148 Wtsco7isin Academy of Sciences, Arts, and Letters. 



sippi river. The depth of the proposed canal is for vessels 

 drawing five feet, and the lengths of the river to be improved 

 to enable boats navigating the canal to cross the river, are, at 

 the upper crossing, 2,300 feet; at the middle crossing, 7,000 

 feet ; and at the lower crossing, two and a half miles. There 

 are two classes of obstructions. Those arising from causes 

 not now operating and which once removed will not return, 

 and those arising from causes now operating and which must 

 be constantly recurring until the cause is removed. In this 

 latter class we find sand the chief obstruction to the improve- 

 ment of the Wisconsin river. What shall be done with it? 

 This problem, John Nader, Assistant U. S. Engineer, under di- 

 rection of the U. S. officer above named, has been endeavoring to 

 solve, since July, 1871. He finds the sandbanks of the upper 

 river the cause of the sandbars below Portage. That wher- 

 ever the river is contracted between narrow banks, the sand- 

 bars will form only behind projections or obstructions, and in 

 this case tend to improve rather than obstruct the channel ; 

 also occasionally, where the river is moderately wide, sand- 

 bars are found to have lodged on one side or the other (prob- 

 ably caused by some obstruction) and preserve a good chan- 

 nel ; but where the stream is straight for some distance, and 

 of considerable width, there will be formed a middle ground, 

 "with but little water over the same, and sometimes a dry bar ; 

 in either case the channel is on one side or the other, never on 

 both. Where the middle ground is flattened out, and extends 

 across the whole width of the river, one side of the bar ad- 

 vances more rapidly than the other, and the crest of the bar 

 is formed obliquely across the river ; the current generally 

 flows at right angles with the line of the crest, and the width 

 of the river is virtually nearly doubled in some cases; in 

 nearly everv case, deep water is found along the crest of the 

 bar. The motion of the sandbars is quite regular, and de- 

 pends not so much on the stage of the water as upon the ra- 

 pidity of change from high to low, and upon the velocity of 



