Dr. Callaway's inquiries into the genesis of the gneissic rocks of the Malvern Hills, and Mr. Peach's experience of the Archæan rocks of Scotland, though not exactly corroborating Mr. Marr's views, are certainly not at variance with them. Dr. Callaway believes that the hornblendegneiss of Malvern is a crushed and modified diorite; that the mica-schists have been similarly formed from felsite, and the mica-gneiss from granite;1 while he considers some of the banded gneisses were produced by earth-pressures acting upon complex interveining of granite in diorite, resulting in a parallelism of the veins; so that he would regard many of the Archæan gneisses and schists as metamorphosed igneous rocks, and not as metamorphosed sedimentary rocks. Mr. Peach says: "The gneisses all bear evidence of having been formed by the crushing and recrystallization of igneous rocks, their schistosity being due to mechanical movement of the particles produced by differential pressure." 2 Without entirely acquiescing in Mr. Marr's hypothesis, there can be little doubt that a large portion of the British Archæan rocks are of igneous origin, and that only a small proportion are sedimentary, and that they were accumulated in a volcanic district which was eventually elevated into lofty mountain ranges. 1 "Quart. Journ. Geol. Soc.," vol. xliii. p. 525. CA CHAPTER II. CAMBRIAN PERIOD. § 1. Stratigraphical Evidence. AMBRIAN rocks rise to the surface in several parts of the British Islands, but they nowhere occupy any very large tract of country, so that we can only compare their isolated exposures, and cannot trace their stratigraphical variations from one district to another. In England and Wales there are five districts where rocks referable to the Cambrian system occur-South Wales, North Wales, Shropshire, Warwickshire, and the Malvern Hills. Wherever the base of the system is exposed, the basement bed is found to be a conglomerate containing fragments of Archæan rocks, and resting unconformably on a very uneven surface of those rocks; the conglomerate being probably in process of formation at different levels throughout the whole of the period, and belonging, therefore, to different stages at different places. The Lower Cambrian rocks are not known to exist east of the Longmynd, but Upper Cambrian shales have been found in Warwickshire, and may occur at intervals beneath many parts of our Midland and Eastern counties. The stratigraphy of the Cambrian rocks is not yet completely understood; there are many unsolved difficulties connected with them, and any account of them must at present be regarded as provisional. İn Wales the normal succession of the Cambrian rocks is as follows: Tremadoc slates, 1,000 to 2,000 feet Lingula Flags, a group of sandy flagstones and slates, 2,000 to 5,000 feet Menevian slates, 200 to 750 feet Harlech Beds, red, green, and purple grits and slates, 3,000 to 9,000 feet Upper Cambrian. Lower Cambrian. The Harlech Beds of Merioneth are more than 8,000 feet thick, and the base is not there seen, but when they reappear on the west side of the Snowdon range, they are much thinner, and are probably not more than 3,000 feet thick, including the basal conglomerate of Llyn Padarn, which lies on the eastern flank of the main Archæan ridge. On the west side of this ridge the basal conglomerate is overlain by grits and slates, and, near Bangor, by sandstones and mudstones, which have been referred to the Harlech series, but it is quite possible that they are shore beds of Lingula Flag age. As they are directly succeeded by Arenig slates (Ordovician), it does not seem probable that the Lingula Flags should be entirely absent, for the lowest beds usually thin out before the higher. If, therefore, we regard the greater part of the Bangor Cambrians as Lingula Flags, it follows that in the space of about two miles the Harlech series has thinned from 3,000 feet to, perhaps, 100 feet of grit and conglomerate. This suggests that the surface of the Archæan rock formed a steep slope, inclining to the south-east, when the Cambrians were deposited against it. In Shropshire the same thing seems to take place on a still larger and more surprising scale; for the rocks of the Longmynd, which are supposed to be of Lower Cambrian age, and which lithologically resemble the Llanberis grits and slates, have a great thickness (possibly 10,000 feet), and are faulted on each side against tracts of undoubted C Archæan rock. But the eastern or Caradoc Archæan ridge is flanked on the other side by a quartzite conglomerate, which is succeeded by sandstones of Middle or Upper Lingula Flag age; the Lower Cambrian being entirely absent. How can this sudden disappearance of so thick a mass of rock be accounted for? There seem to be only three possible ways of meeting the difficulty: (1) That the Longmyndian sediments were raised into a block of land about the epoch of the Menevian Beds, and that this block afterwards sank again to receive the upper half of the Lingula Flags and the Tremadoc Slates. (2) That the Caer Caradoc range was the border of a mass of Archæan land, and formed a precipice or steep declivity about 10,000 feet high, so that it was not over-topped by the Cambrian Sea till the time of the Lingula Flags. (3) That the faults on the east side of the Longmynd and Caradoc ranges are thrust planes, bringing together portions of districts which were originally many miles apart. But the faults appear to be of the normal kind, and there is no evidence of the great crushing which must accompany lateral thrust. If none of these suggestions are considered satisfactory, it can only be argued that the Cambrian age of the Longmynd rocks is an assumption, and that the facts of the case show them to be physically more closely connected with the Archæan than with the Upper Cambrian. It is indeed quite possible that they are pre-Cambrian, and belong to the interval which is elsewhere represented by the gap between the Archæan and the Cambrian. If, however, they should prove to be of Lower Cambrian age, we must select the first of the above hypotheses as the least unlikely, and must suppose that an upheaval took place in the middle of the Cambrian period, but that the land so formed was soon submerged again beneath the sea of the Lingula Flags. In South Wales there are indications of a similarly rapid disappearance of Lower Cambrian sediments. Near St. Davids these beds are only 4,000 feet thick, which is about half their thickness in Merioneth. Whether they thin eastward in the trough between the two principal Archæan tracts is not yet known, but at Trefgarn on the south-east side of the eastern massif no Harlech Beds or Menevian are found, and the Upper Lingula Flags with a basal conglomerate appear to rest on the Archæan. It is true there are some signs of faulting, and Dr. Hicks believes the Lower beds are faulted out, but more evidence of this is required. Dr. Callaway favours me with the following remarks on the lithological composition of the Lower Cambrian rocks: "These sediments show that the adjoining lands were partly built up of granitic, volcanic, and metamorphic rocks. The purple conglomerates and sandstones of the Longmyndian, reaching a thickness of several thousand feet, as well as the broad band of pale-green slates which skirts the eastern side of the Longmynd, are largely derived from the rhyolites of the Uriconian and indicate the proximity of an extensive land area. Rounded fragments of granitic and schistose rocks are not uncommon in the conglomerates, while the grains of quartz and felspar in the grits have probably the same origin. These grits grow less felsitic towards the south, and fragments of metamorphic rocks increase in number—indications which suggest the conclusion that the Malvern Hills are a worn fragment of a mass of land which made a conspicuous feature in the Longmyndian ocean. "The rocks of North and South Wales afford similar evidence. The conglomerates of Moel Tryfaen and the sandstones further west contain a very large proportion of fragments of felsite similar to the Archæan rhyolite near Carnarvon, while the massive grits of the Merionethshire |