ATTEMPT TO MAKE A CENSUS. 187 



settlement at Merges, which is one of the largest in the Lake 

 of Geneva, is 1200 feet long and 150 broad, giving a surface 

 of 180,000 square feet. Allowing the huts to have been fifteen 

 feet in diameter, and supposing that they occupied half the 

 surface, leaving the rest for gangways, he estimates the number 

 of cabins at 311 ; and supposing again that, on an average, 

 each was inhabited by four persons, he obtains for the whole 

 a population of 1244. Starting from the same data, he assumes 

 for the Lake of Neufchatel a population of about 5000. Sixty- 

 eight villages belonging to the Bronze Age are supposed to 

 have contained 42,500 persons; while for the preceding epoch, 

 by the same process of reasoning, he estimates the population 

 at 31,875. 



So far as these calculations rest on the fragments of the 

 clay walls, they must be regarded as altogether unsatisfactory, 

 since Dr. Keller informs me that the largest pieces yet dis- 

 covered are only a foot in their greatest diameter. There is 

 also good reason to believe that the huts were generally not 

 circular, but rectangular. Nor am I inclined to attribute 

 much value to the estimates of population based on the extent 

 of the platforms. M. Troy on himself admits that his " chiffres 

 sont peut-etre un peu eleves, eu egard aux habitations sur 

 terre ferme, dont il ne peut etre question dans ce calcul, et 

 vu qu'on est encore bien loin de connaitre tous les points des 

 lacs qui ont ete occupes," and, indeed, in Switzerland, since 

 his book was written, the number of Lake villages discovered 

 has already been more than doubled. Moreover, M. Troyoii 

 assumes that the Lake villages of the Bronze Age were con- 

 temporaneous, and that the same was the case with those 

 belonging to the Stone Age. This also I should be disposed 

 to question; both these periods, but especially the Stone Age, 

 in all probability extended over a long series of years ; and 

 though in these matters it is of course necessary to speak with 

 much caution, still if we are to make any assumption in the 



