C9f fyxuW ^mmxii §ilr»g THE GIFT OF M..^....fe).n^3:^i>SbudA o.^...Z...'^.3^l>S^.. ^GINEERiNG LiBRARY. ,A..v<^i.i.s.£ '^k/<»3 nwr oen Jl'mS^" University Library QE 262.C28C94 °" 'tie geology of the neighbourhood of C 3 1924 004 054 486 The original of tliis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924004054486 -e QgAETEELr Journal of the Gteolobical SiioiETT/o)' August 1896, Vol. lii.] Oa the Gbolost of the Neighbourhood of Carmarthen. By Miss Margaret C. Ceoseebld and Miss Ethel G. Skeai. (Communicated by J. E. Mare, Esq., M.A., F.R.S., Sec. G.S.) [Plates XXV. & XXTL] Contents. Page I. Introduction and Bibliography 523 II. Physical Structure 524 III. The Succession and Detailed Description of the Beds 525 IV. Comparison with the Deposits of other Areas ... 531 V. Description of some new Species of Trilobites ... 535 VI. Copclusion 539 I. Introduction and Bibliography. The area described in this communication has Carmarthen for its centre, and lies approximately within a 4-mile radius of that town, as indicated in the accompanying sketch-map (PI. XXV.). Our object in examining this district was to trace the continuation of the com- plex anticline which was discovered by the late Mr. T. Eoberts, about 10 miles west of Carmarthen. In the map illustrating his paper ' the anticline is shown to be narrowing rapidly in the neighbourhood of Mydrim, the Didymograptus-heAs seem to wrap round the older Tetragraptus series, and apparently the anticline is dying out. An examination of the beds round Carmarthen has enabled us to establish this conclusion, and to show that in our own district we have a new anticline, the northern limb of which continues in the same direction as that of the anticline farther west, but with a core consisting of older rocks, which we correlate with the Tremadoc Slates. The southern limb of the new anticline is buried under Old Eed Sandstone (which here encroaches farther north than at St. Clears), but in the northern limb we have a regularly ascending series from Tremadoc Slates to Sicranograptus-ahaleB. To the east this regular succession of beds is disturbed by the presence of an extensive series of alternating grits and sandstones, with some shales, the last containing Bala fossils. The beds in question abut on those of Arenig age, but whether their appearance in this locality is due to faulting or to an unconformity we cannot at present deter- mine, owing to the nature of the ground and the exceedingly complex character of the foldings. The clue to the relations of these beds would doubtless be discovered by an examination of the area lying to the east. ' Roberts, T., 'Notes on the Gteology of the District West of Oaermarthen,' Quart. Journ. Geol. Soc. vol. xlix. (1893) p. 166. 524 MISS CKOSriKLD AKD MISS SKEAT ON THE [Aug. 1896, The literature relating to this district is as follows : — Datidsom, T. 'British Brachiopoda,' vol. iii. pp. 232, 233, pi. xxxiii. figs. 17, 21. Memoies op the Geological Suetet, vol. i. (1846) p. 25 &vol. ii. (1848) pt. i. pp. 262 et segq. Mdechison, R. 1. ' The Silurian System ' (1839), pt. ii. pp. 355, 358, 361, 664 & 720, pi. XXV. figs. 3 a, 3 6 & pi. xxxiv. fig. 9. ■ . 'Siluria,' 5th ed. (1872) pp. 55-58. Saltee, J. W. ' A Monograph of the British Trilobites ' (1864), p. 173, pi. xviii. fig. 11 & pi. XXV. fig. 4. Stmohds, W. S. ' Records of the Rocks ' (1872), pp. 99, 100, 125. The district was geologically surveyed by Sir H. De la Beche and Mr. W. E. Logan in 1845, with additional Silurian lines by Mr. W. T. Aveline in 1855-56, and is shown on Sheet 41. II. Physical Sikucicke. The physical structure of the country is sufficiently marked to demand a slight description. The most striking features are the isolated hills with their steep sides, abrupt squarish outlines, and flat tops sloping westward, as, for example, Llangunnor and Trevaughan Hills, and Bryn Merthyn. These hills are part of the system of roughly parallel ridges which cross the country from north-east to south-west. The practically uniform level of the hiU- topB, which range from 400 to 500 feet in height, and the presence of the common westerly slope suggest that we have here an ancient plain of marine denudation. The axes of the hills are formed of bedded grits, sandstones, or slates in a vertical position, while the valleys consist of the softer shales. This structure is primarily due to the earth-movements to which this district, like the rest of South Wales, has been subjected, but the processes of denudation have emphasized it and laid it bare. The streams flowing east and west have worn away the softer shales, forming valleys with gently sloping sides; the dip-valleys, on the other hand, are strikingly deep and narrow, and the streams have cut out steep and rocky gorges. The river Towy enters the district from the east, but makes a sharp southward bend near the town, where it is joined by its tributary the Gwili ; the united streams then take a remarkably sinuous course, and have covered the broad level valley with a thick deposit of alluvium. It seems probable that the river-bed marks a line of considerable geological disturbance. The district is bounded on the south by the long low range of Old Eed Sandstone hiUs, and the undulating slopes of these present a marked contrast to the often abrupt outlines of the hiUs formed of the older rocks. The earth-movements which have afiected this district are so compli- cated that it would be useless to attempt to give any exact idea of their effects within the limits of this paper. They have resulted in the formation of innumerable folds and faults, the latter of which are most noticeable in the north just outside our district. The distribution of the folds makes it possible roughly to distinguish two great periods of earth-movement. The first period gave rise to an immense number of small anticlines and synclines, the axes of which run north and south. In consequence of this folding the beds Quart. Jonrn. Geol. Soc. Vol. LII. PI. XXV. [To face p. 524. Vol. 52.] GEOLO&T or THE NBIGHBOTJKHOOD OP CARMABTHEN. 525 frequently crop out along uorth-and-south lines for a short distance, as is well seen in Mr. Eoberts's district to the west. After this universal crumpling of the crust had taken place a second period of folding on a stiU grander scale set in. This was evidently part of that great movement which produced the Condrozian ridge, and all the beds of this part of Wales were affected by it. In the end a great series of huge anticlines and synclines were formed with their axes running east and west, so that the second movement was at right angles to the first. A single limb of one of these complex anticlines forms the subject of the present paper. The uniform direction (N. to N.W.) of the cleavage throughout the district shows that it was produced either simultaneously with this last folding or at a later period. III. Tub StrccEssiON and Detailed Description op the Beds. 1. Tremadoc Slates. About 2J miles south-west of Carmarthen, exposed in the banks of the Nant-y-Glasdwr, are fine-grained blue-grey shales, which weather a rich yellow or brown, and are often iridescent. Inter- calated with the shales are occasional bands of micaceous sandstones. The beds dip 68° due south, and pass immediately under the Old Eed Sandstone. They have yielded Ogygia marginata, n. sp., ParahoUnella, n. sp., Peltura punctata, n. sp., Orthis, and Orthoceras. Peltura punctata occurs very abundantly, Ogygia marginata much less frequently, and only one specimen of Parabolinella was obtained. Confirmatory Sections. A little west of the Nant-y-Glasdwr, in the Nant Pibwr at Cwm Ffrwd, the shales are more gritty, and assume the character of mudstones. Alternating with them are some hard grey grit-bands. The junction with the Old Eed Sandstone is seen in the bed of the stream, and near it the shales are considerably disturbed, the dip varying repeatedly in the space of a few yards ; elsewhere the dip is south. We found here Ogygia marginata, Peltura punctata, Orthoceras sericeum. Salt., and Modiolopsis. The Ogygia occurs much more plentifully here than in the Nant-y-Glasdwr, but no perfect specimens were obtained. On the western side of the Eiver Towy near Cwrt Hir Farm, in a hole by the roadside, we found fragments of Erinnys, sp., Peltura punctata, and Orthis; and a little east of this, in Cwrt Hir carriage-drive, where there is an alternating series of coarse grits and shales, we found in the shales Ogygia marginata and a species of Orthis. 2. Arenig Beds. A typical section of these beds is seen just below Glan Pibwr Cottage, in the narrow lane which leads down to the stream. They consist of fine-grained, dark, bluish-grey mudstones, finely bedded, and weathering in iridescent shades of brown and red. The rock 626 MISS CBOSPIBLT) AND MISS SKEAT ON THE [Aug, 1896, greatly resembles the Tremadoo Slates, but, on the whole, is harder, and not interstratified with grits or sandstones. The shales in the lane dip south, and thus pass under the older Tremadoc Beds. "We find this inversion of the normal succession in all the strata on the southern side of what we may term the Mount Pleasant and Ystrad ridge. The fossils obtained here were mainly trilobites, hut a few specimens of Phyllograjotus, sp., were also found. The list is as follows : — Ampyx, sp., Ogygia marginata, Ctenodonta, Phyllograptus, sp., of. angmtifolius. Hall. At Pen-y-hanc, about ^ mile eastward, the same species of Ogygia and a Lingula were obtained. Confirmatory Sections. The small stream above Pwntan House, west of the river Towy, has cut its way through similar mudstones of a rather lighter colour. In this rock the bedding-planes are tolerably far apart, and the intervening rook breaks with difficulty. The fossils are abundant and well preserved, but consist entirely of Ogygia marginata and of a few specimens of Ctenodonta. East of the Towy similar mud- stones may be seen in the brook which passes under the road close to Carmarthen Junction, but only a few distorted portions of Ogygia marginata were obtained. At AUt Pen-y-coed, 2 miles nearly due east of Carmarthen ' Junction, the mudstones are exposed in a fine deep gorge ; the beds are not very fossiliferous, but yielded a head and tail of Ogygia, sp., cf. Selwynii, Salt. In the banks of the Nant-y-Caws, on the same strike, is a great thickness of shales lithologically similar to those of Glan Pibwr, but containing, both in the upper and lower parts of the gorge, bands of grit and some conglomerates. As the dip varies from 35° IT., not far from the bridge in the lower part, to 55° S. 30° E. at the water- fall, where the junction with the Old Eed Sandstone occurs, we have probably here a small subsidiary anticline. These shales are very nnfossiliferous, and only yielded Ortlioceras sericeum and Conularia. We have placed these beds provisionally with the Arenig Series, but whether they do not belong rather to the Tremadoc Slates is open to question. Lying to the north of the beds which we have just described are others ot similar lithological and palseontological characters, differing only in the fact that they contain in addition Calymene parvifrons, var. Murchisoni, Salt. The chief localities where the beds containing Calymene crop out are (i.) the Roman Road. — This exposure extends along both sides of the deep lane of this name in the Pensarn district of Carmarthen. The beds are shaly throughout, but vary in character in different parts of the road. In the hard, rather flaggy beds, about halfway np the lane many fossils were found in a very fragmentary con- dition, this being partly due to the fact that the bedding is almost at right angles to the cleavage. Higher up the road the shales are hopeless for purposes of search, being crushed and weathered to Buch an extent that they break always into long biUet-shaped frag- Vol. 52.] GEOLOer op TRE NBIGHBOUBHOOD of CARMARTHEN. 527 ments, without any clear face. This appearance is very character- istic of the beds whenever they occur ia a weathered condition. The fossils obtained were Bellerophon, Ampyx, Oalymene parvi- frons?. Salt., C. parvifrons, var. MurcMsoni, Salt., C. Tristanil, Brongn., distorted fragments of Asaphid trilobites. (ii.) Swansea Road Streams. — From the sloping ground which bounds the southern side of Swansea Road two small streams flow northwards to join the Towy. In the lower part of one of these streams we have the D. hifidus-&h.a\.e:&, but higher up and apparently dipping under these are mudstones which closely resemble those of Nant Pwntan in colour and hardness. We found here Oalymene parvifrons, var. MurcMsoni, Salt., Ogygia marginata, Orthoeeras, and a small Asaphus, sp., Salter; an exposure in the second streamlet yielded a Oalymene. In the banks of the steep lane leading to Cilwaunydd Farm were weathered shales containing Bellerophon, Oalymene parvifrons, var. MurcMsoni, Salt., and Ogygia marginata. Totally distinct from these mudstones, and without a fossil in common with them, are the shales of Hafod-wen spring on the western side of the River Towy. These beds contain a graptolite fauna of Arenig age, but lithologically they resemble the Llanvirn Beds to be next described. We obtained here Didymograptus nitidus. Hall, D. ? constrictus. Hall, D. ? MurcMsoni, Beck, Dictyo- graptus, sp., j^glina binodosa, and ostraooda. 3. Didymograptus bifidus-heis (Llaavirn). These beds consist in some parts of shales alone, in others we have alternating bands of light-grey shales and grey grits, weathering yellow. The grits, which generally are jointed, and often veined with quartz, vary in thickness from a few inches to several feet ; the shales occur in bands of 6 inches or less. The alternating grit- and shale-bands form a well-marked ridge of hills extending for a considerable distance along the southern side of the Pembroke and Tenby Railway. The beds are almost vertical, but have a general northerly dip on the northern side. Five quarries are cut out on the hillsides, those of Ystrad, Llanllwch, Wernddu, Nant-yr-hebog, and Pen Plas. In the last-named a thicker series of shales is seen just outside the quarry, dipping over the grit-beds, and here Phaeops llanvirnensis. Hicks, was found. At Wernddu we obtained Didy- mograptus indentus. Hall ; near Llanllwch and in Ystrad quarry D.hifidus, Hall. North and south of this ridge the beds consist of shales only ; these are of a light-grey colour, fairly soft and well cleaved. South of the ridge, at Ystrad Isaf and Hafod-wen lane, we found Didymograptus hifidus. Hall, D. indentus. Hall, and Diplo- graptus dentatus, Brongn. North of the railway in Trebersed stream a crushed graptolite of the D. patulus-^voxi^ and a fragment of an eye of Mglina were obtained ; at Castell-y-gaer a portion of a Didymograptus, sp., and of an eye of JEglina ; farther east, near Trevaughan Hill, we found Didymograptus indentus. Hall, and Diplograptus dentatus, Brongn. 528 JUSS CK08FIELD ANT MISS SKBiT ON THB [-A^Ug. 1 896, The beds also occur on the eastern side of the Towy, but the grit- bands do not form here such a prominent feature as on the western side. In the small quarry near the beginning of the Swansea Koad, after crossing the river near Carmarthen, the beds are bent com- pletely over, and form a small subsidiary anticline, the axis of which is at right angles to the general strike of the series in the Ystrad ridge; lithologically, the beds are exactly similar to those already described. The shales which overlie the grits may be seen in the road-cutting for about § mile ; they are dark grey or nearly black, and much contorted. In the quarry and along the road we found Climacograptus cwlatus, Lapw., 0. confertus, Lapw., Bendrograptus, sp., cf. flexuosus. Hall, Bidymograptus bifidus. Hall, D. indentus, Hall, D. Murchisoni, Beck, Lingula, and Siphonotreta micula. In the small stream above the road we found Didymograptus lifidus. Hall, and Ampycc, sp. At Cwarebach and Lower Henalt we have apparently an isolated patch of beds of the same age faulted against Dicranograptus-ahalea. The slates are fine-grained, bluish grey, striped with a lighter grey, and well cleaved-; they yielded Dictyograptxis, sp., vel Callograpius, sp., cf. Salteri, Hall, and Climacograptiis, sp., cf. confertus, Lapw. A little south of Cwm P'arm, near the town of AbergwUi, are several small quarries containing hard grey shales, gritty shales, and striped flags. Here Addaspis Buchii, Barr.,' and Ampyx, sp., were obtained. Table of Fossils from the Didymograptus hiJidus-SKk.TJss. J 1 i I— 1 1 » * ■i » 3 i s 1 i ■a 3 Ed 1 s o * * * * ?* » * * . » * CliTnacooroptus confertus, Lapw CiSldtUS f Jj&pw ATn/OVX BD * ^alina SD » * F ... Addaspis Suchii, Barr fU'rykonotrctcL TJiicula Kindly named by Mr. Philip Lake, F.G.S. Vol. 52.J QBOLOar of the NBI&HBOnRHOOD OP CARMiRTHBir. 529 4. Didymograptus Murchisoni-^ales. The outcrop of this series serves as a datum-line throughout tho district, owing to the fact that the fossils are characteristic and ahundant. The beds sweep across the country in a straight line from Felin-wen on the east to beyond Nant-y-ci on the west. Prom this point westward the outcrop takes a southerly bend, owing to the closing in of the anticline, and exposures occur south of Melin Eicket as described by the late Mr. Roberts. The general dip of the beds is magnetic north, at a greater or lesser angle with that of the cleavage. No actual junction with the higher beds is seen except in Pistyll-y-gwion quarry, where a yellow sandy series overlies the shales. The sporadic appearance of this sandstone, and the fact that we have here the higher beds of the Didymograptus Murehhoni-'ionei, render it admissible to suppose that the sandy rock is a local development of Llandeilo Limestone, otherwise absent throughout the district. Just north of the Plough and Harrow Inn higher beds also occur. No certain boundary-line can at present be drawn between these beds and the underlying iifidus- shales, as fossils are so exceedingly rare in the upper part of the latter. An extensive exposure of the Didymograptus Murchisoni-h&di% is seen, in the large quarry behind the Plough and Harrow Inn. They consist of dark-grey shales, with lighter stripes of more sandy material at close intervals. These stripes show the direction of the bedding, which is at an angle with that of the cleavage. The beds contained Didymograptus Murehisoni, Beck, and Siphonotreta micula. Confirmatory Sections. These beds are so well known elsewhere that we need only mention the chief exposures. Proceeding from the west eastward, we bave ; — (i) The small quarry marked on the 6-inch map as Nant-y-ci sandpit ; the purplish slates contained Didymograptus Murehisoni, Beck, and Glimacograptus confertus, Lapw. (ii) Pistyll-y-gwion quarry must be specially mentioned, on account of the overlying sandy series. The sandstone was somewhat coarse-grained and very ferruginous, and in it we found Orthis testudinaria, Dalman. The shales yielded Didymograptus Murehi- soni, Beck, Glimacograptus confertus, Lapw., Lingula, and Asaphus tyrannus, Murch. (iii) In Penlan old quarry and on Penlanflfos Hill the same shales are seen. In the old fiUed-up quarries on the hill some specimens of Diplograptus foliaeeus, Murch., were also found, owing to the fact that we have here a junction with the higher beds. At Cwm-oernant Nurseries specimens of Didymograptus Murehisoni, Beck, were obtained. In the Abergwili district to the north-east of the map we have three principal exposures. At Castell-pigyn are dark finely-splitting slates alternating with harder flaggy beds. At Pelin-wen the 530 MISS CEOSFIBLD AND MISS BKBAT ON THB [-^Ug- 1 896, shales are lighter in colour, brittle, and. extremely fissile : here the fossils are most plentiful. We obtained Didymograptus Murchisoni, Beck, Orthis testucUnaria, Dalman, Lincfula, and Orihoeeras. Near the top of Merlin's Hill is a small quarry of the same shales, which are much weathered, and the fossils are not found on the cleavage- faces. 5. Llandeilo Limestone. There is no Llandeilo Limestone in the district, but, as already suggested, this formation may be represented by the gritty and sandy flags, which are found at the northern limit of the Didymo- graptus Murchis(mi-iha\ss. In these beds we found Asaphus tyrannus, Murch., and Orthis striatula, Dalman ; but we have not attempted to separate them from the lower ones. The limestone that is used for agricultural purposes is brought from a distance, and is mainly of Carboniferous age. 6. Dicranograptus-sh&les. These beds are dark blue or black in colour, soft and well cleaved. They are seen to dip over the Didymograptus Murchisoni- shales, but the actual junction with the lower beds has not been determined. EossOs are rare except in occasional bands, and can be obtained only in a fragmentary condition, as the rocks are generally cleaved at an angle with the bedding. The shales are best seen on the Cardigan Eoad, and are exposed for about | mile from Cwarebach on the south to near Pantau on the north, where they dip under a great series of hard, unfossiliferous, silver5'-grey flags. The fossils obtained from this exposure wore Climacograptus bicornis, Hall, and Diplograptus, sp. Confirmatory Sections. The beds may be traced in several sections westward. At Cwm- du-Mawr and AUt Llyn Teg the beds are black and sometimes very carbonaceous. They yielded Diplograptus foliaceus, Murch., and D. dentatus, Brongn. At Felin Fach quarry and in the lane near the Blough and Harrow Inn, where the rock closely resembles that of Cwm-du-Mawr, the same fossils occur. Near Glan-yr-Afon in the same road, but a mile to the north, Diplograptus foliaceus, Muroh., was found. Between these last-named exposures we have a series of unfossiliferous, soft, pale-grey satiny shales cropping out, which is considerably folded, and of the age of which we are ignorant. In a quarry near Tan-yr-allt Farm, on the western slope of Llangunnor Hill, are pale-grey slates, weathering brown, inter- stratifled with hard micaceous striped slates of a gritty texture. The beds are very barren, but yielded one specimen of Dicrano- graptus, sp., cf. ramosus. Hall. Vol. 52.J GEOLOGY OP THE NEIGEBOTrRHOOD OF CAEMAETHElf. 531 7. Bala Beds. Partly surrounded by shales of Arenig age is a great series of mudstones, grits, sandy flags, sandstones, and conglomerates. These beds extend from near Mount Pleasant, the residence of Sir Lewis Morris, for more than a mile in an easterly direction, and probably farther. They are bounded on the north by the Arenig Beds of the Eoman Eoad and by the Eiver Towy. On the south they rest on the Arenig Beds of Glan Pibwr and Cilwaunydd. In the centre of the series the beds are perpendicular, as is well seen in the flag- stones and pebble-beds of Bolahaul Quarry, and in the alternating grits and shales in the quarry opposite Mount Pleasant Nurseries. Fossils of Bala age have been found at Pensarn, the Star Inn, Ant Cystanog, and Erw-wen. Pensarn is classical ground, for in the dark-blue brittle mudstones which occur near the northern end of the deep lane called the Eoman Eoad, Lady Murchison ' found Stygina [Ogygiri] Murcliisonce, Miireh., which was figured with other fossils from the same locality by Sir E. I. Murchison. This appears to be a small isolated patch of Bala Beds, as it occurs between the Arenig Beds of the upper part of the lane and the Llanvirn Beds to the north. Close to the spot marked ' Spring ' on the 25-inch Ordnance map we obtained 1 Homalonotus, Stygina Murchisonm, Muroh., Orihis alata, Sow., and Bellerophon. Ganfirmatpry Sections. In a lane behind the Star Inn, on the Llandeilo Eoad, the beds are more flaggy in character ; here we obtained Lingula tenuigranulaia, L. Imvis, Orthis alata. Sow., and JVucula Icevis. These flags, which lie on the northern slope of the Mount Pleasant ridge, may be traced for a short distance up the hill, and are seen again at Erw-wen, | mile south-east on the southern slope of the hill. Here we found Orthis alata. Sow., 0. calligramma, Dalm., O.jlahel- lulum. Sow., Lingula, sp., Bellerophon, and Nneula. At Allt Cystanog mudstones lithologically similar to those of Pensarn have been turned out of the new lead-mine. Here we collected Lingula tenuigranulaia, Orthis alata, and 0. jiahellulum. In a quarry opposite Mount Pleasant Nurseries we found a few fragments of Holopella, IV. CoMPAKISOlif WITH THE DEPOSITS OF OTHBK AkEAS. 1. Tremadoc Slates. It is unforttiHate that we have Hot found in these beds a single fossil which is characteristic of the Tremadoc Slates in other parts of Britain. The commonest trilobite that occurs is a new species of Peltura, Milne-Edwards ; but we have also two other Cambrian forms : a Parabolinella closely resembling P. rugosa, Brbg., which is 1 ' Silurian Systeiu,' pt. ii. pp. 358 & 664, & pi. xxv. figs. 3 a, 3 6. 532 MISS CROSFIELD iND MISS BKBAT ON THE [Aug. 1896, found in the Ceratojayge-Limestone, stage 3a'', and Erinnys, sp. Associated with these typical Cambrian genera is the Ordovician genus Ogygia, represented by a large new species, which occurs sparingly in our lower beds, becomes more common as we pass higher, and finally is very abundant in bods of Arenig age. This mixture of Cambrian and Ordovician forms indicates the position of these beds in the stratigraphical succession, and enables us to correlate them in a general way with the Tfemadoc Slates of North Wales and with the Shineton Shales of Shropshire. In the attempt to fix the horizon of these beds yet more definitely we have compared our fossils with those found in Stages 2 and 3 in Scandinavia' ; and we observe that both the subgenera Peltura and Cyclognathus properly belong to Stage 2 (OZemis-Etage), although one species of Cyclognathus (to which subgenus our form, strictly speaking, belongs) is found at the very base of Stage 3 (Asaphus- Etage). As Peltura punctata, sp. n., is by far the commonest fossil in our beds, we are inclined to consider that here we have the equivalents of the lower part of Stage 3 a, which Brogger ^ has correlated with the Shineton Shales. On the other hand, the presence of Orthoceras throughout our beds would suggest that we have the representatives of 3 a P or 3a^, for this genus is not recorded by Brogger as occurring below the upper part of 3a^. Our species of Parabo- linella closely resembles P. rugosa, Brog., which is confined to 3 a ^. It thus seems unwise, if not impossible, to define the position of our beds more closely than by saying that they are of the age of the Tremadoc Slates, and the equivalents of part or of the whole of Stage 3 a ; but as the shales are very rich in fossils, we hope that additional species will ere long be found to throw further light on the question. 2. Arenig Beds. The presence of Phyllograptus, sp., cf. angustif alius, Hall, at Glan Pibwr, and of Calymene parvifrons, var. Murehisoni, Salt., in the Roman Road, and Swansea Eoad upper stream, both associated with Ogygia marginata, sp. n., which has come up from the Tremadoc Beds, indicate the Arenig age of these mudstones ; and that the graptolitic shales of Hafod-wen are of the same period is shown by the presence together of such forms as Didymograptus nitidvs. Hall, Dictyograptus, sp., and ^glina binodosa. These beds are the equivalents of the Phyllograptus-shales of Wales and Scotland, of part of the Skiddaw Slates,- of the Lower Graptolite or Phyllograptus-shaleB of Norway and Sweden, of the Point Levis Beds, and of the St. Anne Zone of Canada. • Brogger, ' Die silurischen Etagen 2 und 3 im Kristianiagebiete,' 1882. ' Ibid. Vol. 52. J aaOLOOr OF THB NBiaHBOUKHOOD OF CARMARTHEN. 533 Lower Arenig. Phyllograptus angustifolius. Hall, is found in Britain, both in the Lower and Upper Zones of the Arenig, and has a wide range in Scandinavia, so that its occurrence at Glan Pibwr is not sufBcifent to determine the horizon of the beds. The absence of Calymene parvifrons, var. Murchisoni, Salt., at Glan Pibwr and Nant Pwntan, and the very great abundance of Ogygia marginata, which is also common in the Tremadoc Beds of Cwm Ffrwd, suggest that possibly these beds are of Lower Arenig age. The evidence which we obtain from their position in the field supports this opinion. The Glan Pibwr Beds are only separated from the soft Tremadoc shales of Cwm Ffrwd by a continuation of low -lying ground, presumably of the same age ; and the Nant Pwntan exposure lies between the Tremadoc Slates of Cwrt Hir and the graptolite-sbales of Hafod- wen, in the latter of which are fossils from the Tetragraptus-'/.oiie. It is impossible to compare these beds with those of other areas until more fossils are obtained. Upper Arenig. In the shales of Hafod-wen occur Didymograptus nitidus, Hall, jD. constrictus, Hall, Diciyograptus, sp., and j^glina hinodosa, which determine the Upper Arenig age of these beds. Although we have not found the characteristic Tetragrapius, yet the similarity of our fossils to those obtained by the late Mr. T. Eoberts, at Talfan etc., suggests that at Hafod-wen we have the same beds reappearing. The shales may be correlated also with the Upper (formerly Middle) Arenig of St. David's, and with part of the Teiragraptus-zone of the Skiddaw Slates described by Mr. J. E. Marr.' We have not been able to determine whether the mudstones of the Eoman Eoad and Swansea Eoad stream underlie the Hafod- wen shales, or whether they are on the same horizon but of a totally different facies. The association of Calymene parvifrons, Salt, (a typical Arenig fossil of North Wales, where it is found along with Ogygia Selwynii), with Ogygia marginata, shows that we cannot class these beds with the Llanvirn Group. We find also Calymene parvifrons, var. Murchisoni, Salt. ; as this fossil is recorded by Dr. Hicks from Perth Hayog, Eamsey Island, together with Didy- mograptus nanus, J), hijidus, D. indentus, and D. patulus, its occur- rence in the Carmarthen beds points to the probability that they are near the upper limit of the series. 3. Didymograptus bifidus-shales (Llanvirn). These beds contain Phacops Uanvirnensis, Hicks, Didymograptus hijidus, Hall, and other fossils common in the Llanvirn (= Lower Llanvirn of Dr. Hicks) of St. David's, with which beds they are exactly comparable. The occurrence of Addaspis Buchii, Barr., connects the Carmarthen beds directly with Stage D d^ of Bohemia, 1 Marr, ' Notes on tbe Skiddaw Slates,' Geol. Mag. 1894, p. 122. 534 MISS CBOBFIELB AND MISS SKBAT ON THE [Aug. 1896, and also with the slates of Angers, in both of which this fossil is found abundantly in conjunction with Placoparia. These forms are characteristic of the well-known Llanvirn fauna, and fix the position of the shales beyond question. 4. Didymograptus MurcMsoni-^aXes. These are too well known to require much notice here ; they evidently correspond to Hicks's Lower Llandeilo or Upper Llanvirn beds of Abereiddy Bay, and to part of Messrs. Marr and Eoberts's Didymograptus-shgles of • Haverfordwest. Very few fossils were found except the characteristic Didymograptus Murchisoni, Beck, and this was abundant. These beds everywhere succeed the Didymograptus 6i;?c?us-shales which constitute the true Llanvirn (Lower Llanvirn of Dr. Hicks) horizon. We see the same relations of the beds in Scania. TuUberg's division E, Middle Graptolite Shales, has at the base a zone = D. MurcJiisoni-zone. ^ ' ^ I with Didymograptus Murchisoni. y. Corresponds to part of our Didymo- graptus bifidus-zone. 5. Dicranograptus-shales. These beds regularly succeed the Didymograptus Murcliisoni- shales : near the base we have a zone containing Diplograptus foliaceus, Murch., and D. dentatus, Brongn. ; in a higher band Climacograptus hicornis, Hall, was found. Dicranogrciptus ramosus. Hall, occurs alone, so that the relationships of the sub-zones to one another cannot yet be made out. Orthis argentea, which is characteristic of the upper bands of the Dicranograpius-shales at Haverfordwest (Quart. Journ. Geol. Soo. vol. xli. 1885, p. 478), has not been found in the Carmarthen district. We have no difficulty in correlating these beds with the Glenkiln Series described by Prof. Lapworth, and placed by him at the very top of the Llandeilo Series. Not only are the chief fossils identical, but their distribution is also strikingly similar, as is shown by the following table : — Glenkiln Beds. Dicranograftus ramosus, Hall. R. Climacograptus bicomis, Hall. 0. Diplograptus dentatus. Brongn. R. „ foliaceus, Muroh. C. Dicranograptus-SHXT^HB. R. C. R. C. The same beds occur in Abereiddy Bay, at Haverfordwest, Builth, Conway, in the Lake District, in County Down, and also in Sweden. 6. Bala Beds. The age of the fossiliferous part of the Mount Pleasant series is determined by the presence of Lingula tenuigranulata, which is found elsewhere in rocks of Bala age, and is not known to occur on Vol.52.] GEOLOGY OP THE NFJOHBOUEHOOD OF CAEMARTHBN. 535 a lower horizon. Further evidence for the Bala age of these beds is the occurrence of Stygina Murchisonce, Murch. For many years Pensarn was the only known locality for this fossil, but it has been found by Messrs. Nicholson and Marr in the DrygiU Shales. A large assemblage of fossils from the same district has been examined by Miss EUes and Miss Wood, and they have recently ^ shown that these beds are of Bala age. The grits, sandstones, and conglomerates have only yielded a solitary fossil, Holopella, so that we reserve our opinion on the age of this part of the series until we can examine them more fully. In this connexion it is, however, interesting to note that Sir R. I. Murchison considered the ' schists,' grits, and sandstones as of one and the same age ; but he compared them with the slates of Angers, and placed them at the very base of the Llandeilo. V. Desceiption oe some new Species oe Teilobitbs. Genus Peltura, Milne-Edwards. Peltura punctata, sp. nov. (PI. XXVI. figs. 1-10.) Oeneralform. — Long oval, slightly narrowing towards the posterior end. Very gently convex. Trilobation distinct. Head. — Rather more than a semicircle, wider than long ; length about I that of the whole body, surrounded by a narrow raised border, which widens slightly in front of the glabella, where it is ornamented by a row of about twenty puncta placed in a groove. Glabella gently convex, parallel-sided, longer than broad, with two pairs of distinct lateral furrows curving slightly backward, a third anterior pair being sometimes very faintly visible. The second and third pairs of furrows occupy rather more than ^ of the width of the glabella. Axal furrows well marked. The neck-furrow cuts off a wide segment, and extends right across the posterior end of the glabella. The neck-segment is crossed on each side by a sloping furrow, which cuts off a triangular piece as in Peltura scarahceoides. In the centre of the segment is a small tubercle. The fixed cheeks are large, triangular, and gently convex ; the neck-furrow on the cheeks is very near the posterior margin, this furrow is continued round the genal angle and a little way up the side. Free cheeks narrow, generally absent. Genal angles rounded. Eyes small, placed very far forward, just below the punctate border. Palpebral lobe prominent. Ocular ridge distinct on some specimens, but not always well marked off. Thorax with twelve segments ; nearly half as long again as the head. Axis wider than the greatest width of the pleura, gently tapering towards the posterior end. In the centre of some of the segments a small tubercle is visible. Pleura nearly straight, sharply elevated as far as the fulcrum, which is placed ^ or less of the 1 Miss G. L. EUes and Miss E. M. B. Wood ' On the DrygiU Shales,' G«ol. Mag. 1895, p. 246. 636 MISS CBOSFIELD AND MISS SKBAT ON THE [Aug. 1 896, distance out, and then gently decurved. Obliquely grooved through- cut their length; ends rounded anteriorly, but with exceedingly short sharp points curving backward, closely resembling those of Cycl.T\ov. A crushed and distorted specimen. Shows that the course of the facial suture is probably intramarglnal. Natural size. Swansea Eoad stream, Carmarthen. 20. 0. marginata, sp. nov. Hypostome ; imperfect. Natural size. Glan Pibwr, Carmarthen. 21 . 0. marginata, sp. nov. Counterpart of fig. 20. 22. 0. marginata, sp. nov. Glabella of a very small specimen. Shows the tubercle on the occipital ring. Natural size. G-lan Pibwr, Carmar- then. 23. 0. marginata, sp. nov. Shows the hypostome and free cheek. Natural size. Nant Pwntan, Carmarthen. 24. 0. marginata, sp. nov. Ditto. Same locality, 25. 0, marginata, sp. nov. Shows the eye and part of the facial suture. Natural size. Glan Pibwr, Carmarthen. 26. 0. marginata, sp. nov. The glabella of a large specimen. Natural size. Same locality.