I/O WALTER LOUIS HAHN. 



The second condition of these experiments, the removal of the 

 ears is also without marked effect on the perception of objects. 



Five M. hicifiigus struck the wires 31.7 per cent, of the 

 chances as against 24.4 for the same individuals when normal, 

 and 25 per cent, for all individuals used. Six M. stibulatus 

 struck 24.6 per cent, of their chances with the ears and tragi 

 removed as against 32.6 per cent, for the same animals when 

 normal. Four P. subflavus struck 20.8 per cent, when operated 

 upon and 26 per cent, when normal. 



The high percentage of strikes for M. hicifngiis after the op- 

 eration is due to a single individual which was injured in the op- 

 eration. When it is omitted, the percentage for the other four is 

 24.3, or about that for the normal individuals. The average per 

 cent, for all three species is 23.2 or 1.8 per cent, less than the 

 total average. This difference is so small that it may be acci- 

 dental and without significance. However, this set of experi- 

 ments shows that the external ears and tragi are not necessary 

 for the perception of objects. These results are in accordance 

 with the conclusions reached by Rollinat and Trouessart ('00) 

 and Merzbacher ('03). 



To stop the external auditory meatus dry plaster of Paris was 

 pressed in lightly with a pair of forceps, and then wet with a 

 drop or two of water. The superfluous water and plaster was 

 wiped out of the ear conch and the bat was held firmly for a few 

 minutes until the mixture hardened. Even then the animals 

 would break the hardened plaster from the ear if they were al- 

 lowed to rest very long. They were examined at intervals and 

 no trials were recorded in which there was a possibility of the 

 plaster having been removed. 



The results were very different from those obtained from the 

 previous experiments. Six M. Incifugiis struck the wires Gj per 

 cent, of the chances. The same six in the normal condition 

 struck but 26.3 per cent. Five M. subulatus struck 65.2 per 

 cent, of chances with ears stopped and the same five normal 

 struck 26 per cent. Five Pipistrelhis subflaznis struck 65.6 times 

 and only 23.2 when normal. The concordance of the figures for 

 these experiments, divided into three groups, is significant and 

 shows that the results are not due to accident but have some 

 common basis. 



