BAROMETRIC PRESSURES ON THE GREAT LAKES 119 



Lake Erie and where the depth is decreasing very rapidly from east to 

 west. 



Similarly, from a study of the depths, the corresponding rates of propa- 

 gation, and the probable points of reflection, it appears that the 2.6 hour 

 seiche, observed at Cleveland is probably a crosswise oscillation back and 

 forth between the 10-fathom curve off Cleveland and the 10-fathom curve 

 on the opposite (Canadian) shore of the lake, about midway between Point 

 Pelee and Point aux Pins. 



Though the hourly observations of elevation of water surface at Mil- 

 waukee, Harbor Beach, and Mackinaw show that seiches of moderate range 

 are prevalent at each of these stations, it was not feasible to make a careful 

 study of these within the limits of this investigation, with the one exception 

 of the study made for Mackinaw, as indicated below. 



SEICHES AT STRAIT OF MACKINAC. 



The Strait of Mackinac constitutes a connection between Lake Michigan 

 and Lake Huron. The Strait for a length of 20,000 feet at its narrowest 

 part has a width of about 20,000 feet and a mean depth of about 68 feet. It 

 appeared to the investigator that there was probably a peculiar seiche hav- 

 ing its nodal line at about the middle point of the length of the narrowest 

 part of the Strait, and that the oscillation would be such that the whole 

 surface of Lake Michigan would rise while the whole surface of Lake 

 Huron went down, and vice versa, as the current ran alternately westward 

 and eastward through the Strait under the influence of inertia, gravity, 

 and friction, after such an oscillation or seiche had once been started. 

 The investigator did not know of any adequate treatment of the problem of 

 determining the theoretical period of such a seiche. Therefore, he started 

 from the known elements of the problem: the areas of the two lakes; the 

 Chezy formulae for the relations between the velocity of a steady current 

 flowing in a channel, the slope of the water surface, and the dimensions of 

 the channel; an assumed coefficient of roughness for use in that formula; 

 and the known relations between mass, force, and acceleration. By a step- 

 by-step process he computed the period of oscillation by main strength and 

 computed the probable rate of damping. The conclusions reached were 

 (1) that the period is about 7 hours for a complete oscillation, and (2) that 

 the damping is at least sufficient to reduce the amplitude of the wave by one- 

 sixth part in each successive half wave. 



After the computation was complete, an inspection was made of the 

 graph of hourly elevations of the water surface on 42 days at the Mackinaw 

 gage, which is at the eastern end of the Strait. The inspection was not de- 

 tailed or complete, as the available time was short. Two cases were found 

 on which a 7-hour seiche was apparently started at Mackinaw by an im- 

 pulse due to barometric pressures, which was peculiarly well adapted to 

 start such a seiche. It was also noted that on May 18-23, 1911, there was 



