496 IRVING HARDESTY 



and cemented with thin asphalt to very minute cork blocks 

 stuck at intervals along the upper surface of the leather tectorial 

 membrane. The apparatus was placed in dii-ect sunlight with 

 the idea that evidences of vibration could be read in the move- 

 ments of the spots of light reflected by the mirrors upon the 

 ceiling of the laboratory, or on specially devised screens tried 

 later, the long light-levers magnifying the movements. This 

 plan, however, was found disappointing because of the very 

 indefiniteness of its results. The light reflected from the sur- 

 faces of the mirrors had to pass through the w^ater in the apparatus 

 and the glass plate of its top side and was so refracted that the 

 boundaries of the spots upon the screen or ceiling were not sharply 

 defined. Continuous vibration could not be distinguished from 

 intermittent, vibrations of higher frequencies were doubtfully 

 discerned and all results seemed exasperatingly confused. The 

 mirrors gave evidence, however, of the first result obtained, 

 namely, that vibrations may be induced by sound waves in a 

 structure imitating the proportions and position of the actual 

 tectorial membrane. 



Distilled water was used to represent the lymph in the cochlea. 

 The actual lymph of course differs in viscosity from distilled 

 water and may be better adapted for the conduction of vibratory 

 motion. But in using the electric current for indicating vibra- 

 tory activity in the model, the fluid used had to be as nearly as 

 possible a non-conductor of electricity. An oil with a low boil- 

 ing point would of course have been most efficient as a non- 

 conductor, but so far oil has not been used because of the difficulty 

 with which its leakage could be prevented with a construction 

 of the apparatus as- here employed and because it was con- 

 sidered that a difference in viscosity between an oil and the lymph 

 would be greater than that of water. It is realized that the 

 ordinarily used distilled water is not absolutely a non-conductor 

 of electricity. Its conductivity is less than tap water and much 

 less than the animal lymph. That the distilled water in use in 

 the model might be as pure as possible, the trough was soaked 

 in distilled water, frequently changed, for about two weeks be- 

 fore the experiments using the electric current began. Further, 



