242 MODIFICATION OF THE GREAT LAKES 



is repeated. The water surface then has some different position, 

 Y Y', and the heights measured are of A above Y and of B above 

 Y'. The difference between the two heights gives again the rela- 

 tive height of A and B ; and if earth movement has tilted the 

 basin toward A or B, the change in their relative height may be 

 shown by the difference in the two results of measurement. 



As the water is in fact not still, but in continual motion, the 

 mere running of lines of level from A and B to the water does 

 not suffice, and it is necessary to determine from observations 

 on the oscillating water surface what would be its position if 

 still. Such observations are made by means of gages. These 

 are of various forms, but each consists essentially of a fixt point, 

 or zero, close by the water, and a graduated scale by means of 

 which the vertical distance of the water surface from the zero is 

 measured. 



Changes in the volume of the lake influence all parts of its 

 surface equally and at the same time. To eliminate their effects 

 from the measurements it is only necessary that the gage obser- 

 vations at the two stations be simultaneous. The effects of wind 

 waves can be prevented by breakwaters. Disturbances due to 

 currents propelled by strong winds can be avoided by choosing 

 times when there is little wind. The effects of light winds can be 

 approximately eliminated by taking the average of many obser- 

 vations, and so can the effects of seiches and tides. The effects 

 ,of differences of atmospheric pressure can be computed from 

 barometric measurements of air pressure, and the proper correc- 

 tions applied. It is also possible, by the discussion of long series 

 of observations at each station, to determine the local tidal effects 

 and afterward apply corrections ; and the land and sea bi'eeze 

 effect may be treated in the same way. 



In the investigation I was able to make, consideration was given 

 to these various sources of error, but it was not practicable to take 

 all desirable measures for avoidance or correction, because the 

 reading of gages was only partly under my control. Gage sta- 

 tions have been establisht on the Great Lakes at various times 

 and at various places, and the records of readings have been pre- 

 served. In some cases the zeros of gages were connected by level- 

 ing with bench marks of a permanent character, and in a few in- 

 stances the gages themselves are stable and enduring structures. ' 

 The most important body of information of this character is con- 

 tained in the archives of the United States Lake Survey, which 

 were placed at my service by the Chief of Engineers, U. S. A. By ^ 



