562 'Journal of Comparative Neurology and Psychology. 



water was admitted to the vessel, or when the vessel was emptied. 

 This third vent was not needed so long as the circulation of water 

 was constant. The outlet was placed at the bottom of the encircling 

 jacket and brought to a level with the upper part of the box. 

 Arranged thus, it could be used to siphon all the water out, which not 

 only prevented the possibility of rust, but also at the same time 

 afforded the opportunity for beginning work immediately on sub- 

 sequent days, by eliminating the necessity of heating a large body 

 of cooled water in the jackets. 



In the bottom of each of these 5x5 inch inner passages small 

 cubical pans were sunk flush with the floor to a depth of one inch. 

 These pans served as food receptacles. They were placed near the 

 end farthest from the opening used as an entrance for the animal 

 (thus avoiding any possibility of the use of vision). The back of 

 each box was covered by a ground glass plate, behind which a 16 

 candle power light (EL) was placed. The boxes were painted 

 throughout a dull black, and every effort exerted to make them abso- 

 lutely alike. The experimenter found this successful so far as his 

 own discrimination was concerned. 



The temperature of these boxes was regulated by forcing hot and 

 cold water of the desired temperature to circulate through them. 

 A triple faucet was used, two vents giving cold and hot water, respect- 

 ively, and the third so attached that the water from the first two 

 could be mixed and any desired temperature of water obtained. 

 The overflow ran into a sink near by. 



The regulation of the temperature was not wholly exact. In the 

 first place, the temperature of the room varied slightly from day 

 to day. The extreme limit of this variation during this series of 

 experiments was 3 degrees. The temperature near the experimental 

 boxes necessarily varied according as the needs of the experiment 

 demanded high or low temperatures. Further, this surrounding air 

 varied considerably during the experiment. It was finally decided 

 that the temperature should be said to be that obtained by a rapidly 

 registering Centigrade chemical thermometer when its glass surface 

 was held against the inner surface of the uniformly heated box, 

 five inches from the mouth or entrance. The mercury of the bulb 



