from the releases on Cruises I, II, and III 

 consistently show a strong current down the 

 eastern shore of the Michigan Thumb (figs. 

 18 A-C); rates were as high as 12.7 miles 

 per day for bottles released on Cruise III. 

 During and for a short period after these 

 cruises in June and July 1956, the winds 

 over the lake were largely from the north 

 and the west. A temporary reversal of the 

 strong surface current clockwise around the 

 tip of the Michigan Thumb is indicated by 

 the travel of bottles released during Cruise 

 IV (fig. 18 D). Apparently this reversal 

 was caused by strong east winds that blew 

 during the first week of August. Of the 

 bottles released on Cruise IV (fig. 18 D) 

 sind V (fig. 18 E), the general direction of 

 those moving 3 miles per day and faster was 

 to the east. This easterly drift coincided 

 with prevailing westerly winds during 

 August and September 1956. Of the bottles 

 released on Cruises VI, VII, and VllI (figs. 

 18 F-H) only two crossed to the Canadian 

 shore at a rate in excess of 3 miles per 

 day. The remainder of the "rapid drifters" 

 traveled to the north (figs. 18 G-H) . Dur- 

 ing the time that these bottles were adrift, 

 the prevailing winds were from the south. 



It might seem that the bottles that 

 had lost their drags would predominate in 

 those classed as "rapid drifters". Without 

 drags bottles presumably come more under 

 the direct influence of wind. However, of 

 the 168 bottles of these studies that 

 drifted at a rate of 3 miles per day or bet- 

 ter, only 9 had lost their drags. 



It should not be assumed from these 

 remarks and figures that no speeds of drift 

 greater than those given were attained. 

 Also, most certainly other bottles would be 

 included as "rapid drifters" if the exact 

 tracks of the bottles as well as their 

 exact landing times were known. 



QUANTITATIVE RELATIONSHIP BETWEEN 

 SURFACE DRIFT AND WIND 



Garstang (1898) believed the relation- 

 ship between winds and surface currents 

 were so precise that he worked out a quan- 

 titative relationship between the two. He 



developed the formula R ~ ^ — where 



2n 



D is the distance traveled in miles and Pn 



is the resultant pressure in pounds per 

 foot determined from n observations daily. 

 The accuracy of this method depends on the 

 assumption that the velocity of drift 

 varies as the pressure of the wind and not 

 directly as its velocity. Pressures were 

 obtained from the following set of values: 



Force, Beauiort scale 1 2 3 45678 9 10 11 12 



Velocity, miles per hour 3 8 13 18 23 28 34 40 48 56 65 75 90 



Pressure, pounds foot 0.05 0.3 0.8 1.5 2.5 4 6 8 11.5 15 21 28 40 



Pressure-equivalents were computed from the 

 velocities by multiplying the squares of 

 the velocities by the factor 0.005 and ex- 

 pressing the results in whole numbers. 

 Garstang did not give proof that these com- 

 putations were valid but remarked that the 

 table had been authorized by the Metero- 

 logical Office in 1875. He admitted that 

 the pressure-ratio is only an approximation 

 to the true law of drift. However, his 

 calculated and empirical travels of drift 

 bottles were fairly close. But he wrote, 

 "Some further examination, however, is 

 necessary before the reliability of my 

 method can be depended upon, because the 

 estimated results depend upon the assump- 

 tion of open water, and this cannot always 

 be conceded." 



R. Witting (Carruthers 1927) decided 



1/2 

 that the formula, V = MW ' , could be 



used to show wind-surface drift relation- 

 ship where V is the velocity of drift in 

 centimeters per second, W is wind speed in 

 centimeters per second, and M is a constant. 

 In observations at Finnish lightships, 

 Witting computed the value of M to be 0.44; 



1/2 

 the equation then became V = 0.44W 



Carruthers in using the same formula with 

 V and W expressed in miles per day, arrived 

 at an M value of 0.45 for the English 

 Channel. Daniel and Lewis (1930) in work 

 on the Irish Sea, expressing V and W in 

 miles per day, arrived at values of M from 

 .04 to 1.29 for different sectors. In 

 later work on the English Channel, Car- 

 ruthers (1930) worked out the formula 

 S = 1/18 W for the wind-surface drift rela- 

 Tionship where ^ is bottle travel in miles 

 per day and W the wind speed in the same 

 units. Days of similar wind conditions, 



24 



