UNUTILIZED FISHES. 21 



light-house. During the first fifteen or twenty minutes several hake 

 and two large pollock Avere caught. Then the dogfish began to appear 

 in great numbers. While one man was trying to land a hake it was 

 attacked by half a dozen dogfish, which followed it clear to the sur- 

 face, biting and tearing it so that all the luckless fisherman got for 

 his share of the catch was the head, and a portion of the spinal column 

 with .a few bits of flesh clinging to it. This experience is one common 

 to all the ground fishermen of that region. Soon all that we could 

 catch were dogfish. In less than tAvo hours' time 78 were landed, 

 which was about as fast as our lines could be baited, thrown overboard, 

 pulled up, and the dogfish taken off; it usually took longer to get the 

 fish off the hook than to get them on it. They grew more and more 

 plentiful as the fishing went on. Often they were seen swimming 

 about the boat close to the surface of the water, but only for a short 

 time, when the} 7 would dart downward out of sight. Captain Kelly 

 and other fishermen said that outside the harbor they were larger and 



much more abundant. 



VARIETY OF FOOD. 



A statistical study of the food of the horned dogfish, as of the 

 other species dealt with in this paper, was not possible in these investi- 

 gations, owing to the scarcity of specimens obtainable during July 

 and August in the vicinity of Woods Hole, where the work was con- 

 ducted. The variety of food taken by the dogfish, however, is well 

 shown in the following instances: Of some 90 specimens taken at 

 Provincetown August 20 and 21, 1003, 31 contained the little jellyfish 

 Cydippe, about 75 to each dogfish; 11 specimens contained squid 

 (Ommastrephes), which were probably stolen bait; 3 contained the 

 remains of 4 small crustaceans; 2 contained annelids; 1 contained a 

 hake ; 1 an insect ; 1 an isopod ; 17 were empty. Three dogfish 

 taken in a trap at Nomans Land August 2, 1904, contained 1 a 

 Nereis, 1 a squid, and 2 contained fish too far digested to be 

 identified. Of 20 or 30 specimens from Gloucester, received at the 

 Harvard Zoological Laboratory during the past two years, all whose 

 stomachs were not empty contained fish. The fish were too far 

 digested to be identified in most cases, but the greater part recogniz- 

 able were herring. 



The recent report of the Massachusetts Board of Fish Commission- 

 ers upon the damage done by dogfish in 1905 records the food of 674 

 dogfish whose stomachs were examined. Fish, apparently market- 

 able species, were found in most instances. A few of the dogfish had 



eaten ctenophores. 



DESTRUCTIVENESS. 



This evil has become greater in the last few years, and the fisher- 

 men state that it is now the most serious they have to contend with. 

 The dogfish in countless numbers overwhelm the nets, mutilating and 



