DUNES AS CHEONOMETERS. 213 



passed ; but the dunes and the ponds constantly advanced, and the 

 inhabitants were again condemned to transport their villages into the 

 midst of the heaths. These were foreseen misfortunes, and the 

 chronicle preserves silence as to the successive emigrations ; it con- 

 fines itself to mentioning the names of some churches which were 

 obliged to be abandoned to the sands and reconstructed far on the 

 plateau of the Landes. Thus we know that the church of Lege was 

 re-built in 1480 and in 1650, the first time at 2g miles, the second at 

 nearly 2 miles further inland ; but the halting-places of other local 

 monuments of the same district are not known in an exact manner. 

 As to the now vanished towns of Lislan, Lelos, and many others, 

 their ancient situation is unknown. After having lost its port and 

 its hamlets, the township of Mimizan, formerly very important, was 

 about to be entirely buried when, at the last moment, they succeeded 

 in fixing the dunes by palisades and plantations. The semicircle of 

 invading hills, like the serrated mouth of a crater, still seems to be on 

 the point of devouring the houses. 



Dunes have often been compared to gigantic sand-glasses measur- 

 ing time by the progressive march of their sandy talus. The com- 

 parison is just, for the western winds, which efiect all the changes on 

 the coast-line of the Landes, obey at present the same laws as they did 

 thousands of years ago_, and very probably their force has not changed 

 during that interval of time. The dunes, the ponds, and even the 

 villages on the shores, may thus be considered as real geological 

 chronometers ; but, unfortunately, the indications that they furnish 

 have not yet been deciphered with any certainty, and now that the 

 dunes are fixed, it. is too late to undertake this study. The illustri- 

 ous Bremontier (whose book, printed in the year 1797,* is still an 

 authority on the question of moving sands) collected during eight 

 years a series of observations which have given an average of from 

 22 to 27 yards for the annual progress of the dunes of La Teste. 

 This result agrees in a remarkable manner with the indications fur- 

 nished by the encroachments of the dunes of Lege during the last 

 four hundred years. In admitting as normal the average calculated 

 by Bremontier, one would arrive at this conclusion, that in the lapse 

 of twenty centuries the dune would be able to invade the entire dis- 

 trict of the Landes and cover the town of Bordeaux. A thousand 

 years would even have sufficed to transform the fair plains of Bor- 

 delais into marshes, for the ^tangs constantly driven back by the in- 

 vading dunes, would have spread on the eastern side after having 

 * Memoire sur les Dunes. 



