SANDUSKY Bay AND CEDAR POINT 185 
peninsula. These have been built up by great northeast storms 
and the approximate age of each of the principal ones has been 
determined from the vegetation upon it. The older ones are 
lower than those formed in the past century. By dividing the 
difference in height between the highest aqueous deposits in two 
ridges by the number of centuries intervening between their 
formation I have obtained the average number of feet the water 
has risen each century. Itis about 2.14. <A detailed description 
of these ridges is given in the part of the paper dealing with 
Cedar Point. 
EFFECT OF VARIATION IN THE RAINFALL. 
The deepening of the water produced by tilting of the lake 
basin has not been noticeable within the past half century 
because of greater fluctuations produced by variations in the 
rainfall. Those who recall the high water of 1858 to 1862 know 
that it has not been so high since. Moreover in 1894 and 1895 
it was lower than they had ever seen it before, unless their 
observations began before the middle of the century. These 
facts seemed to disprove the theory that the water is gradually 
getting deeper. In an article in the National Geographic Mag- 
azine for August, 1903, I showed that, since 1870, when the 
Weather Bureau established stations about the Great Lakes, 
there has been a very close correspondence between rainfall and 
lake level. When the rainfall is above normal the lake rises, and 
when below normal it falls, whether we consider particular 
years, parts of years or periods of years. More recently I have 
learned that at the places in this part of the country where a 
record of rainfall was kept as early as 1857 the rain that year or 
the next was very heavy. Bulletin C., U. S. Weather Bureau, 
gives the rainfall of the United States from the earliest records 
to the end of 1891. In it are twenty-two places within one 
hundred miles of the Great Lakes that have a record for the year 
1858. From data in this bulletin I have made the table, which 
includes besides these, St. Louis, Peoria and Marietta, although 
they are somewhat farther away. In fourteen of the twenty- 
five places the maximum rainfall for the whole period of record— 
in one instance seventy-one years—was in 1857 or 1858. In 
most of the other places the rainfall of 1857 or 1858 has been 
exceeded not more than three times. No wonder there was high 
water in 1858 with such precipitation as that. If the table were 
brought down to 1901, the years given above would still stand 
pre-eminent, for the rainfall from 1891 to 1901 was generally 
light—lighter in 1895 when the lake was also lowest than in any 
other year since the Weather Bureau was established. 
