J. Lubbock on the Ancient Lake Habitations of Switzerland. 187 



and tlie same relative position. The layer of the Stone age was 

 but slightly interrupted, while that of the Bronze era was easily 

 distinguisliable by its peculiar character and color. 



Here, therefore, we have phenomena so regular, and so well 

 marked that we may apply to them a calculation, with some lit- 

 tle confidence of at least approximate accaracy. Making then 

 some allowances, for instance, admitting three hundred years in- 

 stead of one hundred and fifty, for the period since the embank- 

 ment, and taking the Koraan period as representing an antiquity 

 of from sixteen to eighteen centuries, we should have for the age 

 of Bronze an antiquity of from 2900 to 4200 years, for that of 

 the Stone period from 4700 to 7000 years, and for the whole 

 cone an age of from 7400 to 11,000 years. M. Morlot thinks 

 tliat we should be most nearly correct in deducting two hundred 

 years only for the action of the dykes, and in attributing to the 

 Kornan layer an antiquity of sixteen centuries, that is to say, in 

 referring it to the middle of the third century. This would give 

 an age of 3800 years for the Bronze age and 6400 years for that 

 of Stone, but on the whole he is inclined to suppose for tlie 

 former an antiquity of from 3000 to 4000 years, and for the 

 latter of from 5000^ to 7000 years. 



In the settlement at the foot of Mt. Chamblon we have, accor- 

 Q'ng to M. Troyon, a second instance in which we obtam at least 

 sofne approximation to a date. Tlie interest which attaches to 

 tii^ case arises from the fact that Pileworks have been found in 

 we peat at a considerable distance from the lake, whereas it i3 

 evident that at the time of their construction the spot in which 

 t% occur must have been under water, as this mode of building 

 would have been quite out of place on dry land. This however 

 indicates a very considerable antiquity, since the site of the an- 

 ient city Eburodunum must have been, at that time, entirely 

 covered by the lake, and yet the name, which is of Ce.tic origin, 

 denotes that there was a town here even before the Koman pe- 

 Jjod. In order, however,- to form an idea of the time at which 

 f'e dwellings at Chamblon were left dry by the retirement of the 

 l^ke, we must have in the valley a point of determined age to 

 ^erve as a term of comparison, and such a point we fmd in the 

 ^"cietit city of Eburodunum (Yverdon), which was built on a 

 '^'■^)e extending from Jorat to the Thiele. Between th^sdur^e 

 f d the lake, on the site at present occupied by the city of 1 ver- 

 J"'^ no traces of Roman antiquities have ever been discovered, 

 ^yrn which it is concluded that it was at that period under water. 

 5 then we admit that at the close of the fourth century the lake 

 f ^hed the walls of the Castrum Eburodense, we shall have fif- 

 ^;en centuries as the period required to effect this change The 

 f«"e thus uncovered in fifteen hundred years is 2dOO feet in 

 ^I'^^ih, and as the piles at Chamblon are at least 5o00 feet from 

 ^^e water, it may be inferred that three thousand three hundred 



