close of the Upper Llandeilo period in the British area. Eight of these hold their ground in their own area and, as we shall see, pass on into the younger deposits of the Gala group of the south of Scotland (Lapworth). The whole sixteen reappear also in the Coniston mudstones of the north of England (Caradoc). It is quite clear, however, that though eight species of a total of thirty- six, or twenty-two per cent., continued to exist in their original area in the south of Scotland, a great migration must have accompanied an almost equally great extinction of species, marking the close of the Upper Llandeilo period in the Scotch area. Seeing that so large a number of forms, as above mentioned, succeed in retaining their hold on their own area, it is difficult to assign any adequate cause for this migration ; but that such an exodus actually took place seems certain from the two following considerations : — In the first place, all the species of the total thirty-six inhabiting the Upper Llandeilo area of the south of Scotland, which did not actually die out, are found, to the number of sixteen, in the mudstones of the Coniston series of the north of England. Now these beds overlie the Bala or Coniston Limestone ; and the Upper Llandeilo rocks of Scotland underlie the same formation (Girvan Limestone). It seems certain, therefore, that the species of the Coniston mudstones were, to the above extent, derived from the Scotch area by migration ; and traces of such a migration are to be found in the Girvan Limestone itself. In the second place, six species of the Scotch area are found in strata of Caradoc age in Ireland. Five of these, and probably the sixth, are also found in rocks of about the same age in North America ; so that they reached Ireland en route for the American area, and the former country was merely the first halting-place in their migration.
C. Coniston Area of the North of England. — The above considerations lead me to the consideration of the second great Graptolitic period of the north of England, in which the mudstones of the Coniston series were deposited. Whatever may have been the sequence of phenomena at the close of the Skiddaw period, it is certain that the Skiddaw slates are overlain by a great mass of rocks in which no Graptolites are found. That this is really due to the absence of Graptolites from this area, and not to the general igneous activity shown by the nature of the rocks in question (Green Slates and Porphyries), is proved by the occurrence in the series of a dense mass of finely levigated dark-coloured shales, crowded with fossils, and perfectly suitable for the preservation of Graptolites, had such existed. The rocks to which I allude are the " Dufton Shales," which underlie the Coniston Limestone, and which contain no Graptolites. The Coniston Limestone itself, which corresponds with the Bala Limestone of Wales, is also wholly barren of Graptolites. Immediately above the Coniston Limestone, however, occurs a mass of dark mudstones, of variable but never great thickness, to which the name of " Graptolitic Mudstones " was applied by Professor Harkness and myself, owing to the extraordinary abundance in them of these organisms. It is perfectly clear, therefore, that the seas of the area of the Lake- district became abundantly stocked with Graptolites at a time imme-
