76 



EFFECTS OF WINDS AND OF 



On the other hand, the maximum value at any of the three Lake Michigan- 

 Huron stations is 0.95, but little more than one-half of the minimum on 

 Lake Erie. Of the 24 values at the Lake Michigan-Huron stations, 6 are 

 less than one-tenth as large as the minimum (1.72) at Lake Erie stations. 

 Roughly speaking, then, the Lake Michigan-Huron effects, which must be 

 proportional to ~L X if the theory used in this investigation is correct, must be 

 less than one-tenth as large as the wind effects on Lake Erie, at the stations 

 under consideration. Such minute wind effects as those postulated on 

 Lake Michigan-Huron must be very difficult to detect by the observation of 

 water-surface elevations, masked as they necessarily are by the much larger 

 barometric effects and by seiches. 



Table No. 18 shows how small the wind effects are at the fchree Lake 

 Michigan-Huron stations, both in absolute units and in comparison with 

 the wind effects at the two stations on Lake Erie. The values are as 

 computed from formula (69), page 63. 



TABLE No. 18 Maximum value of the wind effect at any hour within the limits of this 



investigation at various gage stations. 



When it is realized that the maximum wind effect at any one of the three 

 stations on Lake Michigan-Huron during the months covered by the in- 

 vestigation was only 0.034 foot even for a single hour, as shown in the table, 

 and that during more than one-half of the time the wind effect is certainly 

 less than 0.010 foot, then it is clear that the large probable errors in table 

 No. 17 are due to this cause. So, also, are the abnormal minus signs on two 

 of the computed values of C x - The wind effects were too small to show 

 through the mask of barometric effects and seiches. 



To what extent is the accuracy of the computed wind effects reduced by 

 uncertainty as to the true value of the exponent of h in formula (51), page 

 39? It has already been stated, pages 54-55, that no theory is known 

 which is deemed adequate to fix the value of this exponent, that it has in this 

 investigation been derived from the observations, that the observations 

 indicate 2.4 to be the most probable value, and that the belief has been 

 reached that whatever error exists in this adopted value has a very slight 

 influence on the accuracy of the computed wind effects. It is appropriate 

 to indicate somewhat more definitely at this point the basis of the belief 

 stated in the last part of the preceding sentence. 



