May 24, 1906] 



NA TURE 



n 



quakes which he has tabulated, and the work before 

 us deals with the records of 171,434 distinct shocks. 

 The labour involved in this compilation would have 

 formed no light task for any man, and when we re- 

 member that, besides being a specialist in seismological 

 statistics, the author is an officer on the active list of 

 Ih^ French Armv, the result seems almost miraculous, 

 in summing up the results of all this compilation 

 the author holds that he has conclusively established 

 the independence of earthquakes and volcanoes, and 

 the greater prevalence of the former along those 

 tracts where the surface relief shows the steepest and 

 longest gradients. Both these conclusions had been 

 reached by Prof. Milne while working in Japan, and 

 the second of them is only an empirical, and not 

 invariable, way of expressing the general principle 

 that earthquakes are most abundant where the crust- 

 movements have been greatest and most recent, while 

 they become rarer as these movements are older and 

 h;ive more or less completely died out; but we must 

 remark that earthquakes seem to be more particularly 

 associated with the changes resulting from, or accom- 

 panied by, compression, for the dropped valleys of 

 the Jordan, the Red Sea, and of Central .Africa are 

 not specially affected by earthquakes. 



Comte de iMontessus attempts to carry his con- 

 clusions still further, and finds that earthquakes are 

 almost confined to certain bands which correspond 

 with the secondary geosynclinals of Haug, and are 

 said to lie along two great circles, making an 

 angle of 67" with each other. We have had the 

 curiosity to plot these bands, as shown on the map 

 accompanying the book, upon a globe, and have failed 

 to find any correspondence between them and the 

 great circles as defined, or, indeed, with any other great 

 circles; approximately, they seem to form a network 

 ot arcs of great circles, joining up in groups of three 

 and four, an interpretation which is more probable 

 than the other, though the departures in detail render 

 the correctness of either view doubtful. However this 

 mav be, the fact remains that nine-tenths of the 

 shocks recorded have originated in regions which 

 cannot cover more than one or two per cent, of the 

 gliibe and are almost all distributed along certain 

 lines, of which the most important are the great 

 girdle of the Pacific, the line which runs up from 

 the Sunda Islands, through .Arracan, the Himalayas, 

 Caucasus, and Alps to the western Mediterranean, 

 ;md another which runs up from the Caucasus 

 through the mountains of Central Asia to Lake 

 Baikal, possibly continuing to somewhere in the 

 neighbourhood of the Bering Straits. 



Though, in the main, the distribution of the more 

 violently shaken regions shows no change from that 

 drawn by Mallet in 1S58, there is a radical difference 

 in the character of the two maps. In Mallet's the 

 frequency of earthquakes was indicated by the depth 

 of tint, and the dark patches shaded off gradually 

 into the white; de Montessus, believing that it is a 

 mistake to treat an essentially discontinuous pheno- 

 menon as a continuous one, has made the limited 

 areas, where destructive earthquakes are known to 

 NO. 1908, VOL 74] 



originate, black, and left the rest of the map blank. 

 This abrupt boundary between the regions classed as 

 seismic and the much more extensive ones classed as 

 peneseismic or aseismic, is held to be a better repre- 

 sentation of what is actually the case than any 

 gradual shading of the one into the other. The 

 difference between the two maps is, in fact, one of 

 principle; Mallet's was meant to indicate the fre- 

 quency with which earthquakes were felt, that of 

 de Montessus the frequency with which they originate. 

 Each of these facts is interesting in itself, but their 

 delineation must necessarily differ, apart from any 

 question of increasing perfection of the data. 



We have indicated some of the conclusions drawn in 

 this book, which do not seem to be so fully established 

 as its author suggests, but this must not be taken in 

 derogation of the value of his work in statistical 

 seismology. We welcome this summary of his re- 

 searches, and regret that he should have followed' 

 the custom, so common in France, of omitting a 

 subject index. 



OUR BOOK SHELF. 



Notes on Shipbuilding and Nautical Terms of Old in 

 the North. By E. Magnusson. Pp. 62. (London : 

 A. Moring, Limited, 1906.) Price li. net. 

 This small volume reproduces a paper read before the 

 Viking Club Society, and its appearance will be 

 welcomed by all who are interested in the history and 

 development of shipbuilding. Although it deals chiefly 

 with Scandinavian records and discoveries, it contains 

 an excellent summary of Greek and Latin references 

 to ancient ships, and does not leave unnoticed much 

 older Egvptian types. In short it is a scholarly per- 

 formance, and the writer has a full appreciation of 

 technical developments which have accompanied pro- 

 gress in shipbuilding. Wide reading and research 

 must have been undertaken to provide the materials ; 

 they have been dealt with in a terse but clear style, 

 and the result is of permanent value as a book of 

 reference and a bibliography of the subject. An 

 excellent glossarial index is appended. The only 

 regret one feels is that there are no illustrations. The 

 rock-carvings of ancient ships found in Egypt, Sweden, 

 and Norway are described and compared ; but simple 

 illustrations would have emphasised the deductions 

 made by the author. Again, the details of methods of 

 construction which Mr. Magnusson gives are readily 

 understood by experts in shipbuilding, but would be 

 grasped bv general readers also if diagrams of a simple 

 nature had been given. The ancient ships found in 

 Scandinavia and preserved in museums might also 

 have been pictured with great advantage. Of course 

 size and cost would be increased if this were done, 

 but that action is well worth the consideration of both 

 author and publisher, as the permanent value of the 

 book would be greatly increased thereby, and its place 

 in the libraries of all interested in shipbuilding would 

 be assured. 



A book so condensed in form and substance must be 

 read to be understood. Mr. Magnusson does not 

 claim originality in discovery or treatment. He starts 

 with the log and raft of the stone age, passes to the 

 canoe hollowed from a single log by the use of fire 

 and flint implements; traces the development of the 

 coracle and other hide-covered vessels, with internal 

 framework; shows how these " skins " were replaced 

 by wood planks, first fastened by thongs or withes. 



