248 REPORTS ON THE STATE OF SCIENCE. — 1916. 



the Vosges and the heights above the Meuse, not in the Meuse, the 

 Moselle, or the Rhine. The annexation of the provinces of Alsace 

 and Lorraine did nothing to damage the efficacy of their national frontier 

 from the military point of view. It rather improved it. That it proved 

 to be a great political blunder is due to German incapacity to appreciate 

 the force of that fundamental consideration which deals with the 

 will of the people and their national incapacity for assimilation. 



Lakes and deserts play approximately the same useful part as 

 barriers between rival States. In Europe, Africa, and America lakes 

 have been largely claimed in support of boundary demarcation and, 

 like deserts, they have on the whole proved efficient, even if the exact 

 position of the dividing line is but ill-defined in their midst. There 

 is, indeed, this gi'eat advantage about both of these geographical 

 features : it is seldom matter of importance that there should be exact 

 demarcation. There may be islands in lakes, or oases and wells in 

 deserts which have to be accounted for in the partition; but beyond 

 them in the great wide sweep of inland water or the sand spaces of a 

 sun-dried wilderness there is seldom the necessity for striking a distinct 

 artificial line. It would be interesting had we time to trace a geographi- 

 cal analogy between a desert frontier and a sea frontier ; and to show how 

 it has happened that through long ages of history a desert-girt land of 

 promise and development has owed continued peace and progress to its 

 environment just as much as a sea-girt island. It may happen that no 

 geographical features of any significance are available for the satis- 

 faction of the boundary maker, and that continuous and obvious arti- 

 ficial means have to be employed to make a boundary plain. Even 

 with the best assistance of nature artificial methods of marking a 

 boundary will always be necessary where man's own artificial impress 

 on the earth's surface is encountered. Passes over the heights and 

 roads traversing less conspicuous divides have to be denoted, and the 

 gateways of a countiy or a State demand careful acknowledgment, 

 but independently of such obvious points, on which it is not necessary 

 to dwell, it very frequently happens that for thousands of miles the 

 natural features (whether divide or river) are not marked enough to 

 advertise the existence of a boundary without a line of pillars or marks 

 of some sort at distances of intervisibility. A divide even may include 

 marshy flats from which rivers drain in opposite directions, or culti- 

 vated areas may intervene, so that at the best of times there is no 

 getting away from artificial expression altogether. It is, however, the 

 employment of means such as are wholly and purely artificial, where 

 nature not only has no hand in the arrangement, but where her gentler 

 efforts are traversed and discarded that so many ridiculously bad 

 boundaries come to grief. The straight line, for instance, whether it 

 represents a parallel of latitude, a meridian, or just a line projected 

 on some particular bearing, is almost invariably bad. It possesses no 

 elasticity, it is often most difficult to determine, it is expensive, and 

 terribly tedious in the process of evolution. It may cut in two local 

 interests of great importance and play the mischief with a well-defined 

 frontier. The worst mistakes in delimitation have occurred where a 

 meridian (undetermined by exact geodetic measurement) or a parallel 



