122 EFFECTS OF WINDS AND OF 



(corrected elevations) there are maxima on October 21 at 4 a.m., 7 a.m., 

 9 a.m., noon, 6 p.m., 8 p.m., and midnight, and on October 22 at 3 a.m., 

 7 a.m., 11 a.m., 2 p.m., 7 p.m., and 11 p.m. These apparently define 12 

 complete waves of the 3.7 hour seiche referred to on page 115. The different 

 values of the period, from the 12 separate intervals between the maxima 

 noted, varied from 2 to 6 hours, with a mean for the 12 waves of 3.6 hours. 

 Similarly, if one judges by the minima on these 2 days, there were 12 com- 

 plete waves between 3 a.m. of October 21 and 10 p.m. of October 22, 

 no single wave being shorter than 2 hours and none longer than 5 hours. 

 The mean period from these minima is one-twelfth of 43 hours, or 3.6 

 hours. 



Examine plate 16, which shows the four graphs for Buffalo on October 

 26-27, 1910, and also in the upper left quarter of the plate the four graphs 

 for Cleveland on October 27. 



On October 26, 1910, at Buffalo a long-period seiche, apparently the 13.1 

 hour seiche, was in progress, with a range of 0.5 to 0.8 foot, and the baromet- 

 ric effects and wind effects were not large nor changing veiy rapidly. 



Between 11 p.m. on October 26 and 11 a.m. on October 27 at Buffalo the 

 barometric effect increased very rapidly from +0.01 foot to +0.50 foot. 

 Between 1 a.m. and 10 a.m. on October 27 the wind effect increased very 

 rapidly from 0.07 to +1.02 feet. The two nearly simultaneous impulses, 

 one from the barometric pressures and one from the winds, together tended to 

 raise the water surface at Buffalo by 1.58 feet. Actually, according to the 

 observed elevations, the water rose at Buffalo 4.65 feet from elevation 

 571.05 at midnight at the beginning of October 27 to elevation 575.70 at 10 

 a.m. on October 27 at the time when the maximum wind effect occurred 

 with a wind of 53 miles per hour from the southwest. 



The wind effect at Buffalo decreased very rapidly from +1.02 feet at 10 

 a.m. on October 27 to +0.14 foot at 4 p.m. on October 27, by which time 

 the wind had died down to 27 miles per hour from the west. Note that the 

 water surface (observed elevations) fell at Buffalo 4.60 feet from elevation 

 575.70 at 10 a.m. on October 27 to elevation 571.10 at 7 p.m. From the 

 facts to which attention has been called in this paragraph and the preceding 

 paragraph it is clear that the great rise culminating in the elevation 575.70 

 at 10 a.m. on October 27 was largely an inertia effect. The barometric 

 effect and wind effect combined, without inertia, would have accounted for 

 only 1.58 feet out of the total rise of 4.65 feet. 



That the great rise was largely due to inertia effects and was the first 

 half wave of a new very large seiche is shown by the continuous graph 

 (corrected elevations). It is such a first half wave of a new seiche, pro- 

 duced by a new large impulse, that is believed to produce the occasional 

 excessively large residuals in the daily corrected elevations which are caught 

 and rejected by the criterion for such rejections which is set forth on page 

 111. Note in table No. 19, page 83, that the residual from the five-day 

 mean for corrected elevations for October 27, 1910, at Buffalo was 0.39 foot, 



