The negative results obtained In the sea tests indicate that the sense 

 of smell plays little or no part in the feeding of skipjack, and perhaps other 

 species of tuna as well, in their natural environment. They further indicate that 

 the pond fish had learned to associate the smell of the juices of cut-up food 

 material with the act of feeding and thus exhibited a conditioned response to the 

 extracts, as already discussed in a previous section. 



SEA TESTS WITH INEDIBLE LURES 



Attention was next directed at a study of the response of tuna at sea to 

 objects which might be visually attractive and which could perhaps be used as chum 

 either alone, with extracts, or with live bait. 



Materials 



The following materials were prepared and tested {others were prepared, 

 but there was either no opportunity for testing them or it was obvious that no fur- 

 ther Information would be obtained by testing them) : 



(1) alumini«n foil: 3/8-lnch squares; used dry (floating) or wet with water 

 and Tergitol (slowly sinking). 



(2) tin strips: approximately 1/4 x 2 inches; cut from tin cans (one test) 

 or tin plate (several tests). 



(3) tin trangles: approximately 2 Inches across, with concave sides (cut 

 from tin plate from which can tops had been stamped out). 



(4) mica flakes: "artificial snow". 



(5) calcium carbide pellets: approximately 1/4- to l/2-inch pieces of irre- 

 gular shape. 



Three "drag lure arrays" were prepared and used: (1) an 8-foot iron bar 

 to which were attached small strips of tin with wire leaders; (2) a 10-foot iron 

 bar, V-shaped, to which were attached tin strips, highly polished sport-fishing 

 spoons with hooks, or rubber fish with or without hooks; (3) two 8-foot l/2-lnch 

 pipes connected at one end to a "T" and bent into a V-shape, to which were attached 

 leaders and lures of various kinds. Extracts could be supplied to this last array 

 through a l/2-lnch garden hose (75 feet long) connected to the "T". The extract 

 was dispersed through holes drilled in the arms of the V-shaped pipe. 



Methods of Testing 



In all teats with inedible objects, the fish were first chummed to the 

 stern with live bait, chumming was stopped, and the materials were thrown to the 

 feeding fish and the response noted. In later experiments a quantitative measure 

 of response was employed: (a) the fish were chummed to the stern, a few were 

 caught by the fishermen, and chumming was stopped although the fishermen continued 

 fishing — the time (in seconds) from the cessation of chumming to the disappearance • 

 of the fish was recorded, as well as the number of fish caught during this period; 

 (b) the school was then re-chumraed to the stern with live bait, a few were caught, 

 chumming with live bait was stopped and chumming with the objects was started — 

 again, the time to the disappearance of the fish and the number of fish caught 

 during this period were recorded. 



In several experiments extracts were introduced at the same time as the 

 Inedible lures were presented to the fish. 



The drag lure arrays were towed behind the ship at various distances (up 

 to about 75 feet) both in chasing a school without using chum and also after having 

 chummed a school to the stern. In some experiments extracts were released either 

 from the stern or side of the ship, or in the case of the third drag lure, through 

 the garden hose by siphoning action. 



15 



