_Atolls or Coral-Islands._—Each of the remarkable coral-islands of the Pacific and Indian oceans consists usually of an irregular ring, part or the whole of which is a few feet above the sea, and which encircles an inner lagoon of no great depth. The outer margin of the reef around each island slopes rapidly, sometimes precipitately, to a depth of, usually, several hundred fathoms. Darwin, taking these facts into consideration, together with the circumstance that no coral-reefs are known to be formed at a greater depth than about 15 to 20 fathoms (90 to 120 feet), showed that all the facts of the case could be explained by the theory that coral-islands were formed in areas of subsidence. This view was generally accepted until Prof. A. Agassiz, Sir John Murray, and other writers brought forward evidence in favour of coral-islands being founded on shoals that may be areas of elevation. Much light has been thrown on this subject by recent exploration. Two instances in especial may be mentioned. The examination of Christmas Island in the Indian Ocean, South of Sumatra, by Mr. Andrews, has shown it to be a raised atoll, founded on a volcanic base, whilst, on the other hand, borings on the atoll of Funafuti, one of the Marshall Islands in the Western Pacific, carried to a depth of over 1000 feet on the ring itself, and to 245 feet below sea-level in the middle of the lagoon, have yielded results which, in the opinion of the geologists engaged, Prof. Sollas and Prof. Edgeworth David, completely confirm Darwin’s theory. It is probable that atolls originate in more than one way, some being formed in rising or stationary tracts, others in areas of depression. The important question, from a geographical point of view, is not so much how isolated atolls were formed as whether the great tracts in the Pacific and Indian oceans in which no islands occur except atolls, for instance, the Marshall, Gilbert and Low archipelagoes in the former, and the Laccadives and Maldives in the latter, have been areas of extensive subsidence during the later geological periods. Further evidence on this question may perhaps in time be furnished by additional borings, for one of which an island of the Maldive group would furnish an excellent locality, since there is in this case independent evidence to indicate that the archipelago occupies part of a sunken tract. Meantime any additional details would be useful, such as careful soundings around those atolls which have not been fully surveyed, so as to give an accurate profile of the sea-bottom in the neighbourhood. MEMORANDUM ON GLACIER OBSERVATIONS. Revised by ALAN G. OGILVIE. The recent movements of glaciers may be noted by the following signs:—