No. 7.] WIIITTLESET — LEVELS OF LAKES. 411 



These results aud those whieh muy follow from the Govern- 

 lucnt observations are of great iniportauce to the commercial 

 interests of this Lake. Harbors, piers and channels to be per- 

 manent must be constructed with reference to low water. Prior 

 to 1854 the chief of the Topographical Corps, under whose 

 direction the surveys of the Lakes were conducted, paid no atten- 

 tion to these fluctuations. 



One of the consequences of this oversight has been the recon- 

 struction of the Canal at the Sault Ste. Marie. Locks that were 

 built to pass vessels of twelve feet draft, would at times allow of 

 only nine feet. Soundings were referred to no fixed plane. 

 There may be a difl'erence of Jive feet ^ between the early sound- 

 ings on the lakes and the true depth of water. Private docks 

 and warehouses are also affected. 



The proper plane of reference is that of mtan water level . 

 Below this, arrangements should be made for a depression of two 

 feet six inches for extreme low water, which may temporarily 

 reach three feet. 



There is another cause of depression which is having its effect 

 continually; but is so small as to be imperceptible during the 

 life of a generation. This is the perpetual wearing away of the 

 channel at the outlet. On the Upper Lakes there arc evidences 

 of a perceptible lowering of the outlets, since they assumed their 

 present general level ; but on Lake Erie itis not yet perceptible. 

 It does not probably exceed an inch in a century. 



According to tradition among the French residents of Detroit, 

 which was settled in 1701, there has not been since that time as 

 high water as that of 1838. 



A conjunction of circumstances such as to cause a state of 

 extreme depression, or extreme high water, will occur only at 

 long intervals, owing to the extent of country drained by the 

 lakes. The tables show that there is no period of seven years, or 

 a multiple of seven. This is a popular belief derived from 

 Indians, by the early settlers, which has not yet entirely died 

 away, but wiiich never had the support of observations. 

 Mean anniKtl rtnctujiti(ju, result of sixteen years' observa- 

 tions 1 It. U ill. 



Ditierenee of k'ujltoit and /o/'v.s/ months within the year. 



Cleveland, 4 years obs - •' - " 



Greatest temporary difference not due to visible storms. 



Cleveland 3 " 2 " 



Greatest permanent difference, 1819 to 1838 5 " :) " 



(Greatest temfjorari/ diftereuce between 1819 aud 1838 .... G " 11 «' 



