94 BULLETIN OF THE UNITED STATES FISH COMMISSION. 



Every trout-angler who has fished in a salmon river knows how readily the yearling 

 salmon or " pan" will take a bait. The fish at this stage has on its sides the "bars" or 

 "finger marks," as they are often called, aud bears a great resemblance to a trout, but 

 presently puts on the livery of the smolt and goes to sea. This change means that 

 it loses the peculiar " finger marks " on its sides, aud also the general trout like appear- 

 ance, and becomes very silvery. It was formerly supposed that this silvery appearance 

 was caused by a fresh growth of scales, but that has been found not to be the case. 

 It is caused by a silvery pigment on the under sides of the scales, and on the opereles, 

 which are scaleless. 



Some years ago a rather interesting experiment was tried. A number of salmon 

 parr were taken from the river and placed in au aquarium tank. In due course about 

 half of them developed into the smolt stage. The others did not. Sea water was 

 then turned on, the supply of fresh water having been previously cut off, and the water 

 in the tank was gradually changed from fresh to salt. What was the consequence? 

 Some people affirm that "parr" will not live as such in salt water. These not only 

 lived, but very rapidly assumed the smolt stage after the salt water was added. 



Now, we know that parrs feed. Anglers know, as I have already said, that they 

 will take almost any bait. They have been found gorged with shellfish, crustaceans, 

 and the larvte of aquatic insects, etc. "Smolts" also feed voraciously; indeed, I have 

 known them completely spoil the sport in a river. 



Can it be supposed that the salmon in its early stages, during which it does not 

 make any abnormally excessive growth, feeds voraciously, and after going to the sea 

 takes either no food or very little, notwithstanding that the smolt, which has left the 

 river perhaps less than a quarter of a pound in weight, returns in a few months as a 

 grilse weighing (> or 7 pounds. It seems unreasonable. It is fouud amongst domes- 

 ticated fish, and this applies to the salmon, that at certain times of the year they feed 

 very freely and at other times they take very little, but at all times more or less food 

 is taken. 



At Stormontfield on the Tay it was noticed on one occasion that the smolts which 

 were to be let down to the sea were of a very much larger size than they had been in 

 previous years. It was fouud on inquiry that the ponds had become charged with 

 minute shellfish, and the liberal diet these afforded accounted for the extra growth of 

 the smolts. Some smolts return from the sea as grilse in two or three months; others 

 in fourteen or fifteen months. Those which return in three months have, taking a 

 very low estimate, attained a weight of some 3 pounds; whereas those which have 

 remained in the sea for the longer period, say from May of one year to July of the fol- 

 lowing year, do not attain a very much greater size than those which return in the 

 shorter period. Exactly the same peculiarity has been noticed in the case of tame 

 trout kept in ponds. Some grow very much more rapidly than others, and the period 

 of a pondful of trout arriving at maturity will often vary a year and sometimes even 

 two years. 



It has been found, in some instances at least, that "parrs" become "smolts" and 

 go to sea, some of them the first year; the great majority the second year; and some 

 not till the third year. This, too, is found to be very much the case amongst domesti- 

 cated trout. It is found necessary at the cud of the first year to take them out of the 

 pond and sort them. If this be not done the larger fish will eat many of the smaller 



