228 DOMESTICATED TROUT. 



is no safe test but the scales. The length is no guide, 

 for his depth and breadth will often in a short trout 

 more than compensate in weight for what is lacking in 

 length, and then again a lean trout in poor condition 

 sometimes actually does not weigh more than half 

 what he would when fat and in his best condition. 

 This is a great difference, it is true, but it is a fact. It 

 is said by medical authorities that a man cannot lose 

 over three eighths of his weight and live. It is not so 

 with a trout; he can lose full fifty per cent and live. 



Section II. — The Commissary Department. 



The question of food for trout is a very important 

 one, and I think, as a general thing, a very simple one 



tioning three, caught in September, 1867, by the subscriber at 

 the outlet of Rangeley Lake, Franklin County, Maine, — this 

 lake being the head-waters of the Androscoggin River : — 



One 10 lbs. male, 

 One 9J lbs. do., 

 One 8.J lbs. female. 



The first and last were transported alive in a box of water, 

 aerated by an air-pump, to my pond in Stanley, Morris County, 

 N. J., but afterwards died in consequence of too high a tempera- 

 ture in the water. The first weighed ten (10) lbs. by steelyard 

 within a half-hour after death. It is now in a glass case in my 

 office in New York. Tie 9^ lbs. trout was sent to General Grant. 

 Two of the trout from these waters I have sent to Professor 

 Agassiz, in 1S63 and in 1S67, and in a personal interview he pro- 

 nounced them real Brook Trout (Salmo fontinalis). 

 Faithfully yours, 



GEO. SHEPARD PAGE, 



Pres't Oquossoc Ang. Ass. 



