90 AMERICAN FISHKRIES SOC lETV. 



SO as to be overlooked by fishermen, caused that method to be 

 abandoned. A metallic taa;, stamped with a recorded number, 

 appeared to offer the greatest promise of success. The first tag 

 tried was of thin aluminum plate, cut about a half inch long and 

 a quarter wide, and attached to a rubber band which encircled 

 the tail of the fish. It is possible that most of the bands 

 slipped off, and tliat those whicli were tiglit enough to stay on 

 cut through the skin, and produced wounds that destroyed the 

 fish. At anv rate, no salmon thus marked were ever recovered. 



The next method employed was the attachment of an alumi- 

 num tag by means of a platinum wire to tlie rear margin (jf the 

 first dorsal fin. This place of attachment was clujsen because. 

 being near the middle of the fish, it has less lateral motion when 

 the fish is swimming tiian anv point nearer the head or tail, and 

 because tiie tag, lying thus in the wake of the fin and close to 

 the back, would be better protected from contact with foreign 

 objects than elsewhere. The attachment was effected by plac- 

 ing the fish upon a narrow table, confining it by straps, and 

 piercing the thin membrane of the fin between the last and next 

 to the last ray, by means of a needle, into the eye of which was 

 threaded the wire already connected with the tag ; tlie ends of 

 the wire were then twisted together, so as to form a loop, and 

 neatlv trimmed with scissors. The tags were stamped with dies. 

 This mode of marking has been adhered to in all subsequent ex- 

 periments of the kind, with no change except that the aluminum 

 tag has been replaced by one of platinum. 



The marking was always done in the fall, after the fish had 

 been relieved of their spawn. They were then liberated, either 

 in tide water or in fresh water whence egress to the sea was 

 easy. 



Of the salmon marked with rubber bands in 1^72, as has been 

 said, none were recovered. In November, 1873, tiiere were 

 marked 391 salmon. In the ensuing year rewards were offered 

 to the fishermen for the return of any marked specimens. In re- 

 sponse, there were sent in to the station twenty salmon, the first 

 in January (taken in a smelt net), and all the others in April and 

 May. All of the twenty retained the wire, by which they we;e 

 with certainty recognized as having been marked in the preced- 



