tion the difficulty of getting soundings sufficiently close together to develop the sub-oceanic formsor what may be called the sub-aqueous orographywe may form some idea of the difficulty of constructing reliable bathymetrical charts.

At present, from the information at hand, we can form but a sketchy idea of the land features at the commencement of Pleistocene times. The details are, however, gradually growing, though it will be reserved for others to benefit by them.

There now remain the oceanic features called 'deeps' to consider. These, as already pointed out by Dana, cannot be volcanic in origin. They occur in non-volcanic as well as in volcanic areas.

They are, in my view, produced by a sagging of the earth's crust, similar to that which originated the Mediterranean basin. They are not necessarily permanent.

Take the Sigsbee' and 'Thoulet' deeps, which are within 100 miles of the continental shelf of the North American continent. Had they existed as 'deeps' in former geological ages, how is it they have not been filled up by the rain of sediment continued through all that lengthened time? They may not be very old and not necessarily more lasting than the deep basins on continental land such as the Black Sea and Lake Baikal.

1 By an expedition sent out by the Vienna Academy of Sciences, the greatest depth of the Eastern Mediterranean found was 3,700 mètres, or over 2 miles, near a great depression which runs between Malta and Cerigo.-Science Gossip, May 1891, p. 114.

The conclusion is forced upon us that movements and interchanges of such magnitude have occurred in the distribution of the oceans and land masses during geologic time that it would be a misnomer to call them permanent.'1

Slow indeed have been these geographic permutations, and, as pointed out in Chapter III., there has probably existed connecting land between them in one place or another through long periods of geologic time.

To sum up in one sentence, the changes are essentially forms of development, the permanence is that of land connection.

1 Dr. R. F. Scharff, in a paper entitled 'Some Remarks on the Atlantis Problem,' read before the Royal Irish Academy, November, 10, 1902 (Proc. vol. xxiv., sect. B, Part 3), brings forward evidence of a varied character to support his contention that there has been at no very remote period a land connection between Europe and America.

CHAPTER XXIV

PLEISTOCENE RAISED BEACHES AND SUBMERGED FORESTS

Description of Plate XXXVII.

I AM indebted to Mr. J. J. H. Teall, F.R.S., Director of the Geological Survey of the United Kingdom, for the photograph of the Raised Beach north of Ballantrae, in Ayrshire, of which Plate XXXVII. is a reproduction. Between Ballantrae and Girvan many examples of Raised Beaches are preserved at different levels, mostly cut in Boulder Clay. The one represented is at a level of 50 feet above O. D. The beach or terrace is covered with gravel and sand.