1 66 WALTER LOUIS HAHN. 



smell, were able to avoid objects in their way and even silken 

 threads stretched so that there was just room for the animals to 

 pass between, and they contracted the wings when the space was 

 too narrow for the expanded wings to pass through. Observa- 

 tions on a large number of captive bats have convinced me that 

 Spallanzani's experiments were, in some way, lacking in scien- 

 tific accuracy. Not a single individual out of more than sixty 

 belonging to five species that I have experimented with, have 

 shown any approach to this degree of skill in avoiding objects, 

 even with the senses all intact. 



The experiments here described were made at the " University 

 Farm " in the spring and summer of 1907, and were checked by 

 additional ones in the laboratory of Indiana University in De- 

 cember of the same year. They were similar in part to those 

 made by Rollinat and Trouessart ('00), with, however, a simpler 

 arrangement of obstacles and with each experiment worked out 

 quantitatively. 



The experiments show that bats are able to avoid objects when 

 flying, but that avoidance is not complete. Several senses may 

 be of use in perceiving obstacles, and air currents perhaps guide 

 the animals to some extent. However, destruction of the sense 

 of sight does not seriously impair their ability to perceive objects 

 nor does the loss of the external ears and tragi. The most important 

 senses are located in the internal ear and any disturbance of these 

 organs seriously impairs the animal's ability to perceive and avoid 

 obstacles. 



The following method was used: The bats were liberated in an 

 unceiled room approximately fifteen feet wide, eighteen feet long, 

 nine feet from floor to eaves and twelve feet from floor to the apex 

 of the roof. Pieces of black, annealed iron wire about one milli- 

 meter in diameter were suspended from the rafters and kept mod- 

 erately tight by fastening the lower ends to a cross wire five feet 

 from the floor. On an average, there was one wire to each eleven 

 inches of space, but they were spaced unequally, the purpose be- 

 ing to determine whether the bats would try to pass through the 

 more narrow spaces or learn to select the wider ones. During a 

 part of the experiments there was an additional row of seven short 

 wires, alternating with the others and placed twenty inches from 



