During 1954 and until August 19 in 1955 the 

 drag was suspended 3 feet below the bottles by a 

 soft iron wire. Thereafter, the drag was susp>ended 

 1 foot below the bottles to lessen resistance to wash- 

 ing ashore. Observations of dye patches revealed 

 that the movement of the bottles was not changed by 

 the shortening of the drag (Smith 1956). Many of 

 the bottles released prior to June 5, 1955, lost their 

 drags. Apparently these losses were caused by re - 

 peated bending of the iron suspension wire to the 

 point of breaking as the bottle was moved by waves. 

 Introduction of a brass ring in the connection at the 

 neck ofthe bottle greatly reduced but did not end the 

 loss of drags (Smith 1956). Losses were reduced fur- 

 ther, however, in Lake Superior and Lake Erie in 

 1958 by the insertion of a brass ring connecting the 

 iron suspension wire to the drag. 



1, 080 drift envelopes in central and southern Lake 

 Michigan between July 9 and September 15, 1954 

 (table 1, fig. 1). In addition, 100 plastic tubes were 

 released July 9 on the Grand Haven -Milwaukee tran- 

 sect; on this date, 6 each of the three types of units 

 were released at all stations except the first out of 

 Milwaukee at which 4 ofeach type were released. Sub- 

 sequent to July 9, 10 drift bottles and 10 drift envelopes 

 were released at each station. Units recovered after 

 January 31, 1955, have not been included in the records 

 of recovery or in analyses. 



The 18 reply cards received from recovered plas- 

 tic tubes were mostly from units released close to shore. 

 This fact suggests that many of the plastic tubes re- 

 leased at stations distant from shore must have become 

 waterlogged and sank. 



Each ballasted bottle contained 40 grams of 

 fine loose sand which caused the bottle to sink in - 

 to the water so that only about 1/2 inch was exposed. 



Plastic Tubes 



The plastic -tube unit consisted of a cylindrical 

 tube of cellulose acetate, 6 1/2 inches long and 

 1 1/4 inches in diameter, into which was inserted a 

 reply card wound around a wooden block (fig. 2B). 

 The unit was sealed with a square piece of cellulose 

 acetate to which was attached a metal drag with a 3- 

 foot suspension wfre. The unit floated with approxi- 

 mately 2 inches of the tube above the surface. 



Drift Envelopes 



The drift -envelope unit consisted of a 3- by 

 5 -inch reply card enclosed within a 3 3/4- by 6 1/4- 

 inch polyethylene envelope of 0. 002-inch thickness 

 (fig. 2C). Water did not adhere to the plastic enve- 

 lopes at first; most of the envelopes floated upon the 

 surface film where they were under the direct pro- 

 pelling force of the wind. Some time after release 

 the plastic envelopes no longer shed water and floated 

 immediately under the surface fUm where they tum- 

 bled end for end in a rough sea. 



RELEASES AND RECOVERIES 



Central and Southern Lake Michigan, 1954 



The U. S. Bureau of Commercial Fisheries M/V 

 Cisco jreleased 1, 080 drift bottles with metal drags and 



Percentage recovery of drift envelopes was low 

 (table 1). The normal seasonal decline in rate of re- 

 covery does not account for the precipitous decrease 

 in reports of envelopes released after August 3. It is 

 believed that the plastic envelopes containing the re- 

 ply cards were improperly sealed and that many units 

 became waterlogged and sank (Beeton, Johnson, and 

 Smith 1959). 



The percentage recovery of drift bottles ranged from 

 47. 8 percent frxthe August 19 releases on the Grand 

 Haven -Milwaukee transect to 62. 9 percent for the re- 

 leases on August 3 on the South Haven -Waukegan tran- 

 sect; for the combined releases the percentage recovery 

 was 54.4 percent (table 1). The recovery rate was 

 slightly higher for units released in July and early Aug- 

 ust than for those released in late August and September. 



Central and Northern Lake Michigan, 1955 



Between April 26 and November 8, 1955, 2, 000 

 drift bottles with dra^and 2, 000 with sand ballast were 

 released (table 2, fig. 1). Ten of eachtype were re- 

 leased at each station. Units recovered after January 31, 

 1956, have not been included in the tables and are not 

 mentioned in the discussion. 



The percentage recovery was slightly hi^erfor bot- 

 tles with drags (56.4 percent) than for sand -ballasted 

 bottles (53. percent). The difference probably is not 

 significant since more sand -ballasted bottles than bot- 

 tles with drags were recovered from 6 of the 16 cross- 

 ings (table 2). 



