86 SMITHSONIAN MlSCfiLLANfiOUS COLLECTIONS VOL. I28 



ice-cutting device (pi. 2, fig. 2). As freezing continued, more holes 

 were made farther from the shore. In order to retard as much as 

 possible the freezing over of the cut surface, as well as to protect the 

 workers, a quonset-shaped icehouse (pi. 4, fig. 2) approximately 8 

 by 10 feet was built over each hole, or the hole was covered with 

 squares of insulating material held in place by boards. Despite this 

 protection, ice formed to a depth of about 3 inches between visits. 

 Also, the ice walls of the hole, which ranged from 2 to 8 feet thick, 

 would grow inward as the winter progressed and had to be kept 

 chipped away to keep the hole large enough for working. As the 

 weather moderated during the spring months it was much simpler to 

 cut holes occasionally than to keep the old ones clear. 



Several types of traps were tried to determine which were the most 

 efficient for various kinds of animals. Window screen was used for 

 small animals such as amphipods and shrimps. These were made in 

 the form of a cylinder with a funnel leading into one end and a hinged 

 door in the side large enough for a hand to be inserted. Another type 

 was made from an oil drum by removing both ends and putting a 

 funnel of -j-inch hardware cloth in one end and covering the other 

 end with the same material. A door was cut with a torch. Traps for 

 catching fish were made of small-mesh chicken-wire netting, and still 

 others were made of hardware cloth with :|;-inch mesh. 



Both rectangular and cylindrical traps were used, but the former 

 were much more difficult to pull up through the hole in the ice than 

 the latter, as they would catch on the lower edge of the ice as they 

 were drawn up from the ocean floor. The diameter of the hole at 

 the bottom was usually smaller than at the surface, for it was more 

 difficult to keep the encroaching ice chipped away from the sides at 

 a depth of 7 feet. 



Crab traps consisting of large metal rings with netting across the 

 circles were also used. These traps were baited with fish or meat 

 fastened in the center of the netting. 



A couple of lath traps, somewhat on the order of lobster traps, 

 were tried, but the laths were too far apart to retain the Arctic cod 

 which were running small at the time this type of trap was tried. 

 Since it was known that Arctic cod feed on amphipods, it was con- 

 sidered desirable to allow some traps to float up under the lower sur- 

 face of the ice in the hope that cod might find their way into the trap 

 if they were feeding on amphipods on the lower surface of the ice. 

 The method or the traps, or both, were unsuccessful. 



If available, some fish or meat scraps were placed in the traps for 

 bait. If this was wrapped in screen or wire netting it was more effec- 



