and finally to assert that these three independent faunas "form by their union an indivisible triad which is the Silurian system." [Bul. Soc. Geol. de Fr. II, xvi, 529-545.] Already, in 1852, in his magnificent work on the Silurian System of Bohemia, Barrande had given to the strata characterized by his first fauna the name of Primordial Silurian. It is difficult to assign any just reason for thus annexing to the Silurian, already augmented by the whole Upper Cambrian or Bala group of Sedgwick, (Llandeilo and Caradoc)-a great series of fossiliferous rocks lying below the base of the Llandeilo, and unsuspected by the author of the Silurian system; who persistently claimed the Llandeilo beds, with their characteristic second fauna, as marking the dawn of organic life.

Up to this time the primordial paleozoic fauna of Bohemia and of Scandinavia was, as we have said, unknown in Great Britain. The few organic remains mentioned by Sedgwick in 1835 as occurring in the region occupied by his Lower and Middle Cambrian, on Snowdon, were found to belong to Bala beds, which there rest upon the older rocks: nor was it until 1845 that Mr. Davis found in the Middle Cambrian remains of Lingula. In 1846, Sedgwick, in company with Mr. Davis, re-examined these rocks, and in December of the same year described the Lingulabeds as overlaid by the Tremadoc slates and occupying a welldefined horizon in Caernarvon and Merionethshire, beneath the great mass of the Upper Cambrian rocks. [Geol. Jour. II, 75, III, 139.] Sedgwick, at the same time, noticed about this horizon certain graptolites and an Asaphus, which were supposed to belong to the Tremadoc slates, but have since been declared by Saiter to pertain to the Arenig or Lower Llandeilo beds, the base of the Upper Cambrian. [Mem. Geol. Sur. III, 257, and Decade II.]

This discovery of the Lingula-flags, as they were then named, and the fixing by Sedgwick of their geological horizon, was at once followed by a careful examination of them by the Government surveyors, and in 1847, Selwyn detected in the Lingulaflags, near Dolgelly, in Merionethshire, the remains of two crustacean forms, the one a phyllopod, which has received the name of Hymenocaris vermicauda, Salter, and the other a trilobite which was described by Salter in 1849 as Olenus micrurus. [Geol. Survey, Decade II.] A species of Paradoxides, apparently identical with P. Forchhammeri of Sweden, was also about this

time recognized among specimens supposed to be from the same. horizon. It has since been described as P. Hicksii, and found to belong to the basal beds of the Lingula flags,-the Menevian group.

Upon the flanks of the Malvern Hills there are found resting upon the ancient crystalline rocks of the region, and overlaid by the Pentamerus beds of the May Hill sandstone (originally called Caradoc by Murchison) a series of fossiliferous beds. These consist in their lowest part of about 600 feet of greenish sandstone, which have since yielded an Obolella and Serpulites, and are overlaid by 500 feet of black schists. In these, in 1842, Prof. John Phillips found the remains of trilobites, which he subsequently described, in 1848, as three species of Olenus. [Mem. Geol. Survey II, part 1, 55.] These black shales, which had not at that time furnished any organic remains, were by Murchison in his Silurian System (p. 416) in 1839 compared to the supposed passage-beds in Caermarthenshire between the Llandeilo and the Cambrian (Bala) rocks; which, as we have scen, were newer and not older strata than the Llandeilo flags. From their lithological characters, and their relations to the Pentamerus beds, these lower fossiliferous strata of Malvern were subsequently referred by the Government geologists to the horizon of the Caradoc proper or Bala group; nor was it until 1851, that their true geological age and significance were made known. In that year, Barrande, fresh from the study of the older rocks of the continent, came to England for the purpose of comparing the British fossils with those of the primordial zone, which he had established in Bohemia and Scandinavia, and which he at once recognized in the Lingula-flags of Sedgwick and in the black schists at Malvern; both of which were characterized by the presence of the genus Olenus, and were referred to the horizon of his Etage C. This important conclusion was announced by Salter to the British Association at Belfast in 1852. [Rep. Brit. Assoc., abstracts, p. 56, and Bull. Soc. Geol. de Fr. II, xvi, 537.] Since that time the progress of investigation in the Middle and Lower Cambrian rocks of Wales has shown a fauna the importance and richness of which has increased from year to year.

The paleontological studies of Salter, while they confirmed the primordial character of the whole of the great mass of strata which make up the Middle Cambrian or Festiniog group of Sedgwick, (consisting of the Lingula-flags and the Tremadoc slates,)

led him to propose several sub-divisions. Thus he distinguished on paleontological grounds between the upper and lower Tremadoc slates, and for like reasons divided the Lingula-flags into a lower and an upper portion. For the discussion of these distinctions the reader is referred to the memoirs of the Geol. Survey [III, 240-257.] Subsequent researches led to the division of the original Lingula-flags into three parts, an upper and a middle, to which the names of Dolgelly and Maentwrog were given by Mr. Belt, and a third consisting of the basal beds, which were se parated in 1865, by Salter and Hicks, with the designation of Menevian, derived from the ancient Roman name of St. David's in Pembrokeshire. It was here that in 1862, Salter found Paradoxides with Agnostus and Lingula in fine black shales at the base of the Lingula-flags, resting comformably on the green and purple grits of the Lower Cambrian or Harlech beds. The locality was afterwards carefully studied by Hicks, and it was soon made apparent that the genus Paradoxides, both here and in North Wales, was confined to a horizon below the great mass of the Lingula-flags; which, on the contrary, are characterized by numerous species of Olenus. These lower or Menevian beds are hence regarded by Salter as equivalent to the lowest portion of the Etage C of Barrande.

Beneath these Menevian beds there lies, in apparent conformity, the great Lower Cambrian series, frequently called the bottom or basement rocks by the Government surveyors; represented in North Wales by the Harlech grits, and in South Wales, near St. Davids, by a similar series of green and purple sandstones, considered by Murchison, and by others, as the equivalent of the Harlech rocks. They were still supposed to be unfossiliferous until in June, 1867, Salter and Hicks announced the discovery in the red beds of this lower series, at St. Davids, of a Lingulella, very like L. ferruginea of the Menevian. [Geol. Jour. XXIII, 339; Siluria 4th ed. 550.] This led to a farther examination of these Lower Cambrian beds, which has resulted in the discovery in them of a fauna distinctly primordial in type, and linked by the presence of several identical fossils to the Menevian; but in many respects distinct, and marking a lower fossiliferous horizon than anything known in Bohemia or in Scandinavia.

The first announcement of these important results was made to the British Association at Norwich in 1868. Further details were, however, laid before the Geological Society in May, 1871,

by Messrs. Harkness and Hicks, whose paper on the Ancient Rocks of St. David's Promontory appears in the Geological Journal for November, 1871. [XXVIII, 384.] The Cambrian sediments here rest upon an older series of crystalline stratified rocks, described by the geological surveyors as syenite and greenstone, and having a north-west strike. Lying unconformably upon these, and with a north-east strike, we have the following series, in ascending order: 1. quartzose conglomerate, 60 feet; 2. greenish flaggy sandstones, 460 feet; 3. red flags or slaty beds, 50 feet, containing Lingulella ferruginea, besides a larger species, Discina, and Leperditia Cambrensis; 4. purple and greenish sandstones, 1000 feet; 5. yellowish gray sandstones, flags and shales, 150 feet, with Plutonia, Conocoryphe, Microdiscus, Agnostus, Theca and Protospongia; 6. gray, purple and red flaggy sandstones, with most of the above genera, 1500 feet; 7. gray fliggy beds, 150 feet, with Paradoxides; 8. true Menevian beds, richly fossiliferous, 500 feet. The latter are the probable equivalent of the base of Barrande's Etage C, and at St. David's are conformably overlaid by the Lingula-flags; beneath which we have, including the Menevian, a conformable series of 3370 feet of uncrystalline sediments, fossiliferous nearly to the base, and holding a well-marked fauna distinct from anything hitherto known in Great Britain or elsewhere.

The Menevian beds are connected with the underlying strata by the presence of Lingulella ferruginea, Discina pileolus, and Obolella sagittatis, which extend through the whole series; and also by the genus Paradoxides, four species of which occur in these lower strata; from which the genus Olenus, which characterizes the Lingula-flags, seems to be absent. To a large tuberculated trilobite of a new genus found in these lowest rocks the name of Plutonia Sedgwickii has been given. Hicks has proposed to unite the Menevian with the Harlech beds, and to make the summit of the former the dividing line between the lower and Middle Cambrian, a suggestion which has been adopted by Lyell. [Proc. Brit. Assoc. for 1868, p. 68, and Lyell, Student's Manual of Geology, 466-469.]

Both Phillips and Lyell give the name of Upper Cambrian to the Lingula-flags and the Tremadoc slates, which together constitute the Middle Cambrian of Sedgwick, and concede the title of Lower Silurian to the Bala group or Upper Cambrian of Sedgwick. The same view is adopted by Linnarsson in

Sweden, who places the line between Cambrian and Silurian at the base of the Llandeilo or the second fauna. It was by following these authorities that I, inadvertently, in my address to the American Association for the Advancement of Science in August, 1871, gave this horizon as the original division between Cambrian and Silurian. The reader of the first part of this paper will see with how much justice Sedgwick claims for the Cambrian the whole of the fossiliferous rocks of Wales beneath the base of the May Hill sandstone, including both the first and the second fauna. I cannot but agree with the late Henry Darwin Rogers, who, in 1856, reserved the designation of "the true European Silurian" for the rocks above this horizon. [Keith Johnson's Physical Atlas, 2nd ed.]

The Lingula-flags and Tremadoc slates have been made the subject of careful stratigraphical and paleontological studies by the Geological Survey, the results of which are set forth by Ramsay and Salter in the third volume of the Memoirs of the Geological Survey, published in 1866, and also, more concisely, in the Anniversary Address by the former to the Geological Society in 1863. [Geol. Jour. XIX, xviii.] The Lingula flags (with the underlying Menevian, which resembles them lithologically) rest in apparent conformity upon the purple Harlech rocks both in Pembrokeshire and in Merionethshire, where the latter appear on the great Merioneth anticlinal, long since pointed out by Sedgwick. The Lingula-flags, (including the Menevian) have in this region, according to Ramsay, a thickness of about 6000 feet. Above these, near Tremadoc and Festiniog, lie the Tremadoc slates, which are here overlaid, in apparent conformity, by the Lower Llandeilo beds. At a distance of eleven miles to the north-west, however, the Tremadoc slates disappear, and the Lingula-flags are represented by only 2,000 feet of strata; while in parts of Caernarvonshire, and in Anglesea, the whole of the Lingula-flags and moreover the Lower Cambrian rocks, are wanting, and the Llandeilo beds rest directly upon the ancient crystalline schists. In Scotland and in Ireland, moreover, the Lingula flags, are wholly absent, and the Llandeilo rocks there repose unconformably upon grits regarded as of Lower Cambrian age. Thus, without counting the Tremadoc slates, which are a local formation, unknown out of Merionethshire, we have (includ ing the Bangor group and Lingula-flags,) beneath the Llandeilo, over 9,000 feet of fossiliferous strata, which disappear entirely in