regards the actual cost of production, I believe England can compare favourably with foreign countries.

Since the 1st January 1885 the sale of Ordnance Survey maps has been entrusted to the Stationery Office. The Director-General of the Survey carries out the survey and the manufacture of the maps, and there his functions and responsibilities cease. As regards the disposal of the maps, he supplies the Stationery Office with such maps as it requires for sale, all arrangements for sale resting with that department.

We are, as I have said, about to commence the revision of Scotland, and there is one point connected with it which I am glad to have this opportunity of bringing to the notice of the Royal Scottish Geographical Society. I allude to a revision of the place-names,-a work in which, I hope, we may have the co-operation of this Society. Very great care was originally taken in the collection of the names, three authorities were obtained for each name, and the orthography was determined by men with a good knowledge of Gaelic. I am afraid, however, that there was too great an inclination to attribute a Gaelic origin to all names, and that in some places, more especially on the West Coast, the Ordnance Survey orthography is sometimes at fault. The following notes on the subject have been given to me by Lieut.-Colonel J. Farquharson, C.B., R.E., the Executive Officer of the Survey, who had very wide and varied experience during the survey of the Highlands and of the Western Islands.

Place-Names of Scotland.

1. In the south-eastern counties the place-names are not materially different from those in the north-eastern counties of Englandi.e. they are probably in the main Anglo-Saxon or Frisian, with minor mixtures of Norse and Celtic.

2. In the south-western counties of Dumfries, Kirkcudbright, and Wigtown, there is a stronger element of Norse (Icelandic or Norwegian) in common with the adjoining English county of Cumberland, where nearly all the names are of Norwegian origin. This Norse element diminishes northward through Ayr to Dumbarton and Stirling.

3. North of Stirling and the Firth of Forth, in the eastern counties, the majority of the names are probably of Celtic origin, in the interior of the country entirely so, but along the eastern coast there is a sprinkling of Norse names.

4. In all the above-mentioned districts the names have been so much corrupted that their original form and their meaning are, and probably always will be, in many cases conjectural. The Norse names are more easily recognised than the other. There is not, as in Ireland, an ancient literature in which many names are recorded in their original forms; and in Ireland also it may, in nearly every case, be assumed that the names, except those applied within a recent period, are Celtic in origin. In the above-mentioned Scottish counties there has been a much

greater mixture or succession of races, with the result that there is greater uncertainty as to the language to which the name may have originally belonged. Until, therefore, the more northern counties are reached, any amount of labour bestowed on endeavouring to trace the origin or meaning of the placenames of Scotland would hardly give any satisfactory result. The only exception might perhaps be in the case of the three south-western counties named above, where such manifestly Norse names as Wigtown, Fleet, Tinwald-(I think there is a place of that name in Dumfriesshire)-Lockerbie, Gretna, Langholm, etc., might be easily enough picked out.

5. From the Firth of Clyde northward through the western counties to Sutherland, the names on the mainland are in the interior purely and manifestly Celtic, and along the coast mainly Norse or Icelandic. Many of the Celtic names of minor features are descriptions of the feature, i.e. the name consists of more than one Gaelic word. In such cases the name has been written in as pure Gaelic as possible on the Ordnance Survey plans. This probably is more consistent, and pleases the eye better than would have been the case if English or phonetic corruptions of the names had been written; but the result must often be very puzzling to people who are ignorant of Gaelic. For instance, Beinn Fhada (Long Hill) is a well-known mountain in the south of Ross-shire, and appears on all the old maps, as it is in fact pronounced, Ben Attow. Sgùrr Fhuaran, another prominent mountain near it, is pronounced Scour Ouran, and so on. In all the old maps, whether of the whole country or of separate counties or localities, the phonetic method of spelling was used, and in the case of villages, farms, or houses, of course the recognised, and not the purely Gaelic form, has been adopted for the name by the Ordnance Survey also.

6 In the case of the Norse names which prevail along the western coasts and islands, and in the counties of Sutherland and Caithness, great difficulty arose as to the proper mode of spelling. All the local authorities, or nearly all, were in favour of spelling the names in the Gaelic fashion. The best illustrations of this are the two Norse words largely used as terminals of names-Vagr or Vogr, a bay, corrupted into the (pronounced) form -vaig, or -vay, or -way; and Fell or Fjall, a mountain, corrupted into the (pronounced) form -val. There is no v and no y in Gaelic, and no k. Therefore the question was should Fiskavaig (Fish Bay) be written as pronounced, or in its Gaelic form Fiscabhaig? Should Stornoway be written as pronounced, or in its Gaelic form of Steornabhaigh, where bh represents v, and gh represents y? Should Roneval be written as pronounced, or in its Gaelic forms of Ronebhal, as written to all mountains (bh for v) in the Lewis survey, or Ronemhal (mh for v) as generally written in other districts? 7. There is another difficulty, namely, that the persons who collected

the names on the ground, or those who were the local authorities supplying the names to the Ordnance Surveyors, were, if enthusiasts for all names being given as Celtic or derived from Celtic words, under great temptation to supply Celtic meanings and forms where they were or should have been manifestly illfounded. Thus, I think, nearly all the names of the most prominent mountains in the island of Rum were originally shown (taking Haskeval as a specimen) in some such form as Aisgemheall. Here the Norse Fjall, Fell, or Val, has been turned into the purely Gaelic word Meal a lump or eminence (of which Mheall is an inflexion, the mh being=v in pronunciation). And no doubt some meaning has been found, or at any rate has been endeavoured to be found, for the prefix Aisge.

=

If this had been done to a large extent, and I think it has been done to some extent, the place-names of these districts would have been quite misleading. There are several words used in place-names which are similar to each other in Norse and in Gaelic, but which are entirely different in meaning, and which are liable to get mixed up in the above way. For instance, the Norse terminal word -bol, a farmstead," becomes in Gaelic pronunciation pol; and if there is anywhere near the place a pool or lake it always is spelt -poll (Gaelic for pool), and assumed to take its name from the lake, although this is by no means always the case. In the island of Tiree there is a loch called Loch Vassapol. The "Vassapol" has evidently been originally "Vatns-bol" the farm on the lake (Vatn the Cumberland "water," as in Derwentwater, etc. = a lake). The farm name has disappeared, and the loch is now given the name Loch "Bhasapoll."

=

=

It seems probable, therefore, that the principal questions as to the place-names of Scotland which could be decided with comparative ease are only two in number, namely, the two questions upon which I should be glad to have the opinion of this Society:

(1) Whether names admitted to be Gaelic should be spelt phonetically with English letters; or, when they apply to natural features, in the manner in which they now appear on the six-inch map, with Gaelic letters and grammatically in Gaelic?

(2) Whether in the case of the coast and island names there should not be a thorough revision to distinguish more clearly the Icelandic and Norwegian names from the Celtic ?

I would also ask whether this Society could not, either by the appointment of a small committee of Gaelic and Norse scholars act as referees, or in some other manner co-operate in the revision of the placenames. Of course the ultimate decision with regard to the orthography must rest with the Director-General of the Survey; but I am personally most anxious, and I feel sure my successors will be so also, to obtain the advice and assistance of those best qualified to pass an opinion on the subject.

STATICAL OCEANOGRAPHY-A REVIEW.

By HUGH ROBERT MILL, D.SC., F.R.S.E.

THE task of writing the first book in any department of science is a serious one, not to be lightly undertaken. Masses of original data which have been accumulated from many sources must be sought out, sorted, and valued. They must be assimilated and elaborated so as to build up a natural system, organised by general laws deduced from many particular cases. A great gulf separates the memoirs of a specialist from a formal treatise on any branch of science. Only second in difficulty to constructing a pioneer work is the authorship of an independent book in a language not hitherto directed to the expression of a particular group of facts, for although similar works already exist in other languages, translations are rarely satisfactory, and adaptations are always clumsy. Professor Thoulet in preparing his treatise on Oceanography 1 was unaided and unembarrassed by any earlier works in French. Indeed the only recent oceanographical books have been produced in Germany, where Boguslawski, Krummel and others had made themselves masters of the subject. In English, even, there are no formal statements of results, although most of the accurate data which make oceanography a science have been obtained by British and American workers, and our language is rich in original researches whence the corner-stones of all future treatises must be quarried.

1

M. Thoulet fully recognises the difficulties with which he has to contend, and sets about the work not only with the scientific enthusiasm common to his countrymen, but with a frank avowal of his shortcomings and a generous appreciation of the labours of foreign workers which command the highest respect. The important nature of the book justifies a somewhat detailed review in this Magazine, from which many of the data for it were taken with due acknowledgment. What has to be said in the way of criticism is meant to be constructive, not destructive: we aim at fault-mending rather than fault-finding.

The volume originated in a course of lectures delivered to naval officers, and first appeared in the columns of the Revue Maritime et Coloniale. Its subject-matter is partly derived from the author's original work, and partly from official visits paid to Norway, Scotland, and Switzerland in order to report on the methods of marine and lacustrine research carried on in those countries. This origin accounts for the compilation being a little scrappy and varying in the degree of detail; although, as the author rightly points out, the facts themselves are not yet equally firmly established in all cases. The division into parts, chapters, sections, etc., is difficult to understand, and becomes confusing when subsequent reference has to be made to any passage. A uniform sub-division into consecutivelynumbered chapters and paragraphs would be a great advantage. The illustrations are of indifferent merit, although most of them are serviceable. We cannot understand the absence of maps or of an explanation of the

1 Océanographie (Statique) par M. J. Thoulet, Professeur à la Faculté des Sciences de Nancy. Paris: Librairie Militaire de L. Baudoin et Cie., 1890, pp. 492. Price, 10 fr.

omission. If the author had kept before him a set of good bathymetrical, temperature, and salinity charts of the oceans, we think that many of the descriptions would have been far more effective, and that he would have been able to take a wider view of general distributions. Indeed the worst fault we have to find, and this we can scarcely mend, is that too much space is given to the description of methods, and too little to the statement and explanation of the results obtained by means of them. How far equilibrium may be restored by the volume on dynamic oceanography, which the title seems to promise, we cannot say, for the distinction between the static and dynamic parts of such a science is very hard to draw. We regret that M. Thoulet had not the courage to introduce the euphonious and descriptive term Thalassology, which he recognises as more in harmony with the genius of the French language, especially since in the form of Talassografia it already adorns Italian, and has been adopted in English by Agassiz as Thalessography. This backwardness in the adoption of new words is unfortunately too prominent in all departments of scientific geography. French, and still more English, writers are constrained to use awkward periphrases on account of their unwillingness to create terms; in the end they too frequently adopt expressions hammered out of definite shape by use with many meanings, while their German contemporaries, rejoicing in the freedom of a boundless language, make a simple and expressive name for each new fact and idea.

It is time to consider our author's definition of the science for which, bowing to the majority, he has retained the name of Oceanography, and to glance at the way in which he has co-ordinated the facts suggesting it :

Oceanography is the science of the ocean; it comprises all the laws applicable to the sea, not only in the domain of chemistry and physics, but also in those of mathematics, mechanics, and astronomy. Oceanography concerns itself with understanding and explaining the form of submarine relief, the nature, mode of deposition and induration of the sediment accumulating in the depths, the chemical composition of the water, the distribution of heat, of salinity and density, the different gases and dissolved solids, the currents which furrow the surface and the ice which covers some parts."

This is comprehensive and satisfactory; but we cannot agree with the statement which follows, that "Oceanography is not physical geography even restricted to the sea." We believe that oceanography is a department of physical geography, because we deny that physical geography is, as M. Thoulet suggests, concerned with the earth only as the theatre of human history. While geography in its widest sense has as its aim the ultimate relations of the earth to man, the department of physical geography is set apart as a description of the earth in its natural conditions with no more special reference to man than to any of the other mammalia.

A brief historical summary in the Introduction might be advantageously extended and carried back at least to the time of Boyle and Hooke, who devised many valuable and ingenious pieces of apparatus for marine research. It is scarcely surprising to find Sir John Ross confounded with Sir James Clark Ross; but an error in enumerating the Challenger staff (p. 15) should be corrected: Captain Tizard was a naval