314 THE CANADIAN NATURALIST. [Yol. vH. 



" 1832. General average 2 feet 10 inclies below June 1838. 



'' 1833. General average 3 feet 2 inelies below June 1838." 



From this date to 1857 many actual measurements are given 

 by Whittlesey, and from these the curve for those years has been 

 constructed. The whole of the observations are reduced as nearly 

 as possible to the average level for each year by comparison with 

 a mean annual curve for about 10 years constructed from monthly 

 averages of bi-five-day means given by the U.S. Lake Survey. 

 1859 to 1869 both inclusive are from yearly means derived from 

 continuous observations at Cleveland by the U.S. Survey. 1871 

 to 1873 are from information kindly furnished by Gen. Comstock, 

 Director of the Lake Survey. I have no data for 1870. 



The earlier and less systematic observers of the fluctuations 

 of the lakes would scarcely give attention to any but the more 

 important changes of level, and it is possible that these in many 

 cases may have been exaggerated in amount. It would seem 

 improbable, however, from the number of observations which 

 have come down to us, that any variations of importance have 

 escaped notice. 



In the upper part of the diagram, the unbroken line represents 

 Carrington's curve founded on the number of sun-spots. The 

 broken line is a reduction of a mean curve based on the area of 

 the spots given by De la Rue, Stewart, and Loewy in the Phil- 

 osophical Transactions for 1870 ; and is introduced as showing 

 the solar periods to a later date. 



3. General Remarhs. — The first four maxima of sun-spots re- 

 presented in the table being separated by long intervals of years 

 with few spots, and not being very intense, would appear to have 

 been closely followed by L. Erie. More especially 1837, the 

 year of greatest known intensity according to both spot curves 

 (333 new groups of spots according to Schwabe), was marked 

 in its eflfects on the lakes, giving rise in 1838 to the highest re- 

 corded level of the waters in Erie and Ontario, and probably also 

 in Superior, though here the data are not so certain. The high- 

 water mark of 1838 has since been employed as the datum to 

 which all the measurements of the Lake Survey are reduced. 



The three last periods of maxima of sun-spots are extreme, and 

 the intervals characterised by their deficiency so short that the 

 lakes seem to have been unable to follow them as closely as before. 

 One period of high water being to a great extent merged in the 

 next, and resulting in a general high state of the lakes for the 



