The surfaces that have been modified by earlier glacial epochs must in general have been long since removed by other denuding agencies. The most important evidence of former ice action consists in the occurrence, embedded in fine sediment, of large boulders, occasionally preserving marks of polish and striation, and usually, though not always, angular. Accumulations of this kind afford evidence of transport by two different agencies, water, which has brought the silt, and ice, which has carried the boulders. If the water had been in rapid movement, and thus capable of moving the boulders, it would have carried away fine silt or sand, instead of depositing it. Evidence of ice action has thus been traced equally in the boulder clay of North-Western Europe, and in the Palæozoic boulder beds of India, South Africa, and Australia, and probably of South America. It is well to search in all mountain ranges for traces of glacial action. In many mountain chains, even in comparatively low latitudes, proofs have been found of the existence of glaciers, at a much lower level than at present, dating from a comparatively recent geological period, whilst in other mountain regions none have been recognised. The question also whether glacial action has been contemporaneous in the two hemispheres is of the greatest importance, and the evidence hitherto adduced is of a very conflicting character. _Deserts._—The great sandy or salt plains, with a more or less barren surface, that occupy a large area in the interior of several continents, have only of late years received due attention from geologists. A great thickness of deposits must occur in many of these vast, nearly level, tracts, for the underlying rocks are often completely concealed over immense areas. The investigation of the deposits is frequently a matter of great difficulty for want of sections; but, where practicable, a careful examination should be made, and exact descriptions of the formations exposed recorded. Some, at all events, of these beds appear to be entirely deposited from the air, and consist of the decomposed surfaces of rocks and the sand and silt from stream deposits, carried up by wind and then redeposited on the surface of the country. Such deposits are very fine, formed of well-rounded grains, and, as a rule, destitute of stratification. The geologist who has especially described these formations, Baron F. von Richthofen, in his work on China, attributes to the loess of the Rhine and Danube valleys a similar origin. It is usual to find beds due to water-action, rain-wash and stream-deposits, interstratified with the subaërial accumulations. Further observations on these formations are desirable. The occurrence of blown sands, the origin of these accumulations, and the peculiar ridges they assume, usually at right angles, but in some remarkable cases parallel to the prevailing winds, are questions deserving of additional elucidation.