ior were adrift during a period of variable winds. Thus, while Harrington's 

 data suggest great constancy in direction of the currents, the data of this 

 study emphasize their inconstancy. 



Without observation of the currents it could be assumed that the trend 

 would be toward the east, because the gradient of the drainage system and 

 the prevailing winds are tovjard the east. This eastward trend is necessar- 

 ily modified locally by the islands and peninsulas in the eastern part of 

 Western Lake Erie. The available data- are inadequate to show all of these 

 modifications, even during a period of westerly winds. Obviously, then, a 

 i tremendous number of drift-bottles would have to be released in order to 

 determine the current system when the wind blows from other quarters. 



Pending further investigation, it will suffice to say that the surface 

 {I currents of Western Lake Erie flow prevailingly toward the east, but may 

 flow in any direction under the influence of winds. 



Meteorological Data 



In order to give the reader some idea of the kind of weather vAiich 

 prevails in the region of Western Lake Erie during the period April to 

 October, inclusive, a brief summary is presented here. Data on air tempera- 

 ture, rainfall, and wind at Sandusky, Ohio,^ are given in Tables h, 5, 6, 

 and 7. Normal values for the Sandusky station are based on records since 

 1877, except those for wind, which are based on records for the period 1921- 

 1930. For the most part the data need no comment, but it may be worth while 

 to call attention to a few points of special interest. 



Considering the period of seven months in question, the years 1928 and 

 1929 were nearly normal with respect to temperature, but 1930 was warmer than 

 normal. It will be shown later that the unusually high temperature in 1930 

 was reflected in higher water temperature in that year as compared with 1929, 

 and that this difference had a noticeable effect on the plankton. 



Rainfall in 1928 was nearly normal, but in I929 it was excessive, while 

 in 1930 it was well below normal. It has not been possible to find any defi- 

 nite relationship between the rainfall and any of the data collected during 

 this investigation. However, the discharge of rivers other than Detroit 

 River was greatly diminished in 1930. 



In this region the wind blows prevailingly from the southv/est. The 

 mean velocity for the period of seven months was not far different in the 

 three years, and in each year the mean velocity was below normal. 



3/ Data for Sandusky were obtained from the Monthly Meteorological Summary, 

 issued by the United States Weather Bureau, and from the Annual Reports 

 of the Chief of the United States Weather Bureau. 



Ui 



