6 ' FLUCTUATIONS OF LEVEL 



Neither general opinion nor tradition can be reduced to feet and inches, and I 

 have, therefore, discarded from the above table whatever depended solely upon the 

 recollection of one person, who had taken no measurements or preserved no memo- 

 randa. There are, however, certain objects, such as roads, wharves, and buildings, 

 that serve as points of reference for high and low water, and tend by association 

 vividly to impress upon the memory facts of this character. The old French inha- 

 bitants of Detroit have no tradition of a water level below that of the year 1819, 

 although Detroit has been occupied since 1702. At Buffalo the year 1810 is 

 remembered as one of low water, nearly or quite as low as 1819. 



In discussing the data here presented, it is apparent that the surface of the Lake 

 is not strictly level, and thus there are discrepancies as to the time of high and low 

 water at different places. The form of the coast at Buffalo is such that the licight 

 of water is affected by it in connection Avith certain winds. Tliose from the east 

 and northeast keep back the waters, and cause a depression that may be observed 

 for one or two months at a time. The reverse occurs with prevailing winds from 

 the west and southwest. The waters driven eastward between two shores, con- 

 stantly approaching each other, are raised above the general surface like the tides 

 in the Bay of Fundy. On the 18th of April, 1848, it appears from the register of 

 Mr. Lothrop that a gale from the northeast reduced the level of the Lake to a point 

 fifteen feet six inches below the surface of October 18th, 1849, when a terrible 

 storm occurred from the southwest. At Cleveland the greatest observed local 

 fluctuation Avas three feet two inches, Avhich took place on the 19 th of November, 

 1845. As the Lake is broad opposite Cleveland, and the place is situated not far 

 from the middle, its surface would be less affected by winds; and here the level 

 during the summer of 1819 is regarded as the lowest. 



But if that year did not differ from other years in the period of the annual rise 

 and depression, it must have been still lower in the winter tlian in the summer 

 months. Dr. Houghton has mentioned one observation, made some time in the 

 winter of 1818 and 1819, by which the water in the Detroit Kiver was six feet eight 

 inches below the flood time of 1838. The winter season, however, has been little 

 noticed, except by those who keep water tables ; and at that time a regular register 

 was not known. 



For want of better data, the well noted low water in the summer of 1819 is com- 

 pared with the great rise of the summer of 1838, two of the most remarkable years 

 in the history of the fluctuations. In 1838, on the shores of Lake Erie, grounds 

 were submerged on which old orchards had come to maturity, and on forest lands 

 trees that were centuries old were killed by the overflow of the Lake water. In 

 the month of June I observed small boats passing from house to house in the streets 

 of the village, at the mouth of the Conneaut river, Ohio. The Avater rose at 

 Cleveland, in the month of July, so as to cover the floor of a warehouse to the 

 depth of one foot. These events served to revive the memory of past times, and to 

 stimulate observation in coming years. 



Among the old inhabitants it brought up afresh the popular idea derived from 

 the aborigines, that the rise is periodical, occurring once in seven years. This 

 belief is very generally entertained, and many persons related the several years 



