628 EEPOET OF NATIONAL MUSEUM, 1888. 



lake. Many times, as at Robenhausen,tlie lake has filled up with peat 

 aud turf, and the piles and other objects are only found by digging. 



Such is always the case with the terremare of northern Italy. These 

 were probably lake dwellings similar to those of Switzerland. The evi- 

 dence of the use of piles is manifest, aud from the relics found it is 

 believed that they were contemporaneous in time and in civilization. 



The greater number of lake dwellings are supposed to have been over 

 the water, although near the shore and where it was shallow, but many 

 (the number can not be estimated, owing to the greater facility for decay 

 and destruction) were on the mainland. 



The era of lake dwelling forms no epoch in itself; they were only the 

 incidents of location. The dwellings on a given spot may have been 

 removed again and again, even in the same age, the preceding settle- 

 ment having been destroyed, possibly by fire, possibly by an enemy. 

 At Eobenhausen, which station has given its name in France to the 

 neolithic age, there were three prehistoric occupations, one on top of 

 the other, and each was destroyed before the next began. The tops of 

 each set of piles are from 3 to 5 feet higher than the earlier set. The 

 number of houses in the first occupation has never been estimated ; 

 that of the second has been estimated at thirty, and the third and last 

 at fifty houses. The settlement covered nearly three acres and con- 

 tained about 100,000 piles. 



Keller reported in 1879 one hundred and sixty-one prehistoric lacus- 

 trine stations, and I can suppose the number discovered has doubled 

 since then. 



The occupation of the lakes for dwellings continued through the 

 bronze and iron ages, as well as during that of stone. These different 

 occupations were not always continuous, perhaps never were. In many 

 places, notably at Morges, on Lake Geneva, there are three different 

 stations occupied by prehistoric man, each independent of the other — 

 all within a space of 500 or 600 yards. The first was called "The 

 Church," the implements of which were all stone— no metal ; the second, 

 Eoseaux — a mixture of stone and the straight flat bronze hatchets be- 

 longing to the earliest period ; the third, the great city of Morges, in 

 which the implements found, to the number of five or six hundred, all 

 belonged to the fine age of bronze — no stone. Here there could have 

 been no contemporaneity— no mixture. Each must have been destroyed 

 before the other began. That this could be, is proved from what we 

 know from history, for the present town of Morges has existed for a 

 thousand or fifteen hundred years, until 1854, without a suspicion that 

 these other three towns had consecutively existed on its site. 



In the Lake of Geneva there are fifteen or twenty stations belonging 

 to the neolithic age and twenty-five or thirty to the bronze age. In the 

 common cantonal map there is shown in Lake Bienne two stations of 

 the stone age, four of bronze, and four of iron — in Lake Morat five of 

 stone, four of bronze, and two of iron — in Lake Feuchatel nineteen of 



