,0 o) o) ri a rt rl 03 ^fl B B oj ^ t- > g 4) .!-! -r-j ^ +J O CO o 00 Sg^Sfl^ar ag S '3 3 3 S ? » *3 o -S 5 -5 a & OS oj JS oj Oj a Oj >,/ a Oj 02 '3 Oj o o Oj Oj 02 05 ^a ai &faa 02 02 a 02 tn ^ ^ t-i Oj Oj 2 Si §§2§ 02 Sh Oj 4) . -°o -a os a 02 +j O rn 4> oj -rt o rt 4) >B 4) oj a a o u O ox O oj a m a a Of a o a o ft a 05 . 02 02 Oj 4) O B 9* 4) o> a ±Z a 75 02 S3 d 4 j' o> a .2 § .3 B 5-s*ft 5 o S aj sj ■a 03 b bp oj o « oj a h'o 2 g-oj s s P r— I Oj fl o w pfl O 4J O 02 flfe 03 VOLCANIC ISLANDS 337 J3 Xi JZ H EH EH ^A--C ^ © ©© © t tH ■* TH ~-^ v — ' " — ' +j -1-3 +J 4J A ■ dd d 60 60 &JD bo '53 0> O) 0) a a a w mm SQ ^ O) O 0) OJ D 0) D ■rt £ .£ .t •& -M -M -M OOOOOO OO a IS d oS s-i 'so > 60 -d ade oli ite. d OS .. fl OS ite site gra a? +j 60 «B O p'S'd 'co o oS fl d "d OS 0) CO .P ' H 'So OS > o d OS j_r w O ' *H -w ■'if "3 .Sn CO • co d «j a 6o a 1SS1 • O Js •£ 60 •73 j,Q +J +J a +. : cs • 0) h ugite-h livine andesi andesi "5 o > •— livin uff f ite. < C ■O EH < o (rt n3 0) P "3 ^ H 5r! £ — - x ~ 1 ,z ■ ri N OS ^s s « oi tn 03 CS oS a q £«2 O oj oS d-d CO CO g 1 OJ O WW a d |f u"Si •r 1 '^ 2 % cs « 03 ^ 0~ 60 d P o> o - d S .—( 03 33 co o o> o ,0 'S'Hi cs oS O d o> O "3 'So To G S3S a d O OS p 6B p d — oS oS d d CO CO oS oS : d ■ >» . +j .. a ■ d . +j : os X : ss . £ ; o • .a ! V • o ; >, ■ 3 • B3 • tf ■ C ^ O '. -i— . >5 . . O d o . tf oS • — ?-J oS u> , 60 o : « V o o S S S S oS oS ^ oS CO CO 0) CO oS oS -w oS gg«°S © 00- -a -d d oi .a .'S oS d d - d o 0) .5 +j 4> *s rt C' "3 X ^ ■CO ^ 71 >. - OPQ^"P5 CO „j OS S 9 2 d d oi u d D d X ■= ? t^ oS d &- o ^2 r >,L5 r HoS'i-i-<-ija;P i -'r; O w .„ a '3 o fl a ^ o-d d d 'd-n >= >> &'§)rt l 2 , § 2-g^-d a s -Jh o M d d o tw a co oS ' g d 8* 3 os a d os o> a> o 338 R. A. DALY PETROGRAPHY OF THE PACIFIC ISLANDS 3-8 '£, W C^ cd n3 A o jh a O V* OO 02 CD • 02 72 ^ !_i (1) ^ ^ oS o3 ? oS oS ft a o - u 020 s-l a a o o _j >w qj a eg > t> oS ft 03 ta a \_ a os" / H a B--S ^ os n _, £>C £ a±e a br a. a 9 r- a 33 cd cd cd cd 1> Sh ££ „ +J a 72 . • a aj • .o-SS • (H O a ft as. CD ft • a - ■ ■F 1 +J o"| _ 02 . 4J~ a • %* : !» CD . a -*-> ■QS • CD . -a •rt ft f-i a • ftaf ' a • ■• 02 • £ a a a= 02 . H a . a3 ft CD a be a a 2 CD a CD oj ll 02 02 2 a 02 WCCO K ~ &J0 ; , ... +3 a +j «-• j'_ BSa a a as a cd -a a 71 a - a w 3; 52 5j cj?S-S a2 43 mS o a -a' a & ft .-S & o 02 -a 02 a >o o -c ^; Em Em ; : ft^aa a m, S • °;a • • • . gg§g£~lp. 3lftjjgg ":|:s a cd a • CM CM 02 U a ^3 -a MH ^C ^A^-A X GO GO iM — - a a 03 a: a) a: 02 02 O; Fh F-i r: a a S §£ +j a S oa a ft ■S ft ^M ^+j O "I- 3 >> o1| a3 a g ftX2 +2 -w g+J a *r* a 02 .> 32 PC OK a£ a a o ft ft If- r5 ^ ft _, S ao^ CD fe O ^«2 "SCQ III a o.S a o a a 02 z 02 VOLCANIC ISLANDS 339 os .p B t*S P I -- -c •* o aT os£ p£ 6J)H t^T itic ephe hau wehi W P ■„. 0) oj . cc ^4 ih Js 3, e roc inol eni -M « S>> •r 1 a; P a> &-p ft •H P a aj r o> r- H +J •> pO +j -5 w £j >h O D OSP 5? 23 .b-H u U P P « 0) p ■ H oS cd aT 9 03 (D -p a> S ^ § P os ■3 .£ jz P ft bt QO OS CO 0) & 03 m 05 CO 03 ? 0) M fl ft-r p^ 03 - s s -s ? b N CO 03 OJ p P 03 >. P^ 03 t3 a flH 03-^ VI >i "£ a 03 P/ • .00 Guppy (11) Guppy (11) Marshall (2 (11) (11) (11) Guppy Guppy Guppy ^ %b o> p bo c 03 „ IP o> Pp? ."S N P S3. SpS •r-l 02 p-p OS p 03 S d S a; So p-O <1 Q 03 a3 P in .ij o3 02 50 o 'P r 1 ■y p To ^ © g 'P 'd& .. i> O *P -- P Pi 03 hO 03 qj bOrg P +3 0> T3 -^ a - P "55 "fj os TJ Pn 1) P P*^ S -S -, 03 P P Pn os P u O P >3 P rp ITS P P 03 h 02 02 P .H OS b0 X 340 R. A. DALY PETROGRAPHY OF THE PACIFIC ISLANDS 0> 00 W • — - r o ^ S3 - Newton. Newton. Newton. Newton. a a 03 a~ Teall and Teall and Teall and Teall and Cohen, Wi Marshall ( 05 4> S ^ S §^ 05 -y 05 rt r n - 3 °s ^3^ * -y ."S a3 o< a? ■ S "3 S s 2 a V 03 o fl 5 T3 a a bO bJO 05 i— i -rt 03 ^h b* :=! d — f- 1 zl S "3 2 o3 § .g c3 £ tH *BQ rgl 05 5° 03 ^j fcjc j3 •rH r-i rH r- 1 ._j 03 «j H * o3 +* '. 03 05 ci •° : o ; +j , cu a : ^ +J ■ 03 05 o3 •° : u . '. 03 OJ . a£ • 05 i—i . -« o . ^ d „pi o - : h 0) •^ •f . .a £? ! 03 £3' be W 3 03 S a; CO — ss n 'bid 0) 0) DOfl n "S o3 _ , m S 03 0) ^j g -r 1 ,-n 05 > 03^ WO o 0) ft §§^Hft gWfr^^&O > ^ 03 OH v2 O =5 O O cjd bibliography 341 Leading References 1. E. C. Andrews : Bulletin of the Museum of Comparative Zoology, Cam- bridge, Massachusetts, volume 38, 1900. 2. Bornstein: Peter. Geog. Mitt, 1914 (1), page 315. 3. E. Cohen: Neues Jahrbuch fur Mineralogie, etcetera, 1880 (2), page 23. 4. L. Cohn: Peter. Geog. Mitt., 1913 (2), page 315. 5. W. Cross : Professional Paper Number 88, United States Geological Survey, 1915. 6. R. A. Daly : Journal of Geology, volume 19, 1911, page 289. 7. A. Diesseldorff : Neues Jahrbuch fur Mineralogie, etcetera, 1903 (1), Referate, page 255. la. F. Dietrich : Untersuchungen fiber die Boschungsverhaltnisse der Sockel ozeanischer Inseln, Griefswald, 1892, page 7. 8. G. C. Frederick : Quarterly Journal of the Geological Society, volume 49. 1893, page 227. 9. J. S. Gardiner: Quarterly Journal of the Geological Society, volume 54, 1898, page 1. 10. F. A. Gooch : Tscher. Min. Petr. Mitt., 1876, page 133. 11. H. B. Guppy : The Solomon Islands, London, 1887. 12. H. B. Guppy : Observations of a naturalist in the Pacific between 1896 and 1899, London, volume 1, 1903. 13. A. Harker : Geological Magazine, volume 8, 1891, page 250. 14. M. Hartmann : Neues Jahrbuch fur Mineralogie, etcetera. 1878, page 825. 15. J. P. Iddings : Igneous Rocks, New York, volume 2, 1913. 16. E. Kaiser : Jahr. k. preuss. geol. Landesan., volume 24, 1903, page 110. 17. A. Kramer: Mitt, aus den deut. Schiitzgebieten (folio), 1908, page 169. 18. A. Lacroix : Bull. soc. geol. France, volume 10, 1910, page 91. 19. E. Lehmann : Tscher. Min. Petr. Mitt, volume 27, 1908, page 181. 20. W. Lindgren : Water-supply Paper Number 77, United States Geological Survey, 1903, page 14. 21. J. J. Lister : Quarterly Journal of the Geological Society, volume 47, 1891, page 590. 22. A. Liversidge : Proceedings of the Royal Society of New South Wales, volume 20, 1886, page 235. 23. P. Marshall : Transactions of the New Zealand Institute, volume 41, 1908, page 98. 24. P. Marshall : Transactions of the New Zealand Institute, volume 47, 1915, page 361. 25. P. Marshall : Transactions of the New Zealand Institute, volume 42, 1909, page 333. 26. P. Marshall: Transactions of the New Zealand Institute, volume 47, 1915, page 387. 27. P. Marshall : Subantarctic Islands of New Zealand, volume 2, 1909. 28. P. Marshall: Oceania, Heidelberg, 1912. 29. D. Mawson: Proceedings of the Linmean Society of New South Wales, 1905, page 400. 30. F. Mohle : Neues Jahrbuch fur Mineralogie, etcetera, B. B. 15, 1902, page 342 R. A. DALY PETROGRAPHY OP THE PACIFIC ISLANDS 31. E. Naumann : Zeit. deut. geol. Ges., volume 29, 1877, page 364. 32. K. Oebbeke : Neues Jahrbuch fiir Mineralogie, etcetera, B. B. 1, 1881, page 451. 33. W. R. B. Oliver: Transactions of the New Zealand Institute, volume 43, 1910, page 524. 33a. J. Petersen : Jahrb. Hamb. Wiss. Anst, volume 8, 1891. 34. P. D. Quensel : Bulletin of the Geological Institute, University of Upsala, volume 11, 1912, page 252. 35. E. Reclus : Ocean et Terres Oceaniques, Paris, 1889, page 389. 36. A. Renard: Report on the Petrology of Oceanic Islands (in Challenger Reports, volume on Physics and Chemistry, Number 2), 1889, page 149. 37. K. Sapper: Verb. XVII Geographentages zu Liibeck, 1909, Berlin, 1910, page 151. 38. K. Sapper : Mitt, aus den deut. Schiitzgebieten, Erg. Heft Number 3, 1910, page 58. 39. K. Sapper: Mitt, aus den deut. Schiitzgebieten (folio), volume 23, 1910, page 193. 40. R. Speight : Transactions of the New Zealand Institute, volume 42, 1909, page 241. 41. R. Speight : Transactions of the New Zealand Institute, volume 45, 1913, page 326. 42. R. Speight and A. M. Finlayson : Subantarctic Islands of New Zealand, Wellington, volume 2, 1909. 43. E. Suess : La Face de la Terre, Paris, volume 3 (iii), 1913, page 1040. 44. J. J. H. Teall : Quoted by D. Mawson, Proceedings of the Linnsean Society of New South Wales, 1905, page 411. 45. J. J. H. Teall : Quarterly Journal of the Geological Society, volume 54, 1898, page 230. 46. J. J. H. Teall and E. T. Newton : Geological Magazine, volume 4, 1897, page 151. 47. A. P. W. Thomas : Transactions of the New Zealand Institute, volume 20, 1887, page 311. 48. C. Velain : Bull. soc. geol. France, volume 7, 1879, page 415. 49. W. W. Watts and E. T. Newton : Geological Magazine, volume 3, 1S98, page 362. 50. M. Weber: Abhand. k. bayer. Akad. Wissen.. Math.-phys. KL, volume 24, 1909, page 287. 51. A. Wichmann : Journal of the Museum Godeffroy, Hamburg, Heft 8, 1875, page 123. 52. A. Wichmann : Neues Jahrbuch fiir Mineralogie, etcetera, 1875. page 658. 53. A. Wichmann : Journal of the Museum Goddefroy, Hamburg, Heft 14, 1879, page 213. 54. A. Wichmann : Neues Jahrbuch fiir Mineralogie, etcetera, 1879. page 663. 55. A. Wichmann : Tscher. Min. Petr. Mitt., volume 5, 1882, page 1. 56. A. Wichmann: Zeit. deut. geol. Gesell., volume 63, 1911, Monatsber., page 77. 57. T. Wolf: Verhand. Gesell. Erdkunde, Berlin, volume 22, 1895, page 253. 58. S. Yoshiwara : Geological Magazine, volume 9, 1902, page 290. NUMBER OP ISLANDS 343 Table III. — Number of Islands from which Eock Species is REPORTED Class Representing feldspar - ba- salt magma. 2. Representing transitions be- tween Classes 1 and 3. 3. Representing pyroxene - an- desite magma. 4. Representing transitions be- tween Classes 1 and 5. 5. Representing peridotitic magma. 6. Representing transitions be- tween Classes 1 and 7. 7. Representing magmas. "alkaline" Names given by authors Number of islands Olivine basalt 32 Feldspar (plagioclase) basalt 16 "Basalt" 26 Dolerite 14 Melapbyre 2 Olivine anamesite 1 Basalt porphyrite 1 Diabase 5 Olivine diabase 3 Diabase porphyrite 2 Gabbro 13 Olivine gabbro 2 Norite 1 - 118 Andesitic basalt 2 Basaltic andesite 3 Olivine andesite 6 Augite-olivine andesite 3 14 Augite andesite 37 Hypersthene andesite 11 Augite-hyperstbene andesite Pyroxene andesite 5 Augite-hornblende andesite 4 Hornblende andesite 12 •'Andesite'' (probably pyroxenie) 16 Augite porphyrite 1 Augite diorite 1 Diorite porphyrite 1 Feldspar porphyrite 1 Porphyrite l Diorite (in part ?) 7 - 102 Picritic basalt 7 Picrite 1 Dunite 2 Peridotite ] Wehrlite 4 Lherzolite l Diallage rock 2 11 Traehydoleritic basalt l Essexitie gabbro 1 Trachydolerite 2 Nephelite basalt 7 Nephelite-melilite basalt ?, Nephelite basantoid 1 Nephelite basanite (bearing haiiynite) 2 Basanite 1 Melilite basalt 1 Haiiynite-hornblende rock 1 Haiiynophyre 1 Magma basalt 1 344 R. A. DALY PETROGRAPHY OF THE PACIFIC ISLANDS Class 7. Representing magmas. "alkaline" 8. Quartzose rocks. 9. Residual types. Names given by authors Number of islands Limburgite (subalkaline ?) 6 Theralite '. 1 Tephrite 1 Monzonite 1 Nephelite syenite 1 Foyaite 1 Nephelite monzonite 1 Syenite (in part ?) 6 Trachyte 9 Soda trachyte 6 Phonolite 6 Nephelinite 1 Trachyandesite 1 Syenite porphyry 1 Tinguaite 1 Camptonite 1 Granite 12 Rhyolite or liparite 5 Quartz f elsite 1 Quartz porphyry 4 Quartz diorite 5 Quartz andesite 2 Dacite 3 Hornblende basalt 1 Hornblende gabbro 1 Bronzite basalt 1 Hornblende-mica andesite 2 Augite-biotite andesite 2 Trap-granulite 1 Kauaiite 1 Pyroxenite 1 Augitite 1 11 64 32 BULLETIN OF THE GEOLOGICAL SOCIETY OF AMERICA VOL. 27, PP. 345-386 JUNE 5, 1916 DOMTNANTLY PLUVIATILE ORIGIN UNDER SEASONAL RAINFALL OP THE OLD RED SANDSTONE 1 BY JOSEPH BARRELL (Presented before the Society December 28, 1915) CONTENTS Introduction 345 Prevailing views regarding conditions of origin 349 Criteria as to modes of origin of sediments 352 Description of the Old Red Sandstone formations 362 General relations 362 Uppermost Silurian — Downtonian formations 364 Lower Devonian — Lower Old Red 366 Middle Old Red — Orcadian formations 370 Relations to older rocks 370 Old Red of the Moray Firth. 371 Formations of the Orcadie basin in Caithness 371 Old Red Sandstone of the Orkney Islands 374 Interpretation of the Orcadian deposits 375 Significance of conglomerates 375 Significance of colors 376 Significance of mud-cracks and rain-prints 377 Significance of intercalated limestones 377 Lowest Upper Old Red of the Shetland Islands 378 The Upper Old Red Sandstones 379 Geography of the British Isles in Devonian time. . 382 General conclusions on Devonian climate in British Isles. . . 385 Introduction The Devonian system is represented in the British Isles, except in the extreme south of England, by a thick series of red sandstones, shales, and conglomerates. Red sandstone is the predominate outcropping rock, the colors of which range from light red to deep chocolate brown, but in places exhibit green, yellow, gray, and mottled tints. The series of for- 1 Manuscript received by the Secretary of the Society January 25, 1916. (345) 346 J. BARRELL PLUVIATILE ORIGIN OP OLD RED SANDSTONE mations which make up this system have been known collectively since the early days of geologic science as the Old Eed Sandstones. In Cornwall and southern Devon the Devonian is represented by slates, grits, volcanic materials, and thick masses of limestone holding marine fossils ; but in Ireland, Wales, Scotland, and the contiguous parts of Eng- land the formations of equivalent age constitute the totally different facies, the Old Eed Sandstones. These show their contrast not only in lithologic character, but in their barrenness of marine fossils- The Devonian age of these rocks was in fact only known at first because of their intercalation between the formations of the Silurian below and Carboniferous above, both identifiable by their marine fossils. What, then, were the geographic and climatic conditions which con- trolled the nature of the Old Eed Sandstone deposits? The question is raised to a higher degree of interest because the formations, though barren of fossils of the sea, yet hold at certain horizons an abundant and varied fauna of ostracoderms and fishes, the latter including forms which lead toward the amphibians. The footprints of amphibians can be traced back to the close of the Devonian period. It was in the Devonian that they doubtless rose from air-breathing fishes. We must look, then, for our own ancestral tree within certain families of these Old Eed Sandstone fossils. . Was it within the physical conditions which determined the nature of the Old Eed Sand- stone, or was it from the open sea that certain fishes grew to breath the vivifying air, to crawl on the solid land, and inaugurate from such humble beginnings that dynasty of terrestrial vertebrates which through all after ages was to lead in the march of evolution and rule over the living things of earth? It can be shown with high probability that it was from "the faunas of the Old Eed Sandstone, molded in adaptation to the physical and climatic conditions which surrounded them, that the amphibians arose. Studies in evolution are commonly regarded as within the field only of comparative anatomists, either students of fossils or of living forms ; but the ancient life history of the earth embraces not only the organic remains, as shown by the fossils, but a study of the environment which surrounded them and to which their life activities made efficient response. The analysis and evaluation of this environment is as important in the complete mosaic of knowledge as is a minute description of the fossils of successive faunas. Yet the student of fossils is not, by virtue of his knowledge of fossils alone, qualified to interpret fully the life surroundings of those former organisms. For this he must draw on his knowledge of physical geology and the habitats of living forms. He must be able to GEOGRAPHY OF BRITISH ISLES IN DEVONIAN TIME 347 — 60 Figure 1. — Geography of the British Isles in Devonian Time The loci of deposition of the Old Red Sandstone of the Lower and Middle divisions are shown and are interpreted as river basins, not as lakes, though subordinate lacus- trine phases would attend fluviatile deposition. The outcrops of Devonian formations are shown in black. 348 - J. BARRELL FLUVIATILE ORIGIN OF OLD RED SANDSTONE interpret the meaning of the sediments which envelop the fossil — sedi- ments which are the record of the environment of the living animal. There is a side, then, to the ancient life history of the earth which belongs to physical geology. It is as a physical geologist, and not as a paleontol- ogist, that the writer has taken up the present problem, but with the purpose of showing the relations between the environment and the evolu- tion of air-breathing vertebrates. This article was prepared in essentially its present form in 190? and made the basis of a paper on the causes of the evolution of land vertebrates presented orally on December 26, 1907, to the American Society of Vertebrate Paleontologists. 2 It was withheld from publication because the writer hoped to make a personal field study of the Old Red Sandstone formations of the British Isles- These plans did not materialize, but in the meantime he has published a paper on the somewhat similar deposits of the Appalachian geosyncline, 3 and he has also published a number of critical studies on the interpretation of the sedimentary record. A broad knowledge of the geologic problems involved and the correctness of the criteria of interpretation are as important for drawing conclusions as are the local facts to be interpreted ; this is the basis of judgment which has led to a decision to publish the following study on the Old Eed Sandstone, chiefly because of its bearings on the problems of the evolution of the amphibians. The descriptions are taken from various memoirs, but experience shows that the features which serve best as criteria of origin have been often overlooked in the field or left undescribed in their reports by the older geologists, partly because of a lack of appreciation of their significance, partly because, even Avhen present, they are often as difficult to find'" as fossils and require similar painstaking search of fresh exposures. The central conclusion reached in this paper is that the Old Eed Sand- stone formations were not deposited in lakes or estuaries, nor are they of desert origin. The analysis of their characteristics and comparison with sediments now forming determines them to be river deposits accumulated in intermontane basins. This is a kind of sedimentation not now found in the British Isles. For a close analogy one may turn to the basin deposits of the western United States laid down in the Tertiary period between the growing ranges of the Cordillera. This reinterpretation of the Old Red Sandstone of the British Isles is in line with that which has gone forward in America during the past fifteen years in regard to the 2 Abstract in Science, vol. xxvii, 1908, pp. 254, 255. 3 The Upper Devonian delta of the Appalachian geosyncline. Amer. Jour. Sci., vol. xxxvi, 1913, pp. 429-472 ; vol. xxxvii, 1914, pp. 87-109, 225-253. GEOGRAPHY OF BRITISH ISLES IN DEVONIAN TIME 349 mode of origin of the continental Tertiary deposits : once looked on as deposited in lakes greater in area than any now existing on the earth, they are now regarded as accumulations made chiefly on river plains. Such a reinterpretation for the Old Eed Sandstone involves radical changes in the conceptions of Devonian geography — no less a change than the substitution of land surfaces, occasionally flooded, as a replacement in the mental vision of wide and permanent bodies of water. If the new interpretation is well founded, it means that such terms as "Lake Cale- donia" and "Lake Orcadie" should be turned into the "Caledonian and Orcadian basins/' Peevailing Views regarding Conditions of Origin Hugh Miller regarded the Old Eed Sandstone as a marine deposit, reaching this conclusion by direct comparison of the structures of the solid rocks with near-by tidal deposits now being made. It was soon per- ceived, however, by British geologists that the sediments and organic contents were of different types from the obviously marine Devonian strata to the south. These distinctions led Godwin-Austen in 1855 to a view, previously maintained by Dr. John Fleming, that the Old Red Sand- stone was laid down in great fresh-water lakes or inland seas. This in- terpretation soon became generally accepted- During the next generation the geologist who gave most thorough study to the subject was A. Geikie. He separated the Old Red Sandstone into several basins of deposition. According to Geikie, Lake Caledonia stretched across central Scotland, and within it were deposited a maximum thickness of perhaps 20,000 feet of strata. Other basins were occupied by the Welsh Lake. Lake Cheviot, Lake Lome, and Lake Orcadie. In the latter area there are exposed as much as 16,000 feet of strata. These views are developed in his paper of 1877-1878 4 and are summarized in his textbook in 1903. Macnair and Reid in 1896 give, however, what they regard as reasons for holding that the fresh-water lake hypothesis of origin is "utterly untenable" and come back to the view of Hugh Miller, that the Old Red Sandstone is marine. 5 In 1904 Good child published a paper which embraces the following statements : "There is no satisfactory reason for regarding any of the Scottish rocks of Devonian age as of marine origin ; and, on the other hand, there is much to be said in support of the view that they were all formed under continental con- 4 On the Old Red Sandstone of western Europe. Trans. Roy. Soc. Edinburgh, vol. xxviil, pt. ii, pp. 345-452. 6 Geol. Mag., Decade 4, vol. iii, 1896, pp. 106-116, 217-221. 350 J. BARRELL FLUVIATILE ORIGIN OF OLD RED SANDSTONE ditions, and under conditions of climate which, though doubtless varying much from time to time, were yet, on the whole, characterized by an annual rainfall decidedly below the average in amount. It is this feature which has imparted a common character to the whole of this series of rocks." "The fauna of the Upper Old Red consists almost exclusively of fishes which probably found their way into the sediments from the rivers of upland origin, whose waters were dissipated by the excessive evaporation when they reached the lowland area." "As regards the mode of origin of the Old Red of the Caledonian area (Lower Old Red), there appears to be evidence of a satisfactory nature that the whole of this vast formation was accumulated under continental condi- tions, partly in large inland lakes, partly as torrential deposits of various kinds, partly as old desert sands, and partly as the result of extensive volcanic action." 6 With these views the present writer is in accord, except that he would give first place to true fluviatile deposition, spreading sediments on broad and flat river plains, a form of deposition intermediate between torrential and lacustrine and yet quite distinct from either. It is one which is not, however, specifically mentioned by Good child. In 1907 the writer, unaware at the time of Goodchild's paper, presented the evidence showing the wide-spread development of floodplain deposits in the Old Eed Sandstone and the presence of a fluviatile piscine fauna. This view was, however, only published at the time in brief abstract. 7 In 1908 Walther published a volume entitled "Geschichte der Erde und des Lebens." In this is a chapter on the Devonian continent which lay northwestward from central Europe and which included Scandinavia and the British Isles. He includes in this same continent Greenland and the Canadian Shield- The name he gives to it is "The Old Eed Northland." The chapter, pages 254 to 271, is largely of general statements and draws an analogy between the climatic and sedimentary conditions of this con- tinent and the present interior of Asia and Australia. Walther states, for example (page 259) : "The northern Devonian land consisted of many parts. Many lines of frac- ture passed through the older mountain structures. Some of the principal lines extending northeast are yet visible. Often the sandstones lie on the steep-walled cores of the ancient mountains, filling up deep basins, so that great regions of deposition, as the 'lakes' of Orcadie, Caledonia, and Lome, can be distinguished. One should not conceive under that term, however, enduring bodies of water, but rather wide basins surrounded by mountains limited by temporary sheets of water, water which converged to shallow lakes of variable 6 J. G. Goodchild : The older Deutozoic rocks of North Britain. Geol. Mag., Decade 5, vol. i, 1904, pp. 591-602. 7 F. B. Loomis : The American Society of Vertebrate Paleontology. Science, vol. xxvii, Feb., 1908, p. 254. PREVAILING VIEWS AS TO CONDITION OF ORIGIN 351 area and depth. These would dry out until the next period of rains filled them." "We attain therefore the conception that the northern continent, already in the Upper Cambrian, again in the Upper Silurian, and further through the whole of the Devonian period, even into the Lower Carboniferous, possessed a hot desert climate whose dry periods were broken only seldom by the down- pours of thunder-storms." AValther considers the red colors as original to the sediments. From this he derives his name of the Old Eed Northland. There has been pre- sented, however, no evidence that the original sediments had the present color tones of the solid rocks. Dehydration and change in color of the iron oxide commonly accompany thorough cementation. Neither has there been presented evidence of pronounced wind action or the existence of evaporation deposits. It seems, therefore, that this picture of a desert climate is overdrawn. A truer view may center on a semi-arid climate and a land whose character lies half way between the permanent lakes of the British geologists and the permanent deserts of Walther. Jukes-Browne published in 1911 a revision of his volume on "The Building of the British Isles." In the chapter on the Devonian period he discusses the evidence which goes to show that the basins of Old Eed deposition were originally much wider than their present limits. The small areas he regards as remnants isolated by erosion. The difference in the faunas and floras of the Caledonian and Orcadian areas must therefore, as Traquair has argued, be due to shifting of subsidence and consequent shifting of deposition during the course of the Devonian, and not to the maintenance of permanent mountain barriers between them. Jukes-Browne recognizes Goodchild's arguments for a climate at times desert in character, but refers to the areas of deposition as lakes. These lakes, under his conception of greater area of deposit, become in fact fresh-water seas of far-reaching extent- Thus the essentially lacustrine conception is maintained and fluviatile deposition is not accorded a place of importance. In -America the trend of opinion in regard to the proper interpretation of the Old Bed Sandstone has been increasingly in the direction of ascrib- ing a larger importance to fluviatile deposition, but these views are mostly unpublished and have naturally therefore carried no weight in British opinion. The earliest American expression of view, in what is here re- garded as the right direction, was, so far as the writer is aware, that by T. C. Chamberlin in 1900. This pioneer in geologic thought, in a paper of philosophic nature on the habitat of the early vertebrates, suggested that the Old Bed Sandstone was deposited under conditions not unlike XXVI — Bull. Geol. Soc. Am., Vol. 27, 1915 352 J. BARRELL FLUVIATILE ORIGIN OF OLD RED SANDSTONE those now found m the Great Valley of California. There was, however, no presentation of detailed argument. Since then Grabau and the pres- ent writer have independently and repeatedly written on the part which the rivers of later Paleozoic time have played in continental deposition, dealing chiefly with the rivers flowing westward from Appalachia. The arguments for them apply with variations to the British deposits also. Grabau studied the latter some years since, in the field, and has stated that he is in full agreement with the conclusions of this paper in regard to the dominantly fluviatile character of the sediments. Approaching the subject from another side, Doctor O'Connell has written an extensive paper, entitled "The Habitat of the Eurypterida," to be published as one number of the Bulletin of the Buffalo Society of Natural Sciences. She has treated the subject of the Old Eed Sandstone at some length, pre- dominantly, though not exclusively, from the standpoint of the faunal evidences. The conclusion is reached in her paper that the eurypterids were preeminently a fluviatile fauna, and their associations in the Old Red Sandstone show that the latter is dominantly fluviatile. This is a supplemental line of evidence, and she states as a conclusion : "I am convinced that the detailed study of the geological and geographical distribution of the eurypterids (and I think the fishes, also) will do more than anything else to prove that they lived in the rivers, and that many peculiar deposits will be easily explained as of floodplain or delta origin when the importance of fluviatile deposition is once realized." 8 In the Textbook of Geology by Pirsson and Schuchert, published 1915, the view is adopted by Schuchert that the Old Red of Scotland is probably wholly continental. The opinions of Goodchild and Walther are quoted, but the basins, following the British custom, are still spoken of as lakes ; and it is stated # that Lake Orcadie may have extended as an estuary north- east to Christiania, in southern Norway, and possibly even to Petrograd, in Russia. The south Irish deposits are regarded by Schuchert as marine.* Criteria as to Modes of Origin of Sediments ° It is evident from the preceding review of opinions that a sound con- clusion regarding the geography of the British Isles during the Devonian period depends not so much on the accumulation of new facts of observa- 8 Personal communication. * Pages 715-717, 1915. 9 This topic is essentially a summary of such arguments as bear on the present prob- lem from articles published by the writer in the Journal of Geology and Bulletin of the Geological Society of America. If this article dealt only with American geology, a very brief restatement of these criteria would doubtless suffice ; but as the subject in hand is CRITERIA AS TO MODES OF ORIGIN OF SEDIMENTS 353 tion as on the soundness of the principles of interpretation- Especially important is the quantitative evaluation of the significance of various characteristics, since it is true that some lacustrine conditions will accom- pany fluviatile deposition and vice versa. Loose sand may be subjected to wind action and build dune structures even in humid climates. Semi- aridity is a condition which will permit the development of features which are also found in the most arid climates, and yet semi-aridity does not exclude the existence of abundant life. Semi-aridity is more widely asso- ciated with seasonal rainfall than with deficient precipitation throughout the year. The rainfall may be as abundant during the rainy season as that of truly humid climates. Fluviatile deposition is vitally distinct from either lacustrine or desert conditions, and semi-aridity is a type of climate equally distinct from either normally humid climates, on the one hand, or arid climates on the other. Yet it is fluviatile deposition under a semi-arid climate which is here held to have characterized the laying down of the Old Red Sandstone. This interpretation is essentially new and lies between that of the lakes of Austen and Geikie and the deserts of Walther and Goodchild. As stated to a considerable degree in a previous paper, 10 many features from which an observer may gain an impression as to the mode of origin of a deposit are really not definite proofs. Thus ripple-marked, cross- bedded, and fossil-barren deposits may be developed either beneath a permanent water cover or on river plains, though doubtless the quantity and quality of their development differ in the two cases. Red beds have been regarded by some, because of their color, as evi- dences of terrestrial deposition ; by others, of seas barren of life. Again, they have been cited as indications of a deeply decayed regolith, of a humid climate, or of aridity. The present writer holds that redness in rocks may in fact accompany any of these conditions, and is not therefore a criterion by itself of any one. 11 Mud-cracks and conglomerates have been cited usually as evidences of ancient tidal flats and beaches, but are now observed to occur chiefly in river deposits of continental interiors. Those formed in the littoral zone, furthermore, except on the fronts of deltas, are rather unfavorably situ- the mode of origin of formations in the British Isles and the special feature of the paper is the application of these criteria, it seems desirable to give a fuller review of the principles on which they rest. This is because it is hoped that the subject will he of interest to British geologists, to many of whom the papers on which this discussion rests may not be readily accessible. 10 Joseph Barrell : The Upper Devonian delta of the Appalachian geosyncline. Amer- Jour. Sci., vol. xxxvi, 1913, pp. 436-440. 11 Relations between climate and terrestrial deposits. (The climatic significance of color.) Jour. Geol., vol. xvi, 1908, pp. 285-294. 354 J. BARRELL PLUVIATILE ORIGIN OP OLD RED SANDSTONE atecl for geological preservation, and are to be distinguished by associa- tions which are commonly absent in ancient examples- 12 Thus it seems clear that it is the character, the quantitative development, and asso- ciations of the stratigraphic features, as well as the nature of the en- tombed organic remains, which are significant rather than their mere occurrence. In general, traditional criteria are liable to lead into error, because they become accepted as axioms and are applied without further thought. They lag behind the development of a science, whereas the very word "research" implies the necessity of continually testing the correspondence of the images of science with nature. Attention may now be turned to what may be regarded in the present- state of knowledge as fairly definite criteria, which will be of use in discussing the origin of the Old Eed Sandstone formations. A fresh-water origin for these deposits will be assumed as established and accepted by practically all geologists, but many are accustomed to thinking of the presence of fresh-water faunas and the absence of marine organisms as implying nothing more than land-locked, brackish-water bodies, perhaps fresh near their heads, as seen in Chesapeake Bay and the Baltic Sea. The deposits, following this conception, would be called estuarine. If clearly separated by a land barrier from the open ocean, then the deposits have in the past been usually regarded as lacustrine and taken to imply the former existence of great fresh-Avater lakes or inland seas. True estuaries must have been of limited development, however, in past times, since, like lakes, they are temporary features and are made by a rising sealevel against a land surface previously dissected by erosion- Deposits made in nearly landlocked water bodies, like the Baltic, should rather be called "bay formations." Where there is some intermingling, interfingering, and gradation of marine and fresh -water deposits with their respective faunas, the relation which is implied is commonly that of the shore of a. delta. Such a shore is marked by shifting growth, by wide advances and retreats, and the inclosure of marginal lagoons of various degrees of salinity. 13 In the Old Eed Sandstone of the British Isles these marginal delta phases are not so important as in the Devonian on the continent-of Europe and in eastern North America. They do occur, however, in northern Devon and in southern Ireland. The problem of the British Old Bed Sandstone really resolves itself into the discrimination between true 12 Joseph Barrell : Mnd-ciackp as a criterion of continental sedimentation. Jour, of Geology, vol. xiv, 1906, pp. 524-568. 13 Joseph Barrell : Criteria for the recognition of ancient delta deposits. Bull. Geol. Soc, Am., vol. 23, 1912, pp. .377-446. CRITERIA AS TO MODES OP ORIGIN OP SEDIMENTS 355 lacustrine deposits on the one hand and terrestrial deposits on the other. Terrestrial deposits may in turn be classified into fluviatile, torrential, glacial, and eolian. Lakes are characterized by being permanent water bodies as measured by the changes of decades, centuries, or longer periods. Constructional alluvial plains, on the other hand, are largely turned into temporary lakes during the annual season of flood and give thereby some similarities to lacustrine deposition, but are drained during the dry season, and their life is that of the land. In the upper part of the alluvial plains the river grades are steeper, the waste coarser, the flooded condition more brief. The marks of sub- aerial exposure are more dominant, as seen especially in the greater oxida- tion of sediments, giving red rocks on cementation, and the greater paucity of fossils. In the lower part of the plains, on the contrary, the grades are flat, commonly less than a foot per mile; the sediments are finer and more impervious to ground waters; the flooded condition is of greater annual duration, and in playa lagoons and swamps may in fact be permanent for a series of years. The rivers, owing to minor climatic fluctuations, will during one series of years entrench themselves and minimize the flooding of the plains. During other series of years they may rise high, flood their plains broadly and for longer periods, depositing an excess of waste- The channels also shift and the natural levees are built upward to above the level of all save the highest floods. In the same geologic section, therefore, river action is characterized by variable relative durations of exposure to flood waters and exposure to the air. True lake conditions, then, as distinguished from the temporary and subordinate water bodies connected with fluviatile deposition, are char- acterized on the margins by wave-formed sands and conglomerates, as distinguished from the current deposits of streams. The basins of per- manent lakes must be sufficiently deep to have the greater portion of their bottom below wave base. The shore deposits of lakes will conse- quently, for any one horizon, be essentially marginal, and the greater area of the lake will be marked by clays and fine silts deposited from suspen- sion. Regular, even lamination, giving rise to paper-thin shales, will be typical. Lenses of sand, but especially beds of conglomerate, will be absent. The presence or absence of sand is in itself, however, of doubtful value as a criterion. Conglomerates, on the other hand, when they are of wide-spread occurrence, are of the highest significance. For floodplain conditions there are a number of definite criteria de- pending on the texture of the deposit and the climate under which they are laid down. Where the deposits are argillaceous and the climate is 356 J. BARRELL FLUVIATILE ORIGIN OF OLD RED SANDSTONE one subject to dry seasons, the most important criterion consists of marks of subaerial exposure, developed both broadly and vertically through mechanical sediments. These consist chiefly of mud-cracks, rain-prints, and impressions of roots in situ. Where these occur in such relations they seem to show conclusively a terrestrial origin, and not to represent the littoral facies of a marine or lacustrine formation. Not only, as stated previously, does the shore form at any one time but a narrow border to the accumulating mantle, but, except on the front of an advancing delta, it is commonly a region of erosion rather than of accumulation- Tidal flats are furthermore flooded and drained twice per day, with the result that they are always limited in width and are cut by deep channels. These conditions are quite distinctive from those which are found inland. For over deltas and playa basins, on the contrary, every part is alternately covered by water and by air for considerable periods of time. Where the climate is suitable mud-cracking is developed habitually and on a broad scale. Mud-cracking in chemical sediments — that is, in limestones^ most typically the impure or "water-limes" — must, however, be distinguished in significance from the cracking in claystones. Lime carbonate is car- ried in solution and its deposition as limestones requires a comparative absence of sand and clay, the mechanical deposits carried by rivers and by waves. The solutions, to have sufficient concentration, may come from permanent water bodies, either lakes or seas. The deposit, therefore, comes not from the direction of the land, but from the direction of the water. The cracking goes on between the extreme levels of high and low water, and the slight shifting of level is not a tidal, but at least a seasonal, phenomenon. Such mud-cracking of limestones is a playa phenomenon, and especially in certain earlier ages, when the lands were baseleveled and lay awash with the sea, broad areas seem to have been at times marine playas. Marine fossils, often of depauperated kinds, occur sometimes in mud-cracked limestones. The nearest approach in the modern world is found, doubtless, in the Bann of Cutch, an area of 10,000 square miles south of the Indus, flooded by the sea for a part of the year, during the period of onshore monsoon winds. In the detection of mucl-eracks in ancient formations reasonable care must be used to avoid mistaking for them a polygonal cracking of the rock arising after its solidification. The two, however, are readily dis- tinguished- True mud-cracks always have a filling; the polygons are irregular, but do not show irregularity constant in one direction. True mud-cracks, although easily separated from simulated features, are, how- ever, often very difficult to detect, as the filling may be identical in nature CRITERIA AS TO MODES OP ORIGIN OF SEDIMENTS 357 with the original stratum; and weathering, furthermore, destroys such strata very rapidly unless the shales are interlaminated with sandstones. Rain-prints must be separated from the pits made by escaping marsh gas and are more usually marked by a spattered surface of the mud than by a few concave depressions. Root-marks should show a branching pattern and finer tendrils given off from the larger marks. Footprints may be very obscure, but the test in that case is the regularity of recurrence on the stratum, owing to the stride of the animal and its regularity. Finally, a criterion of special application to much of the Old Red Sand- stone deposits is found in the nature of the conglomerates. Conglomerates now forming which are both thick and wide-spread, as Blanford and Bonney have shown, are observed to be of fluviatile and not littoral origin. This is because rivers are able to carry gravels far out over a subsiding river plain; but the waves, on the contrary, tend to keep gravel banked in the zone of the shore. During an advance or re- treat of the sea, basal conglomerates may be widely spread, but they are thin and often wanting, reaching their greatest development among islands or along an irregular rocky shore able to withstand the waves for some time during a rising sea. A maximum limit to wide-spread basal marine conglomerates seems to be 100 feet, and therefore broad con- glomerate formations of greater thickness are evidences of terrestrial accumulation. 14 - They may, of course, be of terrestrial origin also when thinner and more limited; but in that case, so far as these criteria deter- mine, they may also be marine or lacustrine. Other characters which are, under certain circumstances, of importance are found in the detailed structures and associations of the conglomerate beds. Gravels which have been carried considerable distances by power- ful rivers may be as well sorted and as thoroughly worn as are the wave- worn gravels of the shore, but weak rivers carry their debris especially during great floods. Coarse and fine are swept along together; sorting and wear are less perfect ; streaks of gravel are swept out from the basin margins and intercalated with dominant sandstones and even with beds of shale. River gravels are shingled by the currents so that the longer diameters of the pebbles dip upstream, giving a faint appearance of false bedding, which on the average, unlike the false bedding of sandstone strata, dips toward the basin margin. Shore gravels, on the other hand, are developed parallel to the shore. The onshore waves have a greater force than the undertow and the shingling dips away from the shore, or runs out laterally from protruding headlands. 14 Joseph Barrell : Some distinctions between marine and terrestrial conglomerates. Abstract. Bull. Geol. Soc. Am., vol. 20, 1908, p. 620. 358 J. BARRELL FLUVIATILE ORIGIN OF OLD RED SANDSTONE Lastly, from a larger point of view, shallow-water deposits, where of great thickness and built of land waste, must very often be dominantly terrestrial, especially on the side of the deposit toward the source of supply. This may be seen by reflection on the broader physiographic relations which attend them. If subsidence were the dominating regional feature of the tectonic movement which results in erosion and sedimenta- tion, then there would be a passage to lacustrine or marine conditions. If, however, uplift in the regions of erosion dominates either areally or vertically over subsidence of the regions of deposition, then the basins will be kept filled to the level of the river grade; there will result an excess of sedimentation. The rate of deposition in any section consisting throughout of shallow-water beds is determined not by the rate of erosion, but by the rate of subsidence. The excess of sediment will be carried farther off, the balance being finally deposited in marine formations. But in all except the most arid regions the great carrying agent is river water. Where sedimentation is in excess of subsidence the deposits will consequently bear usually the marks of fluviatile deposition- The condi- tions of sedimentation can then be divided into two broad classes depend- ing on the ratio of sediment to subsidence. There will only rarely and temporarily occur that delicately balanced condition where subsidence takes place at the same rate, but keeps slightly ahead of deposition, giving rise to permanent yet shallow bodies of water. Nevertheless this balanced state, maintaining a thin yet permanent cover of water, is the one usually assumed to have existed during the Paleozoic when marks of shallow water, of currents, and of exposure to the air are found through great thicknesses of mechanical sediments. It is seen from this analysis that such features must be much more commonty the marks of floodplains and without any necessary relation to shores. The emphasis of discussion has been placed thus far on the distinctions between fluviatile and lacustrine deposition. There remain to be con- sidered criteria of another category — those which serve to distinguish the fluviatile deposits of seasonally dry and semi-arid climates from the dominantly torrential and wind-borne deposits of true deserts. In desert mountains there is a maximum of rock exposure. Eock- breaking is due to sun and frost. The unweathered rubbish washes and creeps down the slopes and is broken finer until it is within the reach and greater power of the occasional cloudbursts. These sweep along coarse and fine materials together, but soon lose power, and torrential action, as marked by this heterogeneity, is confined to the perceptibly steep slopes of streams within a few miles of the basin margin. .On the desert basin plains the wind is the chief agent of transportation. CRITERIA AS TO MODES OP ORIGIN OF SEDIMENTS 359 During the long dry periods the sand is shifted by each high wind; the small grains acquire a millet-seed roundness foreign to the work of water ; the dust is taken up in the air and exported from the desert basin. Where sand blows across stony floors the pebbles become faceted, giving the sharp-edged and polished forms known as dreikanter. Where sand progressively accumulates through wind action the base of each dune remains behind in the forward marching of the dunes. The dunes of the Sahara are frequently 300 feet in height. Owing to the shifting of winds, all of one series of dunes may be removed elsewhere; a hundred feet of the base of others may remain. Cross-bedding of marked irregu- larity on a gigantic scale is developed from this in the growing deposit. The cross-bedded members of dune sands are not limited by parallel and horizontal surfaces, and the lines of cross-bedding are broad sweeping- curves, tangent to the horizontal at the base and tens or even some hun- dreds of feet in their radii of curvature. In true desert deposits there is a marked absence of the argillaceous component, although the sediments of the earth as a whole consist of 80 per cent shale. The constituents of shale are borne from the desert region as dust, deposited hundreds of miles away as loess, or mixed with the deposits of more humid climates, or with the limy oozes of open seas or deep permanent lakes. Where large rivers maintain their way across deserts or terminate within them the alluvial deposits are very largely reshaped by wind, as over the delta of the Indus, and notable deposits of gypsum and salt mark the sites of lagoons or interior seas. Contrast the preceding with the conditions of sedimentation under semi-arid or seasonally dry climates. This is an entirely distinct cate- gory, yet one which is practically not recognized by Walther, the apostle of deserts. Climates of this character are marked by a concentrated rather than a deficient rainfall. If the temperature of the wet season is not too cold, an abundant vegetation may grow and a rich and varied fauna may inhabit the land. Such climates exist over much of the tropics, as seen in Africa and India ; over broad continental interiors in the temperate zones, as seen in the great wheat and cattle lands of the globe. Under semi-arid climates water is the great transporting agent. Dur- ing the rainy seasons the shrunken rivers fill again, flow long distances to the sea, sweep gravel far from the hills, wash out the evaporation deposits of the dry season, cover up the mud-cracked alluvial flats with new lavers of silt and clay- During the dry season the herbaceous vegetation turns brown, the humus is oxidized out of the soil, the slimes left by the last 360 J. BARRELL PLUVIATILE ORIGIN OP OLD RED SANDSTONE flood waters are dried out and cracked, the polygons resulting from the cracking curling up on the edges. From the stream channels sand is blown to leeward and buries, perhaps permanently, the dried and curled plates of mud. The wind plays a much less important part, however, than in true deserts, not so much because of the more restricted period of action as because of the greater efficiency of the water and the binding action of the vegetation of the plains. Clean sands, however, still give rise to dunes, and even in humid climates dune action is not wholly absent. The significance of limestone deposited in fresh water needs to be con- sidered. It is found to occur under a number of dissimilar conditions. First, over semi-arid or arid floodplains evaporation of ground water takes place throughout much of the year. The dissolved material which is most abundantly precipitated because of this evaporation is calcium carbonate. Except under truly arid climates, the salt, gypsum, and alka- line carbonates and sulphates are washed out by the following flood waters ; but the lime, once precipitated, is relatively insoluble and remains as a cement in the silt and sand. If the waters flow from limestone regions, they are correspondingly richer in lime, and this process may occur in relatively humid climates, as in that over the deltas of the Ehine and Ehone. For the precipitation of lime in floodplain deposits from waters flowing from crystalline rocks, it would appear, however, that the climate must be at least semi-arid in its dryness. In India this cement in the alluvial soils forms impervious nodular layers known as kankar; in Mexico it occurs often as a granular impure limestone, called caliche, and in the' western United States such sands and gravels cemented by lime, chiefly at the upper level of the ground water, are known as mortar •beds. Probably the agency of plants is inconsequential in this class of lime deposits. Second, there are found to be deposited in fresh-water lakes and streams lime muds, known as marl, or concretions of nearly pure calcium car- bonate- In springs, salt lakes, and in tropic seas deposits of similar com- position are also found. In recent years, in fact, it has become recognized that Chara mosses, green and blue-green algse, and bacteria are the agents of calcareous precipitation to such an extent that these lowly vegetable forms are now thought by many to be quantitatively the most important geological agents in the making of limestones. For the deposition of fresh-water limestones, chiefly of algous origin, the only condition appears to be that of some degree of warmth and of richness in calcium bicar- bonate. Under the humid climate of the eastern United States such marls have been extensively deposited in the lakes of northern Indiana. 15 15 Blatchley and Ashley : The lakes of northern Indiana and their associated marl deposits. Indiana Dept. of Geology and Natural Resources, 25th Ann. Rept., 1901, pp. 31-321. CRITERIA AS TO MODES OF ORIGIN OF SEDIMENTS 361 Eecently Roddy has made a careful investigation of the lime concretions discovered by him in streams. He shows that in the streams of south- eastern Pennsylvania concretions made chiefly by blue-green algae grow freely where the amount of calcium bicarbonate in the water reaches 300 parts per million. The growth takes place only during the warmer months of the year. 16 In humid climates it is, however, only under exceptional conditions that stream or lake waters could attain the content of calcium bicarbonate which would result in a free growth of such fresh-water limestones. Under conditions such as those of the Old Red Sandstone basins, where the waters came mainly from crystalline rocks, for the formation of broad and abundant nodular limestones, the cornstones of the British geologists, a climate- of at least semi-arid character would seem a necessary postulate. A truly arid climate is, however, not needed; in fact, the absence of gypsum and salt as associated deposits seems to show that aridity was not present. The physical conditions best adapted for the laying down of the cornstones are those of broad, shallow bodies of warm water, lying- in basins beyond the reach of abundant mechanical sediments and largely evaporated during a dry season. Not until the Cretaceous were grasses evolved, and trees even now can not grow in compact forests on semi-arid lands. So far as present knowledge goes, there does not seem to have been during the Devonian a binding of the soil by vegetation adequate to hold the upland waste of semi-arid climates. If that be true, there would appear to have been at that time but little distinction between the weathering phenomena of truly desert and of semi-arid uplands- No effective soil mantle would have held organic acids, atmospheric carbon dioxide, and moisture within its pores. As soon as the rocks were broken fine enough, the waters of the rainy season would sweep the debris away. Lichens, it is true, would live on the bare rocks, and some bacteria in the regolith. In the wet Season these lowly agents would carry on, in a measure, chemical decom- position, but they could hardly be counted as binders of the soil. From these barren uplands there would be a graduation of conditions to the flat parts of river basins. Under a semi-arid climate the level of ground water would not sink far below the surface of these flats. There vegeta- tion could flourish and chemical decay go forward the greater part of the year. Oxidation of the soil would become imperfect: grays and greens as subsoil colors would result. These features are seen to be both quali- tatively and quantitatively different from the typical lagoons of desert climates. 18 Roddy : Concretions in streams formed by the agency of blue-green algae and related plants. Proc. Amer. Phil. Soc, vol. liv, 1915, pp. 246-258. 362 J. BARRELL FLUVIATILE ORIGIN OP OLD RED SANDSTONE These are the kinds of tests which must be used in the study of the mode of origin of the formations of the Old Ked Sandstone. Description of the Old Bed Sandstone Formations general relations In order that each reader may possess the basis for an independent conclusion on the conditions of origin of a series of formations, detailed descriptions must be given of the significant characters. Since the con- ditions may shift from stage to stage, these descriptions must cover the whole series with some degree of completeness. 17 The results of continued investigations by British geologists have shown that the Old Bed Sandstone conditions of deposition began in Scotland and in north England in uppermost Silurian times, the marine rocks of the Ludlow passing conformably, by oscillation of conditions, into barren, red and yellow, cross-bedded sandstones, with some conglomerates, and red to gray or green mud-stones. Goodchild states that a violent uncon- formity separates the base of the true Old Bed of Devonian age from these basal rocks, which the British Survey has named Downtonian, but for which Goodchild prefers the name Lanarkian, from their typical ex- posure in Lanarkshire, in southern Scotland. 18 Above the Downtonian and unconformable to it comes the Lower Old Bed Sandstone, characterized by red and purple sandstones, gray sandy flagstones, and coarse conglomerates. The most characteristic fossils are Cephalasp-is, Pteraspis, Climatius, and Pterygotus. It is typically shown south of latitude 57°, especially in the Caledonian area, according to Geikie, some 20,000 feet of beds being there exposed. Lower Old Bed deposition was closed by disturbances which shifted the regions of great subsidence to the north, especially around the Moray Firth and in Caithness. These are the Orcadian formations, estimated to reach as much as 17,000 feet in thickness. The typical fossils are Estheria, Dipterus, Osteolepis, Homosteus, Mesacanthus, Coccosteus, Ptericlithys, etcetera. The Middle Old Bed is in turn separated by a profound unconformity from the rocks stratigraphically above, which constitute the Upper Old Bed, but recently Flett has shown that the Old Bed of the Shetland Islands bridges to some extent this hiatus. The plants of the Shetland 17 Descriptions of the features significant of origin will be quoted more fully for the benefit of American readers than if this article were to be printed in a British journal, for the reason that much of the original literature may be difficult of access to many American geologists. 18 J. G. Goodchild: The older Deutozoic rocks of Great Britain. Geol. Mag., Decade 5, vol. i, 1904, pp. 591-602. DESCRIPTION OP THE FORMATIONS 363 area appear to represent a distinct flora, but the vertebrate fauna shows relationships to the Upper Old Eed; consequently it may be the lowest, or one of the lowest, zones of the Upper Old Eed. The well determined genera are A sterol epis and Holonema. 19 The typical Upper Old Eed is found resting with angular unconformity on the Middle Old Bed and all older rocks, but passes conformably into the Carboniferous rocks above. The basins of deposition only partly corresponded with those of the older red sandstones. The rocks are characteristically red and yellow sandstones and conglomerates. Bothrio- lepis major and HoloptycMus nobilUsimus are typical species. From this series of unconformities and shiftings in regions of deposi- tion it is clear that great crust movements were intermittently in progress in this zone during latest Silurian and all of Devonian time. These were accompanied by much igneous activity, since lava flows, breccias, and volcanic ash make up large portions of the sediments, especially of the I ower Old Eed. Uplifts along certain axes must have been in progress to have supplied the great quantities of coarse waste. Downsinking of adjacent areas must have accompanied the former in order to permit such thick accumulations. Fault zones and step-faulting are implied ; perhaps also true folding. In these ways may be partly explained the remarkable variations in thickness of formations, the disappearance of others, such as the British Survey has recently shown to exist in southern Wales and in northern Devon and Somerset. Here an east-west axis runs through the Bristol Channel. On the northern margin of this axis in Glamorgan the total thickness of the Old Eed sandstone is reduced to 400 feet, but on the northern outcrop of this southern Welsh area the thickness swells to 3,000 or 4,000 feet. Similar thicknesses are found in northern Devon, on the south side of this axis. It is not easy to assume a narrow and lofty mountain range between these two basins, enduring without rejuve- nescence and gradually buried by the mantle of sediments, for the erosion which would provide this thickness of sediments would also destroy the axial character and height of the mountains. Differential, intermittent uplift of this axis and subsidence of the adjacent belts during sedimenta- tion seems a probable relationship. In the structural complex of ridges and basins we maj recognize considerable resemblance between the British Isles in Devonian time and the Cordilleran area of the United States in the Tertiary. The analogy is instructive not only from the structural standpoint, but from the physical conditions of sedimentation as well. With this introduction on the general character and relationships of 18 John R. Flett : On the age of the Old Red Sandstone of Shetland. Trans. Roy. Soc. Edinburgh, vol. xlvi, pt. ii, 1908, pp. 313-320. 364 J. BARRELL FLTJVIATILE ORIGIN OF OLD RED SANDSTONE the Old Bed, we may turn attention to those detailed stratigraphic char- acters -which go to show the conditions of origin. UPPERMOST SILURIAN — DOWNTONIAN FORMATIONS "Rocks. — The series of strata grouped under the term Downtonian has hitherto been regarded as of Lower Old Red Sandstone age, owing to the prevalence of red and yellow sandstones and shales which are the prominent feature of that formation. The recent discovery by the Geological Survey, in shales and mudstones intercalated in these sandstones, of a marine fauna which in some respects is identical with that of the underlying Ludlow Rocks has led to a revision of the classification hitherto adopted. These passage beds are now viewed as forming the highest subdivision of the Upper Silurian rocks. They may be briefly tabulated in descending order as follows : "Lower Old Red Sandstone, the basal bed being a coarse conglomerate or conglomeratic sandstone, with pebbles composed mainly of greywacke derived from the Southern Uplands. "Unconformability in the Pentland Hills and in Ayrshire, apparent con- formability in Lanarkshire. C 4. Chocolate-colored sandstone. Conglomerate with quartzite pebbles derived from the High- lands. Green and red mudstones with bands of greywacke and brown flaggy carbonaceous shales with fishes and eurypterids. Red and yellow sandstones and mudstones, underlain in the Hagshaw Hills by a fine conglomerate of local occurrence, resting conformably on Upper Ludlow Rocks. "Fossils. — The organic remains, which are restricted to Zone 2 of the fore- going series of strata, consist of plants, ostracods, phyllocarid crustaceans, eurypterids, and fishes. "Among the fragments of plants obtained from this horizon, Mr. Kidston has identified Pachytheca and one specimen as belonging to the genus Parka, though of a different species from P. dccipiens. The ostracods are represented by Beyrichia, a form which is common in the Upper Silurian rocks of the Southern Uplands ; the phyllocarid crustaceans by Ceratiocaris. Most of the genera of eurypterids found in the Wenlock and Ludlow Rocks in Lanarkshire and the Pentland Hills, viz. : Eurypterus, Pterygotus, Slimonia, Stylonurus, have been obtained from the Downtonian fish-band (Zone 2 of above table). "The most striking pala?ontological feature, however, is the remarkable as- semblage of fishes procured from this horizon which are wonderfully complete when carefully extracted from the carbonaceous shales. Dr. Traquair has identified in the collection of the Geological Survey five genera of fishes, four of which are new, and seven new species. One genus (Thelodus) is common to the Upper Ludlow Rocks of Lanarkshire and Wales and to the Lower Old Red Sandstone of Forfarshire and Oban. "Conditions of deposition. — The Downtonian strata indicate a marked change in the phases of sedimentation from those which obtained in Ludlow and Wen- lock time in the south of Scotland. While it is true that the green mudstones, Downtonian.. DESCRIPTION OF THE FORMATIONS 365 greywackes, and brown carbonaceous shales yielding fossils resemble litho- logically Upper Silurian rocks, still the dominant feature of the series as a whole is red and yellow false-bedded sandstones. It seems just to infer that the Downtonian fish-band and the associated mudstones and greywackes are marine deposits, for some of the eurypterids found in the latter strata are the associates of graptolites in the Wenlock Rocks of the Pentland Hills, and of brachiopods (Lingula minima) in the Ludlow Rocks of Lanarkshire. More- over, the occurrence of the Polyzoon, Glauconome, together with Spirorbis and sponges, likewise points to the marine origin of some of the Downtonian strata. The red and yellow false-bedded sandstones, on the other hand, evi- dently herald those conditions which prevailed during Old Red Sandstone time, when the open sea gave place to brackish water or inland lakes." 20 Peach and Home call this fauna marine, but there is to be noted the absence of ccelenterates, brachiopods, echinoderms, and trilobites, repre- sentatives of which are found in the true marine Ludlow rocks. This absence is as striking as the lingering presence of a few marine types. On the other hand, the ostracods, eurypterids, and fishes are groups which are found also as characteristic fossils in clearly fresh-water deposits as well as in brackish water or marine deposits. The conditions of deposi- tion of the fossiliferous zone would seem therefore to represent tbe re- curring but temporary existence of brackish water, permitting for a time tbe incursion of a fauna related more to the open sea than to the land. It seems to have been but a temporary condition, since the other forma- tions, as noted by Peach and Home, are suggestive of the true Old Eed Sandstone conditions. The descriptions suggest the current-laid sedi- ments of streams rather than a condition of lakes or bays. Two hundred miles south, in Staffordshire, limited outcrops of upper- most Silurian rocks have been recently discovered which show brachiopod faunas. 21 One hundred miles northeast of Lanarkshire the Downtonian rocks outcrop again along the shore at Stonehaven beneath the true Lower Old Eed. Hickling gives these basal Stonehaven beds a thickness of 1,500 feet and notes the occurrence at the base of breccia, followed by fine red sandstone, with numerous thin bands of bright red shale. Higher up several beds of the red marly shale, 50 to 100 feet in thickness, are inter- calated with light red or yellow sandstones and fine grits with some bands of gray sandstone and grit. The series of beds is distinctly fine in char- acter as compared with the overlying mass of the Old Red. 22 The base at this place is regarded by B., Campbell as an unconformity and not a 20 Peach and Home : The Silurian rocks of Britain. Mem. of the Geol. Survey of the United Kingdom, vol. i, Scotland, 1899, pp. 67-69. 21 W. W. King and W. J. Lewis : The uppermost Silurian and Old Red Sandstone of South Staffordshire. Geol. Mag., Decade 5. vol. ix, 1912. pp. 437-443. 22 The Old Red Sandstone of Forfarshire, Upper and Lower. Geol. Mag., Decade 5, vol. v, 1908, p. 399. 366 J. BARRELL FLUVIATILE ORIGIN OF OLD RED SANDSTONE fault zone. The thickness is given as nearly 3,000 feet, and from a thick helt of gray and greenish mudstones and shales fossils have been ob- tained — phyllocarids, myriopods, eurypterids, fragments of scorpions, plant fragments, and worm tracks. Further, a thin bed of reddish mud- stone underlying the above series has yielded numerous plates of a new Cyathaspis. 23 Of these fossils the phyllocarids, as an order, are typically marine; the myriopods and scorpions are as typically terrestrial. The phyllocarids belong, however, to the class of crustaceans which has always exhibited a freedom of adaptation to changes of salinity, possibly greater than is found in any other group of organisms. From these descriptions of uppermost Silurian rocks and their fossils, judged in the light of more intensive studies of other delta deposits, it would seem that toward the close of this period delta conditions had developed from the northwest, pushing the littoral zone south of Scotland. The Scottish deposits were dominantly subaerial, forming in the main fluviatile delta deposits in Downtonian times. Occasionally, as is found on the fronts of all large deltas of low gradient, submergence would bring in wide shallow bays of brackish waters with such groups of animals as were least sensitive to changes of salinity. At rare intervals this faun a I invasion reached as far inland as southern or central Scotland. The climate was dry enough to permit drying out and oxidation of the deposits of the floodplains, but there is nothing in the chemical compositions or sedimentary structures to imply the existence of a desert climate. The evidence is not so determinative as could be wished, partly no doubt because it has not been searched bed by bed for the structural features, partly because evidence which may in another decade be re- garded as determinative is not generally regarded as such at the present time. LOWER DEVONIAN— LOWER OLD RED In Cornwall and southern Devon all, or nearly all, of the Lower, as in fact the rest here of the Devonian also, is marine. In southern Ireland, however, the beds are largely or entirely fresh water in origin, since no marine fossils have ever been found there. In south Wales the Lower Old Eed is represented by the cornstone series. This consists below of red marls with bands of nodular limestones (cornstones) and irregular beds of red micaceous sandstones. Above are the Senni beds, consisting of green and dull red sandstones with marls and cornstone conglomerates. 23 The Downtonian and Old Red Sandstone of Kincardineshire. Geol. Mag., Decade 5, vol. ix, 1912, p. 511. DESCRIPTION OP THE FORMATIONS 367 "In this cornstone series cornstones occur sometimes as continuous beds of pale red or green compact limestone, and sometimes as nodular concretions in the marls ; they consist entirely of amorphous carbonate of lime enclosing some sand, and no fragments of organisms except obscure plant remains can be seen in them. They may owe their existence to the agency of some lime- secreting alga?, like that which forms the 'sprudelstein' of Carlsbad (described by F. Colin in 1862) and the travertine of the Mammoth Hot Springs in the Yellowstone Park, U. S. "The Cornstone series has yielded fish of the genera Pteraspis and Cepha- laspis, and there can be no doubt that it is homotaxially the equivalent of the marine Lower Devonian and of the Lower Old Red Sandstone of Scotland. The Senni Beds are included in this series, because they also contain Pteras- pis; they only occur in that part of the area which lies to the west of Breck- nock, where they have a maximum thickness of 1,200 feet. . . . "When the formation is followed westward beyond Llandeilo several changes are found to take place. The base of the Cornstone series ceases to show an upward passage from the Tilestones. the basement beds becoming first sandy and then conglomeratic, at the same time gradually passing across the passage beds and the Ludlow mudstone till, near Carmarthen, they have overstepped the whole Silurian series and rest directly on the Bala Beds. The whole group also becomes thinner, and is not more than 2,500 feet near Llandarog." -* These Welsh deposits have been regarded by British geologists as fresh water in origin and formed in a large lake. Much of the formations may, however, equally well, if not better, be regarded as nuviatile. The great thickness, taken with the arenaceous character, indicates shallow-water conditions maintained by subsidence. It would be difficult to conceive of a continuous fresh-water lake existing throughout and excluding the sea which lay not far to the south of the narrow intervening axis of elevation. If, however, sedimentation were, on the whole, faster than subsidence, nuviatile plains would result and the sea could be readily excluded, at the same time that intermittent lacustrine conditions could exist. Further, although the limestones, where pure and thick, suggest such lacustrine conditions, nodular layers of earthy limestones do not require such an origin. It may be concluded, therefore, that in the Welsh basin the de- posits suggest combinations of nuviatile and lacustrine conditions. The large difference which this means in interpretation is that much of the sediment was deposited on river floodplains, and on paleogeographical maps the area should not be represented as a great lake basin, but as land instead of water. Turning to the Caledonian basin, this, under the interpretation of Jukes-Browne, may be taken as including all that broad region south of latitude 57° and north of 54°. The chief area is that of southern Scot- land, but small outlying remnants occur in both Scotland and northern 24 Jukes-Brovrae : The building of the British Isles, 1911, pp. 108, 109. XXVII — Bull. Geol. Soc. Am., Vol. 27. 1915 368 J. BARRELL FLUVIATILE ORIGIN OF OLD RED SANDSTONE Ireland. The best continuous exposure is that shown on the southeast coast of Scotland from Stonehaven to. the Firth of Tay. A detailed description of these rocks is given by Hickling, and the following is quoted from him, omitting that of the Stonehaven beds, which are now regarded as Downtonian in age. "Table of the Lower Old Red of Forfarshire Feet Bdzell Shales ? 1,000 Arbroath Sandstone 1,200 Auchmithie Conglomerate 800 Red Head Series 1,500 Cairnconnan Series 2,000 Carmyllie Series 1,000 Dunnottar Conglomerate 5,000 Stonehaven Beds 1,500 14,000 "It must be remarked that the subdivisions in the above table are based purely on lithological characters and are only made for convenience of descrip- tion. No breaks in the series exist to my knowledge, and I am far from sup- posing that these subdivisions are likely to be traceable for any great distance ; rapid lateral variation in the character of the rocks is too obvious a feature of the Old Red Sandstone. The names applied to the subdivisions are taken from the localities where the series may be typically seen. The thickness of the subdivisions are estimated from theoretical sections for the most part, and are therefore to be regarded as only approximately accurate. . . . "The Dunnottar Group of coarse red and grey sandstones, grits, and con- glomerates which form the bold coast the whole way from Stonehaven to Johnshaven. As I have only been able to examine the base and the top of this series, I shall add no more than that it forms by far the most extensive series of conspicuously coarse deposits in the district. In its conglomerates pebbles commonly range up to a foot or moi*e in length, and yet are astonishingly well rounded. They mostly consist of quartzite. South of Johnshaven several thin lavas are interbedded with the top of this series, with sandstones and coarse conglomerates of porphyrite blocks between. Beyond this the coast-section is interrupted by the mass of Upper Old Red which is faulted in, extending from East Mathers to Milton Ness (described below), and which covers the junc- tion between the conglomerate series and the great mass of lavas which forms its natural top. These lavas occupy the coast southward by Montrose to Lunan Bay, being hidden, however, almost the whole way to Montrose by sand and alluvium. From Lunan Bay their outcrop strikes inland along the summit of the anticline by Friockheim and Letham, near which latter place they finally die out. About Friockheim and Leysmill are numerous quarry sections of the "Carmyllie Series, which overlies the lavas. Compact red or grey sandstone is the predominant rock of this series, with subsidiary masses of grey flagstone and blue or red shale, termed 'caulm' by the quarrymen. Together with their DESCRIPTION OF THE FORMATIONS 369 interbedded lavas, these rocks form the whole axis of the Sidlaws, all along which they are quarried for building and paving material. The well-known Carmyllie quarries are in the middle of this series. Passing upward, the "Caimconnan Series is reached, distinguished by its coarser materials, prin- cipally dull red or grey grit with bands of conglomerate. The conglomerates are more particularly developed on the north side of the anticline, as at Turin Hill, north of Rescobie Loch. This series should appear on the coast in Lunan Bay, but it is entirely hidden by the sand and alluvium. At the south end of the bay another series of lavas, admirably exposed for study, intervenes be- tween it and the "Red Head Series, which forms the bold cliffs from the promontory of that name southward to Rumness. In its lower part it consists of fine red thin- bedded sandstone with bands of hard bright red shale, while the upper por- tion is made up of thicker-bedded sandstone. Some six or seven miles to the southwest, at Arbirlot, the lower part of this series, as seen in the banks of the Elliott Burn, consists mainly of blue and grey shales, with partings of sandstone, having so strong a resemblance to some of the rocks of Carmyllie as to have led Hugh Miller to consider them as a repetition of that series. This case illustrates very well the rapid lateral variation to which all the beds of the Old Red Sandstone are liable. "The Auchmithie Conglomerate overlies the previous group in the cliffs just north of the village so named. The series consists of three main masses of conglomerate, with intervening sandstones and conglomerates. The pebbles in the conglomerates are well rounded, fairly large (generally 1 to 6 inches, rarely 12 inches), and, as usual, are mostly quartzite. The thickness of this conglomerate series diminishes along its outcrop to the southwest. "The Arbroath Sandstone is the highest series of the Lower Old Red seen on the Forfarshire coast. Coarse, gritty, sometimes pebbly sandstone is its component rock, always red in color. Just above the base of the series, by the Signal Tower at Arbroath Harbor, there is a single band of grey grit and marlstone on the shore, containing nodules of limestone from the size of a pea to 1 foot in diameter. This is noteworthy in view of the almost complete ab- sence of lime from the Lower Old Red System. . . . "The EdzeU Shales . . . overlie the Arbroath sandstones. They are gen- erally bright red fine sandstones, shales, and marls, either hard or soft, fre- quently mottled with small circular patches of pale yellow, grey, or green, or more rarely with bands of the same color. . . . "The volcanic rocks of the district do not call for description here. It may suffice to remark that they are, with very rare exceptions, in the form of con- temporaneous interbedded sheets, and are merely a continuation of the series of the Ochil Hills, where they have been described in detail by Geikie and others (A. Geikie, 1900). "Attention must now be drawn to a point, the importance of which has not, I believe, been hitherto recognized. All the recorded fossils from this dis- trict — and I suspect that the same applies to Perthshire — are from a very limited series of horizons near the middle of the Lower Old Red System, the Upper Old Red being, of course, left out of account. The Carmyllie series is the fossiliferous group par excellence, while a few of the worked localities may lie in the Cairnconnan series (e. g., Tilliewhamland quarry, Turin Hill), 370 J. BARRELL FLUVIATLLE ORIGIN OF OLD RED SANDSTONE or the top of the Dunnottar conglomerates. Odd fossils have occasionally been obtained from other horizons, but they are quite a negligible quantity." 25 To the west, the sections as exhibited in Fife and Kinross show enor- mous quantities of agglomerates, lavas, and tuffs. More or less rude sorting of the breccias shows some rearrangement by water. Some beds of breccia betray, however, no stratification ; others, on the contrary, show a great deal of rounding and sorting of pebbles and boulders. A. Geikie states that the series of deposits in the Caledonian basin everywhere presents traces of shallow-water conditions. 26 He seeks to interpret the whole as laid down in an open lake and sorted by wave action. He notes that — "An objection may arise that the remarkable coarseness of the conglomer- ates, and the large size of many of their included bowlders, are suggestive rather of the powerful breakers of the open sea than of the limited wave- action of a lake or inland sea. But possibly the coarseness of the shingle may rather be some indication of the dimensions of the lake and of the power of the waves; along its exposed shores." 2T It seems, however, that a reinterpretation is called for. The widely extended and coarse conglomerates do not suggest shore action, but re- semble the stream-borne detritus of Eocky Mountain valleys. Their coarseness, thickness, and distance of transportation are all significant of stream rather than wave action. The red or gray sandstones and red shales are similar to the Catskill deposits of Few York and Pennsylvania, and are such as in India and North America have become commonly regarded as fluviatile in origin. MIDDLE OLD RED — ORCADIAN FORMATIONS Relations to older rocks. — During the middle part of the Old Red Sandstone period the chief axes of subsidence and deposition were trans- ferred to the northern part of Scotland. Its locus has become known as the Orcadian Lake, or, better, the Orcadian basin. Disturbance and some erosion of the Lower Old Red may have resulted in retransfer of material. The enormous thicknesses of deposits imply vigorous crust movements and profound erosion. The fossiliferous beds show that the Middle Old Red is younger than the Lower Old Red to the south, but there is no doubt but that the Caledonian deposits once extended northwest of the 23 The Old Red Sandstone of Forfarshire, Upper and Lower. Geol. Mag., Decade 5, vol. v, 1908, pp. 3^-401. 26 Text-book of geology, 1903, p. 1008: 27 Geology of East Fife. Mem, of t!je Geol. Survey, Scotland, 1902, p. 43. DESCRIPTION OF THE FORMATIONS 371 present fault boundaries; so that Lower Orcadian beds may correspond to Upper Caledonian. The basal beds are preserved in certain small areas in Aberdeen and Banff. These show in some cases clearly the form of old filled river valleys. A. Geikie describes that which is now dissected by the River Avon, latitude 57° 20' north, longitude 3° 20' west. "A marked structure in some parts of the conglomerate is a kind of false- bedding of the stones. Between the gently inclined lines which mark the true dip, the stones of each bed of conglomerate are arranged in a rude stratifica- tion obliquely across the line of the valley at angles of 17° to 20°. They thus appear to dip toward the hills (representing the old valley walls), while the true inclination of the conglomerate beds is away from them." 2S This is a highly significant feature. It seems clearly to represent the stream-shingling of gravels, by which the longer axes of the pebbles slope upstream. The situation and character does not suggest beach sorting. Old Red of the Moray Firth, — Passing to the old Eed Sandstones around the Moray Firth, Geikie states : "We have found them throughout that extended tract mainly conglomeratic and arenaceous — the conglomerates in thick masses coming in again and again on successive platforms, while interstratified with them lie bands of grey clay and shale full of calcareous nodules containing fish remains. These fossilifer- ous bands retain their distinctive characters, lithological and palseontograph- ical, throughout the whole district'" (page 446). These strata stand in great contrast to the vast continuous flagstone series of Caithness to the north, though they are considered by Geikie to be equivalent to the upper portion of the Caithness section. Formations of the Orcadie basin in Caithness. — A tabulation of the formations which comprise the whole terrane, with an abstract of such features as bear most strongly on the problem of origin, is based on Geikie's complete report in the Transactions of the Eoyal Society of Edinburgh. 29 Nearly the whole of Caithness consists of Old Eed Sandstone. The interior is almost entirely covered with peat and glacial drift, but the powerful action of the ocean on these rather friable rocks has exposed admirable sections often several hundred feet in height around the entire coastline. These have been compiled by Geikie into a type section of the whole series. 28 A. Geikie : On the Old Red Sandstone of western Europe. Trans. Roy. Soc. of Edinburgh, vol. xxviii, pt. ii, 1878, p. 427. 29 Vol. xxxviii, pt. ii, 1877-1878, pp. 355-421. 372 J. BARRELL FLUVIATILE ORIGIN OP OLD RED SANDSTONE Subdivisions of the Lower Old Red Sandstone of Caithness Feet 9. John o'Groat's Sandstone and Flagstone Group 2,000 8. Huna Flagstone Group 1,000 7. Gill's Bay Red Sandstones 400 6. Thurso, or Northern Flagstone Group 5,000 5. Wick, or Eastern Flagstone Group 5,000 4. Langwell and Morven Sandstones and Conglomerates 2,000 3. Badbea Breccia and Conglomerate 300 2. Braemore and Ousedale Sandstones 450 1. Basement Conglomerate 50 16,200 The significant details of these formations are as follows : 1. Coarse Basement Conglomerate 50 feet This was laid down on an uneven floor of igneous and metamorphic rock. It presents a remarkable granitoid aspect to such an extent that it is often not easy to determine where the conglomerate ends and the granite begins. Its component blocks vary in size up to as much as a yard, or even more, in length. The bowlders are for the most part tolerably well rounded. The thick- ness given was measured at the place where studied, but the basement con- glomerate over the basin as a whole must vary much both in thickness and in age. 2. Braemore and Ousedale Sandstones 450 feet These are dull chocolate-red sandstones with sandy shales or clays. It is noteworthy how abundantly pink orthoclase occurs in the matrix of many of the sandstones. The red shales and sandstones of Braemore are sometimes pitted as if by rainprints, but have as yet yielded no fossils. 3. Badbea Breccia and Conglomerate 300 feet By far the most conspicuous member of the Old Red Sandstone series in the south of Caithness is a remarkable breccia or brecciated conglomerate. It occurs in thick beds, wherein little or no trace of stratification may be found. The stones, considerably smaller than in the basement conglomerate of Brae- more, seldom exceed five or six inches in length. They consist mainly of pink cleavable orthoclase, pink granite, grey quartz-rock, white vein-quartz, and occasional pieces of red sandstone. The feldspar is the predominant ingre- dient, and likewise enters largely, in a comminuted state, into the composition of the paste (pages 37S-379). 4. Langwell and Morven Sandstones and Conglomerates 2,000 feet The Badbea breccias and conglomerate "pass up into a thick series of dull chocolate-red, grey, and yellow sandstones, with layers of dull-red and olive- colored shales and of fine conglomerate. ... In their lower part they are highly felspathic, with a rather coarse texture, and many scattered pebbles, as well as nests and bands of conglomerate. Higher up they are more flaggy, and they finally pass upward imperceptibly into the dull-red and grey flag- stones of Berriedale. Nothing but an arbitrary line can be drawn for their upper limit. . . . The group has as yet yielded no fossils" (pages 379-380). DESCRIPTION OF THE FORMATIONS 373 Tracing these beds toward the north they are found to lose the coarse pebbly and conglomeratic features and to pass into fine flaggy sandstones, with sandy shales, the whole having a prevailing dull chocolate-red tint, through which seams of greenish-gray shales and flagstones occur. In places, as observed by Sedgwick and Murchison, the red tint is local and even superficial, the same stratum being red or green at different parts of its course. Traced inland to the west the beds so increase in the quantity of coarse detritus that on the north side of the Scarabin ridge hundreds of feet may either be called highly conglomeratic sandstones or sandy conglomerates. "The four groups of strata, above described, may be regarded as forming together a red sandy and conglomeratic base, of very variable thickness, on which lies the great flagstone series of Caithness now to be discussed" (page 386). 5. Wick, or Eastern Flagstone Group 5,000 feet This group consists of dark-gray flagstones which are often thick-bedded, thin shales and limestone bands ; the whole passing down into red shales and sandstones. The group is distinguished from that which overlies it by the greater mas- siveness of its flagstones, and by their less calcareous composition and less fissile or shaly texture. Sun-cracks and ripple-marks abound and in various horizons remains of terrestrial plants have been found, one gray shale in particular having its surface covered with carbonized vegetation. Large stems belonging to tree ferns and gymnosperms also occur. An abundant fish fauna has been obtained from this formation. 6. Thurso, or Northern Flagstone Group 5,000 feet Dark-gray and cream-colored flagstones, gray and blue shales and thin lime- stone ; some beds strongly bituminous. This group is more fissile, shaly, and calcareous than the preceding. "The flagstones which, towards the east, retain the usual normal characters of fissile calcareous strata, pass into sandstones and conglomerates as they approach and rest upon the granite and gneiss" (page 391) . "The nest feature to engage the attention of the observer is probably the extraordinary abundance of ripple-marked surfaces and sun-cracks. Though these markings abound also in the lower flagstone group, it is here that they attain their greatest development" (page 392). "More abundant and admirable illustrations of sun-cracks could hardly be found than occur along this coast. Broad gently-inclined sheets of rock again and again present themselves to view so covered with reticulations as to look like tesselated pavements. It may be noticed that the cracks not infrequently descend through many of the fine laminae of deposit for a depth of five or six inches with occasionally a breadth of three or four inches. The material filling up the interstices abounds with small, occasionally curved pieces of shale. These may, no doubt, be regarded as portions of the upper muddy layer which cracked off and curled up during desiccation, as may often be observed on dried-up pools at the present time. Some pittings, occasionally seen on the sun-cracked surfaces, may perhaps represent rain-drops. Altogether, no evi- dence could more conclusively indicate a long-continued, tranquil deposit of 374 J. BARRELL FLUVIATILE ORIGIN OF OLD RED SANDSTONE fine sediment in shallow water, which frequently retired and left wide tracts of muddy shore to be dried and cracked by exposure to the sun" (page 393). The resemblance which the fissile calcareous flagstones "bear to some of the so-called fresh-water limestones and cement-stones at the base of the car- boniferous system of the south of Scotland, cannot but strike any one who is familiar with the latter strata. This likeness includes not only the composi- tion, color, and mode of weathering, but even the minute wavy lamination indicative of intermittent but tranquil deposit. Other shaly layers are strongly pyritous" (page 400) . . . "organic remains abound in the strata exposed on the shore between Dunnet Bay and Reay. Fragments of fish and coprolites are scattered abundantly through most of the flagstones. Some of the cal- careous shales are full of Estheria, while traces of plants occur in great num- bers, though generally in a somewhat macerated condition" (page 393). The groups to which the plants belong are Ferns, Calamites, Lepidodendrids, Stigmaria, and Araucarioxylon. No traces of marine plants have been found. 7. Gill's Bay Red Sandstone 400 feet This formation consists of red, friable, false-bedded sandstones, both its base and top interleaved with seams of flagstone and grading into the flagstone groups. No fossils have yet been found in these sandstones. 8. Euna Flagstones 1,000 feet To the alternation of red sandstone and flagstones succeeds the Huna flag- stone group having characters similar to those of the Thurso flagstones. 9. John o'Groat's Red Sandstones and Flagstones 2,000 feet These consist of a mass of false-bedded red sandstones, with intercalations of flagstones, the whole much resembling the Gill's Bay formation. The sand- stones "occur in successive thicker zones, between which lie many alternations of red sandstone, red and blue flagstones, grey shale, and impure limestone. These latter strata are quite undistinguishable from portions of the older flagstone groups. The highest part of the group consists of a thick mass of false-bedded red sandstone, without flagstone or shale. Fossils occur in some of the blue flagstones and impure limestones" (page 404). Old Red Sandstone of the Orkney Islands. — The greater part of the section previously given is traceable into sandy and conglomeratic facies contiguous to the old margins of the basins. It is important, therefore, in any view of the Orcadie basin as a whole to compare the preceding with the stratigraphic characters as observed throughout the Orkney Islands, since these were much farther from the limits of the basin. In the same report from Geikie from which the preceding statements have been abstracted is given also an account of the stratigraphy as observed in the Orkney Islands. "Almost the whole of the Orkney islands consist of flagstone and sandstones of the Caithness flag series. The only other formations present appear in the southwestern part of this group. A small ridge of the underlying crystalline rocks rises to the surface at Stromness" (page 408). DESCRIPTION OP THE FORMATIONS 375 "The flagstones retain the same features so well marked in Caithness. Sometimes, as at Skaill in Pomona, they are exceedingly hard, fissile, bitumi- nous, and crowded with fossil fish. In other places, as near Kirkwall, they form thicker beds, and can be quarried for building materials, like those which are similarly used at Thurso. Bands of dull red and even yellowish sandstone occur interstratified with the flagstones, as in Scapa Bay, Meal near Kirkwall, Eday, and other places. Where the flagstones rest on the old crystalline rocks they become for a short space conglomeratic at the base'' (page 409). At certain localities occur zones of red sandstone which may be paralleled with the John o'Groat's and Gill's Bay groups of the Scottish mainland. As we advance northwards among the islands the same petrographical characters continue. One "seems to be forever meeting with repetitions of the same rocks. No doubt when these islands come to be mapped in detail, the real thickness of flagstones will be found to be more considerable than might at first have been surmised." The Orkney rocks appear to belong to the upper groups of the Caithness section, since "no equivalents are met with of the massive red sandstones, shales, and conglomerate groups at the base of the Caithness series." "The coarse conglomerate of well-rounded sandstone blocks at Heg- labir on the west side of Sanday, which has been long known (see Barry's "Orkney," page 56, and Neill's "Tour"), seems to occur at a greater distance from the local base, for it is said to overlie sandstones and flagstones" (page 410). "Many of the flagstones of Orkney are charged with organic remains. Espe- cially is this the case with some of the dark, hard, fissile, bituminous bands. On the surfaces of these strata remains of the characteristic ichthyolites are crowded thickly together, and usually in such a tolerable state of preservation as to show that the fishes must have died where their remains are found, or at least that they could not have been subjected to any prolonged exposure and transport before they were buried under the accumulating sediment. As a rule, the fossils have been converted into a brittle jet-like substance, which is so liable to crack and scale off, that unless great precautions are taken, an organism, which at first showed external sculpture in great beauty, becomes eventually a mere black bituminous patch, retaining only the outline of the original specimen. It is not difficult, in most cases, to distinguish an Orkney from a Caithness ichthyolite. "The fossil plants of Orkney include most of the forms found in the Thurso and upper Caithness groups" (page 411). Interpretation of the Orcadian deposits. — The structural features of the strata which are significant of origin have been given bv Geikie in more detail than is found in other descriptions. They permit, therefore, of a more specific interpretation. Significance of conglomerates. — It has been shown that subaerial con- glomerates may be laid down at much greater distances from the sources of supply than subaqueous conglomerates, since streams may transport resistant gravel until they lose velocity or become loaded with finer material. Subaerial conglomerates may also accumulate with much greater thicknesses than subaqueous conglomerates. The only conditions 376 J. BARRELL FLUVIATILE ORIGIN OF OLD RED SANDSTONE necessary for the continuance of deposition of subaerial conglomerates is the maintenance of rapid erosion to supply the material and river volume and grade sufficient for its transportation. The conditions for the con- tinuance of deposition of subaqueous conglomerates, on the contrary, depends on the permanent existence of a shore of suitable materials and a depth of water just such that waves and currents may transport gravel. The great Pleistocene gravel deposits of the continental interiors testify to the magnitude of the first class of conglomerates, while the limited distance to which present gravels are transported from shores, as shown by hydrographic charts, indicates the relative insignificance of the second. The pebbly sandstones far from the base of the Caithness section and at distances of many miles from the regions of erosion are highly sug- gestive of subaerial conditions, while conglomerates, such as those of the Moray Firth, forming thick masses, coming in again and again, and overlying formations of clay and shale, may be taken as fairly conclusive evidence of terrestrial deposition. Significance of colors. — Variegated colors, especially where these alter from one stratum to another and even within a single stratum, are highly characteristic of continental deposits, as Walther has pointed out. Bed sediments have been deposited to a limited extent under permanent water bodies, either salt or fresh, but are characteristically of terrestrial origin. Even where deposited under a water body the color is due to previous thorough oxidation of the iron while exposed to the air without the usual opportunity for later subaqueous deoxidation. The climatic conditions which may give rise to red tones in consolidated deposits are rather broad. As argued elsewhere, 30 the presence of red merely implies the existence at the time of deposition of either temperate or torrid climate with a dry season sufficiently marked to permit periodic aeration of the fLoodplain soil. The dominantly red color and variegated character of the conglomerates and sandstones is therefore strongly suggestive, though not conclusive in itself, of terrestrial origin under rather broadly limited climatic con- ditions. T