October, November, and DeceitiDerj and the high point in March, April, May, 

 July, August, September, and October. For this reason the range of the 

 averages from high to low (1.23 feet) does not give the true mean range for 

 the period. The mean range, determined from the ranges of the individual 

 years, is 1.57 feet. The maximum range (2.5 feet) occurred in 1917, and 

 the minimum (0.87 foot) in 1895. The highest monthly mean stage since i860 

 was recorded for June, I876 (57U.52 feet). The highest stage for which we 

 have a reliable record is that of I838, when the water reached a height of 

 575.11 feet. The lowest monthly mean level since i860 was recorded for 

 February, 1926 (569.90 feet). 



Annual fluctuations 



The annual mean lake level for any one year may be calculated from the 

 monthly means for that year. In Table 3 are shown the annual mean levels at 

 Cleveland for each year of the period l860-1930. For a detailed discussion 

 of the fluctuations from year to year the reader is referred again to Horton 

 and Grunsky (1927). The general subject of lake levels has been discussed by 

 Shuman (1931). 



Fluctuations ir glacial and postglacial time 



Lake Erie has had a complex geological history. Since its origin as 

 Lake Maumee at about 790 feet above sea level, it has stood at no less than 

 26 levels long enough to establish recognizable beaches, the lowest one at a 

 height of about 5U0 feet, or 32 feet below the mean level of recent times. 

 The reader may refer to Leverett and Taylor (19l5) for a detaled account of 

 the lake's history. In passing, it may be mentioned that Moseley (I899 and 

 I90I1) showed that the level of Western Lake Erie has risen in recent times. 

 The evidence rests in part on the existence oT drowned valleys such as 

 Sandusky Bay, and on the presence of submerged stalactites in the caves of 

 South Bass Island. Moseley estimated the rate of rise at 2.lli feet per cen- 

 tury for at least four centuris. He believed that the rise was caused by 

 progressive tilting of the basin toward the west. Taylor (Leverett and 

 Taylor, 1915, p. 333) cast doubt upon this as an explanation of recent changes, 

 stating that the drowning effects, at least to depths of 10 or 15 feet, are 

 probably due to a return of the large volume of discharge to the Buffalo out- 

 let following the Nipissing stage of the Great Lakes. However, in an inter- 

 view with the writer on September 7, 1932, Professor Leverett stated that it 

 is now generally recognized that tilting of the basin is still in progress. 



Transient fluctuations 



If all external disturbing forces were removed, determination of the 

 mean lake level at any one time could be made from one reading of the gauge. 

 In reality, external forces are acting almost constantly upon the surface of 



35 



