Society that may be published after their election, so long as their Annual Contributions are paid; and they may purchase any of the publications of the Society at a reduction of 25 per cent. under the selling-prices. The Library is open on Week Days to the Fellows between the hours of 10 and 5 (except on Saturdays, when the hours are from 10 to 1, and except during the fortnight commencing on the first Monday in September; see also next page).. Under certain restrictions, Fellows are allowed to borrow books from the Library. Publications to be had of the Geological Society, Burlington House. QUARTERLY JOURNAL. (Vols. III to LXXI, inclusive.) Price to Fellows, 13s. 6d. each (Vols. XV, XXIII, XXX, and XXXIV to LXXI, 16s. 6d.), in cloth. GENERAL INDEX TO THE FIRST FIFTY VOLUMES OF THE QUARTERLY JOURNAL (1845-1894). Part I (A-La). Part II (La-Z). Price 5s. each. To Fellows 3s. 9d. each. [Postage 3d.] GEOLOGICAL LITERATURE added to the Geological Society’s Library during the years ended December 31st, 1894-1912. Price 2s.each. To Fellows, Is. 6d. each. [Postage 3d.] THE HISTORY OF THE GEOLOGICAL SOCIETY OF LONDON, by H. B. Woopwarb, F.R.S. Price 7s. 6d. To Fellows, 6s. [Postage 6d.} THE CENTENARY OF THE GEOLOGICAL SOCIETY OF LONDON, recorded by W. W. Warts, F.R.S. Price 2s. net. [Postage 3d.] HUTTON’S ‘THEORY OF THE EARTH,’ Vol. III, edited by Sir ARCHIBALD GEIKIE, D.C.L., F.R.S. Price 3s, 6d. To Fellows, 2s. [Postage 4d.] THE GHOLOGY OF NEW ZEALAND. Translated by Dr. ©. F. - FiscHEeR from the works of MM. HocusruttmR & PETERMANN. With an Atlas of Six Maps. Fellows may purchase one Copy of this Book at 2s. Additional Copies will be charged 4s. [Postage 5d. ] CONTENTS. PAPERS READ. Page 6. Mr. H. G. Smith on the Lurgecombe Mill Lamprophyre and its Inclusions (Blabes VEU & EX) ees ee SANS a aoe eS Liat leat icine Bro ete ote stare lan ae 77 7. Mr. G. W. Tyrrell on the Picrite-Teschenite Sill of Lugar (Plates X & XI). 84 8. Mr. E. B. Bailey on the Islay Anticline (Plate KIT) .................0:ceeee eee ee 132 (The Editor of the Quarterly Journal is directed to make it known to the Public that the Authors alone are responsible for the facts and opinions contained in their respective Papers. | *,* The Council request that all communications intended for publication by the Society shall be clearly and legibly written on one side of the paper only, with proper references, and in all respects in fit condition for being at once placed in the Printer’s hands. Unless this is done, it will be in the discretion of the Officers to return the communication to the Author for revision. The Library at the Apartments of the Society is open every Week Day from Ten o'clock until Five, except on Saturdays (when the hours are from 10 to 1), and except during the fortnight commencing on the first Monday in September, when the Library is closed for the purpose of cleaning. Vol. LXXII. GEOLOGICAL SOCIETY. QUARTERLY JOURNAL : EDITED BY THE PERMANENT SECRETARY: [With Nine Plates, illustrating Papers by Dr. W. R. Jones, Prof. §. J. Shand, and Dr. A. Holmes. | x OO NO ON Np NovemsBer 23rd, 1917. LONDON : LONGMANS, GREEN, AND CO, PARI8:—CHARLUS KLINCKSIECK, 11 RUH DE LILLE SOLD ALSO AT THE APARTMENTS OF THE SOCIETY. Price Five Shillings. g OO eee RRARALRADA A AAA AAA AAO LIST OF THE OFFICERS AND COUNCIL OF THE | A ARAOAO VN ann Vvevvvey Presivent. Alfred Harker, M.A., LL.D., F.R.S. Gice-Presidents. R. Mountford Deeley, M.Inst.C.H. Edwin Tulley Newton, F'.R.S. William Johnson Sollas, M.A., LL.D., Sce.D., F.B.S. Arthur Smith Woodward, LL.D., F.BS., F.LS. Secretaries. Herbert Henry Thomas, M.A., Sc.D. Foreign Secretary. | Herbert Lapworth, D).Sc., M.Inst.C.E. G reasurer. Sir Archibald Geikie, O.M., K.C.B., D.C.L., | James Vincent Elsden, D.Sc. LL.D., Se.D., F.R.S. COUNCIL. Charles William Andrews, D-.Sc., F.R.S. Prof. John Cadman, C.M.G.,D.Sc., M.Inst. C.E. Prof. Charles Gilbert Cullis, D.Sc. Arthur Morley Davies, D.Sc., A.R.C.Se. R. Mountford Deeley, M.Inst.C.E. James Vincent Elsden, D.Se. Prof. Edmund Johnston Garwood, M.A., Se.D., F.R.S. Sir Archibald Geikie, O.M., K.C.B., D.C.L., LL.D., Se.D.. F.R.S. Walcot Gibson, D.Sc. Alfred Harker, M.A., LL.D., F.R.S. Winlay Lorimer Kitcrin, M.A., Ph.D. George William ILamplugh, F.R.S. Herbert Lapworth, D.Sc., M.Inst.€.H. Prof. John Edward Marr, M.A., Se.D., E.R.S. Edwin Tulley Newton, F-.R.S. Richard Dixon Oldham, F.R.8. Robert Heron Rastall, M.A. Prof. Thomas Franklin Sibly, D.Se. Prof. William Johnson Sollas, M.A.,LL.D., Se.D., F.R.S. Sir Jethro J. Harris Teall, M.A., D.Se., LL.D., F.R:S. Herbert Henry Thomas, M.A., Se.D. Samuel Hazzledine Warren. Arthur Smith Woodward, LL.D., F.R.S., F.LS. Permanent secretary. L. L. Belinfante, M.Sc. Librarian. C. P. Chatwin. Clerk. M. St. John Hope. Assistant tn Library. Arthur Greig. STANDING PUBLICATION COMMITTEE. Dr. A. Harker, President. Dr. H. Lapworth. Dr. Herbert H. aoe SECM 2: Dr. F. A. Bather Prof. J. Cadman. Prof. C. G. Cullis. Dr.:J. V Elsden. Dz. W. Gibson. Dr. F. L. Kitchin. Mr. G. W. Lamplugh. Mr. E. T. Newton. Mr. BR. D. Oldham. Prof. T. F. Sibly. Prof. W. J. Sollas. Sir Jethro Teall. Dr. A. Smith Woodward. ORDINARY MEETINGS OF THE GEOLOGICAL SOCIETY TO BE HELD AT BURLINGTON HOUSE. Szssion 1917-1918: LOL? Wednesday December ..................... 5—19* 1918. ss Pa nuary Mysto Sayeed 9*—_23* 5s February (Anniversary, Friday, February 15th). 6*—20* a dni Erbe) see A abel Saree a re 6—20* at ONG 0) al BAU OE Ghat iate gah b 17* Sy NE fs shgrtn Ga a iss Hala REL REALA 1—15* As UNS UE MG Ua hrs Ga G 5—19* [ Busuness will commence at 5.80 p.m. precisely. | The asterisks denote the dates on which the Council will meet. part 38] SECONDARY STANNIFEROUS DEPOSITS IN PERAK. 165 9. The Ortein of the Seconvary STannirerous Deposits of the Kinra Drsrricr, Perak (FEDERATED Manay Srarms). By Wiiiiam Rrcwarp Jones, D.1.C., D.Sc.(Lond)., Ps@S., A.LM.M., Late Assistant & Acting Government Geologist to the F.M.S. (Read June 23rd, 1915.) [Puares XIJI-XV.] CONTENTS. Page Un PrOGUCHIONLS wasramacst ise dle eels ee eo oe 165 Il. General Description of the Kinta District ............... 166 Ill. Geological History of the Kinta Valley .................. 169 IV. Alluvial Origin of the Stanniferous Clays and the boulder k Cl avisi rei ok ates 5 a imtth i otra a rae 175 V. Mode of Hanlting) of the Clays'..)...-....4)--\ t u iy G CHALLI ‘ 6309 + aG.KONBA 6.KONBAR * 4G JasAN Parnes \ +" AG Sucuor > Te ALAM Es, » 53900 |, bo Ms Ti pene Ree, Celie ft +, * 46 CHABANG’, fa Es Selo} + + re saath | a aa: a Sie a sa fpetaer i, tb ope et +t U > we & DANGLt NK fea + 4 + 1AC. PERGANTONG + 4740 GEOLOGICAL SKETCH-MAP OF THE KINTA DISTRICT, PERAK , FEDERATED MALAY STATES. The positions of the GraniteJunctions & Granitic Intrusions are identical with those inthe Government Geologists Geographical Sketch-Map of this District. 4 Gi rt Alavi atone Sete Mee hea itacks, Dykeciocste, overlying Qoartzites InSohists, ete. Limestone Smallorante Floor. pein is The thus®@ represent, in order of production The renee hee Kenta Disorict with the largest oulpue for the Year /913. GHisau. wera ve! G.LaNno. G.CHABANG G SequoR: G IRAU, (saa Fr) (5,010 F9) TALAM. (920) Railway. KINTa RIVER. = Atuvium. Fa @ prea, ii Vertical Scale 1% miles to Goe intl, or 195040 SECTIoN AcRoss Kinva DisTRICT Ea Sehrat, Phyl tee) Horizontal Scale 3 miles to Ue teh, ar 1:180/080. From G.IRav To G.Hrsav. H.&C.GRANAM LY, LITHAA, part 3] OF THE KINTA DISTRICT. 197 two steep granite-ranges, would probably dissolve the limestone to such an extent that no glaciated pavement would exist. He agreed with Mr. Serivenor that no importance should be attached to the absence of such a glaciated pavement. It was very difficult, as Dr. Thomas pointed out, to realize at first what tropical weathering could do; and only after great familiarity with the rocks in the field could one confidently offer an opinion. The speaker’s mapping of the Ulu and Kuala Selangor districts had afforded him much help in understanding the structure of the Kinta district. The only difference stated to exist, between the tin-ore in the ‘clays and boulder-clays’ and that in the admittedly alluvial deposits, is that in the former it is more angular. Mr. Scrivenor published a photograph, in his memoir on ane Kinta district, to illustrate this point.. It was significant, however, that the angular ore there figured came from a mine (Siputeh) where a lode was worked some years ago, and where cassiterite was definitely found in situ in 1914 by the manager and the Author. Mr. Scrivenor, in a lecture on ‘'Tin-Ore in Limestone’! agreed that tin-ore had been found recently 7 szfw at this mine. Mr. Kitson was correct in stating that later developments had yielded much information that was previously unobtainable. Mr. Serivenor still, however, believed in his theory; his memoir and map of this district were only published in 1913, and in a recent paper in the ‘ Geological Magazine’ (July 1914, p. 309) he stated that the glacial theory is the one that best explains the facts seen in the field. In the Annual Report of the Geological Department of the Federated Malay States for 1913, he (the speaker) wrote that in Ulu Selangor he had, in every case, traced the tin-ore from the valleys, up the streams, to schists and granite where tin-ore was definitely 77 s7tuw. With regard to the boulders, Mr. Scrivenor stated that nota » single boulder showing undoubted glacial striz had ever been found. Boulders were being continually washed out of the stanni- ferous clays, and excellent opportunities were given for examining thousands of them in the dump-heaps. He thanked the Fellows for the kind manner in which they had received his paper, and for the interesting points raised in the discussion. 1 See ‘Times of Malaya’ Ipoh, 1914, & F.M.S. Chamber of Mines publica- tion of this lecture. Q.J.G.S. No. 287. Q 198 PROF. 8. J. SHAND ON [vol. lxxu, 10. The PsevpoTAcHYLYTE of Partss (ORANGE FREE STATE), and its RELATION to ‘TRaP-SHOTTEN GNEISS’ and * FLINTY CrusH-Rocxk.’ By 8. James SHanp, D.Sc., F.G.8., Pro- fessor of Geology in the University of Stellenbosch. (Read March 22nd, 1916.) [PLatres XVI-XIX. ] CONTENTS. Page JE da Gata faysurKei nko} he mAanue Incas MareManninndannennoamudacer nods sostaah Soe 198 TO ccurrenceyintheyhiel dimen enna ace ee ceeee 199 III. Microscopic Characters of the Pseudotachylyte ......... 206 IV. The Case for the Igneous Origin of the Veins .....,...... 208 V. Comparison of the Pseudotachylyte with ‘ Trap-Shotten Gneiss’ and ‘ Flinty Crush-Rock’ ......................55 209 VI. The Chemical Composition of the Pseudotachylyte ...... 213 VII. The Case against an Igneous Origin ...................0.055 215 WEEE Conclusion: Oar ntact aoeset Sen aeaee Mase Cao EE eee 216 “EX ui Gera cure sas aia sa cette anne loniomte mada eeen toate Sanches 217 Appendix (Chemical Analyses)..........0..cccecceeeeeseeeer ees 217 I. IyrrRopvction. In the year 1913 I received from a former student of mine, Mr. P. H. 8. De Wet, a consignment of rock-specimens from the neighbourhood of Parijs (O.F.S.). Among these were many pieces of a dense black rock, which, I was informed, occurs very abundantly in veins and networks in the granite. The characters of this rock seemed so remarkable that I took an early opportunity of visiting | Parijs myself, in order to study the phenomena in the field. As a result of my observations, I formed the opinion that the black veins were intrusions of basic magma which had entered the granite by a process of stoping, accompanied by corrosion and solution ; and a paper embodying this conclusion was communicated to the Geological Society in November 1914, an abstract of it appearing in the Proceedings of the Society for that month (No. 964). After the manuscript had left my hands, I became aware for the first time of the work of Sir Thomas Holland and his colleagues of the Geological Survey of India on the ‘ trap-shotten gneiss’ of Salem, Madras, and I could not fail to be struck by the resemblance between some of the phenomena there described and those observed by me. My attention was next turned to the ‘ flinty crush-rocks’ of the Cheviot Hills and other parts of Scotland, which were also at one time held to be of igneous origin. By the kindness of Dy. Flett, the late Dr. Clough, and Mr. E. B. Bailey, I was enabled to study hand-specimens and sections of a large number of these rocks ; and although, on comparing them with my own material, part 3] THE PSEUDOTACHYLYTE OF PARIJS. 199 I found the differences to be at least as numerous as the resem- blances, it became clear that I must reconsider the features of the Parijs occurrence in the light of the evidence derived from other regions. Parijs was accordingly revisited during 1915, and such fresh information and illustrations as were gained have been em- bodied in the account which is now presented. It should be explained that the name pseudotachylyte has been adopted in recognition of the fact that these rocks have a great similarity to tachylyte, also that such rocks have been mis- taken for trap and tachylyte in Scotland and India as well as in South Africa, and for the further reason that no more suitable name is in existence. ‘Tyrap-shotten gneiss’ denotes the entire complex of intrusion and intruded rock, and is misleading in view of the fact that the intrusive part is not certainly ‘ trap’ at all; while ‘flinty crush-rock’ begs the question—as regards the Parijs rocks, at least. Il. OccURRENCE IN THE FTELD. The township of Parijs lies upon the northern portion of the Vredefort granite-mass and on the southern bank of the Vaal River. The granite, which has an outcrop of some 400 square miles, has generally been regarded as Archean (that is, pre- Wit- watersrand), and appears as such on Dr. F. H. Hatch’s map of the Transvaal; but Dr. Molengyraaff, and more recently Mr. F. W. Penny, have held that this granite is in reality intrusive in the Witwatersrand System. No detailed survey of the region has yet been made. The ‘granite’ in the neighbourhood of Parijs is a streaky granitic gneiss, composed of red and grey elements. Sometimes the red forms patches and streaks within the grey, elsewhere the grey matter is similarly enveloped by the red, or again the two elements may constitute alternate bands. The red matter often forms veins and bands of coarse pegmatite which run parallel to the direction of foliation of the grey rock, but in other cases such veins -cut sharply across the foliation. These pegmatites are occasionally very coarse-grained graphic granites. When extensive exposures _are studied, 1t becomes evident that the red portion is of later con- -solidation than the rest of the rock. Isolated ‘ floaters’ of banded grey paragneiss can be found embedded in the red granite; and, to my mind (although I have not made a special study of the gneiss- granite), the matter is susceptible of one interpretation only, as follows :—the grey facies of the granitic gneiss results from impreg- nation, metamorphism, and eventual assimilation of sedimentary country-rock by the ascending magma, while the red is the residual portion of the same magma. Probably neither part reproduces the initial composition of the magma exactly. The red bands are composed essentially of quartz and felspar, with a very small proportion of biotite. Cleavage-pieces of the red felspar show it to be a true orthoclase ; the red coloration is Q 2 200 THE PSEUDOTACHYLYTE OF PARIJS. [vol. lxxii- unevenly distributed, and is especially intense in the neighbourhood of cracks. The larger crystals often pass externally into micropeg- matite. The grey bands contain much oligoclase with practically straight extinction, also a variable but generally large proportion of biotite which often exceeds 50 per cent.. I have not observed any hornblende in the few slices that I have cut; but at the weir, 2 miles above the township, bands of a bright-green amphibolite oceur in the granite. — The special interest of the district attaches, not to the granite- gneiss, but to a remarkable system of veins of apparent tachy lyte which intersect the granitic rocks everywhere throughout: the area north of Vredefort. “Of the southern “portion ot the area, where exposures of the granite seem to be much less numerous, | am unable to speak. ‘The best exposures are seen in the bed and banks of the Vaal, where the scour of the running water cleans and smooths the rock-surfaces ; the veins then show up Jet-black and exhibit a highly polished surface, thus affording a strong contrast to the rougher grey surface of the granite. Hlsewhere they weather grey or greenish, and are then difficult to distinguish from the granite. On each of my visits I took a large number of photo- graphs of these veins, but many of the exposures were so difficult: to ‘catch’ (on account of their awkward situations, irregular sur- faces, slight colour-contrasts, and so on) that I found it desirable to supplement my photographs by scale-drawings, some of which are reproduced herewith. The highly iregular, branching intrusion of Pl. XVI was ex- posed in a deep cutting made for one of the piers of the new bridge; the rock-face shown is about 7 feet high, which will give a measure of the dimensions of the intrusion. This cutting has since been filled up. The photograph reproduced in PI. XV td, fig. 1 was obtained in another excavation near the first; it shows a vein turning from a vertical to a nearly horizontal plane and thinning out. The width of the vein is 6 inches at the centre of the photograph. The detail of the blind end of this vein is shown in text-fig. 8 (p. 203). Pl. XVII, fig. 2 shows a large block which has been thrown out by blasting operations at the weir, above the- town ; the entire block is only a portion of a wide dyke of pseudo- tachyly te. The floaters of granite which are embedded in the dark base are sO numerous and so perfectly rounded that the rock resembles in appearance a sedimentary conglomerate. This is a most remarkable specimen, and, despite its weight, which must be about a ton, one would like to see it removed entire to a museum. The characteristic features of the intrusions, as seen in the field, are summarized in the following paragraphs :-— The veins are utterly irregular in form, direction, and thickness (text- ficures 6, 7, 10-13; Pl. XVI). They have every inclination from vertical (Pl. XVI) to horizontal (Pl. XVII, fig. 1), and they strike towards all points of the compass. They change their direction again and again, and often follow sinuous. Fig. 1.—Blind veins, above the boathouse. riz inches [The one on the right can be traced for about 7 feet; it runs perfectly straight, and is 1 inch wide. Within 5 inches it thins out and terminates. ]’ Fig. 2.— Branching, and a blind vein, seen near the boathouse. Branching and blind veins, near the boathouse. Fig. 3. [The broken lines indicate the foliation of the granite. | Fig. 4.—A thin vein (% to $ inch) of pseudotachylyte following the margin of a pale band in granite-gneiss, in part breaking across it and sending venules into it ; 200 yards above the boathouse. Fig. 5.-—A pegmatite vein faulted 2 inches by the pseudotachylyte. [The broken lines indicate the:foliation of,the granite. | Fig. 7.—Pseudotachylyte a quarter of a mile below the bridge. Fig. 8.—The vein shown in Pl. XVII, feathering out at its end. Cutting below the bridge. Fig. 9.—A wavy band cutting across the foliation of the gneiss, at the end of the blind intrusion illustrated in Pl. XVII. Vaal River Bridge. below the ein Seen r Fig. 10.— .—Veins seen 200 yards above the boathouse. 2 Figs. 11 & 1 part 3] THE PSEUDOTACHYLYTE OF PARIJS. 205 Fig. 18.—Pseudotachylyte 200 yards above the boathouse. - @ 2 4 ‘Gute Inches courses (text-figures 2, 3, 9-13) and their direction is entirely independent of the foliation of the granite (text-figures 3, 5, 6, 8, 10). They thicken and thin repeatedly and rapidly (text-figures 6, 7, 9, 10-12), give off branches at high and low angles (text-figures 2-4, 7, 8, 10-12), and often anastomose in the most complicated way (text-figures 10-12). In innumerable cases they thin out and terminate blindly (text-figures 1-38, 7, 8, & Pl. XVII, fig. 1). The contact with the surrounding granite, as seen with the naked eye, is perfectly abrupt. The granite is never sheared parallel to the course of the veins, its texture remains unchanged, and the felspars continue to show large bright cleavage-faces and straight twinning lamelle, right up to the surface of contact (Pl. XIX. fig. 1). The junction-line is often nearly straight for distances of many feet, especially in the case of those veins which are not more than a few inches thick (text-figures 1, 2, 4, 5, & Pl. XVII, fig. 1) ; in other cases it is strongly serrated (Pl. XVI). The veins vary in width from a fraction of an inch up to 2 or 3 feet; but in the thicker veins there are always numerous inclusions or floaters of granite which occupy a large proportion of the stated width, and the black base is often reduced to the role of a mere cement for the floaters (Pl. XVI, & Pl. XVII, fig. 2). The floaters range from boulders a couple of feet in diameter down to minute grains, and they are far more frequently rounded or turbinate than angular (Pl. XVII). In an average sample about 20 or 25 per cent., but in some cases (Pl. XVII, fig. 2) as much as 80 per cent., of the contents of a vein consists of boulders and fragments visible to the naked eye. . : The black base in which the granite floaters are embedded is a compact black rock, like a microcrystalline basalt or tachylyte. capable of taking on a high polish. Apart from grains of quartz and felspar derived from the granite, 206 PROF. 8. J. SHAND ON (vol. lxxu, no constituent of the black rock can be recognized even with a lens, nor is it possible to say whether the rock is crystalline or vitreous. It is in most cases quite impossible to determine whether the veins act as faults. Sometimes a movement of one side of the vein, amounting to not more than 2 inches, can be proved (text-fig. 5, p. 202), but no proof of great displacements was found. Some of the boulders illustrated in Pl. XVI have- been shifted only an inch or two from the walls which furnished them. IIL. Mtcroscoric CHARACTERS OF THE PSEUDOTACHYLYTE. Sections of most of these rocks must be cut very thin, in order to- secure even moderate transparency. The photomicrograms shown were taken by electric light, with exposures up to 2 minutes, yet the base of the rock appears perfectly black, and only the inclusions exhibit detail. The opacity of the base is due to a multitude of minute black specks (of magnetite), which not only stop a great deal of light, but also make it extremely difficult to. determine the nature and properties of the transparent components. of the ground-mass. Examination of a large number of slides, all made by myself, has shown that three types of ground-mass are represented, as described below. The inclusions are alike in all cases, and will be described afterwards. Type 1.—This is the most opaque type of all, and is found generally in the thinner veins, those less than a couple of inches wide. Clouds of magnetite-grains of irregular shape obscure all detail. Behind these grains one sees a structureless background, which is nct perfectly isotropic, but gives dull-grey tints between crossed nicols. The mass is not homogeneous, since- different parts do not extinguish at the same time, but boundaries between the different portions cannot be made out; the appearance is not in the least suggestive of a powder. Some specimens are nearly isotropic, of very dark colour, and show a concentration of the black colouring-matter at many centres. A slight streakiness of the colouring-matter, suggestive of flow- structure, is rarely seen. Fragments of these black rocks, free from visible- grains of quartz and felspar, were picked out and placed in bromoform (specific gravity = 2°8); they all floated. Type 2.—In this type a certain amount of crystallization has taken place. The magnetite-grains are in part tiny octahedra, and are accompanied by swarms of grains and scales which are pleochroic from yellowish green to pale grass-green (Pl. XIX, fig. 3). These are so minute and so abundant that in the thicker parts of the slides, as also in the specimens of the finest grain, one has merely an impression of greenness over all. Sometimes the prisms form definite collars or mantles around the xenocrysts. The length of the largest prisms does not exceed 0:08 millimetre, and the average is much less ; the breadth is from 0:02 mm. downwards. The extinction of the prisms les between 10° and 20°; they have high refraction and low birefringence (first- order colours), and they show a perfect cleavage parallel to their length. I am not able to distinguish cleavage in cross-sections, but of all the common rock-minerals the characters just enumerated clearly indicate a hornblende. Tiny scales of brownish-green biotite are also present in less abundance. Through the clouds of magnetite, hornblende, and biotite-grains, one sees again a feebly polarizing background which is either indeterminable, as in type 1, or else passes over towards type 3. Type 3.—This type of ground-mass is found in some of the widest veins, but also in some thinner ones, which are not externally different from the- part 3] THE PSEUDOTACHYLYTE OF PARIS. 207 others. It does not seem to constitute entire veins, but to occur in patches within rocks which also show type 2. Considering the invariable blackness. of the rocks, it is surprising to find that the section often shows very little of dark or opaque minerals, but consists either of a honeycomb of polygonal spherulites of a dark-brown colour, showing a perfect cross between crossed nicols (Pl. XIX, fig. 4), or of a felt of felspar-microlites with a subordinate amount of magnetite-dust and a few green scales in the interstices (Pl. XIX, fig. 5). The microlites rise to 0°05 mm. in length and 0:01 in width. They are all twinned, and, although their small size makes the determination diffi- cult, it seems that both Carlsbad and albite twins are present. The extinction. is always straight or practically so, and it is probable that both orthoclase and oligoclase are represented. The arrangement of the microlites is generally radial, not parallel, and it is thought that the stellate groups of microlites. have developed out of spherulites. The granite-walls and inclusions.—I have cut various: sections across the junction between granite and pseudotachylyte, and I find that in general the granite shows no change which can. be traced farther than 1 or 2 mm. from the edge. In some cases. this Junction is absolutely sharp, and the eranite betrays no change whatever. Such a junction is shown in Pl. XIX, fig. 1, where a single plagioclase crystal is seen to be cut sharply across, as if by a knife; a few tiny cracks are observed at right angles to the fractured surface, and along these there has been a very slight. faulting of parts of the crystal; but, with this exception, the twin- ning lamelle are perfectly straight, showing that there has been no. shearing of the granite parallel to the vein. This is not an isolated instance—I have observed the same feature in several slides The general conclusion as to the absence of shearing in the granite holds rigidly for every case that I have studied in the field or in the laboratory. In one case a crystal of titanite was found, half of which is embedded in the granite-wall, while the other half projects well into the tachylyte-vein; here is proof positive of the absence of shearing. In other cases the quartz- and felspar-crystals of the granite, within a space of 1 or 2 millimetres from the edge, are consider- ably cracked, show a very irregular extinction, and are invaded by tongues of pseudotachylyte. Fragments of quartz and felspar actually enclosed in the vein-rock are much cracked, the quartz- grains in particular showing a beautifully minute mosaic structure between crossed nicols. This granulation is unaccompanied by schistosity, and is such as might ‘have been caused by rapid heating of the crystals. Pl. XVIIL, fig. 2 shows granitic fragments floating in tachylyte just at the margin of a vein. The shavlor y grains are felspar, the clear white grains are quartz. All the Srains show fretted, embayed outlines, and the suggestion of active corrosion by the black base is very strong. Actual proof of melting of felspar 1s contained in Pl. XVIII, fic. IL celal, NID.G sie siuielt are taken from the margin of a large vein. In these photographs the quartz- ‘grains stand out in clear relief from a matrix of half-melted felspar which shows viscous flow-structure. Sometimes the included felspars are red, and present a compact, 208 PROF. 8. J. SHAND ON (vol. Ixxn, felsitic appearance. Slices cut through these red felspars. show that the whole substance is minutely granulated, and that in part the felspar has been fused and has afterwards erystallized in the form of perfect spherulites. It is remarkable that only the quartz and felspars of the granite appear among the inclusions; in no single instance is the ete preserved. In view of the evidence of high temperature shown by the melted felspar, this becomes intelligible: biotite is decomposed by moderate heat, and its decomposition has furnished the abundant magnetite of the base of the rock. Although felspar is more abundant than quartz in the granite, yet among the inclusions quartz is more noticeable than felspar. In one case an inclusion of amphibolite was discovered in one of the veins. The bright-green hornblende shows no granulation, and is quite une changed save at the very edge of the inclusion, where it loses its well developed cleavage and takes on a fibrous appearance, although keeping its pleochroism and its uniform extinction. Round about the inclu- sion the pseudotachylyte is coloured green, hence solution is again indicated. The only other xenocrysts that I have observed are a few unchanged erystals of zircon, sphene, and magnetite. The pseudo- tachylyte is always perfectly fresh, and no hydrous secondary minerals have been observed in any case. IY. Twe Case FoR THE IGNEOUS ORIGIN OF THE VEINS. The evidence which suggests an igneous origin for these veins, and led to their beng regarded as tachylytes in the first instance, may now be summarized :— (1) The mode of occurrence in the field, as described above—which is quite consistent with an igneous intrusive origin, but difficult to reconcile with any other known manner of formation. (2) The abrupt contacts between vein-rock and granite, and the absence of shearing in the granite. (8) The common occurrence of blind intrusions. (4) The presence of rounded boulders and fragments in the veins, as also the microscopic evidence of corrosion and eyen fusion of these and of selective destruction of the constituents in a definite order. (5) The presence of spherulitic and microlitic structures that find their nearest analogues in vitreous dyke-rocks and lavas in which crystal- lization has taken place subsequent to consolidation as glass. (6) The sharp distinction which exists between the inclusions and the matrix. (7) Although unquestionable igneous intrusions are not often seen in the granite, yet there is at least one basic dyke of the largest size in the very heart of the area, thus demonstrating the presence of basic magma in depth. This is a dyke of granophyric quartz-dolerite, nearly 600 yards wide, which is exposed in a shallow valley at the weir, above Parijs. Mr. J. W. Penny has described basic intrusions in the Lower Witwatersrand Beds which overlie the granite. part 3] THE PSEUDOTACHYLYTE OF PARIJS. 209 CoMPARISON OF THE PSEUDOTACHYLYTE witH ‘ TRAP- SHorren GNEISS’ and ‘ Prinry Crusy-Rock.’ The term trap-shotten (that is, trap-injected) gneiss was. employed by W. King & R. B. Foote for parts of the gneiss of Salem (Madras), which they describe as being ‘ very largely im- pregnated or shot with strings of dark green or lms ‘black compact trap.’ The supposed trap-velns of this area were sub- sequently studied by Sir Thomas Holland, who found them to consist of an indurated black dust, through which fragments of quartz and felspar are disseminated. He was able to imitate the material by heating powdered gneiss to white heat in a furnace, and he concluded that the production of the veins was due to the brecciation of the gneiss along certain lines, but not to injection of trap; the black malo and indurated nature of the material he ascribed to the action of heat produced during the violent brecciation of the rock. The manner in which these dark veins occur in the field, as. deseribed and illustrated by the above-mentioned writers, recalls many of the features of the Parijs occurrence. nportant points ot difference, however, are that the so-called ‘ trap-shotten ’ bands occur in roughly parallel belts which coincide with lines of dislocation, and that they are often associated with true igneous in- trusions of a basic nature. With regard to microscopic characters, the difference is fundamental; in no single instance that I have observed can the matrix of Ae Parijs rocks be described as a ‘black dust’ or powder. In Scotland rocks of somewhat similar characters to the above have been deseribed as ‘Hinty crush-rocks’ and mylonites. Many references to such rocks occur in the Geological Survey Memoir on the North-West Highlands, and attention may be drawn espe- cially to the descriptions on pp. 124, 221, and 245, It is again significant that the veins are Sears only in highly disleeatsadl regions —especially along, or in the neighbourhood of, ‘well-marked fault- planes and zones oe crushing. In most cases the flinty material is associated with bands ae a more clearly granulitie or mylonitic character, and it has been suggested that by the intensity of the crushing ou fice heat may hee been generated to fuse small portions ce the rock. Some of the specimens in the Geolo- gical Survey collection show the occurrence of black flinty material as streaks within ordinary mylonite, apparently confirming the above view; a good example of this is Slide 12933. Dr. deer Was described crush-lines in granite near Broadford (Skye), and notes that the most marked eitacias are restricted to the vicinity of faulted boundaries. The flinty erush- rocks of the Cheviot Hills, discovered by the late Dr. C. sO Clough, also oceur along crush- lines in granite. Sections of these rocks, which Dr. ‘Clough kindly permitted me to examine, show the clearest. evidence of shearing; but they range from obvious mylonites, 210 PROF. 8, J. SHAND ON [ vol. Ixxi, consisting of angular fragments of quartz and felspar, to very dense, structureless, black bands the origin of which might have remained obscure if they had not been associated with coarser granulitic material. Dr. Clough informed me that these veins were at one time considered to be of igneous origin; in my view, however, their true nature is unquestionably clearer than that of the Parijs rocks. The specimens that exhibited the nearest approach to the characters of the pseudotachylyte of Parijs came from Cunyan Crag (Cheviots). Dr. Clough also showed me a flinty crush- rock from South Ben Lee, North Uist (Hebrides), which, in regard to the extreme opacity of its base, almost defying microscopic resolution, resembled the densest examples of my Paris type 1. Dr. Clough, Mr. Maufe, & Mr. Bailey diseuss flinty crush-rocks at ‘some length in their paper on the cauldron subsidence of Glen Coe, and give a convenient summary of some of the literature mentioned above. They show that at Glen Coe dark, apparently tachylytie, rocks have been produced by the shearing of various types of rock, ‘quite independently of igneous action. Of especial interest is their observation that in some cases ‘the flinty crush-rock has been injected as a fluid away from its source of origin.’ By the courtesy of Dr. Flett and the late Dr. Clough and Mr. Bailey, I was enabled to study a large number of hand-specimens and sections from this area, as well as from other parts of Scotland. I found that the majority of them were unmistakable mylonites, or eyed gneisses on a microscopic scale; but there seemed to be a re- presentative of every stage between these anda type with almost unresolvable dense base, which may show flow-structure or may lack -eyery trace of orientation. Only the most extreme examples of the latter type are comparable to the Parijs rocks, and none of them ‘showed anything similar to the crystallization of felspar and horn- blende seen at Parijs. The points in which these rocks begin to resemble those of Parijs are the following :— (1) The base of the rock is sometimes structureless and opaque, and full of tiny black grains, which are in part euhedral crystals of magnetite. (2) Larger fragments are sometimes rounded, and have what look like reaction-rims, or their margins may appear to have been corroded. Internally they may exhibit a mosaic structure. (3) In rare cases there are beginnings of crystallization in parts of the base, and in one instance (Meall Riabhach, Ross-shire) typical trichites are developed. Out of all the material that I examined I was able to select only four specimens of which I could fairly say that they showed a close approach to my Parijs rocks of the first type.! Typical flinty crush-rocks are developed on a very large scale in the province of Olavarria (Argentine Republic), where “they have 1 The sections in the Geological Survey collection bear the numbers 12332, 13402, 13403, 12934, 4281. part 3 | THE PSEUDOTACHYLYTE OF PARIJS. 211 been studied by H. Backlund. In his memoir on the subject, Backlund traces the mechanical alteration of a coarse porphyritic granite into a black horny rock with relies of felspar-phenocrysts. Optical deformation of the brecciated fragments is said to be visible in every preparation, and as there is no obvious recrystallization it is concluded that the breccia has not been submitted to any very high temperature. Backlund compares these rocks with Scottish flinty crush-rocks, which they appear to resemble closely. Four chemical analyses are tabulated, two of different facies of the granite-gneiss and two of mylonites, The analyses of mylonite do not differ more from the analyses of granite than do the latter one from the other, and, in particular, the analysis of a ‘dense black mylonite with red spots of microcline’ corresponds almost exactly to the mean of the two analyses of granite-gneiss. The only other part of the world trom which I have been able to obtain useful information regarding crush-rocks and pseudotachy- lytes is Little Namaqualand, the north-western corner of the Cape Province. I am indebted to Dr, A. W. Rogers for the following note on the subject :— ‘The trap-shotten rocks in Van Rhyn’s Dorp and Namaqualand are massive gneiss and occasionally the mica-diorites of the copper-bearing intrusions. The cracks are rarely half an inch wide, usually less than a tenth of an inch. They—or, rather, the instances observed—are in groups several feet long, and the individual cracks give off branches and join up with their neighbours. Though they have no constant direction, the cracks in any one group are parallel in a rough way, and at Jubilee Kop and on Mining Lease 81 they cross the foliation of the gneiss approximately at right angles, ‘The strings are black.’ In sections of one of these black veins from Flamink Vlakte (Namaqualand) I was able to determine the following characters. Under a low magnification the appearance seen is that of a powder consisting of sharp angular grains of quartz and felspar of all sizes, with a sort of cement of green scales. Hven the highest mag- nification fails to reveal anything of the nature of a glassy, spherulitic, or microlitic base. The quartz and felspar fragments are mostly free from cracks, and give a perfectly uniform extinction, while the plagioclases show normal twinning lamelle. The green scaly substance lying between the colourless grains is made up of shapeless scales of biotite, with perhaps a few scraps of hornblende of the same colour. This biotite seems to be mainly detrital, just as the quartz and felspar are, but some of it may conceivably have erystallized in place. There is absolutely no suggestion of corrosion, melting, or recrystallization of the quartz or felspar, and nothing whatever to indicate that the temperature was high. Magnetite, which renders the Parijs rocks so dark and opaque, is nearly absent, A few grains of zircon are visible, some of them being enclosed in the larger biotite-scales. As an illustration of this mylonitic type of crush-rock, which appears to be the standard type in all the regions above described, 212 PROF. 8. J. SHAND ON [ vol. ]xxii, except Parijs, I include a photomicrograph of the rock of Flamink Vilakte in Pl. XIX, fig. 6. The essential difference between pseudo- tachylyte and fhnty ‘erush-rock is impressively shown when this photograph i is compared with Pl. XVIII, fig. 2 The facts that have been established so far may conveniently be tabulated in the following form :— (a) Black rocks are composed of crushed materia! without recrystal- lization or any evidence of elevated temperature: Argentine ; Namaqualand ; most Scottish and Indian localities. (b) Black rocks are composed of crushed material, but with evidence of high temperature approaching the melting-points of some of the constituents, and with beginnings of crystallization: some Scottish localities, especially Glen Coe, Meall Riabhach, and North Uist; some Indian localities (?). (c) Black rocks hold inclusions of fragmental material, but lack proof of origin by crushing; the temperature exceeded the melting-point of felspar, and spherulitic and microlitic crystallization teok place : Parijs. Regarding the evidence from a purely qualitative standpoint, it would seem that we have a complete series of rocks connecting up the pseudotachylyte of Parijs with ordinary mylonite, the various links in the series being as follows :— mylonite —> fritted mylonite or flinty crush-rock —> fused mylonite or pseudotachylyte (type 1) — recrystallized pseudotachylyte (types 2 & 3). Arguing on these lines, it is possible to maintain the view that pseudotachyly te is simply an extreme form of flinty crush-rock, the production of which involved a greater generation of heat than usual. Against this view of the origin of pseudotachylyte there are some very weighty arguments. The supposed series of rocks connecting pseudotachylyte with mylonite, although qualitatively complete, is very umperfect when examined quantitatively. In all the regions that have been mentioned here, except Parijs, the first member of the supposed series 1s abundant, the second member much scarcer, the third member is limited to a few minute or even microscopic occurrences, while the fourth member does not oceur. At Parijs, on the other hand, the first and second members are entirely absent (at least a close search has failed to reveal them), while the third and fourth members occur in great abundance throughout a large area. These considerations cast the gravest doubt. upon the actual continuity of the supposed series. In fact, the only conclusion that the study of flinty crush-rocks entitles us to draw is that products similar to pseudotachylyte can be, and in rare cases have been, produced by crushing and heating associated with movement along planes of dislocation. Further objections to the c¢ comppamisen of pseudotachylyte with flinty crush-rock are as follows : part 3] THE PSEUDOTACHYLYTE OF PARIS. 213 (1) The absence of shearing and cataclastic phenomena in the granite. This can only be accounted for by supposing the fractures to have been produced by sudden snapping of the granite in response to a sudden stress; but how, unless there was long-continued friction, could the quantity of fine dust necessary to fill the fissures be supplied P (2) The common occurrence of blind veins seems positively to exclude the possibility of long-continued friction ; the blind veins do not differ in texture or composition from the others. VI. Tur CoemicaL CoMposirion OF THE PSEUDOTACHYLYTE. (See also Appendix, p. 217.) ‘ When I first examined these rocks, I had no means of making a chemical analysis of the material. Subsequently, when the question of the origin of the pseudotachylyte was reopened, my assistant, Mr. R. A. Page, kindly undertook to perform the necessary analyses, and the figures given below are due to him. In justice to Mr. Page, I must state that the work was performed under very unfavourable conditions, with a deficient supply of platinum vessels, with acetylene burners and a spirit blast-lamp for ignition, and without reagents of special purity ; consequently, it was not expected that a high degree of accuracy could be attained. The summation of the analyses is low; but, as we are concerned rather with the general chemical character of the rock than with the very precise determination of individual constituents, the result has proved quite serviceable, and I desire to record my great indebtedness to Mr. Page for his careful labour. The figures are given in the first two columns below; those in the third and fourth columns refer to Backlund’s analyses of black mylonite and grey mylonite respectively, the minor constituents being omitted :— ter ie es Te II. III SHOW 60k Cea eee 62°03 | 62°66 | 65°75 | 68-24 TIO Ls Ath 4 Ge a My GL T EGE 066 0-01 PROM ee es & 1657 | 1626 | 15:56 | 16-08 THO Oy Aye ioe eae 2:51 2°51 0:64 | 0-81 He Ose eee od 2:90 2-89 3:30 | 2-20 | NIC) Ota lee 039 | nd. 177 | 0:40 | CaO per. 3°39 3:99 | 3°75 | 3:42 en Onee ess. 6°37 588 | 811 | 3:34 REO ay es 3-41 275 | 411 | 3:99 EO 2 eae 0:16 \ Oa8 { 008 | 0:08 TEGCOE alt eete eaten 0°72 | 0-69 | 0:83 COM eee ee es: O;O0 i pte 0-74 | 0:59 Motal sees 98:45 | 97:90 | TiO,, ZrO,, MnO, and P.O, were present in small quantities, but were not determined by Mr. Page. Q. J.G.8. No. 287. R 214 PROF. S. J. SHAND ON [vol. lxxii, The rock chosen for analysis shows a combination of the cha- racters of types 2 & 3 as above described—that is, it contains both felspar-laths and hornblende-prisms, with magnetite. The average diameter of grain is 0-01 millimetre, and the usual quartz- and felspar-inclusions are present In comparatively small amount in the hand- specimen, although well-rounded inclusions are very numerous under the microscope. An apparent discrepancy exists between the abundance of hornblende in the ground-mass of the rock and the rather low figure for magnesia in the analysis; but it is possible that the place of magnesia is largely taken by ferrous oxide, and that the amphibole is yich in the or iinerite molecule. The norm of this rock, as calculated from the first analysis, is as follows : yma 9°72, orthoclase 37°81, albite 28°82. anorthite 11°12, diopside 5:02, hypersthene 1°49, magnetite 3°71. The rock, therefore, falls into persalane, subrang pulaskose (1, 5, 2, 3). Conclusions to be drawn from the analyses.—It was recognized from the beginning that a chemical analysis would probably prove inconclusive as regards the origin of the veins, because (a) it is not possible to separate the ground-mass of the rock from the minute granitic inclusions which occur all through it, and (6) whatever theory of the origin of the ves one may favour, it is certain that some part of the granite—especially the biotite, but also some of the felspar—has already been in- corporated in the ground-mass. The analysis, therefore, gives the composition of the ground-mass plus that of innumerable quartz- and felspar-inclusions. No analysis has been made of the granite-gneiss in which the veins occur: this is a composite banded gneiss in which different bands show a very wide range of composition, and no single analysis could give a fair idea of the average composition. From the mineralogical examination of the various facies of the gneiss, however, one may describe the rock-complex as a whole as having the composition of a granodiorite. The real utility of the above analyses, therefore, lies in showing that the composition of the pseudotachylyte is such as might result from the commingling of fragments derived from the different leucocratic and melanocratie bands of the gneiss. My original assumption, that the dykes consist of basic igneous material (approximately of the composition of basalt) which has acidified itself by solution of granite, is now seen to be scarcely tenable. The figures for alkalies and magnesia, in particular, are such as no mixture of granite and basalt could give. Again, in order to account for the proportion of silica alone, it would be necessary to assume that our hypothetical basaltic magma took up in solution and suspension from one and a half to three times its own weight of granite, since, according to Dr. Daly’s figures, the average silica of basalts is 49 per cent. and that of granodiorites and granites from 65 to 69 per cent. The actual proportion of suspended granite-fragments (that is, inclusions) in the analysed part 3] THE PSEUDOTACHYLYTE OF PARLJS. 215 rock was determined by. geometrical measurement of four thin sections; the results were 25-3, 26:4, 28:2, and 43°5 per cent. respectively, giving an average of about 30 per cent. Hence, in order. to account for the observed composition of the pseudotachy- lyte, we should be driven to the assumption that the supposed basaltic ground-mass had actually dissolved and completely assimi- lated about its own weight of granite. But a simple calculation based on the known values of the specific heat and latent heat of fusion of the common silicate-minerals shows that a magma, to dissolve its own weight of foreign rock, would need to have an initial temperature enormously i in excess of any temperature known to be realized in the higher zones of the lithosphere. If, then, the participation of an igneous magma is to be assumed at all it must be a magma considerably richer in silica and alkalies than basalt. But here we are faced by the fact that the only body of magma that is known to underlie the granite-gneiss in this region is the body which gave rise to the enormous dolerite-dyke at the weir above Parijs, and to the various ‘diabases’ in the Lower Witwatersrand rocks. The analysis, therefore, deprives the igneous theory in its original form of an important buttress. If we try to modify that theory by assuming that the invading magma was of andesitic or dacitic composition, we merely beg the question. VII. THe Cass aGatnst AN IGNEOUS ORIGIN. So long as no other than an igneous origin suggested itself, certain peculiarities of structure and composition exhibited by the pseudotachylyte could be overlooked ; but, with an alternative solution in the field, these points of singularity must be regarded _as evidence against the igneous theory. The main points militating against the acceptance of the igneous theory are as follows :— (4) The total absence of macrocrystalline equivalents of the pseudo- tachylyte. (2) The total absence of differentiation between the central and the marginal portions of the veins. (3) The general absence of flow-structure. (4) The extreme opacity of the most glassy-looking rocks, and the abundance of powdery magnetite in them. (5) The absence of perfect isotropism, even in the most structureless types. : (6) The low water-content of the rocks. Mr. R. A. Page determined for me the total loss on ignition in the case of one of the most nearly glassy rocks (type 1); it amounted only to 0°83 per cent., which is much lower than the average figure for pitchstones. (7) The melting (as distinct from dissolution) of felspar contacts. This is a phenomenon for which I can recollect no parallel among igneous rocks ; it seems to demand a temperature higher than one can con- cede to a thin tongue of magma far from its source. (8) The analogy between these rocks and the rare semi-vitrified crush- rocks of Scotland, and through these with other Scottish occur- rences of flinty crush-rocks and with the mylonites of Argentina, India, and Namaqualand. R2 216 PROF. §. J. SHAND ON [vols Tacx, VILL. Conciusion. There are two explanations before us of the origin of these pseudotachylytes: one is, that they represent end-products of the same process as that by which flinty crush-rocks have been pro- duced; the other, that they are true igneous intrusions. I have stated the evidence for and against both hypotheses without prejudice. I must now add that the result of my study of flinty crush-rocks has been to make me chary of reasserting the normal igneous. origin of the pseudotachylyte. 'The only point as to which I am really satisfied is the intrusive relation of the material to its walls. Before rejecting the igneous theory, however, on account of the evidence which has already been cited against 1t, we may pause to consider whether a melt of granite, produced within the granite itself by mechanically-developed heat, would not dtffer in its properties from a normal granitic magma owing to the absence of the usual volatile constituents of a magma. Such a melt, being in all probability extremely viscous, might be unable to undergo. differentiation or to crystallize as a magma would; hence the absence of differentiation and of phanerocrystalline products, although hardly compatible with consolidation from a magma, is, perhaps, not incompatible with consolidation from a rock-melt. That the pseudotachylyte is the product of such a rock-melt, as distinct from a magma, is the hypothesis that best explains the facts. It depends, however, on the assumption that the heat necessary to fuse the granite on the large scale which the intrusions indicate is capable of ‘being produced by the sudden rupture of the granite without long- continued friction or shearing. The validity of this assumption, as well as the properties of such a rock-melt under high pressure, can only be guessed at. Granted the necessary postulates, then this hy pothesis adequately explains the phenomena. Should the above assumption not prove acceptable, there is still one possibility left. Prof. Stanislas Meunier and others still believe in the causation of earthquakes by deep-seated explosions. ' The conditions at Parijs are not such as Prof. Meunier postulates, but there is abundant evidence that the sub-crust of these parts of South Africa has in the past contained highly explosive material : witness the hundreds of tuff-filled pipes which have been drilled through the crust in the Orange Free State, the Transvaal, and the adjacent parts of the Cape Province. The extrusion of the great dolerite-dyke at Parijs itself demands an expulsive force of enormous power. ‘The form of the pseudotachylyte veins indicates. that the granite was shattered by a sudden gigantic impulse or series of impulses. If this impulse were of the nature of an explosion in the sub-crust, it would have as a necessary consequence ‘the outrush of incandescent gases through all the fissures of the granite. In these circumstances fusion of the walls of the fissures might well ensue. Sir Andrew Noble’s experiment, in which cordite was exploded in steel cylinders and the gases allowed part 3] THE PSEUDOTACHYLYTE OF PARIS. 217 to escape through a small hole in a granite stopper, will .at once recur to the memory. Here is not only a possible source of the vein-material, but an explanation of the high temperature to the action of which that material testifies. I hardly dare offer this as a serious hypothesis, since it takes us away from known causes altogether; but, when one is dealing with an extraordinary phenomenon, no possibility is too extraordinary to be worthy of consideration. In brief, the material of the pseudotachylyte seems to be melted *ranite—this is indicated by the chemical characters, and is consistent with the microscopical evidence ; it behaves towards the granite like an igneous intrusion—this is established by the field-observations. The source of the heat and the mechanics of the intrusive process remain obscure. IX. LiverRAatvuRe. F. H. Hatcu.— A Geological Map of the Transvaal’ in Hatch & Corstorphine’s ‘Geology of South Africa’ 2nd ed. 1909. G. A. F. MonpncRAArr.— Remarks on the Vredefort Mountain-Land’ Trans. Geol. Soc. 8. Africa, vol. vi (1903) pp. 20-26. F. W. Penny.— The Relationship of the Vredefort Granite to the Witwaters- rand System’ Q. J..G.S. vol. lxx (1914) pp. 328-34. W. Kine & R. B. Footr.—‘ On the Geological Structure of the Districts of Trichinopoly, Salem, &c.’ Mem. Geol. Surv. India, vol. iv (1864) pp. 223-377. T H. Ho~tutanp.— The Charnockite Series’ Mem. Geol. Surv. India, vol. xxviii (1900). B. N. Peacu & others.— The Geological Structure of the North-West High- lands of Scotland’ Mem. Geol. Surv. 1907. A. HarKer.—‘ The Tertiary Igneous Rocks of Skye’ Mem. Geol. Surv. 1904, p. 167. C. T. CLoueu, H. B. Maurs, & H. B. Batney.—‘ The Cauldron Subsidence of Glen Coe’ Q. J. G.S. vol. Ixv (1909) pp. 611-76. H. Backitunp.—‘ Aleunas Observaciones sobre Rocas Notables provenientes de Olavarrid’ Direccién General de Minas, Boletin No. 2, Buenos Aires, 1913. APPENDIX [September 1916]. i Some fresh Analyses of the Granite and Pseudotachylyte -of Parijs, made by Dr. H. F. Harwood. Through the kind mediation of Dr. Arthur Holmes, Dr. H. F. Harwood, of the Imperial College of Science & Technology, has lately been induced to undertake the analysis of the granite and pseudotachylyte, in order to settle the question of the origin of the latter. For this purpose I supplied him with :— (1) the residue of the material already analysed by Mr. R. A. Page and described by me in the foregoing pages, and (2) a large block from the locality shown in Pl. XVII, fig. 1, consist- ing of part of the tachylyte vein (2a) and part of the granite wall (2 b). 218 PROF. 8. J. SHAND ON [vol. lexii, The analytical results are tabulated below under these numbers :— (1) (2a) 2) So O ede shaa nou naduey ase | 64:87 66°95 Ci2amen AT OF Mv seoledunomneb esos ptace 15°62 15:06 Uee78) | FOO eee ee 1:90 1:58 1:04 | DOW s ane svat erina tame 2°50 2°18 2°33 Mig Oi teeta naan sees | aleae/ 1°38 ee | CaO ees he are eae Wr RUBS OS eh 2:93 @| EN IETE Oo aera acon area ain 4:29 4°32 449 | TRON tic Rae licens wera Zio Aten al 2°85 2°21 —} HeOeey ee ke cmne ee |. OB 0°51 068 | 2 OO) iar acre acetic OOS |. Ole 0-11 ARO VE cohenies Son iaH tase 32 ae |e 1:36 OM ie eeu ee a 0:27 | 0:07 0-11 SO eees eeaones tees maLONe)) || OTG none Diente eee ec on cues eran 0:06 0°12 0°02 IN Bak Oo alee oe ence | 0:05 0:03 0:02 SLOP irew et enmiecee se erly MOEACE Leg Ue enOMe 0:03 Ba Oe ee et reeves ne LOO). OPO DO CON ee le Ae Oe aah | trace trace trace | | Mota Soest ecerne| 99°96 100°22 99°84 ASI Olena sane warns b OOr OF05 | = | eee | 99°94 100717 The analysed specimen of the granite-gneiss is distinctly banded in hand-specimen, the individual bands of alternately hight and dark colour being 1 to 2 mm. thick and neither regular nor persistent when examined minutely, yet sufficiently so to give a distinct grain to the rock. Both white and flesh-red felspars are seen, the latter being the larger and tending to produce eyes. Both orthoclase and oligoclase are present, with biotite as the only dark silicate. Grains of apatite are rather abundant, with some ilmenite and titanite, and large irregular patches of quartz make up about a quarter of the rock. The chemical and microscopical characters determine the rock to be essentially a granodiorite; but, as I have already expressed the opinion that the whole mass is a hybrid granite-gneiss the composition of which varies greatly from place to place, there is no need to lay stress on the name. The analysed specimens of pseudotachylyte are of my type 2, passing towards 3—that is to say, they contain abundant horn- blende-needles and magnetite-grains; but in places, and especially round the larger inclusions, felspar-laths are also prominently developed. Comparing the figures for the two kinds of rock, it is evident that they are so nearly identical as positively to exclude the participation of a foreign magma in the production of the pseudo- tachylyte. The greater content of potash, magnesia, and iron in the pseudotachylyte, as compared with the granite, doubtless arises part 3] THE PSEUDOTACHYLYTE OF PARIJS. 219 from the fact that biotite is the first mineral to fuse. In preparing the powder of 2a for analysis, the larger fragments of quartz and felspar were probably rejected (at least, that was my own procedure) ; but no biotite was rejected with them, hence the enrichment of the base in the constituents of biotite. All three analyses fall, in the quantitative classification, into lassenose (I, 4, 2,4.)—a group containing many plagioclase-granites and dacites. The conclusion to which my re-examination of the rocks, coupled with Mr. Page’s analysis, had already led me is, therefore, com- pletely confirmed by Dr. Harwood’s work. The pseudotachy- lyte has originated from the granite itself through melting, caused (as I have shown) not by shearing but by shock, or, alternatively, by gas-fluxing. I wish to offer my warmest thanks to Dr. Harwood for his kindness in performing the above analyses, and thus enabling the investigation of these remarkable rocks to be completed ; and also to Dr. Holmes for so kindly interesting himself in the matter. EXPLANATION OF PLATES XVI-XIX. PLATE XVI. A large vein of pseudotachylyte exposed in a cutting made for one of the piers of the new bridge below Parijs (Orange Free State). PuATE XVII. Fig. 1, A vein coming up vertically, turning over to a nearly horizontal plane, and thinning out. Thickness of the vein = about 6 inches in the middle of the photograph. Exposed in a cutting at the bridge. (The detail of the blind end of this vein is shown in text-figure 8, p. 208.) 2. A boulder from one of the large veins, thrown out during blasting operations at the weir, above Parijs: shows rounded floaters of granite and gneiss. A foot-rule has been placed alongside, in order to give the scale. Puate XVIII. Fig. 1. Melted felspar at the margin of a vein. The clear white areas are quartz-grains. > 15 diameters. (See p. 207.) 2. Section near the margin of a vein: it shows granite fragments being detached, apparently corroded, and floated away. X 15 diameters. (See p. 207.) PLATE XIX. Fig. 1. Margin of a vein, showing a plagioclase-crystal which has been broken but not sheared. Crossed nicols. » 15 diameters. (See p. 207.) 2. Melted felspar at the margin of a vein. The clear colourless areas are quartz-grains. 15 diameters. (See p. 207.) 3. Pseudotachylyte with ground-mass of type 2, full of minute hornblende- prisms. xX 45 diameters. (See p. 206.) 220 PROF. S. J. SHAND ON (vol. lxxu, Fig. 4. Spherulitic ground-mass, type 3. 80 diameters. Crossed nicols. (See p. 206.) 5. Ground-mass of felspar-microlites, type 3. 100 diameters. Crossed nicols. (See p. 206.) : 6. Flinty erush-rock from Flamink Vlakte (Namaqualand). » 15 dia- meters. (See p. 211.) DIscUSSION. Sir JerHro TEanr considered that the paper was an important contribution to a most interesting and difficult subject. The facts had been very clearly described by the Author. He (the speaker) had examined many specimens of ‘flinty crush’ from Scotland, and had been much puzzled to account for the phenomena. The bits of quartz and felspar which often formed a large part of the rock showed marked signs of crush and strain, while the dark matrix rarely exhibited any signs of crystallization. The flinty material occurred in thin veins which sometimes anastomosed. He found it very difficult to understand how the material was produced, and how it came to be distributed in this way through the roek from which it was largely if not entirely produced. Mr. J. F. N. GREEN said that, whatever might be the case with the occurrences which had been quoted from India and other places outside South Africa, there did not appear to be any evidence for erushing in the description that had just been given. Except for the character of the country-rock, the macroscopic phenomena were of a kind familiar to students of the volcanic areas of this country. He placed on the table two specimens from the Lake District showing, on a small scale, brecciation; coraminution, blind veins, etc., closely resembling the photographs and drawings exhibited. Mr. E. Greenty drew attention to some specimens and shdes of ‘flinty crush-rock’ from the Lewisian Gneiss of Scotland that were exhibited on the table, and had been kindly lent by the Geological Survey. When he first met with these veins m the Loch Maree country, five-and-twenty years ago, he certainly supposed that they could be nothing but igneous ‘injections. Soon afterwards, however, on the other side of the lake, he found that the same kind of material, appearing first as veins in almost uninjured gneiss, increased rapidly in quantity towards a powerful zone of crush near the lake-side, until, in that zone, the gneiss was permeated by it in all directions. "The dynamic origin of the material (already advocated by Dr. Clough) was here upquestion- able. One of the slides on the table showed microlithic structure in a vein-rock, another showed crushed gneiss of the lake-side zone breaking down and passing into ‘ flinty ’ matter. Dr. A. Hones said that, having now heard the revised version of the Author’s paper, he was still “inclined to support the original suggestion that the pseudotachylyte veins were of igneous origin. The. form of the intrusion could be paralleled in Mozambique, where the speaker had seen irregular veins of granulitic granite penetrating gneiss in an equally ‘intricate fashion. If the pseudo- tachylyte were the result of crushing 7 s¢tw, then every gradation Quart. Journ. Geov. Soc. Vor. LXXII, PL. XVI. i VEIN OF PSEUDOTACHYLYTE EXPOSED IN CUTTING BELOw Paris (0.F.S.). ie Sed SA Photo. Bemrose, Collo., Derby. VEIN OF PSEUDOTACHYLYTE EXPOSED IN CUTTING BELOW Paris (0.F.S.). Quart. Journ. Geot. Soc. Vor. LXXII, PL. XVII. Fic.|. VEIN OF PSEUDOTACHYLYTE NEAR Parius (OF S)). : ; we ~ ais, i a a Se abi ~ S.J. S., Photo. Fic. 2. A BouLDER FROM ONE OF THE LARGE VEINS OF PSEUDOTACHYLYTE NEAR PARIS. qe : Fie | f | J Cg A aioe ar S.J. S.,Photo. Bemrace! Collo, Derby ee oe Iie kaye ae By ten’ At CERES vin ae (a is as ray, rears os Quart. Journ. Geov. Soc. Vor. LXXII, Pr. XVIII. S.J.S., Photo. Bemrose, Collo., Derby. PSEUDOTACHYLYTE OF Parius (0.FS,). oa Nei uy me Quart. Journ. Geo. Soc. Vor. LXXII, Pu. XIX. 2. x15 S.J.S., Photo. Bemrose, Collo., Derby. PSEUDOTACHYLYTE OF Paris (0.F.S)), Etc. Sipe . thee nv a . ne te part 3 | THE PSEUDOTACHYLYTE OF PARIJS. 221 between fused material and mylonite ought to be represented. There did not appear to be such a gradation; while, on the other hand, there was abundant evidence that the material had possessed a sufficiently high temperature to fuse the minerals of the invaded rocks. The question then arose: why had the black rock crystal- lized so badly ¥ One explanation might lie in the assumption that the intrusion was feebly supplied with volatile fluxes; and such a condition would naturally arise if some deep-seated rock that had previously crystallized and lost its fluxes, were again to be raised to its melting-point—by severe crushing, or rise of the geotherms, or any other cause. The Author stated, in the abstract of his paper, that the compo- sition of the rock was practically that of a granodiorite, and that it might correspond to an average of the granite-types in which the veins occurred. The speaker said that he could not agree with either of these statements, for the analysis showed about 10 per cent. of alkalies, with soda nearly twice asabundant as potash. The speaker suggested that, if the Author could have specimens of the pseudotachylyte and of the red and grey granites sent to England for analysis, the results could not fail to soon more light on the origin of this most puzzling of rocks. Dr. J. W. Evans thought that it was clear, in the majority of cases, that the material of pseudotachylyte had been injected into the rock in a fluid or pasty condition, as the result of pressure at right angles to the general direction of injection. It did not appear to be derived from an extraneous magma, but at least in some cases, including that described by the Author, the matrix had been in a molten state. It was doubtful whether the heat pro- duced by earth-movements was ever sufficient to cause an appreciable rise In temperature, and the speaker was inclined to attribute the melting to a regional rise in temperature which did not melt the granite because this had lost its magmatic water. 'The intrusive material may have been formed from a basic segregation, more easily fusible on account of the larger proportion of iron and alkahes. Such segregations would in many cases determine planes of weakness in the rock. The brecciation of the granitic material included in the veins might, in some cases, be attributed to the forcible injection of the latter. 222 DR. A. HOLMES ON THE TERTIARY [ vol. Ixxu, 11. Lhe Tertiary Vonicanic Rocks of the Districr of Mozameiqgun. By Arruur Houmas, D.Sc.(Lond.), D.L.C., A.R.C.S., F.G.S. (Read June 28th, 1916.) [Puatres XX & XXI.] CONTENTS. Introduction. Page IL, JEUStOIAy Oe Ig olornAO | ooo goods ooeecd cconca dco sadueonvenad on 222 PES GeneralsReaturesh esc sakes eee eee eset eet 223 Be ThevCoarstall sBeltie cee casey ae eee eran eee neat 224 JIWo- Lav olbioerEROA? — Soocco coo psesacncccooceaecc Setiarciada aaa ae igawae Sot 225 The Volcanic Area. V. The Distribution and Former Extent of the Lavas ... 228 WAL, 1etereioal Cie JOEL ONO es ocn ono 000 co0v0G200b0o000 pnb Fo CA GOOSOs Ade 229 VII. Order of Eruption, and Classification ..................... 231 Petrographic Descriptions and Analyses. Rocks of Group A. VIII. Sélvsbergite..................... EAE Ee ere PS UTADA? Nir ic 233 IDG ya Dleatish aver Mee Kolahyti) an dod can qusebonad bab odd eon soon sso unGRGon oC 235 YG AMY MONG IE OOK coeranvodcbodagacaeoascosoncocanocen eéac- 236 xO RephritiesPumi cela tn es eee oe Cece ee: 239 XO asall Gene ee ence hers ae se dilaars Nawian ciate oa ania eta rey 241 MATTE OPicriteBasaltirca hee cance ose eae eee tee ee erat ea Rocks of Group B. DY, Leaseukns) 9 (Swilley eyaGl ID AKA! ccyoc0s0e sos esc anodoucooonocendcu ver 246 NOY, Ieagralkesig IDeA MONS cocccocosssndounsbbes ceucsuccecodyacouose 250 XeVily Evornblende2Amdesiteleen spree eeeeere eee rere eee 256 KVL. Pyroxene-Andesiter titi once eae erence tae nea ee 259 Relationships and Origin of the Volcanic Rocks. XVIII. Distribution of Similar Lavas in Hastern Africa ...... 260 MOD, WereeynionalDreenenins) {550.0000 eno bocoudsocasoconbncecoboasennoc 264 NOK, Ospiestin OF Wa IES. “ssoscodoocccss anpooonngcasdscconsaveba ne 271. XXI. The Radioactivity of the Lavas, and its Bearing on ilaVeihon Oo miea tala ahve manual E au Nsanaden upon gHmaceb 4c 275 Introduction. I. History or EXPLORATION. Ture Portuguese Hast-African District of Mozambique was, until recently, one of the least known of the East-African coastlands. Livingstone explored the regions on the south-west (1865) and north (1874), but Mozambique itself he did not touch. Sadebeck (1879) was the first to give a geological description of the sedimentary rocks that fringe the coastal plain of the mainland north and south of Mozambique! Island. This work was followed up by Choffat (1903) in an important paleontological memoir on the Cretaceous formations of Conducia Bay. Exploration beyond the coastal zone was initiated by Lieut. H. E. O’Neill, who made a number of successful journeys westwards across the mainland during the years 1881-1884. In his papers (1882, 1884, 1885) he ' The name Mozambique has been used indiscriminately for (a) Portu- guese Kast Africa, (b) the district lying between the. Luli and Ligonia Rivers, (c) the Island of Mozambique, and (d) the city of Mozambique. part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 223: gives a very complete and faithful account of the natives and their customs; but, beyond the suggestion that the prevailing rocks are gneisses and granites, he makes no reference to the geology of the territory. In 1887, Mr. J. T. Last made a journey ‘from Blantyre to the Namuli Peaks, which had for long been considered to be of voleanie origin. This view he disproved (Last, 1887), describing the summit as ‘an almost perpendicular mass of white stone.’ I have since learned from Mr. Last that he referred in this expression to the grey granite and gneiss which builds up the Namuli Peaks, as it does the other mountainous masses of the country. Thus, until a few years ago, Mozambique, with the exception of part of the Cretaceous coastal fringe, was practically a terra imeognita to geologists. In 1910, a prospecting and surveying expedition was organized by the Memba Minerals Ltd., under the leadership of Mr. E. W. E. Barton. The coastal belt was investi- gated during the first season with Memba as the base of operations; but, owing “hs the hostility of the natives, little progress was. sreedle salemel Karly in 1911, a fresh party was formed with head- quarters at Mosuril, and Mr. E. J. Wayland, Mr. D. A. Wray, and myself, were appointed as geologists to the expedition. A short account of the geographical work accomplished, together with the map made during the two seasons’ work, has already been published (Reid, 1913), and in two papers introductory to the present more detailed treatment Mr. Wray and myself have described the geography of the country and the outstanding features of the geology (Holmes & Wray, 1912 & 1913; see also Wray, 1915). The present paper deals with the Tertiar y volcanic rocks of Mozambique, and will be followed by one describing the Pre- Cambrian rocks. Il. Generar FEATURES. With the exception of a coastal zone of variable width, formed by a narrow belt of Cretaceous and Tertiary sediments and flanked on the west by later Tertiary voleanic rocks, the whole territory consists of a complex of gneisses and other foliated rocks with granitic and other intrusions belonging to at least two different periods. Although we proceeded inland for 250 miles, as far west -as longitude 37° E., we met with no signs of unmetamorphosed sedimentary rocks, once the coastal zone had been left behind. The oldest rocks lying on the basal complex are of Lower Cretaceous age, and consequently there is no close stratigraphical evidence for the age of the complex. The petrological details and architecture of the complex suggest for the greater part of it a Pre-Cambrian age, and in a subsequent paper an attempt will be made to show that this view is justified : for a direct measurement of the age of the gneisses, and that of one of the later granites, by the radioactive method, proves that an approximate correlation can be made between the older rocks of Mozambique and those of other countries far removed from the tropics.! ' A. Holmes, Proc. Geol. Assoc. vol. xxvi (1915) p. 805. 224 DR. A. HOLMES ON THE TERTIARY | vol. Ixxu, The geological sequence in Mozambique may be summarized as follows :-— Recent Deposits. Overlying drifts, lateritic deposits, coral reefs, and raised beaches. SA UU Tertiary. 2. Vicloanae rocks. Post-Oligocene. 1. Mozambique Limestone and Olinncene Mochelia Conglomerate. Se oe Cretaceous. 3. Conducia a Monapo Beds. Upper Cretaceous. 2. Mount Meza Beds. Middle Cretaceous. iN Fernao Velloso Brads. Lower Cretaceous. Pre-Cambrian 4, Puronanitar; Complex. 3. Granites and pegmatites. : 2. Gneisses, amphibolites, and eclogites. 1. Schists and crystalline limestones. Ill. THe Coastat Bet. The narrow Coastal Belt is formed essentially of Mesozoic deposits, and extends intermittently along the whole East African coast, broken only where the sea has eaten away the sedimentary formations and reached the more resistant rocks beyond. Through- out the length of Mozambique, no sediments older than Cretaceous are known along this zone, but on the north, in Portuguese Nyasaland and in the German (1914) and British areas, Upper Karoo and Jurassic formations are also found. In Mozambique the coastal zone varies in width up to 10 miles. The shore itself is fringed with loosely-cemented coral-rocks and raised beaches. The underlying sedimentary rocks extend almost continuously from the Lurio in the north to Moginquale in the south, and are nearly all of Cretaceous age. Mozambique Island and the outlying parts of the adjoining mainland are unconformably capped with Tertiary sediments, which were referred by Sadebeck (1879) to the Upper Eocene or Oligocene. Mr. Wray has divided the Cretaceous formations into three main groups, the succession of which, with their lithological characters and distribution, is expressed in the following table :— | Name. | Age. Lithological Character. | Distribution. | | 0. Conducia and Senonian. lLimestones, calcareous Mokambo Bay to Monapo Beds.| | shales and sandstones, the Mikati River. | | with a calcareous con- | | glomerate at the base. | | 2. Mount Meza Albian to Sandstones, often fel- Mount Meza and | Beds. Aptian. spathic,with caleareous the adjoining bands. _ district. | 1. Fernao Velloso Neocomian. Alternating limestones South of Fernao | Beds. ' and shales. Velloso Harbour to Memba. | | part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 225. The prevailing dip is southward or south-eastward, except north of Mokambo Bay, where the Conducia Beds’ appear to be -brought up by a syncline. Folding, however, wherever seen, is very gentle. Evidences of movements of a vertical type are more common, and small north-and-south faults have broken up the Monapo Beds along the northern shores of Mokambo Bay. Mr. Wayland deter- mined their junction with the gneissose rocks on the west to be a faulted one. The Conducia Beds south of Sokoto Hill are faulted against the gneissose rocks, for the conglomerate basement-beds are not seen, and marine limestones are brought to a level with the older rocks. Mz. Wray considers that Fernio Velloso Harbour, with its rugged gneissose cliffs on the west and low- -lying iRsravasbraravers and shales on thes east, represents a line of fault picked out by marine erosion between the Cretaceous sediments and the Pre-Cambrian complex. On linking up all our observations, it seems clear that. a series of nearly north-and-south faults has slowly displaced the Cretaceous beds, generally dropping them down on the eastern side, and locally cutting them off sharply from the ancient rocks beyond. It is a significant feature that the average direction of these faults is parallel to the general trend of the adjoining coast-line and to the basaltic dykes associated with the Tertiary volcanic rocks. The strike of the foliation and banding of the gneisses is truncated obliquely by the coastal belt, and the coast has therefore a typically Atlantic character. IV. BipiioGRaPny. 1879. A. SapEBECK.—‘ Geologie von Ost-Afrika’ in Von der Decken’s ‘ Reisen in Ost-Afrika.’ Leipzig, vol. i11, pp. 35, 36. 1882. C. Mapnes.—‘Makua Land, between the Rivers Rovuma & Luli’ Proc. Roy. Geogr. Soc. vol. iv, p. 79; map on p. 128. 1882. H. WH. ONermu—‘ A Three Months’ Journey in the Makua & Lomwe Countries’ Proc. Roy. Geogr. Soc. vol. iv, p. 193. 1884. H. E. O Neri.‘ Journey from Mozambique to Lakes Shirwa & Amaramba’ Proc. Roy. Geogr. Soc. vol. vi, pp. 632, 713. 1 H. E. O’Neri1.—‘ Eastern Africa between the Zambesi & Rovuma Rivers’ Proc. Roy. Geogr. Soe. vol. vii, p. 480 and Scot. Geogr. Mag. vol. i, p. 337. 1887. J.T. Last.—‘On. the Society’s Expedition to the Namuli Hills’ Proc. Roy. Geogr. Soc. vol. ix, p. 467; map on p. 212. 1900. P. Cuorrar.— Sur le Crétacique Supérieur a Mocambique’ C. R. Acad. Sci. Paris, vol. exxxi, p. 1258. 1903. P. CHorrat.— Contributions 4 la Connaissance Géologique des Colonies Portugaises de l Afrique: le Crétacique de Conducia’ Comm. Serv. Geol. Portugal. 1912. A. Hormes & D. A. Wray.—‘Outlines of the Geology of Mozambique ’ Geol. Mag. dec. 5, vol. ix, p. 412. 1913. R. L. Retp.— Notes on Mozambique Exploration’ Geogr. Journ. vol. xlii, p. 44; map on p. 112. 1913. A. Hommes & ID. A. Wray.—‘ Mozambique: a Geographical Study’ (with map). Geogr. Journ. vol. xlii, p. 143. 1914. A. Hones.— The Lateritic Deposits of Mozambique’ Geol. Mag. dec. 6, vol. 1, pp. 529-37. 1915. E. J. Wayitanp.—‘ Notes on the Occurrence of Stone Implements in Mozam- bique’ (with map). ‘Man,’ vol. xy, p. 57. 1915. D. A. Wray.— Observations snr la Géologie du District de Mogambique ” (with sketch-map). Comm. Serv. Geol. Portugal, vol. xi, pp. 69-84. 1917. A. Houmes.—(q) Picrite from the Ampwihi River, Mozambique’ Geol. Mag. dec. 6, vol. iv, p. 150; (6) ‘The Pre-Cambrian & Associated Rocks of the District of Mozambique * Abs. Proc. Geol. Soc. no. 1010, p. 89; & Geol. Mag. dec. 6, vol. iy, p. 380. p Fig. 1— Map illustrating the distribution of the volcani¢ rocks of Mozambique. VY), Yys df yy / Vy ANI A VG Ye “ Yep Abd AG Brian Sebastian eas UY IES : MOZAMBIQUE @ laa eMor BAG ; St.George or aa Ap. itev= Goals. |~ Sancoul Pt. &St. Iago — up r= E : = fochelia = IN s{~Tephritic Pumice A GEOLOGICAL Mokambo Bay MAP OF A PORTION OF >=! MOZAMBIQUE. aay) MOZAMBIOQU 5 10 ALLUVIUM....... Where Mapped TERTIARY LAVAS Ml TERTIARY CRETACEOUS... ARCHHAN....._ GZ] MILITARY Posts... . Matembo 1520S. 154 Scale of Miles 4 Om aan 5 10 40 40° 40°50" [Based on the topographical survey by the staff of the Memba Minerals Ltd., 1910-11 (Geogr. Journ. vol. xl, 1913, p. 112) and on geological observations made by Dr. A. Holmes, Mr. E. J. Wayland, and Mr. D. A. Wray | ‘xo[duroo uemquey-erg oqo ‘aqmeuded ‘OPIUBLS “SSTOTL ‘9YVAOULO].SU0D SHOeT¥OTBD ° ‘euojsemry * ‘oreyS ‘quoyspueg ° *(eus00SI[Q) euojseuTT : 5 ‘soyAp-qeseg * (euad08T[9 aso) { Ex nN “(snoa0vz0e10) = DOO . eee 2s iS keg 1 JOAL eIonpuoy once esi omryerqy ‘Gd xe) a ooaall = Csaquu TT = yphuay) hog monpuop 0} Oly DIG T WOLY “iQ wo1usag—e “B17 SoS ; Ke 4jne | Jan s9yAC yne.z IOATY METaEe Zon eater uu EMES-EMES odevoy, eysepuy Arepunog HETIL “el (soja Gg = yphuary) hog [4NSOTT 07 1a0ry Yoyrjipy ay} wo.if ‘gp worsay—z “Oly 228 DR. A. HOLMES ON THE TERTIARY [ vol. Ixxn, The Volcanic Area. V. Tue DistRIBUTION AND ForMER EXTENT OF THE LAVAS. The voleanic rocks of the district of Mozambique océur along a narrow belt of country which extends from Mokambo Bay to beyond the Sanhuti River, following a line approximately parallel to the general direction of the coast (see map, tig. 1, p. 226). The boundary-fault between the sedimentary coastal belt and the Pre- Cambrian complex roughly bisects the voleanic area, although in the north the lavas are more abundant on the western side of the fault. The lava-flows are everywhere horizontal, and are clearly the result of fissure-eruptions, no traces of pyroclastic ejecta having been anywhere encountered. The flows are now restricted to innumerable small outcrops surrounded, according to the locality, by gneiss and granite, by Cretaceous or Tertiary sediments, or by more recent lateritic earths and alluvium. Small dykes and sills, representing the feeders through which the magmas reached the surface, penetrate the underlying rocks, and are exposed in many places between the much dissected lava-flows. On the accompanying sketch-map (fig. 1), it has been possible to express only the broader outlines of the various formations which enter into this complicated patchwork. No pretence is made to detailed accuracy, for the inherent difhculties of pioneer mapping, topographical as well as geological, were considerably imereased by widespread sheets of superficial débris, and in places by a dense ‘undergrowth of tropical vegetation. Throughout the area the prevailing lavas are lagelkis of a remark- ably noniesrann type. Allare vesicular, and most of them are coarsely amygdaloidal. That similar flows of amygdaloid originally extended northwards as far as Fernao Velloso Harbour is made clear by the presence of abundant pebbles of agate and jasper around the Nakalla River and along the western shores ot the Harbour. Pebbles of agate and jasper are still being derived from the amygdaloids, and naturally they occur in great “abundance on the coastal side of the exposures, particularly along the lower reaches of the Mitikiti, Monapo, Mikati, and Sanhuti Rivers. Similar pebbles, however, also occur to the west of the exposures, especially in the Monapo district, where the flows appear to have been laterally more extensive than elsewhere. The existence of these relies in places (such as Mitikiti) to which they could not have been transported by the present drainage, makes it certain that the lavas flooded a strip of country from 5 to 10 miles broad, and at least 50 miles long. Beyond Mitikiti’s Kraal, a solitary basaltic dyke, well Bacnaered with laterite, was met with in Maravi’s carefully guarded country. Its position, near Nashologoto, is marked on the map. Ina line with this dyke, a few miles away to the south, a series of hot springs breaks through the gneissic surface, providing the Mitikiti River with a small but permanent tributary. We encountered the steaming waters in a dense forest during the dry season, when all part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 229 the neighbouring gullies were empty. This fact, together with the comparatively high temperature (about 60° C.), suggests that the springs may be of juvenile origin, representing the last flicker of expiring activity on the border of a volcanic field already much reduced by erosion. Nearly a gallon of the water was evaporated to. dryness in a clean prospecting-pan, in order that the residue might be tested by blowpipe reactions. Carbon dioxide was present in just sufficient abundance to be recognized. The residue was very small, and only soda and silica could be detected. The rocks described in the present paper are confined to a district rather more than 80 miles in length, extending from the Mitikiti River to beyond the Sanhuti River. Amygdaloidal and vesicular basalts are not the only types of lava represented. In Fig. 4.—Specimen of amygdaloidal basalt from the Sanhutr- River district. (Photograph by Mr. D. A. Wray.) the neighbourhood of the Monapo River—2 miles north-west of Mochelia—a narrow dyke of hornblende-andesite penetrates a sheet of amygdaloid (H.A. in fig. 1, p. 226), and Mr. Wayland found pebbles of pyroxene-andesite on the débris-strewn flats of the same | district. Here also fragments of tephritic pumice are found, though again not 7m situ. This rock clearly indicates another line of variation which is more strongly represented in the north, for the late Mr. R. L. Reid discovered an interesting occurrence of phonolite associated with basalt and picrite-basalt in the neigh- bourhood of the Sanhuti River. Other rocks not found 7 situ are solvsbergite and zgirine-trachyte. Thus, there occur together, within the limits of what is a comparatively small volcanic field, lavas of both alkali and cale-alkali types. VI. PERIOD oF ERUPTION. The period of eruption of the amygdaloids cannot be definitely ascertained, but three lines of evidence point to their Tertiary, and almost certainly post-Oligocene, age. @); da Gis Sh ING, Wer S FS EE = S55 <= ——— SSS Zane Raised Beach (for diagrammatic purposes vegetation is omitted). Fig. 5.—Sketch of the Mochelia Dyke penetrating Tertiary sandstones, by EH. J. Wayland : (a) Kviden e from Faulting. In the district north of Mo- kambo Bay, between the Monapo and Mitikiti Rivers, sheets of amygdaloid are seen to lie succes- sively on a surface of Cretaceous sandstones and limestones, and on far older crystalline rocks. At Mitikiti’s Kraal, only occasional pebbles of agate and jasper re- main to prove the former extension of the flows. The junctions be- tween sandstone and limestone and between limestone and gneiss are faulted, and it is evident that denudation kept pace with the movement and reduced the various formations to a nearly plane sur- face before volcanic activity com- menced. Unfortunately, it is not easy to determine the period of faulting. The Oligocene deposits of Lumbo, on the southern shcre of Mosuril Bay, are probably faulted down against the Creta- ceous sandstone, but the move- ment there may not have been simultaneous with that within the Cretaceous sediments, nor with that between the latter and the Pre-Cambrian complex. If it be conceded that all the faulting is of approximately the same period, the time of extrusion could not be earlier than the Oligocene ; but, as this assumption is itself not securely founded, the evidence points to a period no more definite than post-Cretaceous. (6) Evidence from the Age of Intruded Sediments. The phonolite of the Sanhuti River and the amygdaloids throughout the area are neces- sarily post-Cretaceous, since they lie partly on a base-levelled Cre- taceous platform. At Mochelia, however, Mr. Wayland found a number of dykes intrusive through sandstone and conglo- merate, which he interpreted a part3] TERTIARY VOLCANIC ROCKS OF MOZAMBIQUE. 231 the shore-facies of the Oligocene deposits of Lumbo and Mozam- bique Island (fig. 5). From this more detailed evidence, a lower limit may be assigned to the period of vulcanicity, determining the age of the vulcanism as late Oligocene or post-Oligocene. (c) Evidence from the Composition of the Oligocene Limestones. In the Oligocene deposits of Mozambique Island and Lumbo, fragments of Cretaceous limestone occur that may be easily identified, although their size is small compared with the similar contained fragments in the Mochelia conglomerate. Now, if the conditions of denudation and deposition were such that pebbles of limestone were preserved, it seems reasonable to suppose that, if the amygdaloids had existed at that time, they would also have played some part in building up the Oligocene formations. At the present day, the rock-detritus that is distributed over the littoral plains at high-water level, invariably contains large numbers of agate- and jasper-pebbles ; and similar, though less abundant, relics of igneous activity are found among the gravels of the numerous raised beaches which fringe the shore from point to point. Hence, since no trace of basalt nor of its siliceous amygdales have been detected in any of the Tertiary sediments, one may conclude with some confidence that the basalts had not then been extruded, and consequently that the period of their extrusion was late Oligocene or, more probably, post-Oligocene. "The phonolite and associated lavas from the Sanhuti district lie across the boundary between Cretaceous and Archean rocks. Their age is, therefore, post-Cretaceous; but their relations to the far more abundant amygdaloids are as yet entirely unknown. VII. OrprEr or ERvprion, AND CLASSIFICATION. The lavas from the Sanhuti mentioned above were collected by Mr, Reid during a journey to Mount Meza. In the case of three of his specimens, the order of eruption can be definitely stated. Near the base of the volcanic series, lying on a floor hidden by lateritic earth and gneissic débris, is a dark compact basalt containing rusty amyedales partly occupied by calcite. Both in the com- position of the amygdale minerals and in its microscopical aspect, the rock differs conspicuously from all the other basalts of Mozam- bique. Ata higher level a lava of lighter colour was met with. This is the phonolite, and it also mannan ealeite as the chief occupant of the vesicles. At the top of the series a black vesicular basalt was found. Owing to its extremely mafic character, it is here designated a picrite-basalt (see § XIII, p. 244). The specimens next collected include a fragment of amygdaloidal basalt rich in chaleedony and heulandite. This rock is identical with the amygdaloidal basalts occurring elsewhere in the volcanic area, and was found in situ on a sedimentary basement. It is interesting to notice that, in Abyssinia, W. T. ianford found two equally contrasted series of lavas. The lower consisted s2 232 DR. A. HOLMES ON THE TERTIARY (vol. Ixxu,. ot amyedaloidal basalts remarkably similar to those of Mozamibtdne, while the upper included extensive flows of trachyte and phonolite.! Thus in Abyssinia the evidence that the amygdaloids preceded a series of alkali-lavas is complete. Unfortunately, in Mozambique the observations hitherto made do not admit of so definite a deduction. Of the remaining alkali-lavas described in this paper, none were found @7 s¢tu. South of the Monapo, the occurrence of a narrow dyke of andesite cutting a flow of amygdaloid leaves no doubt as to the succession. A complete statement of the sequence of igneous rocks through- out the district cannot be given, but in the follow ing synopsis the associations in three representative districts are enumerated. Names marked with an asterisk are of lavas not yet known 7 s¢fu :-— Sanhuti District. Sokoto District. _ Monapo District. ( Sélvsbergite.* Aigirine-trachyte.* Tephritic pumice.* ve Picrite-basalt. *) Phonolite. Basalt. ( Pyroxene-andesite.* | , Hornblende-andesite. B.< Amygdaloidal and Amygdaloidal and Amyegdaloidal and vesicular basalts. vesicular basalts. vesicular basalts. L Basalt sills and dykes. Basalt sills and dykes. The amygdaloidal and vesicular basalts (which extend through- out the area) vary but little in respect of their basaltic frame- work. Such differences as have been noted depend on the pio- portion of brown glass, the sizes of crystals, and the degree of vesicularity. The amygdales are occupied by agate, Jasper, and quartz, and by a variety of zeolites ; moreover, the composition and the relative abundance of the amygdale- noes constitute a second line of variation. With the exception of the dyke of andesite already mentioned, all the dykes and sills encountered in the volcanic area are basaltic, and can be closely matched by rocks occurring as surface-Hows. The reason for this is clear: the dykes and sills in most cases are exceedingly narrow, and they appear not to have been long separated by erosion from the overlying lava-sheets to which they served as feeders. The larger Mochelia dykes are free from vesicles: but all the other specimens collected contain a residuum of glass, and are minutely vesicular and amygdaloidal. A note- worthy feature is the general absence of zeolites from these small intrusions. Taking the basalts as a whole, a continuous series can be arranged, starting with the compact Mochela basalt, passing through various vesicular and amygdaloidal types, and finishing with glassy scoriaceous basalts. The lavas fall naturally into two main groups, designated as A & B in the list of types given below (p. 283). For petrographic description the following order will be adopted. In the cases 1 « Observations on the Geology & Zoology of Abyssinia’ 1870, p. 182. part 3 | VOLCANIC ROCKS OF MOZAMBIQUE. 233 determined, the position and symbol of each rock in the quanti- tative (C. I. P. W.) system of classification is added :— Rocks of Group A. (Alkali series and associated basalts). SOlivsencihenenas eee aeeeeeanee Nordmarkese...... IW yy ik, és Aigirine-trachyte............... n. d. Trachytoid-phonolite ......... Umptekose......... IT, 5, 1, 4: Tephritic pumice ............ Akerose ............ II, 5, 2, 4. Basality eas siete: Bree ithe Camptonose ...... ITI, 5, 3, 4. Picrite-basalithesss sesso. . Auvergnose ...... III, 5, 4, 4. Rocks of Group B. (Cale-alkali series). Basalts: sills and dykes ... Hessose ............ TI, 5, 4, 4. Basalts: lava-flows ......... Tonalose-bandose. II, 4, 3—4, 4. Hornblende-andesite ......... ipandOsee eee. 1 Zin a ale Ground-mass ............ Lassenose ......... I, 4, 2, 4. Pyroxene-andesite ............ ms Gla Petrographic Descriptions and Analyses. Rocks of Group A. VIIL Sétvspereirp. (Pl. XX, fig. 1.) Two specimens (153! & 154) of sdlvsbergite were collected by the late Mr. R. L. Reid, between Miali and the Sanhuti River ; but their source, which was possibly a dyke penetrating gneiss, was not discovered. The rock has a mottled appearance in the hand-specimen, due to the alternation in ill-defined layers of dark-grey and straw-coloured patches. In this respect it bears a close resemblance to the type- rock of Gran, but it is heavier, darker, and more finely-grained than the latter, Small tabular phenocrysts of clear glassy anorthoclase are sparsely scattered through the rock, reaching 1 em. in their largest dimension. Under the microscope the following minerals can be distinguished. Their approximate proportions by weight, measured by the Rosiwal micrometric method, are added :— Estimated Minerals. Percentage by weight. Anorthoclase 7 Cossyrite (accompanied in small amount by katoforite and arfvedsonite) ............... 13 Agirine and egirine-augite..................... 8 Magnetite (and ilmenite?) ..................... 2 Freer a bibeupie ki Heya iy Duin We NAPUS eE ne ens coh n.d VAR COY daa COR A Reo Ratt sinc ou dat Roaah Sanaa n. d. Motalienreere 100 Anorthoclase determines the texture of the rock, which is trachytic. The crystals are clear and unaltered, and have a 1 These numbers in parentheses refer to specimens in the collection of Mozambique rocks studied by the author. The collection is at present in the Geological Department of the Imperial College of Science & Technology. When its investigation is completed, it is proposed to present the collection to the Mineralogical Department of the British Museum (Natural History). 234: DR. A. HOLMES ON THE TERTIARY [ vol. xx, prismatic habit, elongated parallel to the @ axis. The three refractive indices are respectively slightly above, equal to, and slightly below 1°526 (monochlor benzol) according to the direction. Carlsbad twinning is common, but the fine striations due to albite twinning often seen in anorthoclase have not been detected. The extinction, however, is mottled and undulose. Small inclusions are present, frequently with a peripheral arrangement. Most of them are minute prisms of egirine, but a few highly refractive blebs, which exhibit high polamey ion colours, are probably zircon. Cossyrite occurs in ragged patches, and in thin wedges occupying the angles between adjacent felspars. Except in especially thin sections and near the eages, when the deep-red and brown pleochroism becomes visible, the mineral is practically opaque. Small grains of katoforite and arfvedsonite are associated with the more absorbent amphibole. Aigirine and egirine-augite lie among the felspars in shreds and granular “aggregates. The pleochroic scheme of the egirine is as follows :— x (nearly=c)...... blue-green. Wea lN cassoueacaease yellow-green. z(nearly=a) ... yellow-brown. Magnetite, probably titaniferous, is present in masses of about the same size as those of cossyrite. Itcan readily be distinguished from the latter by observation in reflected ight. Red ferruginous stains accompany the amphiboles in some cases, and probably represent an excess of ferric oxide over that required for their erystallization.! Chemical Composition. Analysis of a perfectly fresh specimen (153) gave the fee results :— Molecular Mineral Composition Percentages. Proportions. (Norm). S1Os nan 61:02 1:017 Qmart7 2 ee 1:74 AO REALE. Ti diatlal 168 Orthoclase...... 26°69 | Salic IEKO, coecso | RB 027 AM OMG r00 265 560852 55:02 ( =87°62. HeO ee) or42 047 Anorthite ...... 4:17 Mic Olena 0°63- ‘016 ‘ ; CaO 1:20 "021 Diopside......... 1:40 Wes@® cooose 6°51 105 Hypersthene... 3:28 Femic AONE ei: 4°47 048 Magnetite ...... 6:26 ( =12:00. TELO)se con, OFSL oe Tmenite ......... 1-06 HAO seen Ose au TWAS) THO, soces 0°59 007 o2N02 WENO). oe sare 0°34 005 Waiter ctacca-se 0°64 ovale 100°30 Mo tallest 100°26 Classiiay Persalane. Ra=2:26 X10—) orms. per grm. Order 5 ... Canadare. Specifie gravity =2°58. Rang 1... Nordmarkase. Subrang 4 Nordmarkose. 1 For ie primary character of hematite and limonite in certain igneous rocks, see G. M. Murgoci, Amer. Journ. Sci. ser. 4, vol. xx (1905) p. 145. part 3] VOLCANIC ROCKS OF MOZAMBIQUE, 235 Several analyses of sdlvsbergite from other localities are tabu- lated below for comparison with the Mozambique example. All the analyses bear a close family resemblance, indicated in the quantitative classification by thei falling into divisions I, 5, 1, 4. or I1,5,1,4. The Mozambique sdlvsbergite i is richer in 1ron- aitles. both ferric and ferrous, than any other hitherto analysed. The abundance of ferrous iron, which is the more noteworthy, is clearly related to the dominance of cossyrite over the more usual egirine, and in part to the presence of magnetite :— | ante B. C. D. E. F. KO) Sovade | 61:02 63°74 65°46 64°28 64:92 58°90 JMO) oeoace | Weil 17°86 17:40 15:97 16°30 17°70 WW@XOs ococ0- | 4°37 4°27 3°00 2:91 3°62 3°94 Ro Ou as 3°42 0°30 1:60 3:18 0°34 2°37 | RUB O oon ece | 0°63 0:10 0:09 0:03 0°22 0°54 CRYO) obo cel! aleA0) 0°33 0:76 0°85 1:20 1:05 | EO} be.cns 6°51 7:23 6°51 7:28 6°62 7:39 HaCGS OB Eerea | 447 5°19 4°74 5:07 4°98 5°59 IE Os eee | 0°64 0°83 0:87 0:20 0°50 1:90 Mai Oy ae | 0°59 tr. 0:24 0:50 Pe 0:40 Je Oarieen nee | ilo Gl. we Eco 0:08 Be eae MnoO ...... | 0°34 0-19 ee tr. 0:40 0°55 | Totals....... 100°30 | 100:54 | 100-67 | 100-35 99°60 | 100:33 A. Between Miali and the Sanhuti River, Mozambique (an. Holmes). B. Edda Gijorgis, Abyssinia (an. Prior). C. Camel’s Hump, Victoria, Australia (an. Bailey, Lewis, & Hall). D. Andrews Point, Essex County, Mass., U.S.A. (an. Washington). EH. Sélvsberget, Gran, Norway (an. Schmelek). F. Kjose, Aklungen, Norway (an. Schmelck). IX. Moetrine-Tracuyte. (Pl. XX, fig. 2.) Fragments of a porous lava with small phenocrysts of felspar (anorthoclase) and strongly developed flow-structure were found north of Sokoto Hill (151 & 152). Myr. Starey and I worked in the district for several weeks, but failed to discover the rock in situ. The weathered surface of the specimens collected -has a rusty-red colour, broken by white-lined vesicles, the contents of which prove to be opal (refractive index between 1446 = chloroform, and 1:472 = nut-oil). Within, the rock presents a mottled appearance, similar to, but more finely Srna than, that of the sdlvsbergite just described. Specific gravity = 2°33. Under the microscope the minerals that can be distinguished are anorthoclase, wgirine, cossyrite, magnetite, and opal. In addition, a great deal of an obscure brown alteration- product is. present ; similar material in the trachytes of the Great Rift Valley has been called ‘ferrite’ by Dr. G. T. Prior. As in the Rift- Valley rocks, so here, the ‘ferrite’ seems to be an alteration-product of cossyrite and other soda- -amphiboles, for its mode of occurrence 236° DR. A. HOLMES ON THE TERTIARY [vol. Ixxu, is exactly paralleled by that of the soda-amphiboles in the other rocks of Group A. Flow-structure is of two kinds, which merge one into the other. The first is determined by parallel laths of anorthoclase accom- panied by radial growths of that mineral with xgirine, cossyrite, and ‘ferrite,’ occupying the wedge-shaped interstices. The second type of flow-structure is that which imparts to the rock its mega- scopic aspect. It is determined by the alternation of bands rich in prisms of egirine, with bands containing long drawn-out vesicles now occupied by opal. The opalized bands are further character- ized by abundant ‘ferrite.’ A subsidiary type of band is mainly composed of parallel laths of anorthoclase with cossyrite. Hlse- where the laths of anorthoclase, while presenting a general trachytic aspect, have crystallized in radial aggregates. The lava bears a close resemblance. to the phonolitic trachyte from Lake Baringo described by Dr. G. T. Prior in 1903.1 The radium-content of the rock was determined in 1913, and proved to be 2°83x107~! grms. per grm. The silica percentage was found by the late Lieut. G. Kirby to be 64°51. X. Tracuyror-PHonorire. (Pl. XX, fig. 3.) The phonolites of the Sanhuti-River district (149 & 150) differ from the other lavas of Mozambique in two noteworthy respects. Large phenocrysts of clear glassy anorthoclase, an inch or more in length, are distributed through a grey ground-mass somewhat coarse in grain for a phonolite. Amygdales, averaging half an inch in diameter, are also scattered through the rock, and these are occupied by natrolite and calcite. The former mineral lines the cavities in aggregates of white fibres, leaving the bulk of the interior partly or wholly filled with calcite. Under the microscope, the minerals recognized and _ their approximate percentages, determined by the Rosiwal micrometric method, are as follows :-— Estimated Minerals. Percentage by weight. ANTOTUINOGINRG): "de coonsscadusansacsanes 33°8 Natrolite and isotropic material... 24°7 INephelime ss ein: boi aba be oe 6°6 Aigirine and egirine-augite ...... 1syAl Katoforite Arfvedsonite ; Ee ae eee aoe 11195 Barkevikite Cossyritenn wun oe reenter re 4°3 @all cite eee tosren ee caee te San phar 2°0 eee and ilmenite ............ 2°0 Apatite .. Bs SEE eee pottery hy CAOTHLO, No taller ne 100 The phenocrysts of anorthoclase are prismatic, elongated 1 Min. Mag. vol. xiii (1903) p. 2441. part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 237 along the @ axis. Carlsbad twinning is common, but no twin striations can be detected.. The refractive indices are the same as those of the sdlvsbergite felspars. Small inclusions of the ground- mass are present. in which egirine and magnetite can be seen. An analysis of the mineral for alkalies, made by Lieut. Angus McIntyre in 1913, yielded the following results :—soda = 6°68 per cent. ; potash = 4°51 per cent. The phenocrysts are surrounded by a narrow zone of kaolimiza- tion, which in places penetrates the interior along cleavage-cracks. Wherever kaolin is present, calcite is also found, sometimes in granular crystals, sometimes in cryptocrystalline aggregates with kaolin. In the ground-mass small laths of anorthoclase are distributed in a mesh, which in texture falls between the trachytic and the intersertal types. Prisms of nepheline of similar dimensions are also present. Soda-pyroxenes and amphiboles occupy the wedge-shaped spaces between the felspars, but not completely. The ‘ultimate base of the rock is a mineral, the refractive index of which hes between 1-482 and 1:478, and is identical with that of the natrolite found in the vesicles. Associated with this interstitial natrolite is an obscure isotropic substance. Its refractive index approximates but is higher than, that of the natrolite: it contains no chlorine, and is probably analcime or glass. With the exception of cossyrite, which forms occasional ‘small phenocrysts, the coloured minerals are all of later crystallization than the anorthoclase and nepheline. Grass-green egirine is very abundant; but in most cases the colour is patchy, indicating association with zgirine-augite, which is pale green and gives a high extinction-angle. A few grains of augite, sometimes zoned, are also present. The pyroxenes are frequently included in the monoclinic amphiboles. Cossyrite, however, itself containing inclusions of ilmenite ame apatite, is often surrounded by egirine. The pleochroism of katoforite, as exhibited in this rock, is variable; but the following scheme represents the usual distribution of colour (and also the maximum extinction-angle ) :— xO) Nao ee straw-yellow to colourless. NYS Sle aa smoky red to purple-brown. bi J\ C= 38° soons0 greenish-yellow. The birefringence is very low, and the refractive index is approxi- mately that of monobrom- naphthalin, 1:66. Some of the grains and wedges of katoforite pass into arfvedsonite, orientated in parallel growth. Its pleochroism is in tints of lavender-grey X ; violet Nee and yellow-green Z. The extinction-angle 16° differs markedly from that of katofor ite, while the birefringence remains about the same. A few individuals that look like “katotorite in ordinary light are probably barkevikite, for their polarization- colours are high. The extinction-angle could not be measured, but the refractive index was found to lie between those of monobrom- naphthalin = 1-66 and methylene iodide =1°74. 238 DR. A. HOLMES ON THE TERTIARY [vol. lxxi, Owing to the smallness of the grains and the lack of cleavage, the optical orientation of cossyrite could not be determined. The refractive index is near 1°66, but slightly higher. Pleochroism is. in tints of deep brown, reddish-brown, and lighter brown, and can only be observed in a very thin section. The first minerals to crystallize out were magnetite, ilmenite, and apatite, followed by phenocrysts of anorthoclase and cossyrite. After partial resorption of the phenocrysts, during which the anorthoclase suffered peripheral alteration to kaolin, the crystalli- zation of the groundmass began. Nepheline and anorthoclase were followed in order by cossyrite, wgirine, katoforite, and arfvedsonite. Finally, natrolite, a little calcite, and an obscure isotropic substance filled up the still-unoccupied spaces ; also natrolite and calcite were deposited in the vesicles. Chemical Composition. An analysis of fresh portions of No. 149, free from visible amyedales, gave the following results :— Molecular Mineral Composition Percentages. Proportions. (Norm). SHO), scseop aes 56°34 939 Orthoclase ...... 27°80 ALO) Se 17°86 175 Allloubes se ssaceane 46 11 | Salic HesOy eek 443 . 1028 Nepheline ...... 8°52 | = 84°38. FeO........ 3°81 053 Anorthite ...... 1:95 IMIEO) seoc00 0°52 013 CaO Pea c Ie 036 Diopside ......... 3°07 | NBEO) Goceao 7:33 ‘118 Olivine) eee eB) | Renee KOH ers 4°72 ‘050 Magnetite ...... 6°50 F 14-28 TEUO TE aug ie Ilmenite ......... 1225 |e H,O— ... 0°82 Se Calcite meaerereces 1:60} CORE: 0-71 "016 — TO seisoce 0°67 008 i 98°66 P,O5 daboou tr. 6p IMGKO) \sooans 0°28 004 Water ios... 2°04 Motalleeeeeee 100°68 Mota pees 100-70 Class II...... Dosalane. Ra=4:10X10—! grms. per grm.. Order 5 ...... Germanare. Specific gravity = 2°47. TMM Wo joc Sac Umptekase. Subrang 4... Umptekose. Analyses of similar rocks from neighbouring regions are quoted. below for comparison. The Mozambique example is almost, identical with the phonolite from Mount Kenya analysed by Dr. G. T. Prior. The similarity extends equally to the mineral composition. Dr. Prior has kindly allowed me to examine the phonolites from British East Africa in the Natural History Museum, and, except in texture, they bear a striking resemblance to the phonolite from the Sanhuti River. The latter rock, hke the associated sélvsbergite, is unusually rich in iron-oxides, this: being due to the presence of magnetite as an accessory. part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 239 ? | | | Blof 1B C. D. |) WE. F, | SHO). sooede | 5634 54°94 | 58°37 57°81 | 58°62 61:49 AU Oy -ssoces 17°86 19°34 16°65 18°74 21°50 18°25 lniies OF eae 4°43) 1:80 4:09 5°76 0:47 1:77 FeO... 3°81 4°52 | 3:08 0:42 3°65 3:13 MgO ...... Ors ) atoll | OBy tr. 0:56 0:41 Ga Oye WB BOR fe 8G 1:28 0°88 1°65 | SOO sac ece 733 8:39 | 7:28 San | 7Os 6°78 ESS Oe | A772 5°93 5°46 4:52 | 5:47 SAT | PalelO) Ge aine 2°04 | 0°32 2°36 Wes0) |. el 0:26 | WARIO Beanies. 0:67 0:67 0-21 .. | 0:06 O51 | RC OMe col): pa pes ni Lae ae MEO! eee tr. 0-18 0:08 gee ae ig 0:09 | GTO) feet. 0:28 tise 0:43 | tr. pt : | Totals...... 100-68 99:90! 99:99 | 99:38 | 100-28 99°81 ' Including ZrO, =0°38, and SO, = 0°27 per cent. Phonolite. Near the Sanhuti River, Mozambique (an. Holmes). . Glassy katoforite-phonolite. W. Kibo, B.E. Africa (an. Hyme). Phonolite. Base of Mount Hohnel, Mount Kenya, B.H. Africa (an. Prior). . Tinguaite. Hdda Gijorgis, Abyssinia (an. Prior). . Phonolite. Nosy Komba, Madagascar (an. Pisani). . Phonolitic trachyte. Ravin des Fleurs Jaunes, Réunion (an. Boitean). ida XI. TepHritic PUMICE. Among the detritus that litters the northern bank of the Monapo, near its entry into Mokambo Bay, numerous fragments of dark-grey pumice occur. Unfortunately the rock has nowhere been seen 77 situ, although it may be traced for a few miles along the northern bank of the river. It must, therefore, have come from the district between the Monapo and Murimatigri—an area originally covered by flows of amygdaloid, but now greatly dissected by erosion. In chemical composition, the rock differs strikingly from the basalts and andesites of the neighbourhood, whereas it appears to be closely related to the alkali and associated lavas of the Sanhuti- River district. Under the microscope, the rock is seen to be made up of a dense brown glass containing in places irregular patches of less intense colour, in which crystallization had just made a_be- ginning. Within these local clearings, very minute crystals of magnetite or ilmenite and microlites of felspar can just be seen under a quarter-inch objective. The felspar shows no trace of twinning, but a maximum extinction-angle of 25° points to its identity as andesine. ‘This result suggested that the rock might be an andesite-pumice. Against this view, however, was the fact that the radium-content of the lava was 259x107!" grms. per erm., a ‘figure which would be surprisingly large for a rock of andesitic composition. In these circumstances it was thought 240 DR. A. HOLMES ON THE TERTIARY [vol. lxxu, desirable to have the rock analysed. With the exception of the alkali determin4tions the analysis was made by my friend Dr. H. F. Harwood (Chemical Department, Imperial College of Science & Technology), and for his assistance I wish to express my grateful thanks. Chemical Composition. Molecular Mineral Composition Percentages. Proportions. (Norm). SHO idpacodse 52°27 ‘871 Orthoclase ...... 18°90 INO), Veonace 17°70 174 #Mlovtns) ps2teacuonse 37°31 Salie 1WE,Ojs onenes 701 044 Nepheline ...... 6°98 ( =75°42. 12050) seevpanee 4°48 ‘063 Anorthite ...... 12°28 WSO) seccos 1:79 “045 Diopside......... 10°90) CHO soccssces OCB 105 Olivine me eee 0:62 | Wore INELO) daeens 5°95 096 Magnetite ....... 10°21 + 93-99 IKE Olena 3°20 034 mentees. asl ee IBEOse ohn OSI ae Calcite =...-.-.. 1:20) EO eae 0:43 se a COME Reena Oro 012 . pial MUO) else ge 0°56 007 WiM@) es cae tr Warten. ea: ceanect 0-80 Mopalleeeser 100:°20 Mo Gauls eae eter 100-21 Class ler. Dosalane. Ra= 2°59 10-7 grms. per grm. Oxdertomeeseen Germanare. Specific gravity = 2°55. TRAN? 2) sei doc Monzonase. Subrang 4... Akerose. A B. C2 aye) SHO conve 52°27 52°78 51°80 53°04 TANK @ Reese 17:70 19°08 17°90 17°34 INELOS wooase 701 3°63 3°10 2°12 FeO ...... 4°48 3°79 4:36 6:96 Vic Okeaaee 1:79 1 | 3°72 2°49 CaGOinen kt 598 509 | 6°59 5°86 WEO cosoos 5°95 G95) Ac 74 5°61 KE OU: 3°20 Bois) | 3°65 3°00 HOM fake 0°80 0-44 | 2°87 0:37 COP wie: 0°51 0-10 oe PPI @ Se ove 0°56 1:50 | 1°41 2°12 PLO PER ahac n.d. 0:63) 57 0°83 Whn@) scocas tr. | tr. tons). so ebe 100-20 100°75 ! 100714 99-74 i Including 0°33 per cent. of chlorine. A. Tephritic pumice, Monapo River, Mozambique (an. Harwood & Holmes). B. Trachydolerite, Mount Meru, British Hast Africa (an. Maritz). C. Essexite, Bekinkina, Madagascar (an. Pisani). D. Olivine-trachyandesite, Vellouve, Réunion (an. Boiteau). The most notable feature indicated by the analysis is the high part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 247 percentage of iron, and especially of ferric oxide. This character- istic distinguishes the rock from the trachyandesites and trachy- dolerites, and suggests that the name teph rite would be somewhat more appropriate, According to Dr. H..S. Washington,! the name trachydolerite should be restricted to rocks containing labra- dorite or a more basic plagioclase, as well as allali- felspars ; ‘and the name trachyandesite to reels containing andesine in addition to alkali- felspars : both types of rock, moreover, should be rather rich in potash. But for the high iron-content, the composition of the rock now deseribed would Fulfil moderately well the conditions for a trachyandesite. Its close similarity to the trachyandesite from Réunion (analysis D) bears out this opinion, but the chemical characters of the rock are probably expressed more accurately if it be named a tephritic pumice. XII. Basanr. (Pl. XX, fig. 4.) The lava-flow represented by specimen 156 was found by Mr. R. L. Reid at the base of the volcanic succession in the Sanhuti- River district. It is exposed on the north side of the river near the boundary-fault, where it les on a basement of eneiss, and probably extends to the Cretaceous sediments on the east. The basalt a dense dark-grey rock holding rusty amygdales that. average 3 inch in diameter, ‘The amygdales are lined and some- times enitnelly filled by palagonite, which in thin section varies in colour from yellowish green to reddish brown, though in the hand-specimen, ‘beneath the rusty shell of iron-oxides, it is nearly black. The specific gravity is about 2°42, thus agreeing closely with that of the palagonite from Cape Flora (Franz Josef Land) examined by Sir Jethro Teall.2 The same material occurs in the main body of the rock, wedged in the crystalline network of the ground-mass. The only other minerals found in the amygdales are thomsonite, easily recognizable by its moderately-high double refraction, and calcite. Thomsonite oceurs in radiating aggregates, each acicular crystal giving straight extinction. The refractive index, about 1:52, is low for thomsonite, and points to a high content of soda. Excluding the amygdales, the mineral composition of the rock, determined by the Rosiwal method, is approximately as follows :— Estimated percentage Minerals. by weight. Labradorite ........... BBCI oH GUERIN CLy in a ta isi le ne Sa ODEN 56 SANT OUDO BI: eras avisliite csp cent MTA tation Nee tay tee ERT et S 24 1 EEE eXONaLI ays MERE Ieoncualdawaohaaebooueed nec Aan ad cane marr apeT: 15 Magnetite, ilmenite,® and decomposition-products ...... 5 ADS HEH Jee ntabietel Ac 100 Journ. Geol. Chicago, vol. v (1897) pp. 366-68. Q. J. G. S. vol. liti (1897) p. 486. ° The percentage of these minerals is greater than 5, for only a few of the inclusions in felspar and augite could be measured. Loy 24:2 DR. A. HOLMES ON THE TERTIARY [ vol. lxxu, The rock is for the greater part very fine-grained, but in places small ophitic patches occur, never more than ; inch long, in which the texture is much coarser (see Pl. XX, fig. 4). On an average, two of these patches may be seen in a section of half a square inch. In the direction of the greatest extension the ophitic texture seems to have been broken up, perhaps by flow, for porphyritic crystals of labradorite and augite stream away into the fine ground-mass. At right angles to this direction similar porphyritic crystals occur, but less abundantly. In the ophitic patches, as usual, the felspar was the first mineral to crystallize (except for a few skeleton- crystals of magnetite or ilmenite which form rare inclusions). The porphyritic crystals are sometimies composite, and here too the felspar preceded the augite. In the ground-mass, on the contrary, the augite began its crystallization betore the felspar (and continued beyond the latter), and the texture is minutely intersertal. It is clear, therefore, that two distinct periods of crystallization are represented, and that these took place under conditions sufficiently different to affect not only the grain and texture, but also the order of crystallization of the two chief minerals. The felspar of each generation is labradorite, but the early erystals are richer in the anorthite molecule than the later. In the former, which are characterized by. broad albite twinning and absence of zoning, the average refractive index is 1°57. All the felspars are clear and glassy, and are nearly free from inclusions except around the borders of the smaller laths, where innumerable dark globules associated with tiny green blebs of augite obscure the edges. Only in the neighbourhood of the inclusions is there any trace of alteration, the body of the felspars being perfectly fresh. Augite also shows a slight difference in composition, according to the time of its crystallization. ‘The ophitic plates and larger crystals are grey-green, while the small granules of the ground-mass frequently exhibit a brown or even a purple madder tint. Magnetite and ilmenite began to crystallize before the other minerals of the rock, but continued beyond them. ‘They occur in skeleton-crystals and in small irregular masses, and probably the dark globules that border the felspars and fringe the edges of the palagonite wedges represent the final attempts of the same minerals to crystallize: for, where they occur in palagonite, the colour of the latter is paler than elsewhere. Besides palagonite, there is a small amount of obscure inter- stitial matter, some of which is colourless and feebly birefringent, thus suggesting that it may be a zeolite or possibly orthoclase. A green alteration-product also occurs, and may be serpentine or chlorite. There is nothing to suggest that palagonite has been produced by the weathering of an original glass. Glass does not occur in this rock; but, in many of the vesicular basalts belonging to the second stage of eruption, glass occurs abundantly, and .when it part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 243 weathers 1t changes in the direction of laterite, not of palagonite. The latter substance appears generally to be associated with amygdaloidal basalts, not with vesicular types. The formation of palagonite, like that of amygdaloidal infillings, is therefore to be ascribed to a concentration of water (and possibly of other volatile fluxes) in the residual liquors left in the erystal network of the nearly consolidated rock. In the present case, the oceur- rence of thomsonite with palagonite is significant, on account of the light that it throws on the origin of the alkali rocks which succeeded the basalt. 'Thomsonite may contain as much as 7 or 8 per cent. of soda, and the low refractive index of the present example suggests that it is a soda-rich variety. Moreover, palagonite often contains 4 per cent. of soda or more. In contrast with these figures, the rock itself (ncluding a certain amount of interstititial palagonite) contains only 3°29 per cent. of soda, showing that the progress of crystallization led to an enrichment of the residual liquors in soda. The same conclusion would be deduced from the characters of the two generations of plagioclase. Thus, even within the limits of a lava-flow that has rapidly cooled, it is possible to discern a process of differentiation which, on a larger scale and with slower cooling, should result in the production of a magma relatively rich in soda. It is, surely, more than a mere coincidence that the magma erupted next in succession was that which crystallized as the phonolite already described. Chemical Composition. Portions of specimen 156, visibly free from amygdales and weathered crust, were crushed and analysed. The results obtained are as follows :— Molecular Mineral Composition Percentages. Proportions. (Norm). Si0,,. ,. 48°82 814 Orthoclase ...... 8:34 Salic AVOe eee 15:93 "156 . PAUlll ose haeenanneader 27°77 : — 60°57 HepOry kos. 718 045 Anorthite ...... 24°46 } HeOi cn ote 6°63 “092 WIEO)~ osodoc 6:02 pisyl Diopside......... 13°79 ) CaQmisace: 8°67 155 Hypersthene ... 7°84 iTlesaniie EO cooves 3°29 0538 Olivine ......... rit] 39.98 IKE Oren sce 1-41 015 Magnetite ...... 10°44 | ~~ H,O+...... 0:98 ma Iimenite ......... 1°52 HAO = n.- 0°54 an ——— RIOR ee 0:83 ‘010 98°95 MnO ...... 0:32 004 Wraitericteacescc-« 1°52 Notalieerss 100-62 Motel scesec 100°47 Class III Dosalane. Ra=1'38 x10—- grms. per grm. Order 5 ...... Gallare. Specific gravity =2°78. Rane TONE Camptonase. Subrang...... Camptonose. 244 DR. A. HOLMES ON THE TERTIARY [ vol. xxi, | AS ines ©: Dit eee SiO, )....|| 48-82 47°20 46°36 50°22 |. 49-43 | IO e a aon USB 15°29 16°70 16720. |g c2c7 30am TMEOs concn: 7-18 9°57 7°53 3:13 45 5106 TMSO!" shea: 6°63 6-13 6:23 8:07 |. 847 MgO ...... 6-02 5:90 584 754 | 6:96 | CHO) acenen 8:67 8:93 O77 te 8:57 ee 59 PINE t Om 3°29 B61) e822 336 | 3:50 FEO cisbee gf. Teel OSS | 1a 138) emer REL ORn eee 1°52 1:12 0°35 220 i) eerie CORA ES mt ah - 4 | 0:28 TON ena 0°83 114 1:88 I I SHS | MnO! 0°32 ea ‘ sine Seas ONO Plu re a 0:46 0°58 | 0-49 | Totals...... 100°62 100-20 99°58 100°64 | 100-90 | ee ——— =e A. Basalt, Sanhuti River, Mozambique (an. Holmes). B. Basalt, Philippi Lake, Kerguelen Island, South Indian Ocean. C. Basalt, Mount Drygalski, Heard Island, South Indian Ocean. D. Basalt, Glass, Cockburn Is., Louis-Philippe Land, Antarctica (ar. Prior). BE. Basalt, Yandina, Queensland (an. Jensen). The composition of the basalt is in no way abnormal, and can be closely matched by that of other basalts from many other localities. Some of these are cited above for comparison with the Sanhuti example. The closest resemblance is found in the case of basalts from the islands of the South Indian Ocean. XII. Prerrrs-Basanr. (Rl. XX, fig, 5.) Overlying the phonolite, described above, is a nearly black vesicular rock. Most of the vesicles are empty, but a few are occupied by a powdery form of phillipsite. The retractive index of the mineral is 1:53, its double refraction is very. low, and the cleavage is good. In thin section the minerals present and their relative proportions, determined by the Rosiwal method, are as follows :— Minerals. Percentage by volume. ; Titaniferous augite ............ 21 Phenocrysts | Oivine ee ee ae ca 14 Ibe orpaeloratns) js5ss000000n se sensoce ; Grannies Titaniferous augite ............ 65 Magnetite and ilmenite ...... TESoY TOO ISEIE@) son osaaabanccse acc The rock is conspicuously porphyritic, though augite alone is represented in two generations. Olivine, in clear colourless erystals, which have suffered some resorption but practically no serpentinization, is present only as phenocrysts. Labradorite, part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 245, on the contrary, occurs only in tiny laths distributed sparsely through the ground-mass. Purple-brown augite is the most abundant mineral present, and bears a close resemblance to the augite described by Dr. A. Scott! from the Arenig augite-andesites. of Bail Hill (Dumfries-shire). The crystals have a tendency to group together in a roughly radial fashion about a central nucleus. of olivine. In the ground-mass small grains of deep-purple augite are very abundant. Whether the purple coloration of certain augites is due to the presence of titanium or manganese cannot yet be definitely stated.. The usual belief that it is due to the former element is supported by the analyses tabulated in this paper, but the subject is one that still awaits systematic investigation. The ground-mass is very fine-grained, and dark owing to the abundance of magnetite and ilmenite, which are present in well-shaped crystals and in a dendritic network of branching rods. Tron-ores, titaniferous augite, and a sprinkling of labradorite-laths. are crowded together in a residuum of grey-brown glass. In sporadic patches labradorite becomes sufficiently abundant to: control the texture, which is then not unlike that of the inter- sertal basalts described below. Chemical Composition. Portions of specimen 155, free from phillipsite, were analysed with the following results :— Molecular Mineral Composition Percentages. Proportions. (Norm). Si@yeree eee 44°05 734 Orthoclase ...... 4°00 Sali AT Oe ia. 13°39 131 TG Ee 16°83 ; Lea Fe,O, ...... 7°64 048 Anorthite ...... 25°44 | HeO eens 8°42 117 MgO ...... 10°51 263 Diopside......... 21°06 ) CaOre ro: 10°46 187 Hypersthene... 3°79 idler WEL) bent 2:01 032 Olivine ee 13°46 . Fomie LGA Oey an 0°67 007 Magnetite ...... re a cs H,O+...... e 0°44 ais Ilmenite ..,...... 3°65 HO 2. 0°47 a TIO Ase 1:93 024 99°37 Mn Opene 0°25 ‘003 Watery eee 0-91 Mow” Sconce 100°24. Motel verses 100°28 Class III ... Salfemane. Ra=0°47 x10-12 germs. per grm. Order 5 ...... Gallare. Specific gravity =2°78. Rang 4 ,...., Auvergnase. Subrang 4... Auvergnose. The rock is similar, both chemically and mineralogically, to the Hillhouse or picrite type of the Scottish Carboniferous basalts.” Chemically, it bears some resemblance to the camptonite of Kjose Aklungen (Norway), and this is the more interesting because the 1 Min. Mag. vol. xvii (1914) p. 100. 2.G. W. Tyrrell, Trans. Geol. Soc. Glasgow,:vol. xiv (1912) p. 228 (Analysis II, p. aa1)) 3 W. C. Brégger, ‘ Hruptivgesteine des iGeisHoniaeeinettes: TIT’ 1899, p. 51. Q. J. G.S. No. 287. 7 246 DR. A. HOLMES ON THE TERTIARY [ vol. 1xxi1, frequent association of camptonite and bostonite is in the Sanhut- River district paralleled by the association of picrite-basalt and sdlvsbergite. The rock is also very similar to the limburgites of the Macedon District, Victoria! (Tertiary alkali series), which, like the Mozambique picrite-basalt, are associated with anorthoclase- bearing rocks. The main outstanding feature of the analysis is the high pereentage (for this class of rock) of ferric oxide. This is due to the preponderance in the rock of augite, and to the abundance of magnetite. 'Titaniferous augites frequently show an excess of ferric over ferrous iron; the Bail-Hill example is a case in point. Rocks of Group B. XIV. Basauts: Srims anp DyKkzs. (CP) KOC fie Grd PIE UXOXGIe fie i) The specimens collected? are slightly vesicular rocks from the following localities :— 128, 129, 180. Sill, west side of Sokoto Hill, at the base. Dae 142. Dyke, near Ibrahimo. BES IB TE: 141. Dyke, Mochelia (reddish brown). These rocks are of even grain and fine texture, and consist essentially of a.mesh of lath-shaped felspars, between which groups of granular augites are contained. Olivine has in all cases been alter all to ser pentine, and the latter is only a minor constituent. Filling up the interstices of the crystalline network is a grey- brown glass swarming with dark globulites, or with branching sprays of magnetite. Examined closely, specimens from Sokoto Hill and Ibrahimo are seen to be pitted with minute green-lined vesicles or amygdales. The first stage in the infilling of the latter is represented by a fibrous green mineral. The same mineral occurs in greater quantity in the amygdaloids; and, since it contains alumina and has a refractive index greater than 1°576,.1t may be referred to one of the chlorites. The main body of the amyg- dale occasionally consists of brown or greenish-brown palag gvonite; but more usually it is occupied by chalcedony, ahi nee opal is sometimes associated. The minute amygdales of the Mochelia dyke appear to be free from a chloritie lining. The felspar occurs chiefly in small tabular crystals accompanied by a sprinkling of larger individuals, indicating an approach to porphyritic texture. This appearance is intensified by the bunching together of some of the smaller crystals in parallel or divergent series. Hxtinction-angles, twinning, and refractive index Gs 560) show that the species is labr adorite. Many of the larger crystals enclose patches and shreds of chlorite, and lines of globulitic glass-inclusions immediately within their outer borders 1 H. S. Summers, Proc. Roy. Soe. Vict. n. s vol. xxvi (1914) p. 278. 2 A number of dark compact dyke-rocks collected by Mr. Wray (from the district of Fernio Vellosa Harbour and from Makulia farther north) were lost during transit to England. part 3 | VOLCANIC ROCKS OF MOZAMBIQUE. 247 are common. Otherwise, the felspars are clear and glassy, and show no traces of alteration. While most of the felspar came early in the order of erystalli- zation, there is another member of the felspar family which is -allotriomorphic, and was the last of the minerals to crystallize. It is free from twin lamelle, exhibits lower polarization-colours than labradorite, and has a refractive index of 1522. Its identity is thus determined as orthoclase, which is further verified by the relatively high percentage of potash contained in the rocks. Augite is pale yellowish-green (iron-stained round the borders in the Mochelia rock), and exists in rounded grains or sharply- defined wedges determined by the disposition of the felspar-laths. It is probably a member of the enstatite-augite series. The intersertal texture thus produced is a marked characteristic of nearly all the Mozambique rocks. Magnetite, perhaps accompanied by ilmenite, is evenly ‘distributed in small octahedra, irregular grains, and skeleton- crystals. It tends to cling round the augite-crystals, but rarely forms inclusions in them, and never does so in the labradorite. In the Mochelia rock, magnetite is very abundant inside dark patches of glass or palagonite. The mineral certainly belongs to .a late stage in the crystallization of the rock. Dark interstitial matter occupies nearly a quarter of the volume of the rocks, being, however, much less abundant than in the lava- flows. “Some of this material is an isotropic glass packed with globulites, or penetrated by a network of iron-ores. In other cases, microlites of felspar and obscure aggregates of dark augite are buried in a pale glass. Palagonite is also present in deep-brown or greenish patches, which under crossed nicols are seen to have _a fibrous cryptocrystalline texture. By means of a diffusion-column of methylene iodide and benzine, the constituents of the Sokoto-Hill rock (128) were separated into. the following groups. The approximate proportions of each amineral, measured by micrometric analysis, are added :— Mineral Composition of No. 128. Specific Gravity. Mineral. Percentage by weight. PPPS ek Sees Ojpailiotpaneny reentrant a gute 3 ase Glass and palagonite 20 OAR [oe FONG. pee ae Poe Settee a On eRe ae: Orthoclasenwe muse ashen wea ceecaccee 2°62 P G5 (oc Chalcedony and chlorite ...............406 2-68 i O70 [ce Waloradoriter ery rarcsecrncctaieer ener sacar er 40 Ae ORD HSH YSLERTEN “Soca sacs ccbueuted ces 18 Samikergeene ener Magnetite and ilmenite with glass...... 6 Totals nese. 100 Taking the specific gravity of the material that sank as 5, the specific gravity of the rock should be 2°74, The actual value, 2°70, corresponds with the slightly vesicular character of the rock. m 2 245 DR. A. HOLMES ON THE TERTIARY Chemical Composition. (vol. lxxi, The Sokoto-Hill and Mochelia basalts were analysed with the results tabulated below. The analysis of the former rock was. made under my direction by my former students, Lieut. J. L- Harris, M.C., the late Lieut. G. Kirby, and Mr. E. Spencer :— Percentages. si0, 52°42 INO), couase 17°40 WOO sy sencoe 4°21 Be Ox shea 5°16 IWUEO) sooner 5°30 CaOR eer 9:08 INa3 OP eee 2°65 KO ner sute 1-77 EOL eee 1-26 TiO, 0°67 WERO) ogous 0°24 Motaleeeere 100°16 Ra=0:98 X10—} erms. per grm. Sokoto Hill, 128. Molecular Proportions. 874 sali gal 026 072 132 163 043 ‘018 ‘008 0038 Specifie gravity =2°70. Percentages. SIO) soossouee 51°21 ; JNO) bop se 17:99 Hes Ole 4-63 WEO) sonsaes00 581 WIEO conve 6-11 CaO ae 9°36 INGO) cosas 2°92, Kei Oren ae 1:03 FRO aE aa: 0°32 HO ae 0-72 MH ORs peck 0:42 PEOe ees tr. WEXO) Seoece 0:27 Motel eeeeee 100°79 Mineral Composition QUBRATA og tadacdoun Orthoclase ...... JNM) 55. de 00600 Anorthite ...... DMiopside: 4.4.5. Hypersthene ... Magnetite Imenite Class II Order 5 Rang 4 ...... Subrang 4 ... Mochelia, 14: Molecular Proportions. 854. 176 029 081 153" 167 047 ‘O11 005 004. Ra=0'85 x10-! orms. per grm. Specific gravity =2°72. (Norm). Bul) 10:01 Salic 22°53 ( =66°87. 30°58 Her) 12:66 | Femic 6:05) (9 =327035 cre 100716 Dosalane. Germanare. Hessase. Hessose. Mineral Composition Quartz ITT oc. Diopside......... Hypersthene ... Magnetite Ilmenite Rang 4 ...... Subrang 4 ... (Norm). 0:99 Salie = 64°54. 10:96 16°75 6:7 Femic = see Dosalane. Germanare. Hessase. Hessose. For comparison, the analyses are repeated below with others of the same type. The occurrence of a similar basalt in Somaliland ! is interesting, for it preceded the eruption of a series of alkali 1 A.de Gennes & A,-Bonard, C. R. Acad. Sci. Paris, vol. exxxi (1900) p. 196.. part3 | VOLCANIC ROCKS OF MOZAMBIQUE. 249 rocks of the same type as those found in the north of the Mozambique area. Microscopically, the east-and-west dykes of Northumberland—Crookdene, Collywell, and Tynemouth 1 —are very similar to those described here. A. B. | Cc. 1D). E. Sie ne 51:21 52-42 49-47 5l-3t |) eb 1620 Al,0, 17:99 17-40 18°70 14:55 18°89 Fe,0, 4-63 4:21 15-05 757 THEO) eyes 5°81 Kong | pet 9-02 ne MgO ...... 6-11 5-30 | 5:34 6:85 6:75 CaOre ea 9:36 9:08 | 8:90 11°61 10°52 NagOUG2 2-92 SS Maen ORE ETS 1°71 ike Osea! 1-03 1:77 0-79 0-60 0:51 TSO ee 1:04 1:26 | 0:81 1-41 1:70 TO), ess 0-42 0-67 | Kae 1:00 1:14 Ps Ob es tr. pane Sci ug MnO ©... 0:27 0-24 | Bey 0:47 COs | 1-47 Totals...... 100-79 10016 | 101-93 | 10008 , 99:99 . Dyke, Mochelia (an. Holmes). . Sill, Sokoto Hill (an. Harris, Kirby, & Spencer). . Basalt, Kallele River, 56 miles from coast, Somaliland. . Crookdene Dyke, Northumberland (an. Smythe). . Tynemouth Dyke, Northumberland (an. Stead). HoOane Contact-Metamorphism. The metamorphism of the sediments through which the dykes and sills imtrude is everywhere of a simple character. Usually, the changes are restricted to bands extending only a few inches on each side of the contact. Limestones are recrystallized to granular calcite, and in places crystals of quartz are developed, indicating the percolation of silica-bearing solutions from the lavas. Mr. Way- land found that the upper part of a dyke, intrusive into calcareous sandstone near Mochelia, had been removed by weathering, thus exposing the sandstone walls which had enclosed it. Upon these well-marked hexagonal jointing had been superinduced. Near the contact of’ this dyke steam-holes have been developed in the sandstone, and these are now filled with opal and chalcedony, the latter being pale brown owing to the incorporation of limonite which was present in the original cement. Examination of the unaltered sandstone in thin section shows that it consists of slightly rounded grains of quartz, orthoclase, and microcline, cemented together by calcite and limonite. Small nummulites are found embedded in the cement. Near the dyke, calcite is replaced by opal, the metasomatism having equally affected the nummulites. In the amyegdales of the sandstone, opal was first deposited, chalcedony followed, and the infilling was completed by a clear mosaic of quartz. (See Pl. XXI, fig. 2.) 1M. K. Heslop & J. A. Smythe, Q. J. G. S. vol. Ixvi (1910) p. 3. 250 DR. A. HOLMES ON ‘THE TERTIARY [ vol. lxxii,. XV. Basarts: Lava-Frows. (Pl. XXI, fig. 3.) The specimens collected include the following :— 124. Brown amygdaloidal basalt, near the bend of the Monapo River,. ws south side. 125. Red amygdaloidal basalt, between Mochelia and Mitikiti (penetrated by an andesite dyke). 131. Purple vesicular amygdaloidal basalt; half a mile south of Sokoto Hill. 132. Purple vesicular basalt; three-quarters of a mile west of Sokoto Hill. 134. +} Brown amygdaloidal basalt; south-east of Ibrahimo. 135. 136. Purple vesicular amygdaloidal basalt; east of Murimatigri. 145. Purple-brown amygdaloidal basalt ; near the Sanhuti River. 146. Grey-black vesicular basalt; near the Sanhuti River. One of the best and most continuous exposures of amygdaloidal and vesicular basalts is to be found between the Monapo and Mitikiti Rivers. Numerous small streams and long alluvial fats, connected with the sea by way of mangrove swamps, cut up the flows, and sometimes expose the underlying floor of shale er lime- stone. In places, hard dykes stand out in ridges running nearly due north and south. The basalt is a rough purplish-grey rock, sometimes dull and earthy at the surface, but often glazed by the deposition of a thin film of lateritic constituents. Along the military road (made in 1910) which skirts the northern shore of Mokambo Bay, the freshly-broken rocks could be admirably seen in situ. The colour of the unweathered rock is there purple or reddish-brown, variegated by the green linings of the amygdales and brightened by “the sparkling “cleavage: surfaces. of innumerable crystals of zeolites, such as heulandite. Sunilar exposures were examined east of the forts at Murimatigri and Tbrahimo, and in the Sokoto-Hill district. The thickness of the flows is now very variable and much reduced by erosion, but it was nowhere seen to exceed about 40 feet. The lavas are characterized by a high proportion of glass—dark red-brown to black—and by a strong development of vesicles and amygdales. The chief minerals present in the latter are chlorite, eo) Hes, and various forms of silica. It is noteworthy that calcite is extremely rare. Chalcedony is very abundant, and oceurs in a great variety of forms including waxy carnelian, red and green jaspers and bloodstones, and beautifully banded agates. The interior of many of the siliceous amygdales is occupied by a mosaic of quartz. In other cases, where a hollow cavity has been left, well-formed crystals of quartz, or zeolites, project from the sides. Among the agates and jaspers found in the alluvium flanking the Monapo, Mr. Wayland discovered a number of worked stenes and implements. These are figured and described in an interesting paper, published in ‘ Man,’ July 1915, p. 57. Under the microscope the minerals seen are identical with those of the dyke-rocks already described. The felspars in some part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 251 Specimens are microporphyritic, and the smaller crystals are frequently grouped together in glomeroporphyritic clusters. Where individuals of markedly different size come into contact in parallel positions, the Becke test indicates that the larger crystals possess a slightly higher refractive index than their smaller neighbours. ‘This evidence, implying that crystals of the second erop are slightly richer in the albite-molecule than those of the first, is interesting, in view of the fact that the zeolites of the amyegdales are predominantly lime-bearing varieties. Some of the larger crystals of labradorite—especially in the neighbourhood of well-filled amygdales—are replaced by chalcedony, and veined and streaked with chlorite. Signs of weathering, however, are rarely met with. Augite is present in colourless or yellowish-green grains and agereoates, the colour tending to deepen towards the outer borders, which are often ill-defined and merge imperceptibly into the surrounding glass. In the dark reddish- or greenish-brown glass that occupies the interstices of the crystal network, minute dendritic sprays of magnetite and innumerable globulites are present, particularly around the borders of ill-defined felspar- and augite-crystals. It is worthy of notice that the glass around the larger vesicles is frequently very dense and undifferentiated. ‘This effect is doubtless due to the local cooling set up by the sudden expansion of the volatile constituents of the erupted magma. Only the amygdale minerals remain to be considered. The vesicles in which they occur’ are of two strongly-contrasted types :— (a) Well-shaped and smoothly-lined spaces easily visible to the naked eye in thin sections (Pl. XXI, fig. 4), and in hand-specimens extending in dimensions up to the size of the closed fist. (6) Minute branching interspaces with obscure and irregular borders, connecting sometimes with the larger vesicles, and sometimes with small cracks and fissures that cut abruptly across the texture of the rock. The minerals found in the amygdales are chlorite, heulandite,. laumontite, natrolite, opal, chalcedony, quartz, calcite, and gypsum. In good exposures, such as that on the Mochelia road, it can be seen almost at a glance that zeolites are by far the most abundant of the amygdale minerals, and that of the zeolites heulandite is the chief member. In the associated surface-débris, the zeolites must soon fall a prey to decomposition, and consequently forms of silica greatly predominate, giving the erroneous impression that the parent amygdales were mainly siliceous. In hand-specimens the order of infilling is found to be (1) chlorite (which appears also in later stages), °(2) zeolites, and (3) silica. Palagonite sometimes occurs in minute vesicles ; but, with the exception that it is occasionally replaced by chalcedony, it is unaccompanied by zeolites or other minerals. Chlorite is the usual lining material, and in vermicular aggregates it is often present in Remlandice and chalcedony. 252 DR. Av HOLMES ON THE TERTIARY [ vol. Ixxii, Heulandite (CaA1,8i,0,,5H,0), with its bright cleavage- surfaces and pearly lustre, occurs in great abundance throughout the area, from Mokambo Bay to the Sanhuti River. Cleavage- flakes give good biaxial figures, which are optically positive. The refractive index is 1°5 and the extinction-angle practically straight. Heulandite is frequently followed (for imstance in 125a) by downy sheaves and radiating granules of yellow laumontite (CaAl,8i,0,,4H,0). Individual er ystals are so small that a quali- tative analy sis was necessary to complete the identification. The mineral gelatinizes with hy drochlorie acid, and contains alumina, iron-oxide, lime, and much water. The aoheciire index averages 1°52. In one specimen divergent aggregates of stilbite were found on heulandite, and were distinguished from the latter by the larger extinction-angle, and by the fact that cleavage-flakes do not show a complete biaxial figure. All the zeolites were carefully tested for soda, but of the three already mentioned it was detected only in stilbite; natrolite (Na,A1,8i1,0,,+2H,O), on the other hand, is free from lime. Slender needle-like erystals of this mineral are found on heulandite, andin No. 1366 radiating sheaves of natrolite complete the filling of vesicles already lined with heulandite. The refractive index is 1°47, and the extinction is straight. Heulandite is frequently replaced by opal, chaleedony, and quartz, and in some sections (as, for example, in 136, Pl XXI, fig. 4) it is clear that the change began at the walls of the amygdales and developed inwards. There is no evidence that the other zeolites have been similarly replaced, but both laumontite and natrolite are sometimes succeeded by chalcedony and quartz. Calcite is exceedingly rare, and it has only been found in one specimen (No. 1366), in which it was deposited in minute crystals on quartz. Gypsum is also rare, but was met with in some of the sections in narrow veins cutting through everything from heulandite to quartz. The order of deposition in the vesicles may now be stated for a number of typical cases. Chlorite is omitted, as it is liable to appear at any or every stage up to quartz. No. 125 a. No. 133. No. 136 a. No. 186 b. No. 125 b. Heulandite. Heulandite. Heulandite. Heulandite. Laumontite. Stilbite. Natrolite, _ Opal. Opal. Chalcedony. Chalcedony. Chalcedony. Quartz. Quartz. Quartz. Gypsum. Calcite. Gypsum. The most striking feature of this suite of minerals is the pre- dominance of lime and silica, and the rarity of alkali and carbonate minerals. The sequence in No. 1364 is significant, for natrolite free from lime follows heulandite free from soda, while finally calcite appears. In the Sanhuti basalt (p. 241). thomsonite rich in soda is followed by calcite. In the phonolite from the Sanhuti district, natrolite is followed by calcite. There is thus a notable part. 3 | VOLCANIC ROCKS OF MOZAMBIQUE. 253 difference, which may be of great significance, between: the vesicle- minerals of the Sanhuti lavas and those of the amygdaloids. On the one hand, is a series of lavas eluding types rich in alkalies in which soda-zeolites and calcite occupy the vesicles; on the other, is a series of typically calc-alkali lavas in which the vesicles are occupied by lime-zeolites and forms of silica. These two associations, which can be matched in other regions, seem to indicate that among the volatile fluxes of the alkali series carbon dioxide was a dominant member, whereas among those of the cale-alkali series water was the chief constituent, carbon dioxide being absent, or of minor importance. The order of infilling is closely parallel to that determined in many other localities where the same minerals are represented. As long ago as 1849 Breithaupt! recognized the sequence from heulandite to stilbite (given in the original nomenclature as stilbit to desmin). In the Watching basalt, New Jersey, C. N. Fenner? shows that heulandite may be Fallioned by laumontite, stilbite, or natrolite. At the well-known locality of Carnmoney Neck (Antrim), the order of infilling is given by J. Strachan? as (a) hullite and dellesite, (4) zeolites, (c) chalcedony and quartz. The circulating solutions of the Mozambique amygdaloids appear to have ‘become saturated in alkaline constituents only towards the end of the period of zeolite deposition. C. Deelter has shown * that natrolite is deposited from solutions of soda, alumina, silica, and water at temperatures below 190° C. Above that temperature analeime is formed. The presence of gypsum among the end-products gives another point on the temperature- -seale, Van’t Hoff® investigated the relations of gypsum and anhydrite, and found that the latter mineral is deposited ia place of gypsum if the temperature exceeds 63°5° C. In the presence of alkali salts the temperature of the transition-point is lowered. The stage of silica deposition falls between those of natrolite and gypsum, and at such low temperatures it is difficult to understand why the silica was not all deposited as opal, unless it is assumed that alkalies were present. G. Spezia ® has shown that opal is deposited at moderately low temperatures from siliceous solutions, but that, when alkalies are present, quartz forms instead. Opal already deposited may be transformed into quartz by the action of a hot solution of sodium silicate. The order commonly observed in siliceous amygd: ; seems, there- fore, to imply a growing concentration of soda cin the percolating waters. Interesting in this connection is the composition of the hot springs at Nashologoto, for they may be regarded as a feeble ‘Die Paragenesis der Mineralien’ Freiberg, 1849, p. 105. Ann. N.Y. Acad. Sei. vol. xx (1910) p. 169. Proc. Belfast Nat. Field Club, vol. i (1906) Dao. vil & vill. Tschermak’s Min. Petr. Mitth. vol. xxv (1906) pp. 99, 102.. Zeitschr. Phys. Chem. vol. xlv (1903) p. 257. 6 Atti Accad. Sci. Torino, vol. xxxiii (1898) pp. 289, 876, & Journ, Chem. Soe. vol. Ixxvi, pt. 2 (1899) p. 300 (abstract). Oo fp &S we 254: DR. A. HOLMES ON THE TERTIARY [ vol. Ixxii, solution containing soda and silica in water charged with carbon- dioxide. In explaining the origin of amygdale minerals, appeal has been made to two chief processes. By some writers they have been regarded as decomposition-products due to weathering; by others they are regarded as hydrothermal products of the final stages of consolidation. 'The view is now gaining favour that these minerals represent, not the first stage of destruction of the lavas in which they occur, but the last stage in their construction.! C. N. Fenner? takes an intermediate position: for, although he concludes that the vesicles of the Watchung basalt were filled while it cooled, he thinks that the water concerned in the process was drawn from the underlying sediments. In Mozambique the amygdale minerals are clearly i the result of weathering. In the freshly exposed surfaces on the Mochela road and on the Ibrahimo road, weathering products are typically absent. The felspars are generally fresh, and, where altered, they are replaced by chalcedony and penetrated by streaks of chlorite—a change clearly to be correlated with the repiacement of heulandite by silica and chlorite. If due to atmo- spheric weathering, the distribution of zeolites and agate should be uniform: on the contrary, it is very irregular. In some localities the basalts are vesicular and free from infillings; in others the walls of the vesicles are lined toe a varying depth, while along certain belts the vesicles are all tightly packed. The view suggested by Fenner cannot be adopted, for the amygdales occur as “abundantly when the basalts he on a platform of ‘granite and gneiss as when they cover a sedimentary basement. There seem oh be, however, no “objections to regarding the minerals and the water that accompanied them as the last products of consoli- dation. The regular sequence which, so far as can be seen, is one of decreasing temperature and of increasing concentration of soda and silica; the instability of heulandite towards silica; the absence of zeolites from the sills and dykes (because the material that would have formed zeolites was enabled to crystallize as plagio- clase); the general absence of calcite, limonite, and other weathering products ; the conspicuous freshness of the pyrogenetic minerals ; “dl finally, the occurrence of hot springs on the border of the volcanic field—all favour the hydrothermal or late magmatic origin of the amygdale minerals. 1 For a recent discussion and references to many other localities, see W. F. P. Mclantock ‘ On the Zeolites & Associated Minerals from the Tertiary Lavas around Ben More, Mull’ Trans. Roy. Soc. Edin. vol. li, pt. 1 (1915) p. 13. 2 Ann, N.Y. Acad. Sei. vol. xx (1910) p. 106. part 3} VOLCANIC ROCKS OF MOZAMBIQUE. 255 Chemical Composition of the Sokoto Vesicular Basalt (No. 132). Molecular Mineral Composition Percentages. Proportions. (Norm). SiOke ea: 56°42 943 @) ant Zieee ere 8:97 AUS OR cea: 17°33 170 Orthoclase ... .. 9°45 | Salic eR OR parr 2°78 018 Albite ...,.....-:. 24°89 ( =73:19. HeOne es 3°23 044 Anorthite ...... 29°88 MgO ....... 4°67 calli : CAO Rae 8°61 154 Diopside ......... 10°30 Na,O ...... 2°95 047 Hypersthene ... 10°09 | Femic KO ay 1°58 017 Maenetite ...... 4°18 ( =25:08. H,O+...... 1:32 wee Ilmenite ......... 0°46 HO es 0°44 oe TOMI Mima: 0°25 003 No tells pee, cee 98°22 Min Oo 0:17 002 Waterine en tee Lae otal 99°75 WOW cooven ce 99°98 Class II...... Persalane. Ra=0°94x 10-1" orms. per grm. Order 4...... Austrare. Specific gravity =2°43. Rang 3-4... Tonalase-Bandase. Subrang 4... Tonalose-Bandose. Comparing this analysis with that of the Sokoto sill, the most striking difference lies in the excess of silica in the vesicular lava. This is clearly referable to the presence of chalcedony in the microscopic vesicles of the latter. However, even if free silica be subtracted from the analysis, and the remaining figures recalcu- lated to 100, there still remain noteworthy differences. The lava contains a smaller relative proportion of iron-oxides and magnesia than the sill, and the compositions of the normative plagioclase in the two rocks are :— Sooo Solo onccoesossaacvonn sacodooat An,, Sokoto vesicular lava ............... An. It thus appears that the lava is deficient in the constituents of chlorite and of lime-zeolites. The absence of these amygdale minerals from the rock analysed and their abundance in the neighbouring lavas confirm this result. If we regard the Sokoto sill as representing an approximation to the parent magma of the surrounding flows of vesicular and amygdaloidal basalt, then we may follow the progress of differentiation (controlled by the presence of abundant water and perhaps of other volatile fluxes) and reconstruction through various stages that may be represented as follows :— 256 DR. A. HOLMES ON THE TERTIARY [vol. Ixxu, Water and other volatile fluxes. We Vf ResrpuAuL Lime-zeolites and chlorite, / Liquors. Soda-zeolites and chlorite, - \ Silica and chlorite. Sx SoKoTo . AMYGDALOIDAL (Containing BASALT. : BasauT. anyor all of XS yA the amygdale N VA minerals. ) VESICULAR (Relatively enriched in | Wa BASALT. soda.) XVI. Hornprenpu-Anpesire.. (Pl. XXI, fig. 5.) The specimen (No. 126) was collected from a small dyke penetrating a flow of amygdaloid on the south side of the Monapo, about 25 miles north-west of Mochelia. In the hand- specimen white phenocrysts of felspar (averaging 4 mm. x 2 mm.) are seen distributed in a dark-grey ground-mass. Smaller crystals of hornblende and flakes of biotite can also be disting: wished. The rock is compact, but a few very small amygdales containing well-crystallized quartz are present. Under the microscope the chief minerals.seen and their approxt- mate proportions by weight, determined by the Rosiwal method, are as follows :— Estimated Percentage Minerals. by weight. Andesine and inelusions........................ 26 h 2 a Hormblendepanen see eee ee 9 Phenocrysts ormblende ..........0. 01. i. IBS TOLLS a Ra ere ne EE Re titre ene ec: 5 TRIB EHOYCIENSE) oo ccoonn ccocosonn =) See BPAY cla not determined l Ground-mass SERS i Re RAE LOANS separately =(H0), BIO tIGC Wh Beno co yan eee | Eornib lend cyeeenene ee rer yj [otal] Wee 100 Part of the rock was crushed, and by means of bromoform, subsequently diluted with benzine, three groups of minerals were separated :— (a) Hornblende and biotite sank in bromoform; in Thoulet solution diluted with water, the average specific gravity of the crop was found to be 2°97. (b) Plagioclase sank in bromoform diluted with benzine at 2°67. (c) The remaining material—mainly ground-mass—was collected for analysis. Its average specific gravity was found to be 2°61. Mineral Growp. Percentage by weight. Specific Gravity. Andesine and inclusions ...... 26 2°67 Hornblende and biotite ...... 14 2:97 GrOUNG=MaASSeeeceeeeeeeeeeeee 60 2°61 part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 257 With the data thus ascertained the specific gravity of the rock is calculated to be 2°67. The value actually found was 2°68. The phenocrysts of plagioclase average andesine in com- position. They are often strongly zoned, and the interior is, as is usual, richer in the anorthite molecule than the outer zones. Inclusions are rare, except for occasional wisps of chlorite. Alteration is common to a microcrystalline aggregation in which kaolin, perhaps sericite, epidote, and a very little calcite can be distinguished. Green hornblende is present in good idiomorphic erystals modified by slight resorption. Patches of ground-mass, shreds of biotite, grains of magnetite, and small crystals of apatite are found as inclusions, and in one case a rounded crystal of zircon. Biotite also exhibits slight resorption, and its fibrous texture suggests a certain degree of alteration. Magnetite and apatite are present sparingly as inclusions. The ground-mass contains. small laths of andesine or oligoclase surrounded by fan-like inter- growths of colourless minerals having refractive indices respectively above and below that of Canada balsam. The minerals in question are probably quartz and orthoclase: for the intergrowths exhibit an approach to microgranophyric texture, and the chemical composition ,of the ground-mass (tabulated below) leads to the same conclusion. Wisps of biotite largely altered to chlorite are distributed through the ground-mass, and irregular grains of horn- blende are also present. Chemical Composition (Hornblende-Andesite). An analysis of part of specimen 126 was made, with the following results :— Molecular ' Mineral Composition Percentages. Proportions. (Norm). SiO Meee cua teae insti: 60°76 1:013 @uartzeeeereen 14°16 JO) ea roeepepatneasc 16°81 165 Orthoclase ...... 14°46 Salic Bek Oar. ty 2:07 013 INO s-cooavoscoe 27°77 ( =80°30.. De O) tee nantes ae 2°34 032 Anorthite ...... 23°91 Mio Oba a soto 3:27 ‘082 (ORNO 4 Mumewanenerana 6°57 117 Diopside......... 6°89 NETO) eae ten scngeay 332 053 Hypersthene ... 7:28 | Femic LOE, saelale SHAUN 2°42 026 Magnetite ...... 3°02 ( =17:80. Lossabovyell10°C. 1:19 ee Ilmenite ......... 0-61 Loss at 110° C.... 0°48 es - FIA Oe yeereyeiaiat Sacer 0°31 004 Motaliieree ae 98:10 Miia Oy ase Bee aerate hac 0°22 004 Loss on ignition 1°67 AOpen sanucea vec ees 99°76 Ro tpaileerey ner, 99:77 Class Il 5500 Persalane. : : ‘ Order 4 ...... Austrare. SISEEUEE BUNTY SS 208. Rangr3 =. fu. Tonalase. Subrang 4... Tonalose. The analyses of some other andesites from other Hast African localities are quoted below (B, C, D, E) for comparison with the 258 DR. A. HOLMES ON THE TERTIARY [vol. xxi, Mozambique example. The high silica-percentage of the latter is probably due to the presence in the specimen analysed of small quartz-bearing amygdales, and also to the presence of quartz in the microgranophyrie groundmass :— | TS Aa Sarah Rosa s-10) BAS gil Gen ciate eSiOpe eames 60-76 51°32| 56°65 51-97 5614 || 65-32| 66-40 62-07, | Al,O, ...... 16°81 | 16°62) 22:11 | 16:21 | 17-81 16:08 | 1291 17-15 | eRORneeee USP oi Peasy all a SUZ 69 | een) PoSrauleea | lweo | 234] 228) 382l" gsi] 467 |. 282). 256! ai | MgO ...... | 3°27 | 5:36) 3°42 ee 415 2°13] 0:53{ 2:39) @a@) 2.) (657 19°62) (66a) 7-67) |m6-80 3°47| 2°32 4-84. |Na,O......| 332 |: 215] 186) 3-41) 359 | 433] 3:30). 4:75 \SKe OM ee | 242 | 2:96| 4-10 75. 1°63 | 3°62) 3-09)" 215 EON Ir 1:67 | 2:60| 2-20] 1:97) 1:16 | 1:06] 4:25! 118 TOs eee | 031 | usd) mid, |, 0:22)" 1:28 0-17| 0-54, 0-21 | Mn0 ... .. 0:22 | 0:55| 0:16/ n.d.| 0:08 | n.d.| 0-09) 0:26] EBS Oo aa mph s (ORAS eae, doen, Gh We (OA ned,)| O1Giie stra Motel sider 99°76 100-71 100-48 99-76 | 100:13! | 99-07 100-02 100-44. 1 Including 0-01 per cent. SrO and 0:04 per cent. BaO. . Hornblende-andesite, Monapo River (an. Holmes). Olivine-augite-andesite,! Mount Meru. Amphibole-biotite-andesite,? Mount Meru. . Andesite, Kordofan? (an. Linck). Andesite, near the Lucalla River, Angola’ (an. Harwood). Base of hornblende-andesite, Monapo River (an. Holmes). Toscanite, Indulawane Hill, Zululand (an. Prior). . Quartz-porphyry, Kadero, Kordofan® (an. Linck). Tass van > Chemical Composition of the Gronnd-mass. The heavy-liquid separation of the ground-mass already men- tioned was not quite clean, for a greater proportion of biotite and andesine accompanied the residue than appeared to be visible in thin section. However, it was thought advisable to analyse the material thus obtained, on account of the light that it might throw on problems of differentiation.’ The ground-mass would be expected to have the composition of a latite or rhyolite, and although in Mozambique no such rocks have been discovered. in association with the basalts, corresponding dyke-rocks are known to exist at Sinjal in the extreme west of Zambesia, among the volcanic rocks south of the Zambesi and the Lebombo, and in Uganda and ‘Kordofan. * The separated material was air-dried at 110° C. Owing to 1 B. Pinkert, Féldt. K6zl. vol. xxxvii (1907) p. 292. 2 Id. ibid. p. 296. 3 Neues Jahrb. Beilage-Band xvii (1903) p. 428. 4 A. Holmes & H. F. Harwood, Min. Mag. vol. xvii (1916) p. 69. > See Sir Jethro Teall, Q. J. G.S. vol. lvii (1901) p. Ixxxii, for the applica- tion of this method to the rocks of the Cheviots. part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 259 the lack of material, ferrous oxide was not separately determined, nor were several important minor constituents. The analysis gives a low summation, but is sufficiently full for the immediate purpose. Molecular Mineral Composition Percentages. Proportions. (Norm). Si@seaeis cr 65°32 1:089 Quartz 92 -yes 16°86 AIO esan 16:08 _ 158 Orthoclase ...... 21°13 Salic FeO, ...... 2°89 ‘018 Albite ............ 36°15 ( =88°32. HeOsn tees n.d. ve Anorthite ...... 14:18 MgO ...... 2:13 053 Diopside......... 2°38 CaO rn hee: 3°47 "062 Hypersthene ... 4°20 | Femic WE O) Sasoss 4°33 ‘069 Hematite ...... 2°72 ( =9°60. Ke Oren as. 3°62 038 Ilmenite ......... 0°30 Loss on — ignition... 1:06 aes Motaloeceenee 97°92 ANNO) lato adabo 0-17 002 Loss onignition 1:06 Total ...... 99:07 Noy Eee ace sean 98:98 Classen Persalane. a : : Order 4 ...... Brittanare. Specific gravity 2°61. Raney) ween Toscanase. Subrang 4... Lassenose. Analyses of African rocks of similar composition and associated in the field with andesites are quoted on the preceding page (G & H). XVII. Pyroxense-Anvesire. (PI. XXI, fig. 6.) Pebbles of pyroxene-andesite (No. 144:) were found by Mr. Way- land in the bed of the Monapo River and on the alluvial flats to the south, at distances of 3 to 4 miles from the mouth of the viver. The rock is dark grey in colour, and is speckled with phenocrysts of felspar, some of which are transparent, others being opaque. It lacks the prevailing reddish tints of the basalts, but weathers to a dark brown. The rock is singularly compact and free from vesicles—a feature suggesting that, like the neighbour- ing hornblende-andesite, it may be derived from a dyke. No known lava from the Mozambique volcanic area is entirely free from vesicles or amygdales. The specific gravity is 2°72. Under the microscope the rock presents no unusual features. It consists essentially of phenocrysts of andesine, augite, and hypersthene, and occasional pseudomorphs of serpentine after olivine, embedded in a fine-grained ground-mass of hyalopilitic texture, in which tiny laths of oligoclase, shreds of biotite, grains of pyroxene and iron-ores, together with a little interstitial glass, are the chief constituents. The phenocrysts of plagioclase are generally clear glassy andesine, with characteristic twinning and zoning, and an average refractive index of about 1556. The opaque felspars of the hand- specimen owe their colour to the presence of large inclusions of 260 DR. A. HOLMES ON THE TERTIARY [vol. lxxii, biscuit-coloured glass or of innumerable blebs that suggest a frozen emulsion of felspar and glass. Large pyroxenes are also wholly or partly included by the larger felspars, and apatite, magnetite, and serpentine are accessory inclusions. The glass- filled felspars are surrounded by a clear mantle of oligoclase, which may represent a final period of growth simultaneous with that of the ground-mass. The average refractive index of the outer borders 1s approximately 1545; that of the laths in the ground-mass appears to be a httle less. Hypersthene and augite occur in somewhat rounded forms, having rarely retained their idiomorphic outlines. Hypersthene, which is beautifully pleochroic, frequently acts as a nucleus around which augite crystals are built. Both pyroxenes contain magnetite, serpentine, and glass as inclusions. Peripheral wisps of biotite occur around many of the rhombic pyroxenes, and also disseminated sparsely through! the ground-mass. In several cases the change from hypersthene to biotite is gradual and continuous, but against felspars the biotite always terminates sharply. _Serpen tine is not abundant, and is present as pseudomorphs after idiomorphie phenocrysts of olivine, and as confused aggregates in the ground- mass. It is of the golden- yellow variety found in the Tyas. Like hypersthene, it is bordered with wisps of biotite which have clearly grown at the expense either of the serpentine or of the olivine which preceded it. The potash-bearing molecules of the magma seem to have reacted with the ferromagnesian silicates already precipitated—hypersthene, olivine, or serpentine—in such a way as to produce biotite.! The order of crystallization is as follows :—apatite and magnetite, olivine, hypersthene, augite, andesine, oligoclase, biotite, and, finally, interstitial glass. Relationships and Origin of the Volcanic Rocks. XVIII. Distrreurion oF Stminar Lavas IN HKASTERN AFRICA. The outstanding feature of the volcanic district of Mozambique is the development of two strongly-contrasted series of rocks— one mainly composed of amygdaloidal basalts and including ande- sites, the other including alkali-trachytes and phonolites. Amygdaloidal basalts similar to those of Mozambique are found in many localities of South and Central Africa, but nowhere else south of Mozambique are they known to be of Tertiary age. At Sinjal, near the border of Nyasaland and Portuguese East Africa, Dr. A. R. Andrew and Mr. T. E. G. Bailey ? found basalts, associated with andesites and rhyolites, which they refer to the close of the Karoo Epoch. Dr. Prior has kindly allowed me to see the specimens from this locality (now deposited in the Natural History * N. L. Bowen, Journ. Geol. Chicago, Supplem. to vol. xxiii (1915) p. 46. * Q. J. G. S. vol. Ixvi (1910) p. 220. part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 261 Museum), and it is clear that in no essential respect do they differ from the Mozambique examples; even the amygdale minerals and the order of their deposition are the same. South of the Zambesi, basalts, largely amygdaloidal, are found associated with late Karoo sediments, flanking the plateau of erystalline rocks for hundreds of miles. In the north they have been mapped by Dr. E. O. Teale & Mr. R. C. Wilson.! Between the Pungwe and the Sabi (Save) rivers they were traced by Col. A. Freire d’Andrade.? Some 50 miles west of the mouth of the Busi, he collected melaphyres in which chlorite-lined amygdales containing zeolites and chaleedony and quartz are abundant. Similar flows of melaphyre occur in Lourengo Marques. I have examined specimens from the Komati River that appeared to be identical with the basalt of the Sokoto sill. The Lebombo amygdaloids of Swaziland and Zululand are associated with andesites and rhyolites, and some of these have been analysed.? Occurring farther from the coast, the Drakensbergen lavas,* the Bushveldt amygdaloid of the Tyransvaal,> the Tuli lavas of Southern Rhodesia,® the Batoka basalts of the Zambesi,’ the basalts of the Forest Vale between Bulawayo and the Zambesi,8 and the Negami amygdaloids of the Kalahari,? all show a marked likeness one to the other and to the Mozambique lavas. Towards the south the resemblances become less marked, though the differ- ences are limited to the incoming of olivine (generally serpen- tinized) among the pyrogenetic constituents, and of calcite among the amygdale minerals.!? Throughout this vast region, west and south-west of Mozambique, no alkali series of rocks is known to occur in association with the late Karoo amygdaloids (except doubtfully those of the Bushveldt complex, where the age of the intrusions that preceded the amygdaloids is still uncertain). It seems, however, worthy of remark that the igneous rocks that followed the Upper Karoo amygdaloids and the later dolerites of South Africa are generally melilite-basalt, alnoite, or kimberlite. These rocks occur in the Cape Province, the Orange Free State, the Transvaal, Southern Rhodesia, Northern Rhodesia, and Katanga. 1 Geogr. Journ. vol. xlv (1915) p. 29. 2 P. Choffat, ‘Contributions 4 la Connaissance Géologique des Colonies Portugaises d’ Afrique’ Comm. Geol. Portugal, ii (1905) p. 33. 3 J. McC. Henderson, Trans. Geol. Soc. 8. Africa, vol. xii (1910) p. 24. 4 A. W. Rogers & A. L. Du Toit, ‘The Geology of Cape Colony’ 1909, . 225, F > #. T. Mellor, Ann. Rep. Geol. Surv. Transvaal, 1901, p. 31. 6 A. J. C. Molyneux, Q. J. G. S. vol. lix (1903) p. 266. 7 G. W. Lamplugh, ibid. vol. lxili (1907) p. 162. 8 F. P. Mennell, ibid. vol. lxvi (1910) p. 369 ; and Spec. Rep. 2, Rhodesia Museum, 1904, p. 15. ° S. Passarge, Zeitschr. Gesellsch. fiir Erdkunde, Berlin (1904) no. 3, .179. Pi A. W. Rogers & A. L. Du Toit, ‘The Geology of Cape Colony’ 1909, _ p. 226; see also P. A. Wagner, ‘The Diamond-Fields of South Africa’ 1914, p. 99. Q, J.G.S. No. 287. U 262 DR. A. HOLMES ON THE TERTIARY [vol. lxxii, Melilite-basalt also occurs among the alkali rocks of the Great Rift Valley, where it appears to stand towards them in a comple- mentary relation, as a mafic end-product of differentiation specially rich in lime and magnesia. In South Africa the ‘alkaline affinities’ of melilite-basalt and kimberlite have been noted,! and it is interesting to find that the amygdale minerals of these rocks include zeolites and calcite, but never chalcedony or quartz.” Although the lavas of Mozambique, and of the greater part of Africa south of that latitude, are predominantly amygdaloidal basalts characterized by siliceous and lime-zeolite amygdales, the lavas known in the north are not commonly of this type, except in Abyssinia. The majority of the lavas belong to an alkali series (accompanied by mafic complementary types), to which those of Mozambique bear a close resemblance.? The distribution of the various types has been well summarized by Prof. J. P. Iddings in his ‘ Igneous Rocks,’ vol. 1 (1913) p. 576 et seg., with references to the literature, and therefore the details need not be repeated here. It will be sufficient to draw attention, as Rosiwal did many years ago,* to the frequent association of a calc-alkal series (including amygdaloidal basalts, andesites, and rhyolites) with a more strongly-developed alkali series. North of the Songwe River in German Kast Africa, Bornhardt found basalts, andesites, and trachytes, while south of the river, in the extreme north of Nyasaland, A. R. Andrew & T. E. G. Bailey discovered tuffs containing trachyte and phonolite.» In Masai- land, along the Uganda Railway towards Victoria Nyanza, in the Kavirondo country ® adjoining the Nyanza, in the islands of Lake Rudolf, in Kordofan, and especially in Abyssinia and Somaliland,’ the two series are well represented. Wherever the succession has been noted® the alkali series, often accompanied by normal basalts, is the later. In the islands of Lake Rudolf, in Italian Somaliland, and in Abyssinia, the amygdale minerals of the earlier series are chiefly quartz, chalcedony, and opal, accompanied by heulandite and other lime-zeolites. Blanford draws attention to the striking similarity between the amygdaloids of Abyssinia and those of Southern Arabia and the Deccan. In each of the two last-named regions an alkali series is also present.? No detailed petrographic account of the Deccan traps has yet been published, 1 Pp, A. Wagner, ‘The Diamond-Fields of South Africa’ 1914, p. 106. 2 A. W. Rogers & A. L. Du Toit,‘ The Geology of Cape Colony’ 1909, p. 348. 3 Especially in the presence of such soda-amphiboles as katoforite and cossyrite. 4 Denkschr. K. Akad. Wissensch. Wien, vol. lviii (1891) p. 465. > Q. J. G. S. vol. lxv (1910) p. 222. 6 BF. Oswald, Q. J. G. S. vol. Ixx (1914) p. 140 et seq. 7 W. T. Blanford, ‘Observations on the Geology & Zoology of Abyssinia’ London, 1870, p. 180. 5 As, for example, in Abyssinia, Lake Rudolf, and the Kavirondo country. 3 See J. W. Evans, Q. J. G. S. vol. Ivii (1901) p. 38. In Kathiawar the nepheline-bearing rocks appear to be intrusive into the Deccan traps, and therefore of later age. j part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 263 but several specimens in my possession bear the most detailed resemblance to the basaltic dyke-rocks of Mozambique. In the adjoining table, the distribution of the two series of lavas is tabulated in as complete a form as the present state } ‘ Calc-Alkali’ Series. ‘ Alkali’ Series. co | nD AB oe Aes District. ae at : g 2 | of on reales = [oe |e ay a P| oa | wm | © wr) =a m3 See, for example, J. Joly, Phil. Mag. vol. xxiv (1912) p. 694; and A. Holmes, Sci. Progress, no. 33 (1914) p. 15. 2 Foid. p. 17. 3 Geol. Mag. ser. 6, vol. i (1915) pp. 68-69. Silbid. py lulls > Ibid. p. 67. part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 27 there were a considerable relief of the pressure appropriate to such depths. In this connexion, the boundary-fault along which the lavas are aligned is of particular interest, for it seems to mark a zone along which pressure actually was relieved to an extent and depth sufficient to promote fusion. The boundary-fault system is probably a consequence of isostatic readjustment between the uplifted mainland and the sunken region of Mozambique Channel. Long denudation of the crystal- line plateau should, on the hypothesis of isostasy, naturally lead to its uplift, and that such a movement has taken place is demon- strated by the presence of a coastal belt of Cretaceous and Tertiary sediments, and by the existence of still more recent raised beaches at various levels along the coast (see fig. 5, p. 280). More- over, the heavy denudation of the crystalline rocks implies the removal of an important section of the ‘radioactive layer.’ This process therefore leads, like uplift, to the cooling and contraction of the lithosphere, and thus additional tension-stresses are developed that can only find relief in fissures and faults. Doubtless the actual formation of fissures was helped by regional movements of « tensional character, associated in a complementary way with the compressive movements of Tertiary mountain-building elsewhere. In conclusion, I wish to place on record the unfailing courtesy and hospitahty with which we were everywhere received by the _ Portuguese officials in Mozambique, who were always ready to help us in every way that lay in their power. For permission to publish the geological observations made during the expedition, I owe my thanks to the Directors of the Memba Minerals Ltd. ; in particular, I have to thank them for placing at my disposal a large collection of rock-specimens, including those collected in the Sanhuti-River district, and many of those collected by the staff of the Company during the 1910 season. From my friends and former colleagues Messrs. E. J. Starey, E. J. Wayland, and D. A. Wray, I have received notes and sketches referring to the districts in which they worked, and I desire to express my indebtedness to them for help thus afforded. In connexion with the investigation of the rocks in the laboratory, I wish to thank Prof. W. W. Watts for granting every possible facility for work at the Imperial College of Science & Technology; Prof. the Hon. R. J. Strutt for continuing to lend me his apparatus for determining radium in rocks and minerals; Prof. C. G. Cullis and Dr. J. W. Evans for examining many of my sections, and offering valuable suggestions ; and Dr. G. T. Prior and Lieut. W. Campbell-Smith for looking through my sections, and allowing me to compare the Mozambique lavas with specimens of similar rocks in the Natural History Museum from British East Africa, Abyssinia, Nyasaland, and other African localities. In making the analyses published in this aper I received considerable help from Dr. H. F. Harwood, Lieut. J. L. Harris, M.C., the late Lieut. G. Kirby, Lieut. A. McIntyre, and Mr. HE. Spencer, and for their assistance in an Q. J. G.8. No. 287. x 278 DR. A. HOLMES ON THE TERTIARY [ vol. Ixxii, arduous task I am particularly grateful. Finally, I owe my thanks to Mr. G. 8. Sweeting, who kindly cooperated with me in the preparation of the photomicrographs reproduced in Plates XX & XXI. EXPLANATION OF PLATES XX & XXI. PLATE XX. Fig. 1. Sélvsbergite (153) found between Miali and the Sanhuti River, Mozam- bique. Chief minerals: anorthoclase, cossyrite, and soda-pyroxenes. x 20. (See p. 233.) 2, Aigirine-trachyte (152) north of Sokoto Hill. Chief minerals : anortho- clase, cossyrite, and egirine. X 20. (See p. 235.) 3. Phonolite (150) near the Saahuti River. Chief minerals: anorthoclase, nepheline, natrolite, soda-pyroxenes, and amphiboles. The dark phenocryst is cossyrite surrounded by a green chlorite-like mineral. x 20. (See p. 236.) - 4. Basalt with ophitic patches (156) near the Sanhuti River. Chief minerals: labradorite and augite, with some palagonite in the ground- mass. xX 20. (See p. 241.) 5. Picrite-basalt (155) near the Sanhuti Ree. Chief minerals: purple augite and olivine; labradorite, purple augite, and giass in the ground-mass. xX 20. (See p. 244.) . Micro-vesicular basalt (128) from a sill at the western base of Sokoto Hill. Chief minerals: labradorite and augite, in a base of dark-brown glass. The vesicles are occupied by Chalcedony and chlorite. x 20. er) PuatTEe XXI. Fig. 1. Basalt (141) from the dyke at Mochelia. Chief minerals: labra- dorite and augite in a base of dark-brown glass. X36. (See p. 246.) . Amyedaloidal sandstone (112) from contact with the Mochelia dyke (text-fie. 5). Grains of quartz, orthoclase, and microcline are present in the sandstone, cemented by calcite which is replaced near the dyke hy opal. Inthe steam-holes opal, chalcedony,and quartz are present. x 12. (See p. 249.) : . Amygdaloidal basalt (124) south of the Monapo River. Chief minerals : labradorite and augite in a base of dark-brown glass. x 20. (See p. 250.) 4. Amygdale in amygdaloidal basalt (136) east of Murimatigri Quartel. Tilustrating the replacement of heulandite (dark areas) by chal- cedony. xX 20. (See p. 252.) 5. Hornblende-andesite (126) from a dyke penetrating amygdaloid south of the Monapo River. Chief minerals: andesine, hornblende, and biotite, in a ground-mass consisting of the same minerals, together with quartz and orthoclase. x 20... (See p. 256.) 6. Pyroxene-andesite (144) found near the Monapo River. Chief minerals : andesite, augite, and hypersthene, in a ground-mass of hyalopilitie texture. x 20. (See p. 259.) bo ee) _ Discussion. My. F. P. Mernnetr said that he had listened to the paper with the greater interest, in that he had spent more than ten years ip a neighbouring part of Africa. He was quite in accord with [November 23rd, 1917. | Quart. Journ. Geor. Soc. Vor. LXXII, Pr. XX. ik ae ie or at ARAL 7 AH. & 6.S.S,, Photomicro. Bemrose, Collo., Derby VOLCANIC ROCKS oF THe DISTRICT o— MOZAMBIQUE. Quart. Journ. Geot. Soc. Vor. LXXIl, Pi. XXI. AH. & G.S.S.,Photomicre. Bemrose, Collo., Derby VOLCANIC ROCKS oF tHe DISTRICT oF MOZAMBIQUE. ail if Jans part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 279 the Author, that alkaline and non-alkaline rocks were to be found along the African coast in intimate association. It was not clear, however, from what had been said, on what principle the Author separated the rocks that he put into his alkaline group from the others with which he had dealt, and it was not easy to see why they should be treated apart. It was interesting to note that. ‘although alkaline voleanic rocks were so often Sonndicn the coastal regions, there were no occurrences known in the inland districts of this part of Africa. Dr. J. W. Evans congratulated the Author on his presentation of the results of several years’ work on some very interesting rocks. He had not confined himself to simple descriptions of structures and minerals, but had endeavoured to ascertain what light the facts disclosed threw on the evolution of igneous rocks. especially those rich in alkalies. The speaker enquired whether there was any evidence that the alkali-rocks were later than those of normal types in this area. Mr. A. E. Kirsoy congratulated the Author on the paper. He asked whether there was any general linear arrangement of the dykes, and if so, whether the ‘lines were parallel w with the coast. He said that on both sides of Africa, wherever late volcanic rocks occurred, it seemed that they had been extruded along lines of fracture roughly parallel with the coast, or in Peon -zones inland, and that on the Gold Coast volcanic necks and plugs occurred in several places along the Volta River; but he had not found similar evidence in that colony outside of zones of fracture. The AvTHoR, in reply, thanked the Fellows for their kind reception of his paper, and stated that most of the points raised in the discussion were Exily dealt with in the paper itself. Answering Mr. Mennell~he said that the division of the lavas into—40_ series _we< based on analyses, varlation-diagrams, the composition of the amygdale-minerals, and in part on the sequence. so far as this was known. In reply to Dr. Evans, he was sorry to say that there was no definite field-evidence from which the relations of the two series could be deduced. Referring to Mr. Kitson’s remarks, he observed that, as in many other of the African coast- lands, the main faults were parallel to the prevailing trend of the coast, and most of the basaltic dykes followed the same direction. ADMISSION AND PRIVILEGES OF FELLOWS OF THE GEOLOGICAL SOCIETY OF LONDON. Every Candidate for admission as a Fellow must be proposed by three or more Fellows, who must sign a Certificate in hisfavour. The Proposer whose name stands first upon the Certificate must have a personal knowledge of the Candidate. Fellows on election pay an Admission Fee of Six Guineas. The Annual Contribu- tion paid by Fellows is Two Guineas, due on the 1st of January in every year, and payable in advance; but Fellows elected in November or December pay no Contribu- tion for the current year. The Annual Contribution may, at any time, be compounded for by a payment of Thirty-Five Pounds. The Fellows are entitled to receive gratuitously all the volumes or parts of volumes of the Quarterly Journal of the Society that may be published after their election, so long as their Annual Contributions are paid; and they may purchase any of the publications of the Society at a reduction of 25 per cent. under the selling-prices. — The Library is open on Week Days to the Fellows between the hours of 10 and 5 (except on Saturdays, when the hours are from 10 to1, and except during the fortnight commencing on the first Monday in September; see also next page). Under certain restrictions, Fellows are allowed to borrow books from the Library. Publications to be had of the Geological Society, Burlington House. QUARTERLY JOURNAL. (Vols. III to LXXI, inclusive.) Price to Fellows, 13s. 6d. each (Vols. XV, XXIII, XXX, and XXXIV to LXXI, 16s. 6d.), in cloth. GENERAL INDEX TO THE FIRST FIFTY VOLUMES OF THE QUARTERLY JOURNAL (1845-1894). Part I (A-La). Part II (La-Z), Price 5s. each. To Fellows 3s. 9d. each. [Postage 3d. | GROLOGICAL LITERATURE added to the Geological Society’s Library during the years ended December 31st, 1894-1912. Price 2s.each. To Fellows, 1s. 6d. each. [Postage 3d. | THE HISTORY OF THE GEOLOGICAL SOCIETY OF LONDON by H. B. Woopwagp, F.R.S. Price 7s. 6d. To Fellows, 6s. [Postage 6d. } THE CENTENARY OF THE GEOLOGICAL SOCIETY OF LONDON, recorded by W. W. Warts, F.R.S. Price 2s. net. [Postage 3d.] HUTTON’S ‘THEORY OF THE EARTH,’ Vol. Ill, edited by Sir ARCHIBALD GuIkix, D.C.L., F.R.S. Price 3s. 6d. To Fellows, 2s. [Postage 4d.] THE GEOLOGY OF NEW ZEALAND. Translated by Dr. C. F. FiscHER from the works of MM. HocusteTtTer & PETERMANN. With an Atlas of Six Maps. Fellows may purchase one Copy of this Book at 2s. Additional Copies will be charged 4s. [Postage 5d. ] 3 JUST OUT. VOLCANIC STUDIES In Many Lands.—Being reproductions of Photographs taken by the Lare TEMPEST ANDERSON, M.D., B.Sc. (Lond.). Second Series. The text by Prof. T. G. Bonney, F.R.S., Sc.D. Crown 4to. lds. net. Contains the results of further visits to the Lipari Islands, Vesuvius (after the outburst in 1906), and to Etna, together with photographs taken after the great eruptions of the Soufriére in St. Vincent and Mont Pelé in Martinique, which occurred in 1902. Dr. Anderson again visited these islands in 1908 and obtained illustrations to show the return of vegetation and other changes. The present volume gives photographs of some Mexican volcanoes, taken in 1908, and of others from Guatemala, New Zealand, Matavanu, in the Samoan Islands, and Kilauea. It concludes with a selection from the photographs obtained by Dr. Anderson on his last journey—that to Java, Krakatau, and the Philippine Islands—so as to form with its predecessor an exceptionally large collection of Volcanic Studies from all parts of the globe. JOHN MURRAY, Albemarle Street, LONDON, W. 1. CONTENTS. PAPERS READ. Page 9. Dr. W. R. Jones on the Secondary Stanniferous Deposits of the Kinta District (Plates ADL OKV) 522 a caus cee areas eee eae so arnch s eteewis seu ataee te 165 10. Prof. S. J. Shand on the Pseudotachylyte of Parijs, Orange Free State (Plates 2X ViTSXDRe) ences oe Ee eedocle eh ier tare ac eae gates. ete ieee meee 198 11. Dr. A. Holmes on the Tertiary Volcanic Rocks of the District of Mozam- bigue (Plates Kexe & XN) I Sere nie se esac oupty aaa eer sie ee ee ae OemeecirmEs 222 [The Editor of the Quarterly Journal is directed to make it known to the Public that the Authors alone are responsible for the facts and opinions contained in their respective Papers.| *,.* The Council request that all communications intended for publication by the Society shall be clearly and legibly written on one side of the paper only, with proper references, and in all respects in fit condition for being at once placed in the Printer’s hands. Unless this is done, it will be in the discretion of the Officers to return the communication to the Author for revision. The Library at the Apartments of the Society is open every Week Day from Ten o'clock until Five, except on Saturdays (when the hours are from 10 to 1), and except during the fortnight commencing on the first Monday in September, when the Library is closed for the purpose of cleaning. Vol. LXXII. No. 288. Parr 4. THE QUARTERLY JOURNAL OF THE GEOLOGICAL SOCIETY. EDITED BY THE PERMANENT SECRETARY. [With Three Plates, illustrating Dr. Stanley Smith’s Paper. | DrcEMBER 31st, 1917. LONDON : LONGMANS, GREEN, AND CO. PARIS :—CHARLES KLINCKSIECK, 11 RUE DE LILLE. SOLD AL8O AT THE APARTMENTS OF THE SOCIETY. Price Five Shillings. RRR RII I IPD IRIN GLI LL IIS II IEEE DO OOO PRADEEP ODE LIST OF THE OFFICERS AND COUNCIL OF THE GEOLOGICAL SOCIETY OF LONDON. Blected February 16th, Ue President. Alfred Harker, M.A., LL.D., F.R.S. Gice-Prestvents. R. Mountford Deeley, M.Inst.C.H. Edwin Tulley Newton, F.R.S. William Johnson Sollas, M.A., LL.D., Se.D., F.R.S. Arthur Smith Woodward, LL.D., F.R.S., F.L.S. Secretaries. Herbert Henry Thomas, M.A., Sc.D. Foreign Secretary. | Herbert Lapworth, D.Sc., M.Inst.C.E. Creasurer. Sir Archibald Geikie, O.M., K.C.B., D.C.L., | James Vincent Hlsden, D.Sc. LL.D., Se.D., F.R.S. COUNCIL. Charles William Andrews, D.Sc., F.R.S. Prof. John Cadman, C.M.G.,D.Sc., M.Inst. C.E. Prof. Charles Gilbert Cullis, D.Sc. Arthur Morley Davies, D.Sc., A.R.C.Sc. R. Mountford Deeley, M.Inst.C.H. James Vincent Elsden, D.Sc. Prof. Edmund Johnston Garwood, M.A., Se.D., F.R.S. Sir Archibald Geikie, O.M., K.C.B., D.C.L., LL.D., Se.D., F.R.S. Walcot Gibson, D.Sc. Alfred Harker, M.A., LL.D., F.B.S. Finlay Lorimer Kitchin, M.A., Ph.D. George William Tamplugh, F.R.S. Herbert Lapworth, D.Sc., M.Inat.C.E. Prof. John Edward Marr, M.A., Se.D., E.R.S. Edwin Tulley Newton, F.E.S. Richard Dixon Oldham, F.R.S. Robert Heron Rastall, M.A. Prof. Thomas Franklin Sibly, D.Sc. Prof. William Johnson Sollas, M.A.,UL.D., Se.D., F.R.S. Sir Jethro J. Harris Teall, M.A., D.Sc., LL.D., F.R.S. Herbert Henry Thomas, M.A., Se.D, Samuel Hazzledine Warren. Arthur Smith Woodward, LL.D., F.R.S., E.L.S. Permanent Secretary. L. L. Belinfante, M.Sc. Librarian. C. P. Chatwin. Clerk. M. St. John Hope. Assistant tn Library. Arthur Greig. STANDING PUBLICATION COMMITTEE. Dr. A. Harker, President. Dr. H. Lapworth. Dr. Herbert H. ieee Secretaries. Dr.\F. A. Bather Prof. J. Cadman. Prof. C. G. Cullis. Dr. J. V_ Hisden. Dz. W. Gibson. Dr. F. L. Kitchin. Mr. G. W. Lamplugh. Mr. HE. T. Newton. Mr. R. D. Oldham. Prof. T. F. Sibly. Prof. W. J. Sollas. Sir Jethro Teall. Dr. A. Smith Woodward ORDINARY MEETINGS OF THE GEOLOGICAL SOCIETY TO BE HELD AT BURLINGTON HOUSE. Szssron 1917-1918. 1918. Wednesday, January .. 9*—23* ss February (Anniversary, Friday, February 15th) . wee eee eee see neneee Se a 6*—20* [Business will commence at 5.30 p.m. precisely. | The asterisks denote the dates on which the Counoil will meet. part 3] VOLCANIC ROCKS OF MOZAMBIQUE. 279 the Author, that alkaline and non-alkaline rocks were to be found along the African coast in intimate association. It was not clear, however, from what had been said, on what principle the Author separated the rocks that he put into his alkaline group from the others with which he had dealt, and it was not easy to see why they should he treated apart. It was interesting to note that, although alkaline voleanic rocks were so often found in the coastal regions, there were no occurrences known in the inland districts of this part of Africa. Dr. J. W. Evans congratulated the Author on his presentation of the results of several years’ work on some very interesting rocks. He had not confined himself to simple descriptions of structures and minerals, but had endeavoured to ascertain what light the facts disclosed threw on the evolution of igneous rocks, especially those rich in alkalies. The speaker enquired whether there was any evidence that the alkali-rocks were later than those of normal types in this area. : Mr. A. E. Krrson congratulated the Author on the paper. He asked whether there was any general linear arrangement of the dykes, and if so, whether the lines were parallel with the coast. He said that on both sides of Africa, wherever late voleanic rocks occurred, it seemed that they had been extruded along lines of fracture roughly parallel with the coast, or in fracture-zones inland, and that on the Gold Coast voleanic necks and plugs occurred in several places along the Volta River; but he had not found similar evidence in that colony outside of zones of fracture. The AvurHor, in reply, thanked the Fellows for their kind reception of his paper, and stated that most of the points raised in the discussion were fully dealt with in the paper itself. Answering Mr. Mennell, he said that the division of the lavas into two series was based on analyses, variation-diagrams, the composition of the amygdale-minerals, and in part on the sequence, so far as this was known. In reply to Dr. Evans, he was sorry to say that there was no definite tield-evidence from which the relations of the two series could be deduced. Referring to Mr. Kitson’s remarks, he observed that, as in many other of the African coast- lands, the main faults were parallel to the prevailing trend of the coast, and most of the basaltic dykes followed the same direction. Q. J. G.S. No. 288, ———— 280 DR. STANLEY SMITH ON AULINA ROTIFORMIS, (vol. xxii, ~ 12. Avrivna ROTIFORMIS, gen. et sp. nov., PHILLIPSASTR.EA HENNAHT (Lonsdale), and OrronasTR#£.4, gen. nov. By SrantEy Suira, B.A., D.Sc., F.G.S. (Read November 8th, 1916.) [Puatres XXII-XXIV.] 7 CONTENTS. Page dBRL bai ndoleuneinlorabymandesnhocwaucsbcooatepetooceddiccanost aabane oncanb sd yoDe so a5. 280 Il. Definition of certain Terms employed ...... ........:::.:ecee eden 281 III. The Astreeiform Corallum ........ ee Sete A ohh ge cane ae 281 IN EI ACUUC SECRETE cane aun catnabaosphbdnnaboncanesscaracunenwevescagas0tpoan. 284 Ai, HISTO Is e313 Oi “soonSesanaco sdesnd =JoSec ono nnbar> sods cedisbancuococbasue 290 VI. Orionastred, GON. NOV. .....6c.csceccscee cece etc e ete eet ene tee neneet ees 294 VII. Notes on Cyathophyllum regium and on Koninckophyllum sp. 304 “ I. IyrRopvucTIoN. PHILLIPSASTRZA is a genus of Devonian Corals possessing - certain well-marked characters. The genus Ovionastrea has been established to include certain Carboniferous species very closely related to Lithostrotion, but which have been regarded by many writers as congeneric with those of Phillipsastrea. O. phillipst (McCoy) is commonly known as Ph. radiata (Martin). Aulina is a new genus, found at the horizon of the Millstone (Grit. One species only has been recognized: this has been confused hitherto with O. phillipsi, and hence recorded as Ph. radiata. I vegard Phillipsastrea as the ancestor of Aulina, but do not consider that these are related to Orionastrea. All three genera, however, are colonial in habit, and possess a similar type of corallum: namely, that in which the individual corallites have lost their epitheca, and consequently are united by their dissepiments—a type of colony which may be termed ‘astreiform.’! The foregoing statements explain my reasons for including the history of the name Phillipsastrea and the description of its genotype in a communication primarily concerned with Car- boniferous genera. I am deeply indebted to the late Prof. T. MecKenny Hughes, to Dr. A. Smith Woodward, Dr.’ F. L. Kitchin, Dr. W. G. Lee, and Mr. Peter MacNair for the loan of material (including various type-specimens) preserved in the Sedgwick Museum, 1 The term is frequently used by Edwards & Haime in their well-known Monograph, as also by other writers, to indicate the type of corallum in which the septa of the corallites are confluent, and does not in any way imply rela- tionship to the Astreide. I here employ the word in a slightly more extended sense, so as to include any Madreporarian colony (Rugose or Aporose) in which the corallites are united as above described. : , io y part 4] PAILLIPSASTR 4A HENNAHI, AND ORIONASTR4A. 281 Cambridge, the British Museum (Natural History), the Museum of Practical Geology, the collection of the Geological Survey of Scotland, and the Glasgow Museum & Art Gallery; and to Mx. Charles Edmonds, Prof. E. J. Garwood, and Prof. T. F. Sibly for use of the specimens that they had collected. My thanks and acknowledgment are also cordially extended to Dr. F. A. Bather, Mr. W. D. Lang, and Mr. C. D. Sherborn for their ever-ready advice and kindly assistance. IL. DEFINITION OF CERTAIN TERMS EMPLOYED. On account of the unfortunate lack of uniformity in the nomenclature of Corals, it seems advisable, before proceeding farther, to define clearly certain terms here employed. I denote the skeleton of a single individual, whether solitary or member of a colony, by the term ‘corallite,’ and use the term ‘corallum’ to mean the skeleton of the whole colony. In the ease of solitary or ‘simple’ corals the terms are synonymous. I divide the Rugose corallite into two regions: (1) the intrathecal region, which is built up of tabule or of tissue ontogenetically derived from tabule, and (2) the surrounding extrathecal tissue, which is built up of dissepiments. The junction of the tabular with the dissepimental tissue constitutes an annular wall—the theca. The corallite is usually clothed with an outer mural investment—the epitheca, the nature and develop- ment of which I have discussed in a previous publication.! The terminology applied to the septa will be found in a paper by Mr. R. G. Carruthers,? or in my own paper ‘On the Genus Aulophyllum. ® IIL. Tue Astr#1rorm CoraLLumM. Among the Rugose colonial corals there are certain genera, or certain species within a genus, in which the corallites have lost their epitheca, and are united by their dissepimental tissue.* The septa of adjacent corallites in the ‘astreiform’ colony tend to become confluent; but all stages of this development, from that in which it is incipient to that of perfect confluence, are to be found. In some cases, on the other hand, the. septa of one corallite do not extend to those of another, and leave an inter- vening space to the sole occupation of the dissepiments. Astreiform colonies make their appearance in widely divergent stocks and at different periods of time: being, it would seem, the ultimate terms in a progressive deve- lopment along a well-defined line. The steps towards this end are :— 1Q. J. G.S. vol. lxxi (1915) pp. 228-29. 2 Ann. & Mag. Nat. Hist. ser. 7, vol. xviii (1906) pp. 356-63. 3 Q. J. G.S. vol. lxix (1913) pp. 60-62. 4 An epithecal growth external to the whole corallum may, nevertheless, be retained in the astreeiform colony. Nee 282 DR. STANLEY SMITH ON 4UL/NA ROTIFORMIS, (vol. lxxu, 1. The simple or solitary coral. 2. The fasciculate colony, in which the corallites are not contiguous or do not press on one another, and are therefore cylindrical in form, as (for example) Lithostrotion martini. 3. The basaltiform colony, in which the corallites are contiguous, and are prismatic through mutual pressure. Although the corallum is massive, the corallites are defined by an epitheca, as (for example) Lithostrotion basaltiforme. And, finally, 4, The astreiform colony: for instance, Orionastrea. The examples quoted are from a single lineage. The number of genera represented by the astreeiform type of colony among the Rugosa does not appear to be numerous at an period, although the individuals may be extremely abundant.! Strombodes is fairly common in the Upper Silurian ; Phillips- astrea is very plentiful in the Upper Devonian; while, among the British Carboniferous Rugosa, the only examples of astreeiform genera and species that have come under my notice are Awlina and Orionastrea, Cyathophyllum regiwm Phillips, and a species of Koninckophyllam (or Diphyphyllum). All these forms are from the uppermost part of the Carboniferous Limestone Series—D, and higher beds. Certain Characters assumed by the Astizeiform Corallum in the Rugosa, and Factors that lave produced them. The coralla of Ovionastrea and of Aulina exhibit certain interesting characters—namely, a flattened form, an extensive development of extrathecal tissue, and epithecal growth external to the colony, even in cases where there is no trace of epitheca within it. (i) The form of the corallum is correlated with the mode of growth of the corallites. These may arise at any point on the surface of the corallum, but the region of most active growth is the margin. The peripheral corallites at first grow outwards, away from the centre of. the colony, and then bend upwards into a more ver teal position, as is shown in fig. 1 (p. 288). The extreme case of this form of colony is to be found, not in the Rugose but among the ‘tabulate’ corals—for example, Heliolites, where an almost horizontal base is often developed. 1 This type of corallum, while exceptional in the Rugosa, is characteristic of the massive Aporose corals—in fact, exceptions are rare.—but it must be remembered that the epitheecal covering to the skeleton (always present in Rugose simple corals) is often lacking in the case of Aporose simple corals. The mezandriform colony (due to repeated but incomplete fission so common among the Aporose stocks) is unknown in the Rugosa, and necessarily so, on the assumption that Rugose colonies increase by gemmation only, but never by fission. (See my notes on the subject in Q. J. G.S. vol. Ixxi, 1915, p. 233, and W. D. Lang, ‘Homceomorphy in Fossil Corals’ Proce. Geol. Assoc. vol. xxviii, 1917, p. 90.) : part 4] . PHILLIPSASTR.4:4 HENNAHI, AND ORIONASTR_EA. 283 Fig. 1.— Longitudinal section,of Orionastrea placenta (MeCoy), holotype of Sarcinula placenta, showing the corallite in its earliest stage. x 5 diameters. W. Tams photo. (ii) The development of dissepimental tissue can be shown to be progressively greater, as we follow the passage of dendroid and fasciculate into basaltiform, and these into astreeiform colonies. The case is not merely one of phylogenetic change, but may be observed to take place wherever Fig. 2.—Group of corallites two or more corallites of a den- of Lithostrotion martini. droid-fasciculate colony approach x 2 diameters. each other to the extent of con- tiguity. This is illustrated by the accompanying figure of a group of corallites of Lithostrotion martini Edwards & Haime. The phenomenon may be accoun- ted for, by supposing that two neighbouring polyps mutually ex- tend one. towards the other until they are in contact. This results in a stretching of the area of dis- sepiment-secreting tissue where the two touch, and in a_ consequent widening of the dissepimental area of the skeleton when secreted. (ii) The external epitheca, which stretches round the corallum as a continuous surface, usually displays pronounced rugosity. It differs neither in its origin nor in character from the epitheca that clothes a simple coral or an individual corallite ; yet its presence is often emphasized in descriptions of corals (especially if the corallum in question is discoid in form), and 284: DR. SLANLEY SMITH ON AULINA ROTIFOKMIS, (vol. lxxil, it is frequently defined as the ‘basal epitheca.’ It represents — merely the epitheca of the outer side of the peripheral corallites. * IV. Pururipsasrrma VOrbigny. Summary of Research. William Lonsdale (1840), in Sedgwick & Murchison’s ‘ Physical Structure of Devonshire, & on the Subdivisions & Geological Relations of its older Stratified Deposits, etc.’ Trans. Geol. Soe. ser. 2, vol. v, pt. 3, p. 697, pl. lvili, figs. 3, 34, & 346, described, under the name of ‘ Astrea (Stderastrea de Blainville) hennahir (sp. n.),’! a type of colonial coral since proved very character- istic of the Upper Devonian. The specimen (or specimens) upon which the species was established was stated to have been found at Barton, north-west of St. Marychurch, and to have been in the collection of Daniel Sharpe.? This collection is now in the Museum of Practical Geology. The specimen represented by fig. 3 can be recognized with absolute certainty in No. 6185; but the source of fig. Ba 1s. uncertain: it was probably another fossil. Edwards & Haime assumed such to be the case, although Lonsdale made no mention of there being two specimens. John Phillips (1841), ‘ Figures & Descriptions of the Paleozoic Fossils of Cornwall. Devon, & West Somerset’ p. 12, pl. vi, figs. 16aa, 16/36, 16), & pl. vu, fig. 15 pb, ustrated Bh hennahi by a drawing (fig. 16a a) clearly intended to represent the weathered surface of Lonsdale’s actual specimen, and by two figures—16 8b (enlarged) & 16 ¢ (natural size)—ot a polished specimen in his own possession. His description of the species does not differ materially from that given by Lonsdale-—in fact, Phillips appended to it the following statement :— ‘The above description is almost verbatim from Mr. Lonsdale. I have added the words in brackets, from a beautiful polished specimen in my possession (fig. 16c¢), which may be distinct.’ Phillips’s specimen unquestionably belongs to the same genus as Lonsdale’s; but the figure is not sufficiently exact to enable one to pronounce a certain opinion as to the species. Pl. vu, fig. 15 p, is that of a specimen which Phillips found in the Carboniferous Limestone of Flintshire, and figured under the impression that it was allied to, if not identical with, A. hennaht Lonsdale. As a matter of fact, the species bears no relationship at all to the Devonian coral, but is the form that I here deseribe as Ov7on- astrea phillipst (McCoy). Untortunately, this figure has been the cause of much subsequent confusion. It must be borne in ? J.B. Lamarck, ‘Systéme des Animaux sans Vertébres’ 1801, p. 371. ? Daniel Sharpe’s Collection was presented to the Geological Society by Henry Sharpe, his brother, in 1856, and passed into the possession of the Museum of Practical Geology in 1911, when the Society divided its collection between that institution and the British Museum—the British material going to the former and the foreign to the latter. e part 4] PHILLIPSASTR4#A HENNAHJ, AND ORIONASTR 4A. 285 mind, in connexion with the controversy which followed, that Phillips did not state that it was ‘A. hennahii, but, in his own words, a ‘specimen from Mountain Limestone much allied to Astrea hennahii. HK. A. Roemer (1843), ‘Die Versteinerungen des Hartzge- birges’ p. 5, pl. ii, figs. 13 a-136, & pl. iii, fig. 1, contributed to the literature of the genus a description and figures of two species allied to ‘Astrea hennahii, both from the Devonian ot the neighbourhood of Griind in the Harz, namely ‘ Astrea hennahii’ Lonsdale? and ‘ Astrea parallela N.’ :— 4 Astrea hennahii (pl. ii, figs. 13a & 1386).—The figure indicates a form differing from the type-specimen in regard to the smaller size of the corallites and the apparent presence of an epitheca. It suggests Phillipsastrea pentagona Goldtuss (Acervularia penta- gona Goldfuss auctt.). Astrea parallela (pl. iti, fig. 1).—In the figure of this species the intrathecal regions are very small and widely separated, and the septa are markedly confluent. Rcemer identified this species with Phillips’s polished specimen, also with the coral from North Wales, for which he assumed a Devonian age. Alcide d’Orbigny (1849), ‘Note sur des Polypiers Fossiles’ p- 12, introduced the generic name ‘ Phillipsastrea, and he defined the genus in the following words :— ‘G. Phillipsastrea, @’Orb., 1847.1 Le dessin qu’en a donné M. Phillips pourrait faire croire que ce sont des calices petits, 4 cloisons rayonnantes, entourées-de cloisons costales communes. On en connait deux espéces de Vétage déyonien. Ex.: Astrea parallela et hennahit, Phillips.’ In the ‘Prodrome de Paléontologie’ published the followmg year (1850), on pp. 106-107, he tabulated under various generic names most of the species that I have previously mentioned, namely :— : Inthostrotion hennahii (Lonsdale, pl. lviii, fig. 3). Actinocyathus hennahii (Phillips, pl. vi, fig. 16 a). Phillipsastrea parallela (Roemer, pl. iii, fig. 1). Phillipsastrea hennahii (Phillips, pl..vi, fig. 16 3, and pl. vii, fig. 15 D). Furthermore, he re-defined the genus Phillipsastrea thus :— ‘Ce sont des Siderastrea dont la columelle, au lieu d’étre styliforme saillante, est large et divisée en cloisons rayonnantes, comme chez les Columnastrea.’ H. Milne Edwards & J. Haime, in 1850, in the Introduction to their monograph on ‘British Fossil Corals’ pp. lxx—lxxi, defined the genus Phillipsastrea, and quoted as the type-specimen ‘ Astrea hennahi’ Lonsdale. Later, however (1851, ‘Mon. Polypiers Fossiles’ pp. 417 & 421; and again in 1852 & 1853, ‘ British Fossil Corals’ pp. 203 & 240), they restricted the name Phillipsastrea to a genus represented by Phillips’s Carboniferous species, which they identified with ’ Martin’s species Hrismatolithus radiatus (see p. 296), and quoted 1 This date is untenable, since the tract was not published until October 10th, 1849, fide C. D. Sherborn. 286 DR. STANLEY SMITH ON AULINA ROTIFORMIS, [vol. lxxii, “A. hennahii Lonsdale’ fig. 3 (excluding fig. 3 a1) as the type of “a new genus—Smithia.. Phillipsastrea, they stated, possessed a columella, and Smthza did not. While acknowledgment is due to the service rendered by these ‘authors in separating the Devonian and Carboniferous forms, the legitimacy of their use of the name Phillipsastrea must be challenged: because, although the figure of the Carboniferous species was quoted by d’Orbigny, he clearly intended the name Phillipsastrea to be applied to the Devonian species, since he framed his definition of the genus upon the figures and descriptions of those corals. It is quite certain that, in quoting fig. 15D, he did so under the impression that the species was identical with that represented by figs. 1666 & ec. It is also true that he mentions the presence of a columella as one of the generic characters of Phillipsastrea; but, here again, he was misled by the figures that he quoted, and interpreted as a coral structure the interstitial calcite which filled the intrathecal region and stood out from the weathered surface as a boss. The name Phillipsastrea must, therefore, be retained for the Devonian genus, and a new one given to the Carboniferous species. The actual genotype of Phillipsastrea is the species to which Phillips’s specimen be- longed, but there is nothing in the figures to warrant the assumption that it was not specifically identical with the type- specimen of Astrea hennahi—it may, or may not, have been. In either case the two specimens belong to the same genus, and, even supposing that they represented different species, Phillips’s specimen is lost, and we are at liberty to closeé’a neogenotype-— besides, in so doing, we rectify an anomaly. ~~ Edwards & Haime also described under the name Acervularia 2 (‘Polypiers Fossiles’ 1851, pp. 416, 421, and ‘ British Fossil Corals’ 1853, pp. 236-41) certain species of Devonian corals, which differed from Phillipsastrea (Smithia of these authors) only in the presence of a thin but distinct epitheca. \ T agree with Frech in regarding the two as congeneric—in fact, careful examination of material shows that certain species of the one are identical with certain ‘species’ of the other, and that the development (or, rather, loss) of the epitheca is one of degree only. Indeed, in a single corallum, epitheca may be present in one region and absent in another. Frederick McCoy (1851), ‘ British Paleozoic Fossils’ p. 72, 1 They founded upon this figure a new species, Smithia pengellyi, and __ thereby definitely fixed for us the type of A. hennaht. ~ 2 The genus Acervularia was founded by Schweigger, ‘Handbuch der Naturgeschichte’ &c., 1820, p. 418, upon the figure and description of a Silurian coral from Gothland by Henry Fougt, ‘ Corallia Baltica’ in ‘ Ameeni- tates Academic’ pp. 92-98, pl. iv, fig. ix and No. 2. Fig. ix bears a strong superficial resemblance to weathered specimens of the so-called Devonian Acervularia, but it is not at all likely that these are congeneric with the coral figured by Fougt. No. 2 (also fig. viii, referred to as a variety of the same ‘ madrepora composita’) undoubtedly represents Acervularia luxurians. —— part 4] PHILLIPSASTR.£A HENNAHI, AND ORIONASTR EA. 287 assigned Astrea hennahi of Lonsdale to Dana’s genus drachno- phyllum. Fritz Frech (1885), ‘Die Korallenfauna des Oberdevons in Deutschland’ Zeitschr. Deutsch. Geol. Gesellsch. vol. xxxvu, pp- 44 et segg., has anticipated me in merging Smethia and the Devonian species of ‘Acervularia’; he also questioned the right of including Phillipsastrea radiata in the same genus. Rudolph Schifer (1889), ‘On Phellipsastrea WOrb., with special reference to Phillipsastrea radiata 8. Woodward sp. & Phillipsastrea tuberosa McCoy sp.’ Geol. Mag. dec. 3, vol. vi, 1889, pp. 398-409, pl. xii, reviewed the literature of Phillipsastrea, and redescribed the species of Orvzonasti¢a mentioned in the title of his paper. He examined McCoy’s types at Cambridge and specimens of Orionastrea in the British Museum, and came to the conclusion that these forms did not possess a columella, and therefore the basis upon which Edwards & Haime separated Phillipsastrea and Smithia had no real existence. I must modify Schifer’s assertion, that the Carboniferous species do not possess a columella,! by the statement that a columella may be present or absent; although in most cases it is present, the columella is admittedly merely the dilated prolongation of the counter-septum, but in this origin it does not differ from the columella in other Rugose genera. Even if Schiafer had been correct in his statement concerning the absence of a columella in Orionastrea, there still remains the very striking character of Phillipsastrea that is not developed in the former: namely, the peculiar septal dilations already described. Several other paleontologists have discussed the genera Phil- lipsastrea and Smithia; but, in every case, they have contented themselves with the question of the identity of the two ‘ genera,’ and have neglected to inquire into the legitimate application of the names. In fairness to these authors, it must be stated that they decided that the genera were identical, and consequently the other question did not concern them. Generic Characters. The corallum is composite and massive; the corallites are united by their dissepimental tissue, or are separated by a thin epitheca only : in the former case the septa are often confluent. In the 1 Tt should be noted that some of Schiafer’s own figures (5, 6, & 7) show the columella, and that, although in figs. 1 & 2 no columella is represented, yet in the actual specimens (R. 541 & 56740) the columella is prominent. Schafer quoted Kunth (Zeitschr. Deutsch. Geol, Gesellsch. vol. xxii, pp. 30-87, pl. i, figs. 4a—4.d) to support his contention that there was no fundamental difference between the Devonian and the Carboniferous species of Phillipsastrea; but, as a matter of fact, Kunth’s figure of Ph. hennahi bears very little resemblance to theactual type, and may be dismissed from the argument. The major septa are shown in these figures to unite in the centre of the corallite, and so form a ‘columellarian tubercle’; moreover, they do not’ appear to dilate at the theca. 288 DR. STANLEY SMITH ON AULINA ROTIFORMIS, { vol. lxxu, mature corallite the septa assume 2 radial symmetry, although the bilateral symmetry of its early growth-stages is sometimes traceable in the adult state through the presence of a cardinal septum shorter than the rest, and is indicated in some species. by a prolonged counter-septum. Both the major and the minor septa dilate at the theca, and the latter terminate there ; but the major septa, in an attenuated form, advance into the intrathecal region, and frequently dilate again at their axial edge. The central part of the corallite is occupied solely by the tabule. Confluence and dilatation of septa are characters comparatively rare in the Rugosa, but of very frequent occurrence in the Aporosa. Genotype: Phillipsastrea (Astrea) hennahi (Lonsdale). Type-specimen of Ph. hennahi: No. 6185, Museum of Practical Geology. The type is possibly the holotype, but is much more probably one of two syntypes. Edwards & Haime assumed the Jatter to- be the case, and chose it as the lectotype. Description of Type-Specimen of Phillipsastrea hennuhi (Lonsdale). (Pl. XXII, figs. 1—4.) No. 6185 & Sections 28348-28349, Museum of Practical Geology (‘ Daniel Sharpe Collection’—part of the ‘ Geological ee Collection’). For literature and history, see p. 284. The specimen is a weathered and rounded block measuring some 10 cm. along its greatest length and about 4 cm. through its thickest part. There are two plane polished surfaces, transverse and longitudinal respectively to the direction of growth. The material is translucent and of a grey colour, except near the surface, where it is bleached and opaque. Superficially, it is iron-stained. Weathering has reacted more vigorously upon the coral tissue than upon the interstitial calcite, and so has engraved the coral structures into the stone. The calcite occupying the interseptal spaces stands out from the surface, and counterfeits actual septa, while that filling up the intrathecal region projects as a central prominence. External Characters. None present. Internal Characters. (Pl. X X11, figs. 2-4.) Transverse section.—Epitheca almost but not entirely absent, weak traces being detectable here and there. Confluence on the part of the septa is not highly developed: the septa may be confluent; but, for the greater part, those of one corallite abut against those of another, or intermingle in a confused network. The thee are of uniform size, measuring 3 mm. in diameter; part 4] PHILLIPSASTR_£A HENNAHI, AND ORIONASTR_EA. 289 but the extrathecal tissue is very unevenly developed, so that the intrathecal regions may lie closely together or widely apart. There are twenty-six septa present im the majority of the corallites, and the dilation of these at the theca extends for a distance of about 1mm. ‘The dilation at the axial edge is but feebly developed in this specimen. Longitudinal section.—The dissepiments are small and crowded ; the tabule are closely set. Shallow concave tabule occupy the central part of the intrathecal region, and these are supplemented near the theca by smaller proximally-inclined plates. The figure should be compared with that of Aulina, p-292. The peripheral plates do not appear to be present in all species of the genus. Notes upon Lonsdale’s Figures and Statements. Fig. 3 combines a view of the weathered surface and the ibveranalle: -polished face of specimen No. 6185. The features displayed have been most accurately drawn,! and the portion included in the figure can be precisely defined by a line. It is evident that Lonsdale (and others after him) mistook the etched- out interstitial calcite of the surface for actual coral-tissue. Fig. 36 shows, on an enlarged scale, the longitudinal section through a few septa as seen in the polished surface. It supplies no useful information. Fig. 3a illustrates a transverse section, but, as has been pre- viously mentioned, this figure has probably been supplied by another specimen. The corallites are represented as being larger, and the septa as being more numerous and more perfectly confluent, than those in the specimen figured in fig. 8 a—the type. I have searched through the Daniel Sharpe Collection, and, among those examined, the specimen with which the figure most closely agrees is No. 6192, a thin polished slab of irregularly pentagonal shape. In this specimen the thece are larger (5 mm. in diameter) and the septa are more numerous (about 36) than in the type-specimen ; and, furthermore, the dilation of the septa at the theca is less localized and consequently not so obvious. The characters displayed agree with Lonsdale’s statements that there are 36 septa (or rays, as he termed them), ‘ unequal in length and breadth, and of a crenulated structure.’ The last remark is of some importance in identifying the specimen: it refers to a feature not untrequently found in mineralized corals, and is strongly shown in the present case—a feature due to the obli- teration or partial obliteration of the dissepiments, except where these meet the septa. Despite his assertion that the number of septa is 36, Lonsdale only shows about 30 in this figure. The origin of his fig. 8@ must, nevertheless, remain uncertain. 1 The drawing is reversed on the plate, so that the right side in the original is the left in the figure—a common occurrence in plates of the period. 290 DR. STANLEY SMITH ON AULINA ROTIFORMIS, [vol. xxii, V. AULINA, gen. nov. Family ? Phillipsastreide. | The structure of Awlina is in most respects similar to that of Phillipsastrea, but it appears to carry to a further stage of development the septal characters peculiar to the latter genus. Generic Characters. The corallum is massive, and the corallites are united by their extrathecal tissue. All the septa dilate at the theca, and those of the major cycle again dilate at their axial edges, in such a manner as to fuse together and so to form a cylindrical wall or tube within the theca (see fig. 4, p. 292). Stability of character is a feature strongly maintained. Genotype: -dulina rotiformis 8. Smith. Type-specimen of 4. rotiformis. The holotype has been cut in two: one half is in the British Museum (R. 17497), and the other in the Sedgwick Museum, Cambridge. AULINA ROTIFORMIS, Sp. Nov. 1910. Phillipsastrea radiata (Martin), S. Smith, Trans. Nat. Hist. Soc. Northumberland, &c. n. s. vol. iii, pt. 3, pp. 629-30. a 1912. Phillipsastrea radiata (Martin), E. J. Garwood, Q. J. G.S. vol. Ixviii, pp. 542-43. : 1916. Aulina rotiformis S. Smith, Abs. Proc. Geol. Soc. No. 995, pp. 2, 3. Aulina rotiformis was recorded as Phillipsastrea radiata from South Northumberland by myself and from Hurdorthwaite Moor (North Yorkshire), by Prof. E. J. Garwood, in the publications above cited. No description of the form was given in either paper. I remarked that the form was plentiful in and charae- teristic of the Fell Top Limestone around Harlow Hill, while Prof. Garwood stated that it was somewhat abundant in the Botany Beds on Hurdorthwaite Moor and that he had not found it ¢ sttw at any other horizon.! iExternal Characters. The corallum is massive, and is typically depressed in form. The corallites are not defined by epitheca ; consequently, the only epithecal development present is that external to the whole colony. The upper or distal surface of the corallum is generally flattened and fairly uniform. The thecze measure about 1:5 mm. in diameter; they are near together and regularly arranged, as compared with those of Orionastrea, lying 1 to 3 mm. apart.® 1 He recorded the finding of an angular and unworn specimen of Phillips- astrea radiata resting upon the Tyne Bottom Limestone in High Cup Gill and of a cast of the same from the Drift of Ravenstonedale. The High-Cup Gill specimen was correctly identified, and therefore not Aulina. but Orion- astra. * In one of the specimens collected from the Botany Beds the calices are larger and proportionately more widely separated, and the surface more mammiferous, than in the Harlow-Hill material. part 4] PHILLIPSASTR£ZA HENNA, AND ORIONASTR AA, 291 The general surface of the corallum (as will be noticed in the text-figure) rises towards the thece, so as to form mound-like borders to the calicular depressions, suggesting the idea of minute Fig. 3.—Distal surface of the type-specimen of Aulina rotiformis, sp. nov. x 2 diameters. A. H. Harrow photo. [Fell Top Limestone, Harlow Hill (Northumberland). One half is in the British Museum (Natural History), R. 17497, and the other half in the Sedgwick Museum, Cambridge. | craters. Within the calice the tubular wall terminates as an elevated ring, forming a hollow axis against which the major septa abut like the spokes of a wheel. Internal Characters. (PI. XXII, figs. 6-11, & text-fig. 4.) The most distinctive feature is the tube traversing the tabular tissue. Within this tube the tabule are horizontally disposed, and are remarkably flat; externally to it, they take the form of cones perforated by this axial tube. The dissepiments are very small and regularly formed. All the septa dilate at the theca, and there the minor series abruptly terminate. In these respects Awlina is similar to Phillipsastrea. The major septa, in an attenuated form, advance towards the tube, to which their axial edges appear to be fused. Actually, the septa dilate along their axial edges in such a way as to present in cross-section the letter T, and form the tube by the mutual fusion of these peculiar dilations into a perfect wall. Confluence of the septa is well developed. Transverse section.—In this section the septa display the 292 DR. STANLEY SMITH ON AULINA ROTIFORMIS, [ vol. Ixxii, characters above mentioned, and the tube appears as a polygonal ring. The number of septa in each cycle is usually eleven; very rarely does it exceed Fig. 4.—Diayram combining a transverse twelve or is it less and a longitudinal section of Aulina than ten. The theca rotiformis. x about 18 diameters. is emphasized by a very slight secondary thickening of the dis- sepiments which form it. The diameter of the theca is about 1°5 mm., and the diameter of the tube about 0°5 mm. ‘The dilation of the septa at the theca does not extend far in a peripheral direction (as indicated in the figure ). _ Longitudinal section (Pl. XXII, fig. 8)—The tube is represented by two vertical walls limiting the horizontal tabule ; the tabule external to these walls are seen as inclined plates. Ontogeny. In the earliest growth-stages observed in duw/zna the septa then present meet in the centre of the corallites (Pl. XXII, figs. 9-11). Later, these leave the centre and form the initial stage of the tube, while new members are being added to their number. Although, by the time that it reaches the adult stage, the form displays a radial symmetry, the insertion of septa follows the general rule appertaining to the Rugosa as a class. A cardinal fossula (in some cases very conspicuously, but in others less pro- minently displayed) and a distinct pinnate symmetry may be observed in the immature stage. Phylogeny. The morphology of Awlina clearly points to its derivation from Phillipsastrea: the axial tube in the former and the dilated edges of the septa in the latter are probable stages in a continuous line of development. The interval of time, however, separating the epochs in which the two genera lived respectively is consider- able—practically the whole of the Lower Carboniferous Period, and no intermediate forms have as yet been recorded from the beds of that age. Horizon and locality.—Awlina occurs abundantly in the Fell Top Limestone in the vicinity of Harlow Hill,! where the ! Harlow Hill lies on the Neweastle and Carlisle Road, 11 miles west of the former city (Quarter Sheet 105 N.W. New Series 14). : : part 4] PAILLIPSASTR.ZA HENNAHI, AND ORIONASTRAA. 293 outcrop of that limestone is exposed by a chain of old quarries extending from the south side of the village to Stob Hill, near Cheesburn Grange, some 2 miles in a north-easterly direction. In that locality the Fell Top Limestone lies somewhere about 1000 feet above the Great Limestone, and is between 20 and 30 feet thick (I have not been able, however, to expose the actual base nor find the exact top). The lower portion consists of well- bedded crystalline limestone; the middle is much more argillaceous, and is divided up by bands and partings of shale; but the upper part, again, is more massive and more purely calcareous. The fauna does not materially differ from that of the lime- stones between it-and the Great Limestone, except in a few (but noteworthy) particulars —namely, the presence of Aulina and the abundance of Lithostrotion ct. junceum (but approaching L. irregulare), which oceurs as small stunted colonies of tortuous growth-habit. These colonies form in places a band at the top of the lower and more massive portion, and appear to have been buried as they lived by the overlying silt (now calcareous shale). Dibunophyllum near y Vaughan (small, conical and often twisted forms) is plentiful, and Lithostrotion portlocki is common. The most striking feature of the brachiopod-fauna is the great abundance of Productus latissimus. Fenestella is present in large quantity. The other locality in which Aulina has been found is Hurdor- thwaite Moor, from a thin series of limestones and shales inter- calated in the Millstone Grit, to which series Prof. E. J. Garwood has given the name of Botany Beds,! after the farm north of their outcrops. Prof. Garwood? defines the stratigraphical position and geo- graphical situation of the beds, and describes their character in the following passages :— ‘The beds lie some distance [200 feet or so] above the base of this series [the Millstone Grit], and occupy a tract extending for 2 miles in a general east-and-west direction on Hurdorthwaite Moor, south of Botany Farm. They are repeated by a strike-fault with a southward upthrow, which causes an extension of the outcrop round the south of the moor, and affords additional exposures. ‘The beds are well seen in several old quarries at Scoletree, Howgill Head, and Greenhill. They consist of a few feet of compact crystallized lime- stone at the base, which is overlain by some 10 to 15 feet of impure calcareous and ferruginous shales’ [and I may add—especially towards the top—lime- stone-bands]. The sequence suggests at once the Fell Top Limestone, and the resemblance is still further strengthened by the characteristic appearance of the pale crumbly shale packed with Fenestella. Faunistically also the two series bear a striking similarity, since, in addition to Aulzna, the Botany Beds have yielded the same 1 Botany Farm lies 23} miles west-south-west of Romaldkirk, and about half a mile north of the Reservoir (Quarter Sheet 102 S.E. New Series 31). 2 Q. J. G.S. vol. Ixviti (1912) p. 542. . 294: DR. STANLEY SMITH ON AULINA ROTIFORMIS, [ vol. \xxii, stunted and twisted forms of Lithostrotion junceum, the same dwarfed and conical forms of Débunophyllum, and also Litho- strotion portlocki.| 'The list of fossils from these beds published by Prof. Garwood? is very similar to that of the Fell Top Limestone which | previously published.? Thus, in their agreement in faungl contents, lithological cha- racter, and stratigraphical position there is strong evidence for correlating the Botany Beds with the Fell Top Limestone.* This correlation being accepted, the bed represents the highest known limestone of standard conditions ® occurring at a horizon occupied in most places by arenaceous deposits (‘ Millstone Grit’) or argil- laceous beds (‘ Pendleside Series’). Assuming that subsequent investigation does not discover Awlina in beds lower than those which it is known to occupy, I suggest that the horizon in which it occurs be removed from the Dibunophyllum Zone to con- stitute a new zone—the Auwlina Zone, which, in accordance with Vaughan’s system, would be indicated by the initial letter A: a definite limit would thus be given to the Dibunophyllum Zone, which at present tends to extend indefinitely upwards. The Postdonomya Zone® in Gower and elsewhere is approximately equivalent to the Aulina Zone, but represents a shallow-water phase, and, appearing earlier in time in some localities than in others, must also represent beds belonging to ‘ D,’ as well as ‘A.’ The question as to whether the Avonian Stage should include the Aulina Zone, or be restricted so as to end with the Dibwno- phyllum Zone, is worthy of future consideration, but cannot be discussed here. VI. ORIONASTRHA, gen. Nov. Family : LirHostRorronvTip az. Species.— O. phillipsi (McCoy). Genotype. O. placenta (McCoy). O, ensifer (Edwards & Haime). The species grouped together under the name Ovzonastrea are closely allied to the species of Lithostrotion, and appear to . 1 Although I paid a visit to these beds in the summer of 1915, in company with Mr. C. T. Trechmann and Dr. D. Woolacott, and found Awlina and other forms, the principal source of information for the above facts was the material collected y Prof. Garwood, which he kindly allowed me to examine. 2 Q. J. G. S. vol. Ixvili (1912) p. 542. 3 Trans. Nat. Hist. Soc. Northumberland, n. s. vol. iii, pp. 630-31. 4 The identity of the two limestones was suggested long ago by J. G. Goodchild, but he does not appear to have ever published these views— fide Prof. G. A. Lebour. ® Standard conditions as opposed to a phase: that is, any interruption in a sequence of deposits formed under conditions which have been fixed for the purpose of zoning. See A. Vaughan, Proc. Bristol Nat. Soe. ser. 4, vol. i (1906) p. 79. ° B. E. L. Dixon & A. Vaughan, Q. J. G.S. vol. lxvii (1911) p. 494. i i ert alee Pe + gal a part 4] PHILLIPSASTR4ZA HENNAHI, AND ORIONASTR ZA. 295 represent the phylogerontic (or senile) stage in the history~of that stock. The separation of Lithostrotion ensifer from the other species of that genus and its inclusion here may encounter reasonable objection, since the differences that it exhibits are cer- tainly not great; but, as 1t possesses (although in a less-developed condition) those features which distinguish the genotype of Orionastrea, I believe that the step is justifiable. Generic Characters. The corallum is composite and massive, and the corallites are either defined by a thin epitheca, or, in the more typical forms, by no epitheca at all; in this latter case the corallites are united by the dissepimental tissue, and the septa are confluent. Both major and minor septa are well developed, and a columella is present (except in O. placenta). The characters of Orionastrea are essentially those of Litho- strotion, but in a modified form, and their structures are more unstable and variable. Even in the same corallum the difference between the corallites is often very marked, as in Pl. XXIV, fig. 2. From a large quantity of material examined three distinct types of Orionastrea can be readily isolated, represented respectively by the three species here described; but a large proportion of the specimens falls between two species rather than within one of them, and might, therefore, be ascribed to the one species equally as well as to the other. To O. ensifer specific rank can with little hesitation be accorded, and it is easily distinguished from O. phillipsi and O. placenta. Furthermore, it represents almost exclusively the genus in the Bristol area, whereas it is verv rare in North Wales, where O. phillipsi is so abundant and where O. placenta also occurs. I have considerable doubt as to whether, in the strict zoological sense, it is correct to regard O. phillipst and O. placenta as separate species: both occur together, and the difference may be nothing more than that between individual colonies of the same species. Since, however, in the absence of sufficient evidence this question cannot be settled, and since it may prove useful to have separate names for the two types, 1t seems permissible and even desirable to recognize MeCoy’s species rather than keep them merged under Martin’s name ‘ radiata.’ Genotype: O. phillipsi (McCoy). Type-specimen of O. phillipsi: No. 2138.a, Sedgwick Museum, Cainbridge. (Lectotype chosen from two syntypes.) Summary of Research. The earliest figures of Orionastrea are those of a highly siliceous specimen from Winster (Derbyshire), published -;by William Martin in 1809, ‘ Petrificata Derbiensia’ pl. xvii, figs. 2 & 3. Q. J.G.S. No. 288. Z 296 “DR. STANLEY SMITH ON AULINA ROTIFORMIS, (vol. xxi, Martin described the specimen under the name ‘ H7ismatolithus tubiporites (radiatus1)’, and stated that the coral was built up of ‘straight tubes connected by transverse dissepiments or partitions ... embedded in a calcareous cellular mass.’ From his fig. 2 these ‘tubes’ can be readily recognized as cherty casts of the intrathecal region of corallites, and the ‘transverse dissepiments ’ as silicified layers in the extrathecal tissue. The confluent character of the septa is mentioned by Martin, and is well shown in the figure. John Fleming (1828), in ‘A History of British Animals ’ p- 529, waeonan Martin’s species as ‘ Tubipor a radiata. This citation appears to have been overlooked hitherto, and those paleontologists who refuse to acknowledge Martin’s claim to the authorship of the species which he named, on the grounds that his terminology did not conform to the laws of Linnzan nomen- clature, credit the trivial name ‘radiata’ to Samuel Woodward, ‘Synoptical Table of British Organic Remains’ p. 5 (1830). John Phillips’s ‘ Figures & Descriptions of the Paleozoic Fossils of Cornwall, Deron, & West Somerset’ (1841) contained the figure (pl. vii, fig. 15 p) of the coral from the Carboniferous Lime- stone of Flintshire which Phillips believed to be allied to ‘ Astrea hennahii’ (see p. 284), but which actually belonged to the same genus as Martin’s ‘ Hrismatolithus tubiporites (radiatus). ‘The form is common, and very characteristic of the Lower Carboniterous of North Wales. Frederick MeCoy, in 1549, ‘On some New Genera & Species of Palzonoic Corals & Foraminifera’ Ann. & Mag. Nat. Hist. ser. 2, vol. iui, pp. 124-25, and in 1852, ‘A Systematic Description of the British Paleozoic Fossils in the Geological Museum of the University of Cambridge’ p. 110, pl. 118, figs. 8 & 9, described, under the generic name Sar cinula,2 three ‘species’ of Orion- astred, namely, S. tuberosa, S. placenta, and S. phillipsi. he last he rightly identified with the coral from North Wales figured by a Phillips. Brief notes upon the ieee specimens will be found on pp. 299-800. H. Milne Edwards & J. Haime, ‘Monographie des Polypiers Fossiles’ 1851, pp. 448 & 449 and * A Monograph of the British Fossil Corals’ 1852, pp. 203-204, pl. xxxvu, fig. 2, merged S. phillipsi and S. placenta, and identified these with Martin’s coral, but retained as a separate species S. tuberosa. Now that IT have cut tie type-specimens, I consider it more desirable to recognize S. phillips: and S. placenta as distinct species, and to merge S. twberosa® with the former. Moreover, on grounds 1 The name ‘radiatus’ has not been retained, for reasons subsequently explained. 2 J. B. Lamarck, ‘ Histoire Naturelle des Animaux sans Vertébres’ vol. ii (1816) p. 222 (for two recent forms, one of which he confuses with a Silurian fossil). 3 H.M. Edwards & J. Haime, ‘ Polypiers Fossiles des Terrains Paléozoiques’ p. 447, include an American species, P. vernewili in the same genus; but, arguing from the species so named in the British Museum, I do not consider the form congeneric with the British species. part 4] PHILLIPSASTRLA HENNAHI, AND ORIONASTR-EA. 297 already adduced (pp. 285 e¢ seqq.), these authors restricted the use of A. dOrbigny’s. generic name Phillipsastrea to these species—by which name they have since been known. In the same works (pp. 442-43 and p. 193 respectively) Edwards & Haime described a new species of Lithostrotion (L. ensifer), which I propose to include in the genus Orionastrea, although acknowledging its equal claim for inclusion in either genus: it constitutes, in fact, the link connecting Orionastrea phillipsi with the typical forms of Lithostrotion. James Thomson, in 1883, ‘On the Development & Generic Relation of the Corals of the Carboniferous System of Scotland ’ Proce. Phil. Soc. Glasgow, vol. xiv, pp. 394-96, pl. iv, figs. 1, 1 a, 16, Le, & 2, recorded Orionastrea (as ‘Phillipsastrea’) from Blackridge (Linlithgowshire), and thus widened its known area of distribution to include Scotland. Moreover, he demonstrated its instability of character by enumerating the varieties that he ob- tained from a very limited horizon at that locality, which he stated was the only one in Scotland where the genus had been found. He specified four varieties, distinguished by the following characters :— Ist variety. Columella absent, tabule regular. 2nd do. do. do. do. irregular. 3rd do. do. present. 4th do. do. do. mammilliform surface (=S. twherosa (MeCoy)). He illustrated his description by figures of : (a) a form in which the counter-septum alone invades the intrathecal region, and constitutes a slender and inconstant columella ; (b) a form in which several septa unite in the centre of the intrathecal region. I consider it probable that his varieties 1 & 2 are O. placenta, and that 3 & 4 are O. phillipst. Rudolph Schafer’s (1889) contribution to the literature, ‘On Phillipsastrea dOrb., with especial reference to Phillipsastrea radiata 8. Woodward sp. and Phillipsastrea tuberosa_ McCoy sp. Geol. Mag. dec. 3, vol. vi, pp. 898-409, pl. xii, has already been discussed (p. 287); wherefore here I merely desire to draw attention to the text-figures on p. 403, representing calices over the floors of which a number of septa have straggled: the figures are intended to demonstrate the absence of a columella. I admit the accuracy of the figures; but unfortu- nately they illustrate selected calices, while others of the same specimens in which the columella can be detected have been overlooked. Schafer followed Edwards & Haime in merging McCoy’s species S. phillipsi and S. placenta, and retaining his S. tuberosa. A. Vaughan (1903), ‘ Notes on the Corals & Brachiopods obtained from the Avon Section & preserved in the Stoddart Collection’ Proc. Bristol Nat, Soc. n. s. vol. x, pp. 109-10, included 72 -~ 298 DR. STANLEY SMITH ON AULINA ROTIFORMIS, [ vol. Ixxti, remarks on Lithostrotion ensifer Edwards & Haime; and in 1905, in ‘The Paleontological Sequence in the Carboniferous. Limestone of the Bristol Area’ Q. J. G. S. vol. xi, p, 199, he stated the exact paleontological horizon at which the species. occurs—namely, the upper part of the Dibunophyllum Zone: that is, Subzone of Lonsdaleia floriformis (D,). ORIONASTRMHA PHILLIPSI (McCoy). 21809. Hrismatolithus tubiporites (radiatus) W. Martin, ‘ Petrificata Derbiensia” pl. xviii, figs. 2 & 3. [O. phillipsi or O. placenta. | 21828. Tubipora radiata (Martin), J. Fieming, ‘A History of British Animals’ p. 529. [O. phillipsi or O. placenta. | ? 1880. Tbonor radiata (Martin), S. Woodward, ‘A Synoptical Table of British Organic Remains’ p. 5. [O. phillipsi or O. placenta. | 1841. ‘Specimen from the Mountain Limestone much allied to Astrea hennahii Lonsdale’ J. Phillips, ‘ Paleozoic Fossils, &c.’ p. 12 & pl. vii, fig. 15 D. 1849. Sarcinula phillipsi FE. McCoy, Ann. & Mag. Nat. Hist. ser. 2, vol. ui. p. 125. Sarcinula tuberosa. Id. ibid. p. 124. 1850. Phillipsastrea hennahii Lonsdale, partim A. @Orbigny, ‘Predrome de Paléontologie’ vol. 1, p. 107. 1851. Phillipsastrea radiata (Martin), H. Milue Wdwards & J. Haime, ‘ Polypiers Fossiles des Terrains Paléozoiques’ p. 448. Phillipsastrea tuberosa (McCoy). Id. ibid. p. 449. 1851. Sarcinula phillipsi KF. McCoy, ‘ British Paleozoic Fossils’ p. 110. Sarcinula tuberosa. Id. ibid. p. 110, pl. iB, figs. 8 & 8a. 1852. Phillipsastrea radiata (Martin), H. Milne Edwards & J. Haime, ‘ Mono- graph of the British Fossil Corals’ Pal. Soc. p. 203, pl. xxxvil, figs. 2 & 2 a. Phillipsastrea tuberosa (McCoy). Id. ibid. p. 204. 1888. Phillipsastrea radiata (Martin), J. Thomson, Proc. Phil. Soc. Glasgow, vol. xiv, p. 394, pl. iv. figs. 1, la, 16, & 2. 1885. Phillipsastrea radiata (Martin), F. Frech, Zeitschr. Deutsch. Geol. Gesellsch. vol. xxxvii, p. 48. Phillipsastrea tuberosa (McCoy). Id. ibid. p. 48. . 1889. Phillipsastrea radiata Martin, partim R. Schafer, Geol. Mag. dec. 3, vol. vi, p. 401, pl. xii, figs. 1, 3, 4, 7, 8, & 9, text-tigs. 1-3 (p. 403). Phillipsastrea tuberosa (McCoy). Id. ibid. p. 407, text-figs. 4-6 (p. 403). Phillipsastrea radiata (Martin) includes, in all cases quoted, Sarcinula phillipsi and S. placenta. External Characters. (Pl. XXIII, figs. 1, 3, & 4; Pl. XXIV, tig. 2.) The corallum is typically depressed, and the distal surface is usually flat. Nevertheless, the species occurs in more tumular masses, and may display a mamiillate surface (see Pl. XXIII, fig. 4). “The dissepimental tissue, which constitutes the greater part of the skeleton, is often unevenly developed ; consequently, the cali- cular depressions may be near together, or widely separate: these depressions are not infrequently “pounded by a sharply-elevated border (Pl. XXIII, fig. 1). A columella is present, and is often prominent. Hpitheca clothes the lower surface of the corallum. ENNAHI, AND ORIONASTR.EA. 299 (Pl. XXIII, figs. 2 & 5; XIV, fic. 1.) septa are confluent; the inequality ber number of the major septa is the latter advance far into the r major septa may extend to wh is prolonged into the centre to form a conspicuous columella: en becomes isolated from the rest of (Many of these features are is In section. ) s from 30 to 40 in all, 15 to ough varying in size, measure Transverse seétion between the minor and @ ‘not marked, since on intrathecal region; bu the columella. The & of the corallite, and the the swollen axial portion of the septum in the ephebic stage. displayed equally well ingthe alice The number of septa pres 20 in each cycle, and the theee on the average 2°5 mm. in @ Longitudinal section tabulze are conical in form, the columella is persisten dissepiments are fine. The © absence of the epithecal boun to the corallites, and perhaps to a small extent the less uniform disposition of the tabule, distinguishes the species from those of Lithostrotion. 2 to more typical examples. t to much variation: a less with O. placenta, and less epta merges the members of The foregoing description applies on The characters enumerated are subj well-developed axis leads to eom perfect confluence on the part this species with those of O {[ Sarcinula phillipse Notes upon the Type-Speceimens | berosa McCoy in the McCoy (genotype) and 8. 7 Sedgwick Museum, Cambr: 1. SARCINULA PHILLIPSI: Specimen Merionethshire. (PI. The species was described, bu figured, by McCoy. ‘There are two syntypes mounted on the same tablet, and of these I have chosen the first (213 a, Pl. XXIII, fig. 1) as the lectotype. ~ Photographs of both specimens are reproduced in Pl. XXIII, which illustrate sufficiently well the external characters to obviate the necessity of detailed description. The two types of calices ~ present in the specimens shoulc be observed, namely, those in which the theca is marked by a raised border (fig. 1) and those in which the extrathecal region simply rounds off at the intra- thecal depression (fig. 3, and certain calices in fig. 1). In the ‘ease of 213 (fig. 3) the intrathecal depressions are filled with matrix. The internal characters of the lectotype agree with the general description, and need not here be repeated. ‘The number of septa present ranges from 32 to 40. & 2136; trom Corwen, 3.) é 300 _ DR. SPANLEY SMITH ON AULINA ROTIFORMIS, [vol, xxu, 2. SARCINULA TUBEROSA: Specimen 212; found in Derbyshire, and presented by W. Hopkins. (Pl. XXIII, figs. 4 & 5.) The holotype is a silicified specimen, perforated by several large drusy cavities. The coral structures are nevertheless well pre- ‘served, and the interstices are but incompletely filled with mineral deposit. The distal surface, which is in a fair state of preservation, displays in a somewhat exaggerated form a mam- millate development. The thecz are ‘slightly larger than usually observed, and the columella is not very prominently shown in the transverse section (Pl. XXIII, fig. 5). The number of septa present is smaller than is generally the case, being only about 30: otherwise the characters of this form are in agreement with those of O. phillipsi. Its one pronounced and distinguishing feature—the mamimilli- form calices—may be found associated in a single corallum with calices of the types illustrated as typical of O. phillipsi. It seems therefore inadvisable to regard the form, simply because of its external characters, as a separate and distinct species. McCoy illustrated (from a fragment of the type) what he con- ceived to be the internal characters of the species (‘ British Paleo- zoic Fossils’ pl. ins, fig. 8a), but he omitted the columella, and showed the tabulz as concave—an idea probably due to the tangential nature of the sections, which missed the former, and incompletely displayed the character of the latter. ORIONASTREA PLACENTA (McCoy). 1849. Sarcinula placenta K. McCoy, Aun. & Mag. Nat. Hist. ser. 2, vol. iii, p. 124. 1851. Sarcinula placenta F. McCoy, ‘ British Paleozoic Fossils’ p. 110 & pl. 118, figs. 9, 9a, 9b. See also synonymy and notes on O. phillipsi. O. placenta differs from O. phillipsi solely in the absence of a columella, and consequently in the form of the tabule, which, deprived of their axial support, are concave instead of being conical. The major and the minor septa are almost equal in length, the former advancing very little farther beyond the theca than do the latter. O. placenta bears the same relationship to O. phillipsi as the non-columellate forms of Lithostrotion (Diphyphyllum and Stylastrea)! do to the columellate forms. At certain horizons (for example, at the base of the Great Limestone around Chollerford, South Northumberland) ‘ Diphy- phyllum’ not infrequently occurs to the exclusion of Lithostrotion locally. It is conceivable that similarly O. placenta may subse- quently be found, to the exclusion of O. phillips?, at some particular place or level. 1 See W. Lonsdale, in R. I. Murchison’s ‘Geology of Russia, &c.’ vol. i (1845) pp. 619 & 624: Diphyphyllwm refers to the fasciculate, and Stylastrea to the massive, variety. part 4] PHILLIPSASTR ZA HENNAHI, AND ORIONASTRZA, . 301 Notes upon the Holotype of the Species in the Sedg- wick Museum, Cambridge, No. 211, presented to the Collection by W. Hopkins. (PI XXIII, figs. 6 & 7.) The type-specimen of Sarcinula placenta McCoy is part of a depressed corallum measuring some 6 by 4 cm., and is about 2 cm. thick. Both the upper and the lower surfaces are re- markably flat. For the greater part, the coral interstices are free from mineral deposit, and consequently the material is very friable. Along certain planes, however, it is highly silicified, bemg then converted into thin bands of chert. Superficially, the fossil is stained a deep red. The theecze measure about 2 mm. in diameter, and are fairly regularly spaced, the interval between them averaging 4 to 5 mm. None of the septa invade the intrathecal region, and there is no columella. Im consequence of the absence of a columella, the tabulz are concave, but are somewhat irregular in habit. In this specimen it may be observed that the thecz are some- what smaller than in the average examples, and the number of septa correspondingly less—about 25. Attention must be drawn to the striking resemblance of this type to Martin’s figure of Hrismatolithus radiatus. In mode of preservation it would seem identical with the specimen illustrated in ‘ Petrificata Derbiensia.’ Unfortunately, since we know nothing of the intrathecal character of Martin’s coral, it is impossible to identify McCoy’s type with Martin’s figure, although the pre- sumption is that they are the same. ORIONASTRHA ENSIFER (Hdwards & Haime). 1851. Lithostrotion ensifer H. Milne Edwards & J. Haime, ‘ Polypiers Fossiles | des Terrains Paléozoiques’ p. 442. 1852. Lithostrotion ensifer H. Milne Edwards & J. Haime, ‘ Monograph of the British Fossil Corals’ Pal. Soc. p. 193 & pl. xxxvil, figs. 2, 2a. 1889. Phillipsastrea radiata (Martin), partim R. Schafer, Geol. Mag. dec. 3, vol. vi, pp. 401-407 & pl. x1, figs. 2, 5, ?6. 1903. Lithostrotion ensifer Kdwards & Haime, A. Vaughan, Proc. oe Nat. Soc. n. s. vol. x, p. 109. 1905. Lithostrotion ensifer Edwards & Haime, A. Vaughan, Q.J.G.S. ‘Fak Ixi, p. 199. The specimens upon which Edwards & Haime established the species came from Bristol, and were preserved in the British Museum. The identification of these is not possible; nevertheless, there are several specimens which were in that collection at the time when these authors were engaged upon their researches. The figures and descriptions given by them leave no room for doubt as to the identity of the species. They drew attention to its resemblance ue ‘ Phillipsastrea’ (that is, Orionastrea), and remarked that ‘in this fossil the columella is more prominent than in any other species of the same genus, and the walls much thinner.’ I cannot quite agree with the statement concerning the columella, although 302 DR. STANLEY SMITH ON AULINA ROTIFORMIs, ETC. { vol. 1xxu. J admit that the prominence of this axis is a characteristic feature of the species. The material examined by Schiafer included specimens of Orionastrea ensifer from Bristol, as well as of O. phillipsi and O. placenta from North Wales and Derbyshire. The Bristol specimen (No. 56740, British Museum), which he figured as ‘P. radiata, could be referred almost oe as well to O. phillipst as to O. ensifer; but I consider it to belong to the latter, rather than to the former species. O. ensifer is considerably less differentiated from the species of Lithostrotion than is the genotype O. phillipsi, and constitutes, as previously suggested, the passage-form between Lithostrotion and Orionastrea. Characters not described here may be taken to be in agreement with those of the genotype. External Characters. (Pl. XXIV, fig. 5.) The corallum has the same general form and characters as O. phillipst, except in regard to the distal surface; in this it differs very shghtly from Lithostrotion basaltiforme. The indi- vidual calices are clearly defined, and each is divisible into an intrathecal depression and an extrathecal platform. 'The latter is broad and sloping, and meets that of the contiguous corallites in a sharp ridge. In Lithostrotion this ridge is surmounted by a distinct wall of epitheca, and it is in the absence or weak development of this that O. ensifer is distinguished from the massive forms of Litho- strotion. Calices of the two genera are figured in juxtaposition for the purpose of comparison (Pl. XXIV, figs. 5 & 6). Internal Characters. (Pl. XXIV, figs. 3 & 4.) A thin epitheca may divide the corallites; but, if present at all, it is usually reduced to a palisading of isolated rods which in transverse section present a line of dots. The septa never attain a confluent condition. .The number of septa present is a fairly constant character: the number, including both series, closely approximates to 86. The major septa extend farther into the intra- thecal region than those of O. phillips generally do, and are much more uniform in their length. The minor septa, on the other hand, are but feebly developed. The columella is persistent and stoutly built, but this axis is not found swollen to the extent sometimes observed in O. phillipsi. The thecze usually measure about 2°5 mm. in diameter, but I have examined a specimen in which this measurement attained 4 mm. and in which the extrathecal region was proportionately wide, yet the number of septa present was the same as in the forms of more. usual size. —— a ‘MNEsN, MOSsr]H) 9} Ur wor140es (100d) 9]SUIg ‘ds DYLISHUOLO *(QALYS-SertyuIn( ) stqtexoory FO ysva-Y4nos “odpliqezAry “TosmIOy,y, sommep Aq masts *[NjLQUET *SUILOT sv ‘suIIO, oY} JO SUOT}dIIMsep puv plOdet oYy d}RIPOULISZUL puv fq paryddns uoneuraoyuy “puepjoog Jo Aoaang pyuaanjd °C “wsdeypyd -O ‘(aMYSMOSTIULT) Ioplcyoryq [arsojoan ay} Jo wOIaToQ ay} UT suatntoedg OyoneiTa Maya] POS Pe ET NAS UL PUNOT Jou Inq (au0 SOUL ‘poomaey “¢ “oT ‘Jorg Aq paqdaT[oo uawmoods atG TaIB YY asdinpyd *O V9 dup ys = u0yj0q; euT,, ‘pueptoungsayy Coa oy ‘spuourpsy sapreyD “Ay Sq paqoatjoo wauttoads ouc “OLB “afisua *O “FUOWLAAS “[[BE{ loy[Vg ; euwoJseuL] YANO “purprequing beet eee tee ee at : “UMOUY JOU ST TOMA fo sjnoqraroyM oy “UTzIPTY “AA Aq pansy tiemiroedg “ds DYLISNUOLLOC) *TOYSUL AA ‘surydopy “A “IN Aq poytesoad QSpuquiry “UMesny Yormspeg ey} Ul yIOg - =) = = WISN! NVINOSHIINS S3luve dg LIBRARIES SMITHSONIAN INSTITUT! Eas Ws Ww Z @ 2 Bios a ys fa Qe mG a 1B YE. s = 4 WRAY = fas SEN \; = = ss We G Zz ond ee 2) bes ‘ARIES SMITHSONIAN INSTITUTION NOILALILSNI NVINOSHLINS S3I1YVd' z= ia = lo : = S) is S) seed Mr Cee S) Fe Z whe i w - SS Ss = =) uy / sap bi > a | S a EY 4 oy ce Ee NLILSNI NVINOSHLINS S3J1IYVYd!IT LIBRARIES ry SMITHSONIAN INSTITUT NVINOSHLINS S3S1YVUsa SMITHSONIAN NVINOSHILIWS SMITHSONIAN SMITHSONIAN * RARIES SMITHSONIAN INSTITUTION NOILANLILSNI NVINOSHLIWS w > wa S (eee op) a +2 c a: Ceo # = : : 2K = ioe oy ra = INS ro iy i: Pa e = a Ea da | NLILSNI NVINOSHLINS S3IYVYUSIT LIBRARIES SMITHSONIAN INSTITUT aa ~ Pats Be ¥ = Sey iy i |] = YyGgG 7 Sf cn Ne a> WO CUCU wy) 5 Gp = ey 5 Gu 2 SY 5 Gs Dey 2 ‘ Yl? m Qn Oc w TSS m 4 oa Lagiiss = ee) =— n * = JLILSNI NVINOSHLINS S3IYVYUGIT LIBRARIES SMITHSONIAN INSTITUTIC Zz w rad She 22) } Zz Re = = » = = = cl ae yy rs — ‘ht 5 = pip S aa S bp an OYA iG 2 vi D Yj, 2 iy E Z F Yi = > ea = > = : ” z= op) iy FA ” ARIES SMITHSONIAN INSTITUTION NOILNLILSNI we. = oc ar ox = S & < re < fad = & ar o ca 5 = 3 = 5 Zz i 2 SNe ie SJIYVYGIT LIBRARIES SMITHSONIAN INSTITUTIE Ss ss us = a w = w = ow ~ eS) a 283) a > ke > = %, po) = Po) 7 es) Hae ep) ad nm? ran ; eo ee uD ie wo ARIES SMITHSONIAN INSTITUTION MOTE SNC NVINGSH TINS eee “ye ie 2) ; = ae = < = ~ z a vgs 4 ae Oo at Sa Coy ee Snacks yy WG) Bane. Ww (e) ae Wate} as oO = Fe 2, os =. A > = ~ Sy = > ss pa a aks a, 7p) oz LILSNI NVINOSHLINS S3IYVYS!I1 LIBRARIES INSTITUTI¢ S) VLILSNI NVINOSHLINS S3IUVUGIT LIBRARIES SMITHSONIAN INSTITUTI ARIES SMITHSONIAN INSTITUTION NOILNLILSNI NVINOSHLINS S3IYVY SS ay SMITHSONIAN z 7p) _ pea “ Md = LW e ies! 4 ane 4 4 : = = us ay e SS ~ S a O ay 2 ARIES SMITHSONIAN INSTITUTION NOILNLILSNI Sa3i1y¥VvuE = a ia c z= (S) Pan oO oe ‘ol E sd ie w fa =D) rag re | 70 = Ey = im aii cs a0 +4 wa a aw “aE ee) ee o Ea a z a z wo z n = | = oH Oo z > z NVINOSHLIWS Cy X SMITHSONIAN Y-y G Yj A te, Z aa » SMITHSONIAN pot hs oP] ews! op) RS SN ook i, et oc ti \\ S oc 4 ce 5 aN oc ior (a0) —_ . fai) S) ct O Beige z po | we ¥ Be i a vee cw ena res C7 Tatras ktiwo81ws 1riRrenmpmaAmtrr. PaLIWVWLICMALIIA AT Rt CO Soe pie? edly len Wal gt ry NON It on yen He ON ow of Oh tipo Pompe yeh ed marie tend nt GAY aoe ys Cjae aby VE ban & iA tru tart Ay VATA CU uNUS Caryerk ante 4 Abs oaisps» : Dern Sade NP : Ents ht FARES GAN, He Hy ( Ussivt'g Ward Bal be 4) Ba WDE 6 Ke) ye > toes ne eay Vik syste Cnn Wain ULES Sa A tits my, © RG iene qeryaus yee Cab at pve Iss Weta aS untise Se wiicey uh FU! Ah a Wy NUE Ea a Ua HOD BL FW aethe a ene ONAL nears, 0h sr ost has rede nN j i Kielce net Baht WE WML Aye baa Oe SMAI | Bi Ora i t ASOT UD UEW TAL a ELPLRGINS AL OFA) Patent ona ta Heusticw HOWEN ‘ Run BEN ANE EY tet cette WINS. SUG be my fF TER aa & ‘ Mien ty) atRneehe mgt ISN HW MEW Sey satires Ry ! Diy , Whteens SH Nhe Fh iP Gee Oe igh Asie urn LS Vise i PEW nse eBasoittie ie RAED EUS U NE % Poo a anes ASIP e gay cate i Hay NE Wer ak Oden Nd ems the TAT hy Nabe be cee y TEE Ora ‘ a SL TAMRON Rea nue A Weale p ave Dae AA MAURIE HL PAU delomativ ae eae (PuMasataoy \sin te Shree CTA USL AVENE R UM AUAN eee Tu Par y ere eae i reat Tec tig © POV AARNE SE af RS Ue te ae ne aclu Cele ROruryy sera. Raa gy wee i ee HMO eet Sap da OTM TINS Tee edict ? Le 7A Cue EASY ‘ rans at ROD ow “Emhanke tata ui Tui Puree ai Fe Re Cer HERS Rout Ne es eas hie We WNez URNS bowiLe) . Se A SUE Gs ‘ VSR es Me Mee tas Gethin gabe le aa tua Tnith a raha AMES RT raat u NaN tate . i hag an ma } te e Wtb aiee ‘S beh re LUO ATR new he Pty § pa SMITHSONIAN INSTITUTION i ll 0 nett Urieae utara