many of the past changes of our earth, explains also this undoubted feature of the Arctic climate in the Miocene warm period.

The difficulty of associating such luxuriant vegetation with the long polar night and its accompanying severe cold, has led many writers to maintain the necessity of a considerable decrease in the obliquity of the ecliptic, while others have supposed that the position of the pole itself may have changed. But astronomers and physicists both deny that such changes have occurred to anything like the requisite amount, while the latter change would not produce the desired effect, since there are proofs of a nearly simultaneous mild climate all round the polar area.* The changes of eccentricity, and of the phases of perihelion and aphelion, are, on the other hand, universally admitted and calculable phenomena; while their effects on the accumulation of snow and ice, on the winds and on the ocean currents, have been worked out in so masterly a manner by Mr. Croll, that they are rapidly taking their place among the established deductions of physical science. If we add to these such moderate changes in the distribution of sea and land, the outline of coasts, and the elevation of mountains and plateaux, as we know to have occurred again and again during the Tertiary period, we obtain a combination of causes which seem fully adequate to have brought about those wonderful changes of climate, manifested on the one hand by the recent glaciation of our own islands, and on the other by the luxuriant Arctic vegetation of the Miocene age.

Indications of Glacial and Mild Climates throughout Geological Time. It will of course occur to the reader that, if the extreme eccentricity 850,000 years ago was the cause of the mild climate in the Arctic regions and its accompanying luxuriant flora, there must have been also, each alternate 10,500 years, a glacial epoch of extreme intensity; and it may be asked, where are the proofs of such periods of glaciation? The answer is that there are some indications of such a glacial epoch, and, though very scanty, they are such as we can alone expect to find. The evidence of the last glacial epoch, which is more especially convincing, is that of the superficial effects produced by the ice-the striations, the roches moutonnées, the moraines, the travelled blocks, and the 'till.' But these have only been preserved to us here and there, because their formation is so recent, and because they once covered the whole country. The greater part of them have been destroyed; and that any traces of them still remain is probably

* Rev. Samuel Haughton's 'Notes on Physical Geology.' Read before the Royal Society, and published in Nature,' vol. xviii. p. 266.

due

due to the fact that, since the last glacial epoch passed away, there has been a period of very low eccentricity, and consequently great stability of climates and comparatively little denudation. No fragment of any such evidence from the remote Miocene glacial epoch could possibly reach us, because there has been, in almost all the time intervening between it and our last glacial epoch, an amount of eccentricity always much greater than now, and on the average nearly double. The whole lapse of time has therefore been a continued series of alternate periods of mild and severe climates, often culminating in lesser glacial epochs and necessarily leading to a great denudation and an almost complete remodelling of the earth's surface. There are only two kinds of evidence of these remote glacial epochs that can possibly remain to us,-beds containing glacial fossils, and travelled angular blocks imbedded in marine or lacustrine deposits. The former have not yet been recognised, though they probably exist, and it is quite possible that some deposits of Arctic shells, classed as glacial, may really belong to this period. The shells of warm seas would undoubtedly suffer modification and extinction by change of climate, but those of cold regions might perhaps undergo little change; for though the polar seas of mild epochs might never be frozen, neither would they ever rise in temperature much above the freezing-point, except in comparatively shallow water. Many of the Arctic mollusca live at considerable depths and have a wide range, and it is not improbable that the Arctic shell-fauna has continued almost unchanged from the Miocene period. It is therefore quite possible that some of the numerous deposits with shells, which have been found in the Arctic regions at elevations of from 100 to 500 feet above the sea, may belong to the Miocene period. The same difficulty occurs in the interpretation of the crag and associated deposits of our own country. The presence of a large proportion of living and Arctic shells has been held to prove a more recent origin, whereas it may indicate an intercalated cold period at an earlier date. Geologists almost always form their conclusions as to the age of these fragmentary deposits on the assumption that there was a continuous deterioration of climate from the Miocene through the Pliocene to the Glacial epoch. If, however, as we believe, there have always been alternations of warm and cold periods, a very different interpretation may be placed on the facts, and some difficulties may be overcome.

The other kind of evidence, that of travelled blocks, is however found in the Miocene deposits of Central and Southern Europe. In Northern Italy, near Turin, there is a sandstone formation full of Miocene shells, intercalated among whose strata

are

are beds of conglomerate, containing huge angular blocks of serpentine and greenstone, sometimes more than twenty feet long. Some of them are partially striated and polished, and similar rocks occur in the Alps about twenty miles off.

When we go back to the Eocene period, we find indications of a decidedly more tropical climate than the Miocene, in fact quite as warm as that of any earlier period, so far as we can judge by organic remains, for palms, turtles, large snakes, and crocodiles, then inhabited England. Yet in this period we also have indications of ice-action, in an extensive deposit of finely stratified sandstone several thousand feet in thickness, extending from Switzerland to Vienna, and quite destitute of organic remains, but which contains in several places enormous blocks, either angular or partly rounded, and composed of oolitic limestone or granite. Near the Lake of Thun one of the granite blocks of this formation is of huge dimensions, being 105 feet long, 90 feet wide, and 45 feet thick. The granite is red, and of a peculiar kind, which cannot now be matched anywhere in the Alps. Similar erratics have been found in beds of the same age in the Carpathians and in the Apennines, indicating probably an extensive inland European sea, into which glaciers descended from the surrounding mountains.

Going back from the Eocene all through the Secondary formations, the organic remains are such as to indicate warm European seas, and there are no satisfactory proofs of ice-action. But when we reach the remote Palæozoic formations, we lose all clear evidence of very high temperatures, since even the wonderful coal-flora is now generally admitted to be indicative of a mild or warm uniform climate, but by no means necessarily of a tropical one. Here again we meet with unmistakable signs of ice-action in the Lower Permian conglomerates of the west of England. These contain partially-rounded or angular fragments of various rocks, with striated or polished surfaces just like the stones of the 'till.' These blocks lie confusedly bedded in a red unstratified marl, and can often be traced to Welsh rocks, from twenty to fifty miles distant. This remarkable deposit was first adduced by Professor Ramsay as indicating a remote glacial epoch, and, after a personal examination of it on the ground, Sir Charles Lyell agrees that this is the only possible explanation that, with our present knowledge, we can give of it.

Numerous examples of erratic blocks, which seem to indicate ice-action, occur in the Carboniferous formation in Scotland, in France, in Nova Scotia, in Ohio, in India, and in Australia. The Old Red Sandstone and Silurian formations contain similar blocks and boulders; while Principal J. W. Dawson, of Mon

treal,

treal, who is not an extreme glacialist, and whose opinion is therefore unbiassed, believes that we have evidence of ice-action in temperate latitudes as far back as the Huronian age-that is beyond our Cambrian formation.

The only evidence wanting to complete the proof of glacial epochs having occurred repeatedly throughout all geological time, is the discovery of deposits of arctic marine shells similar to those of the drift, which have sometimes been raised more than a thousand feet above the sea. But here again it is a question whether such deposits can be recognised if they exist. The alternate periods of about ten thousand years of mild, and of glacial conditions, are so short geologically, that the marine deposits formed during a series of such changes may be represented by alternate bands or strata in one deposit, and the fossils of both periods may be more or less mingled together. Even if the deposit formed during a phase of glacial conditions is sufficiently distinct in composition to be separated from adjacent beds, its comparative poverty in organic remains, or their small size, will not be imputed to cold, because geologists have not yet recognised the constant alternation of short periods of mild and glacial climates as an established fact, but are accustomed to consider a glacial epoch to be one long-continued period of generally glacial conditions. Again, it must be remembered, that during the cold periods denudation will be greatly checked, so far as the carrying of the denuded matter to the sea by running water is concerned; while during the succeeding warm period it will be at a maximum, the melting of snow and ice being added to the normal rainfall, while there will be abundance of loose materials ground off by the ice, to load the rivers with great volumes of sediment. It follows, that marine deposits representing periods of glaciation will always be very scanty, as compared with those of the succeeding warm periods, and this is another reason why any deposits of this kind which do exist may easily have been overlooked.

But if the evidence of remote glacial epochs is rare and fragmentary, that of the occurrence of warm periods in the Arctic regions is frequent and ample. Besides the wonderful Miocene flora already described, there is a somewhat older one of the Upper Cretaceous age in Greenland, containing besides abundance of dicotyledons-such as figs, magnolias, myrtles, and the subtropical genus Myrsine-cycads, conifers, and numerous ferns, one of which is a tree-fern with a thick stem which has been found in the Greensand of England.

In the same locality in Greenland (70° 33′ N. latitude, and 52° W. longitude), and also in Spitzbergen, a more ancient flora

of

of the Lower Cretaceous age has been found, differing widely from the preceding in the great abundance of ferns, cycads, and conifers, and the scarcity of dicotyledons, which are represented by a single species of poplar. Among the ferns many belong to the genus Gleichenia, now entirely tropical.

Proofs of a mild Arctic climate in Jurassic times are found in the rich flora of this age in East Siberia and Amur-land, with less productive deposits in Spitzbergen, and at Ando in Norway within the Arctic circle. But more remarkable are the ammonites and the vertebræ of Ichthyosaurus and Teleosaurus found in the Jurassic beds of the Parry Islands in 77° N. latitude.

In the still earlier Triassic deposits of Spitzbergen, species of nautili and ammonites also occur.

True coal of the Carboniferous period has been found at Spitzbergen and at Bear Island, on the north coast of East Siberia, and it contains Calamites and Lepidodendrons, with large spreading ferns. Marine deposits of the same age contain large stony corals; while the more ancient Silurian limestones, which are widely spread in the high Arctic regions, contain abundance of corals and shells of Cephalopodous mollusca, like those of the same formation in the temperate zones.

It must undoubtedly be admitted, that this connected series of records of the animal and vegetable productions of the Arctic regions, extending over the whole vast period from the Silurian to the Miocene, inclusive, and widely scattered over the Arctic and sub-arctic zones, indicates rather a constant mild climate than an alternation of warm and cold periods; and it is therefore not surprising that this view should be generally adopted. We must remember, however, that there are yet vast gaps in the record, representing long ages of which we know nothing; while the condition of the Arctic regions during cold periods would certainly be such as not to favour the formation of stratified deposits, or the preservation of animal or vegetable remains. Denudation by water would be almost wholly checked, except perhaps where great northward-flowing rivers brought down the products of warmer lands; while the masses of debris carried by glaciers and icebergs, even if preserved to our times, would rarely contain fossils.

In view of the important part played by extensive tracts of high land in producing glaciation, it is worthy of consideration whether, in the absence of such conditions, anything like a severe Arctic climate could exist, even during periods of high eccentricity with winter in aphelion. Mr. Croll has well remarked, that the influence of the lofty ice-clad mass of Green

land