Figure 2. --Post card, 3i by b\ inches, 

 included in bottle. 



ranged from 46.2 to 95.1. A grand total of 

 745 bottles were deposited, and 517 bottles 

 or 69.4 percent were recovered. Most of the 

 release stations were approximately 2 miles 

 apart although a distance of about 1 mile was 

 employed on three dates. 



Unknown numbers of bottles landed many 

 weeks or months before their recovery; many 

 of them no doubt remained at about the same 

 location, but some could have been moved by 

 winds, waves, or ice. However, the largest 

 monthly numbers and percentages (table 2) 

 followed the month when the bottles were 

 released except August, when the major re- 

 coveries occurred in the month of release. The 

 majority of the bottles had been recovered 

 by the end of August 1931 and of July 1932. 

 For the combined years, the greatest number 

 of recoveries was in July. As the season 

 progressed in both 1931 and 1932 the monthly 

 figures decreased and reached zero as early 

 as September (May release of 1932). 



The numbers and percentages of the bottles 

 recovered were recorded also according to 

 the number of days (table 3). During the first 

 9 days 6 bottles (2.1 percent) were recovered 

 in 1931 and 46 bottles (10.0 percent) in 1932— 

 a total of 52 bottles (7.0 percent) for both 

 years. From the 10th until the 50th day the 

 percentages for the 10-day period varied 

 from 9.9 to 13.4 in 1931, 8.2 to 10.8 in 1932, 

 and 9.7 to 11.8 in the 2 years combined. 

 They then decreased until the 80th day, and 

 thereafter averaged below 2 percent (including 

 those beyond 129 days). The highest 



percentages were between 30 and 39 

 days. 



Many recovered bottles had lost their drags. 

 In 1931, 120 (64.5 percent) and in 1932, 133 

 (40.2 percent) of the bottles had no drags, a 

 grand total of 253 (48.9 percent)~see table 3. 

 The percentages of the lost drags of 1931-32 

 bottles increased generally to a maximum of 

 80.0 during the period of 90-99 days but 

 subsequently averaged about 59 percent. Bot- 

 tles that held their drags amounted to 58 

 (31.2 percent) in 1931, 169 (51.0 percent) 

 in 1932, and 227 (43.9 percent) in both years. 

 Thirty- seven (7.2 percent) of the cards gave 

 no report on the drag. 



Only 16 drift bottles, at least 10 of which 

 had lost their drags, were found more than 



1 year after release (table 3)— 13 in 1932-36, 



2 in 1938, and 1 in 1942. Only two bottles 

 were recovered beyond Lake Michigan. One 

 released June 22, 1931, 43.5 miles from 

 Frankfort, Michigan, was found April 8, 1935, 

 at St. Martins Point, St. Martins Bay, Lake 

 Huron. Another, released June 24, 1932, 6.7 

 miles from Kewaunee, Wisconsin, was re- 

 covered June 6, 1942, near Jefferson Beach, 

 Lake St. Clair. 



RELEASES AND RECOVERIES 

 BY TRANSECT AND COURSE 



The following sections include 19 graphs 

 of recoveries showing the directions followed 

 by recovered bottles. Tables 4-24 list the 

 stations of release and recovery of drift 

 bottles, number of days, minimum miles 

 traveled, and records of drags. 



Chicago to St. Joseph 



On June 17, 1931, 27 bottles (table 4) were 

 released along the transect (fig. 4) and 23 

 (85 percent) were recovered in 17-76 days, 

 an average of 43.7. Eight (40 percent) of the 

 20 bottles held their drags. The bottles had 

 covered minimum distances from 9-144 miles 

 an average of 61.5. Sixteen bottles, including 

 six on the western part of the transect, scat- 

 tered widely on the east coast from 8 miles 

 south of St. Joseph to 7 miles south of Manistee. 



