LIGHT DISCRIMINATION IN THE ENGLISH SPARROW 83 



II. METHOD 



(1) Care of the birds. — The writer found it very difficult to 

 keep sparrows alive in captivity. The birds were kept in a large 

 cage before a large, open window in the experiment room. They 

 were provided with an abundance of Spratt's mixed bird seed, 

 fish bone, and clean water. The adjustable floors of the cage 

 were cleaned frequently and covered with coarse sand. But 

 for some unknown reason many sparrows died, usually in spasms. 

 The experimenter found that the best plan was to cage the 

 sparrows some weeks before beginning an experiment. If they 

 died it was usually very soon after being placed in confinement. 

 If they survived the first few weeks it was reasonably safe to 

 begin an experiment with them. The four birds used in the 

 final experiment kept in good physical condition until near the 

 end of the investigation. One died in spasms shortly before 

 the end of the experiment, and two others shortly after. The 

 cause of death was not apparent. The fourth bird is still alive, 

 after a confinement of one and a half years. 



(2) Method of handling the birds. — In the preliminary experi- 

 ments the writer tried handling the birds. This seemed to make 

 them wilder, instead of taming them. To avoid the necessity 

 of handling the birds the writer constructed a box 24 x 19. 5 x 15 

 cm., and covered it with wire mesh. It was provided with a 

 handle, and with a door at the end, and was used to transfer 

 the birds from the cage to the apparatus. Against the back 

 wall of each half of the cage was an inside adjustable wall of 

 wire which, when pulled forward, forced the bird out at the 

 door at the front, and into the portable box. Then the door 

 of the box was closed and the bird carried about at will by the 

 experimenter. This seemed to be a very good method of hand- 

 ling the birds as it did not frighten them after the first few times. 



(3) Description of the apparatus. — The apparatus used in the 

 investigation consisted of the Yerkes-Watson brightness appa- 

 ratus, 10 and an experiment box modelled after the one described 

 by these authors. 11 



10 Robert M. Yerkes and John B. Watson. Methods of Studying Vision in Ani- 

 mals, pp. 17-24. 



11 Ibid, pp. 24-25. 



