PECULIARITIES OF BEHAVIOR OF WELLS 137 



except in what are known as the drilled wells, just noted. It is 

 therefore a common practice to set the cylinder at as great a depth 

 as possible and where practicable to surround it with tightly 

 packed earth to shut out the air. 



An investigation of the wells of Maine, a large part of which 

 are in granites, slates, shales and other hard rocks that are free 

 from openings, showed no instances of deep freezing. In Minne- 

 sota, North Dakota and Nebraska, on the other hand, large 

 numbers of wells that penetrate porous deposits or cavernous 

 limestones freeze every winter. In Wisconsin and Michigan freez- 

 ing, though less common, occasionally occurs, especially in some 

 of the wells in the porous gravelly hills and ridges. Even in 

 Pennsylvania freezing, apparently due to the same causes, occurs 

 in oil wells at depths of thousands of feet. In states farther south, 

 especially in Iowa, Missouri, Kentucky and Indiana, wells occa- 

 sionally freeze, both those in the porous surface deposits and those 

 in limestones. 



Cause of Phenomena. -- The foregoing phenomena, including 

 the fluctuations of head and flow, breathing, freezing, etc., are all 

 found, on careful study, to be dependent on a single general 

 cause. The relations to temperature and wind direction, with 

 which the phenomena have been correlated by some, are found to 

 be only casual. On the other hand, it is found that the peculiar- 

 ities of behavior are very intimately connected with barometer 

 changes, the relation of the blowing to storms being recognized 

 by nearly every owner of a blowing well. Freezing, indraft of 

 breathing wells, low water level, small discharge and clear water 

 all occur in clear weather during periods of high barometer; 

 while the thawing of the well, the melting of the surrounding 

 snow, blowing, high head, strong discharge and milky or dis- 

 colored water occur during periods of low barometer. 



Since, in the breathing wells, the blowing is commonly asso- 

 ciated with a falling barometer and the sucking with a rising 

 barometer, it seems certain that they are caused by the variations 

 of atmospheric pressure. The essential conditions, in most in- 



