July 30-31, 1954, but recorded the Lake 

 Michigan bottles that were recovered in the 

 bay, 6 reported in 1892 by Harrington (1894) 

 and 11 in 1955 by the Great Lakes Research 

 Institute. These bottles indicated that the 

 main current flowed into the West Arm and 

 a small portion deflected into the East Arm. 



Ayers et al. (1958) reported the release of 

 1,297 drift bottles in Lake Michigan during 

 four synoptic cruises (15 round trips) on 

 June 28 and 29 and August 9 and 10, 1955; 

 239 bottles or 18.4 percent were recovered. 

 Eighteen percent of the recovered bottles 

 were found in the southern half and 82 percent 

 in the northern half of the lake. Nearly all 

 bottles recovered from releases in June had 

 traveled northeastward but most of those of 

 August floated southeastward. Fifty-seven per- 

 cent of the retrieved bottles were from the 

 eastern area of the lake, while 17, 16, and 10 

 percent, respectively, were recovered from 

 the western, central, and northern areas. 

 Twenty-one bottles were discovered in Lake 

 Huron. 



Johnson (1960) described surface currents 

 in Lake Michigan during 1954 and 1955. From 

 July 9 through September 15, 1954, 1,080 

 drift bottles, 1,080 drift envelopes, and 100 

 plastic tubes were released along two transects 

 (six trips) in the southern area and one 

 transect (one trip) in the central region. 

 Returns of these drift units amounted to 587 

 bottles (54.4 percent), 77 envelopes (7.1 per- 

 cent), and 18 tubes (18.0 percent). In general, 

 the 1954 drift was from west to east. Recov- 

 eries toward the north and south were about 

 equal. 



During April 26 to November 8, 1955, 2,000 

 drift bottles with drags and 2,000 bottles with 

 sand ballast were released along two transects 

 (six trips) in the central region, two transects 

 (nine trips) in the northern area, and one 

 cruise between Manitowoc and Sturgeon Bay, 

 Wisconsin. Returns included 1,128 or 56.4 

 percent of the bottles with drags and 1,060 

 or 53.0 percent of those with ballast. The 

 general drift of the 1955 bottles was from west 

 to east. Bottles released in the eastern areas 

 moved mainly to the north, but those dropped 

 to the west showed no prevailing pattern. 



MATERIALS AND METHODS 



Our drift bottles released in 1931 and 1932 

 had a capacity of 13.9 ounces and were made 

 of heavy glass with the initials U.S.B.F. blown 

 in the side (fig. 1). A metal drag suspended 

 by 3 feet of wire from the bottom of the 

 wired bottle could rotate within a restricted 

 arc. The drag, composed of 22-gauge galvan- 

 ized iron, was bent in such a manner ths.i the 

 bottles resisted direct effects of wind and 

 thus gave a true indication of surface cur- 

 rents. Each bottle contained a numbered post 

 card (fig. 2) that requested the following 

 information: locality, date, and hour of re- 

 covery; whether or not a metal drag was 

 attached; name and address of the finder. 

 The cards were addressed to the U.S. Bureau 

 of Fisheries, Washington, D. C, and included 

 a notice that the finder would be informed 

 where and when the bottle was set adrift. The 

 bottles also contained a strip of red muslin 

 about 1-inch wide, part of which extended 

 outside and floated along the surface, to 

 attract attention. The corks and muslin that 

 stoppered the bottles were heavily waxed. 



Series of bottles (table 1; fig. 3) were re- 

 leased across Lake Michigan between the 

 following ports: Chicago, Illinois, and St. 

 Joseph, Michigan; Waukegan, Illinois, and St. 

 Joseph, Michigan; Racine, Wisconsin, and 

 Grand Haven, Michigan; Grand Haven, Michi- 

 gan, and Sheboygan, Wisconsin; Manistee, 

 Michigan, and Two Rivers, Wisconsin; Frank- 

 fort, Michigan, and Two Rivers, Wisconsin; 

 Frankfort, Michigan, and Kewaunee, Wisconsin; 

 and Manistique, Michigan, and Manistee, Mich- 

 igan. In addition, bottles were released along 

 the courses between ports of the same coast, 

 between ports and net stations, and at various 

 locations where gill nets were set. 



In 1931, 293 bottles were released from 

 June 17 to August 17 in the open lake south 

 of a line from Frankfort, Michigan, to Algoma, 

 Wisconsin (table 1). One hundred eighty-six 

 bottles or 65.7 percent were recovered. The 

 percentages varied from 25.0 to 100.0. From 

 May 9 to August 25, 1932, 462 bottles were 

 deposited south and north of the above-defined 

 line and in Green Bay. The recoveries totaled 

 331 bottles or 71.6 percent. The percentages 



