BULLETIN OF THE UNITED STATES FISH COMMISSION. 377 



collecting headquarters to Orland village, a distance of about 5 miles, 

 the route is in brackish water, and the tow is favored by the flood 

 tide. At Orland is a dam which is surmounted by means of a lock, and 

 thence, two miles further to Dead Brook, the route is through the tide- 

 less fresh water of Narramissic River. The sudden change from salt to 

 fresh water does not appear to trouble the fish except when the weather 

 is very hot and the fresh water is much the warmest. The cars are 

 towed directly into the inclosure, where the fish are at once liberated. 



The inclosure is formed by placing two substantial barriers of wood- 

 work across the stream 2,200 feet apart. The lower barrier is provided 

 with gates which swing open to admit boats. Within the inclosure the 

 water is from 3 to 8 feet deep, the current very gentle, the bottom part- 

 ly muddy, partly gravelly, supporting a dense growth of aquatic vege- 

 tation. The brook has two clean lakes at its source, and its water is 

 purer than that of ordinary brooks. 



The collection of salmon usually continues from the first ten days of 

 June until the beginning of July. During the early weeks of their im- 

 prisonment the salmon are extremely active, swimming about and 

 leaping often into the air. After that they become very quiet, lying in 

 the deepest holes and rarely showing themselves. Early in October 

 they begin to renew their activity, evidently excited by the reproduc- 

 tive functions. Preparations are now made for catching them by con- 

 structing traps at the upper barrier. If the brook is in ordinary 

 volume, these means suffice to take nearly all, but a few linger in the 

 deeper pools and must be swept out with seines. About October 25 

 the taking of spawn begins. After that date the fish are almost always 

 ripe when they first come to hand, and in three weeks the work of 

 spawning is substantially finished. 



Although the salmon are taken from the fisherman without any 

 attempt to distinguish between males and females, it is always found 

 at the spawning season that the females are in excess, the average of 

 four seasons being about 34 males to G6 females. This is a favorable 

 circumstance, since the milt of a single male is fully equal to the im- 

 pregnation of the ova of many females. 



The experiment has several times been tried of marking the salmon 

 after spawning and watching for their return in after years. After 

 some experiments, the mode finally fixed upon as best was to attach a 

 light platinum tag to the rear margin of the dorsal fin by means of a 

 fine platinum wire. The tags were rolled very thin, cut about half an 

 inch long and stamped with a steel die. The fish marked were dis- 

 missed in the month of November. Every time it was tried a consid- 

 erable number of them was caught the ensuing spring, but with no essen- 

 tial change in their condition, indicating that they had not meanwhile 

 visited their spawning grounds. In no case was a specimen caught in 

 improved condition during the first season succeeding the marking. 

 But the following year, in May and June, a few of them were taken in 



