348 M. JEFFERSOK — -LATERAL EROSIOX ON MICHIGAN RIVERS 



beaches south of Port Austin, on the Thumb of Michigan, are not per- 

 ceptibly out of level. Inference has been made from their results that 

 "no earth movements have affected this region since the earliest Algon- 

 Ician stage, and that if an)^ deformations are now in progress in these two 

 hike basins, the}' are limited to the more northerly portions." Gilbert's 

 result, on the other hand, seems to give an entire preponderance of evi- 

 dence that the region was tipping southward along the line mentioned 

 above, at the rate of about 5 inches per hundred miles per century, 

 between the years 1858 and 1896^ Gilbert's work rests on three inde- 

 pendent pairs of stations on lakes Michigan, Erie, and Ontario, two of 

 which reach across the region of the present study of rivers. Bench- 

 marks regarded as stable at these pairs of stations were compared with 

 the observed height of water in the Great lakes during a considerable 

 period of suitable weather, identical period for the two stations of each 

 pair — in 1876, lake Michigan; 1858, lake Erie, and 1874, lake Ontario. 

 The comparisons were repeated in 1895 and 1896. Doctor Gilbert found 

 the southern station seemed to sink in the interval an amount that, when 

 allowance is made for distance and time elapsed, is equivalent to a de- 

 pression at the south end of a line running south 27 degrees Avest of 

 0.43, 0.46, and 0.37 feet per hundred miles per century — three inde- 

 pendent values that strongly confirm each other. Using the southern 

 lake Michigan station over again, paired with Port Austin, in lake Huron, 

 a comparable value of 0.39 foot is obtained; but this of course is not inde- 

 pendent of the lake Michigan pair. The accuracy of the'result rests on 

 the stability of the bench-marks. Any one of them may have changed. 

 But that all three southern marks shall have changed by amounts that 

 lead to results so accordant and systematic as these is very improbable. 

 The lake Michigan pair lie north and south of our area, the lake Erie 

 pair reach into it. Examination of the evidence cited in Goldthwait with 

 regard to beach levels reveals nothing incompatible with this result. 

 An}'' individual determination of the elevation of a beach may be several 

 feet in doubt. A beach is a somewhat indefinite object to determine with 

 precision. We note in the descriptions of writers such phrases as "about 

 20 feet," "10 to 15 feet" above the lake, and so on. That the lake sur- 

 face is incessantly fluctuating by amounts covering several feet is not ex- 

 plicitly noted by geologists. Perhaps they recognize those fluctuations 

 as less than the uncertainty involved in deciding at just Avhat point they 

 are to measure the elevation of the beach. If this be true, it is possible 

 that a tilting to the amoimt of two or three feet actual sinking of the 



