of surface water into Saginaw Bay; a strong 

 southerly littoral current flows at times 

 along the eastern coast of the Michigan 

 Thumb area (figs. 18 A-C, see page 22). 



RATE OF DRIFT 



Reliability in computing rate of drift 

 of bottles is impaired by a lack of knowl- 

 edge of how long a bottle was ashore before 

 it was discovered, and by a lack of informa- 

 tion on the course a bottle followed from 

 its release point to the recovery point. 

 Effects of the first of these two factors 

 might be minimized by using in calculations 

 only those bottles that are actually ob- 

 served washing ashore or those found still 

 drifting in the water. However, comparisons 

 of records for bottles from the same lot 

 that traveled similar courses revealed that 

 many discovered still floating or washing 

 ashore exhibited a lower rate of drift than 

 those recovered from the shore. Most prob- 

 ably some of the former had actually been 

 beached and refloated by changes in winds, 

 water level, and wave actions. For this rea- 

 son computations of rate of drift were not 

 restricted to recoveries of bottles found 

 still floating or washing ashore. 



The second factor, lack of any means 

 of determining the exact course of a bottle 

 from its release point to its recovery point, 

 is not to be eliminated. The straight-line 

 distance from release point to recovery 

 point has been used in calculating the rate 

 of transport. 



Saginaw Bay crossed the lake east to the 

 Canadian shore at a minimum average rate of 

 0.95 miles per day, the lowest averate rate 

 of bottles that crossed the lake for all 

 the cruises (table 7). The average rate of 

 movement increased for bottles that were 

 released at the mouth of Saginaw Bay on 

 Cruises II and III and later recovered in 

 Canada. During Cruise II, however, only 

 one bottle made this crosEing--not enough 

 to give a fair value for this cruise. The 

 number of bottles crossing the lake that 

 were released during Cruises IV and V was 

 more than twice the number that crossed from 

 releases of the six other cruises. During 

 August and the first weeks of September 

 when bottles released on Cruises IV and V 

 were adrift, the prevalence of westerly 

 winds was greater than at any other time 

 during the investigations. Recovery of 

 bottles from Cruises VI, VII, and VIII that 

 crossed the lake was low and no trend is 

 apparent in the average drift rate. Cover- 

 age of shore area and consequently returns 

 from these last cruises were much lower 

 than for the earlier ones. Had better 

 coverage existed, perhaps, more bottles 

 would have been found on the Canadian shores. 



Rapidly moving bottles, those drifting 

 at an arbitrarily chosen 3 miles per day 

 cuid faster (10 percent of the bottles re- 

 covered) have been used to give some indi- 

 cation of actual drift rates approached 

 along assumed straight-line courses (figs. 

 18 A-H, see pages 22 and 23). The large 

 number of rapidly drifting bottles recovered 



A bottle exhibiting a high 

 rate of drift along a course 

 should be a better indicator of 

 the actual drift rate than the 

 average of the rates of all of the 

 bottles traveling this course. 

 The average rate might include 

 bottles that were on the shore 

 many days before they were found. 

 The average drift rate can be 

 of value, however, not in indicat- 

 ing actual speed, but in making 

 certain comparisons between 

 cruises (table 7). 



During and for a short time 

 after Cruise I in June 1956, 

 winds were variable but they 

 did display some tendency to be 

 from the west. Six bottles of 

 110 dropped at the mouth of 



Table 7. — Number of bottles released at stations at the mouth of 



Saginaw Bay each cruise in 1956, total number of recoveries of 



these bottles crossing Lake Huron, and the average minimum 



rate of drift bottles made in the crossing 



21 



