necessary to pick out the essential facts from the map without the aid of a written description, one wonders if the assumption is not too bold, so far at least as the average reader is concerned.

Curiously enough also, one fact of much general importance is not mentioned in this chapter at all. This is that the area covered by the sheet includes parts of the three states of Peru, Bolivia, and Chile. It is true that this is clearly shown on the sheet, in a diagram entitled Synoptical Index, but this might be overlooked. A full discussion of the frontier problems of the three states is, however, given in an Appendix.

Now there are still geographers who remain unconvinced, or incompletely convinced, of the value of the regional method, believing that it leads to a piecemeal type of study. Even the appearance of leaving a region "hanging in the air" increases the scepticism of these, and we feel that as a matter of principle every regional survey should begin with a precise account of the position on the earth's surface, and a summary description of the main land-forms, with notes on their extension beyond the field of study. Even where, as in the present case, the emphasis is laid on physical features, the statement of position should, we think, include political relations, if for no other reason because further information must be sought in the publications, official or other, dealing with the states concerned.

Another point of much general interest, and one in regard to which there is as yet no general consensus of opinion, concerns the names to be given to "natural regions." How far is it desirable that local terms should be used for these? One might suggest, as a basis for discussion, that a local term, especially when it is in a foreign language, should not be used unless no equally vivid English term is available; that if used it should be precisely defined at the outset, and used only in the sense defined, whatever may be local usage. In the list of natural regions given in the explanation of the diagram already mentioned as forming Fig. 5, two regions only are designated by Spanish terms. These are the Puna, defined as "high, dry cordillera with intermont basins in the south, unforested," and the Altiplano, "an intermont basin with Puna' characteristics." In the text the Altiplano is defined as the central depressed portion of the Andean peneplane, which forms a basin of interior drainage. A very large amount of space is devoted to this region, owing to its importance as an area of settlement, but the name does not appear on the La Paz sheet, and though the region is shown on Fig. 5, the scale there is too small for any details as to limits to be made out. Is much gained by the use of the Spanish word as against High Plateau? On the other hand, another area in Fig. 5 is designated by the cumbrous title of "Broken mountain slopes, moist and forested." This area is, we are told in the text, known locally as Yungas, on account of the altitude and temperature conditions, and the term is frequently used in the descriptions. In this case the use of the native term would have seemed more justifiable.

The chapter on Natural Vegetation is remarkably full and detailed, a matter of very great importance in view of the closeness of the relation between plant life and human activities. It is interesting to note that. the zones of vegetation recognised in the text and accompanying figure do not correspond in their entirety to the "natural regions" shown on Fig. 5. This point also seems to us of considerable general importance. It is one of many gradually accumulating indications that the late Dr. Herbertson's conception of natural regions requires to be rediscussed in view of the amount of material which has accumulated since he first brought it forward.

If our account of the book has been chiefly critical, it is, we may repeat, because its prime importance appears to us to lie in that it is. an attempt to formulate a model scheme for a geographical account of an area. It does quite definitely show that such an account may constitute a narrative of sustained interest, rising to a climax in a final study of human interrelations with the physical environment. But just because so much success has been attained, we have thought it well to indicate some of the problems which still remain to be solved by the geographer of to-day. It may be said in conclusion that the mass of material included, the critical account of sources, the full bibliography, and the admirable illustrations, including the two striking folded maps, render the book invaluable to all those with any interest in the geography of the Central Andes, and make it a notable addition to the publications of the American Geographical Society.

ORDNANCE SURVEY OF SCOTLAND, "POPULAR" EDITION.1

"I AM told," wrote Stevenson, "there are people who do not care for maps, and find it hard to believe. The names, the shapes of the woodlands, the prehistoric footsteps of man still distinctly traceable up hill and down dale, the mills and the ruins, the ponds and the ferries, perhaps the Standing Stone or the Druidic Circle; here is an inexhaustible fund of interest for any man with eyes to see or twopence worth of imagination to understand with." To those who do possess this "twopence worth of imagination it will be a real pleasure to welcome and to pore over the first sheets issued of the new One-Inch "Popular' Ordnance Survey Map of Scotland. These sheets are Nos. 90 (Stranraer), 91 (Wigtown), and 92 (Castle Douglas and Kirkcudbright).

Many improvements are at once to be signalled, largely due no doubt to experience of the third revision of the sheets of England and Wales, which has recently been completed. In the first place, while the Scottish small-scale maps have previously been constructed on Bonne's projection, with 40° W. as the base meridian, the new edition is to be

1 Ordnance Survey of Scotland, Popular" Edition. One-Inch Map. Sheets Nos. 90, 91, 92. Ordnance Survey Office, Southampton, 1924-1925, VOL. XLI.

Р

on the same projection (Cassini's) as the English map, and with the same base meridian (Delamere Forest), so that there will in future be no confusion at the Borders of the two countries. This change has been made possible financially by the policy of producing the maps by heliozincography instead of by hand engraving; and this has also made it possible to add an inch each way to the Scottish as compared with the English sheets, making them normally 28 x 19 inches instead of 27 x 18. This gives the great advantage of an overlap with the adjacent sheets. It is anticipated that the whole map will be complete in 1928.

As regards details, what strikes one first is the admirable clearness and legibility of the sheets. The type of all the names is most readable, of good size and well spaced out. Land relief is represented by contours in brown at 50-feet intervals, "reinforced" at every 250 feet. This is a useful guide to picking out individual contour lines. Woods are shown in green, with the addition of the conventional sign for trees; the limits of rough pasture are given; and a special feature of these first three sheets is the attractive representation of the sands and sand-hills at the mouths of the rivers and along the coasts. The employment of a very pale buff tint with black stippling gives admirable results. Parishes are outlined and named.

There are four points only where a question for consideration might be raised. The first is the emphasis given to the roads by the use of red colouring for all "fit for fast traffic." It is safe to prophesy that these one-inch sheets will be little used by motorists. They will be used by the "foot slogger," and for studying the country as a whole.

It seems a pity to let these prominent red roads relatively dwarf the importance of the general lie of the land. Next, there is the general question of sheet limits. The Wigtown and Castle Douglas sheets overlap each other by eleven inches; but the land on the Wigtown sheet is cut by the estuary of the Cree into two detached portions, absolutely without connection on the sheet. Let it be admitted at once that there are few harder tasks than arranging sheet limits which will satisfy every one, all over the map. But it may be suggested that if the Wigtown sheet had been turned the other way on the paper, so as to run northward to Newton Stewart, it would likely have had a better sale.

The two remaining queries both refer to the blue outline, representing water. Not on all but on many of the outlines of lochs, the lines indicating the north and west shores of the loch seem to be considerably broader than those for the south and east shores, giving a "shadow" effect. To some this may give a slightly antiquated impression. It is surely not a device much in use on modern maps. Lastly, the register" of black and blue in printing has been rather less good than that of brown and blue. This is perhaps due to the fact that the Ordnance Survey practice is to draw water and contours on the same sheet, making two photographic negatives from this, which give water and contour plates respectively, while the black name and outline plate is drawn separately. At any rate, in many cases the gap left in the blue

course of a stream fails to fit the black lines of a road which crosses it. One wonders whether it is not attempting unnecessarily much to break streams in this way, and whether they might not equally well be left continuous, as contours are left.

There

These points are probably only matters of individual opinion. will at any rate be nothing but universal agreement that the new sheets are very fine specimens of the cartographer's art, and a cause for hearty congratulation to the Ordnance Survey. They are priced at 2s. each, paper, flat, and 3s., mounted on linen and folded, in an attractive cover, "featuring " (as one must, I suppose, say nowadays) the Lion Rampant. W. R. K.

THE INTERNATIONAL GEOGRAPHICAL CONGRESS

AT CAIRO.

UNDER the auspices of the new International Geographical Union an International Congress was held at Cairo from April 1 to April 9 last, with a preliminary meeting at Alexandria, which began with the arrival of many of the European members on Saturday, March 28.

The choice of Cairo as the meeting-place was due to the initiative of the Royal Geographical Society of Egypt, whose secretary, M. Adolphe Cattaui Bey, carried out much of the laborious work of organisation. We regret that we have to record the sudden death of M. Cattaui, which took place on June 11, not many weeks after the close of the Congress. His loss will be severely felt by the Egyptian Geographical Society, for which he did much, and will be greatly regretted by all those who came into contact with him during the meeting of the Congress. S. E. Adly Yeghen Pasha was the president of the Organising Committee, and took an active part in the proceedings throughout. The Congress coincided with the jubilee of the Royal Geographical Society of Egypt, which was celebrated by a formal meeting in the afternoon of Friday, April 3. H.M. King Fuad has done much to promote the activities of this Society, which is housed in a fine building in the garden of the Ministry of Public Works. Here and in some adjacent buildings the separate Sections of the Congress found accommodation, while the general meetings were held in the Society's Great Hall. Within this there was on view a number of large relief and geological models illustrating conditions in the Nile valley, while an exhibition of fresh and dried specimens of some of the chief economic plants of Egypt was set out in another part of the building.

The president of the Congress was General C. G. Vacchelli, one of the official representatives of the Italian Government, who was assisted by Sir Charles Close, the secretary-general of the International Geographical Union.

The attendance, both of local and of foreign members, was large.

It is difficult to give precise figures, as the Provisional List issued, if one may judge from the section devoted to the British Isles, contained the names of many who had expressed a desire to attend, but were unable to carry out their first intention. This list contains nearly 600 names. Many nationalities were represented, France, Italy, and Poland sending large contingents, but there were no delegates from the Scandinavian countries, nor from Austria, Germany, Hungary, and Russia; the United States also was not officially represented, though Dr. E. L. Stevenson of the Hispanic Society was present as a private individual. The British party was comparatively small, including only about a dozen representatives. The Royal Scottish Geographical Society had appointed Mr. Alan Ogilvie and the Editor as delegates, but at the last Mr. Ogilvie was unable to go. Professor Gregory and Mr. A. Stevens of Glasgow University, who had also hoped to be present, were also both prevented.

The Congress was organised in five Sections, which were as follows: (1) Mathematical Geography, Geodesy, and Cartography; (2) Physical Geography; (3) Biological and Human Geography; (4) Anthropology and Ethnography; (5) History of Geography and Historical Geography. The number of papers was so large that any attempt even at a list is impossible. A few notes may be given in regard to some of those heard by the writer, mostly delivered before Sections 2 and 3. Professor E. de Martonne exhibited a map, compiled by him and his students, showing the world distribution of basins of inland drainage, and read a paper on the facts brought out. He finds that, apart from the polar areas, some 28 per cent. of the land surface of the globe has no regular outlet to the sea. The percentage is greatest in Australia (60 per cent.); Africa comes next with 49 per cent.; Eurasia follows with 32 per cent. ; while the percentages in North and South America are 14 and 7 respectively.

A paper which aroused much interest, and gave rise to an animated discussion, was one read by M. Demangeon on the distribution of rural dwellings, and the causes which determine concentration, such as occurs, for example, in the Mediterranean area, as compared with the isolated habitations found elsewhere. His conclusion was that the distribution depends primarily upon the original type of land tenure, fertile areas tending to show dwellings crowded together, while isolated farmhouses tend to occur in the poorer areas, which were later in being occupied and cultivated. His conclusions did not command universal acceptance, and there is a prospect that fuller discussion may take place at the 1928 conference, which it is proposed to hold in London.

Dr. L. W. Collet presented his volume on the lakes of Switzerland and made some observations on his results, and also communicated a paper by one of his students on sedimentation in the Lake of Geneva. A paper of great interest, both local and general, was that by M. Victor Mosséri on the fertility of the soil of Egypt. The special point brought out was the effect of the older method of basin irrigation in conserving soil fertility, as compared with the dangers of perennial irrigation.