282 THE AMERICAN NATURALIST. [Vol. XXXVIIL 



responded well to the tuhing fork, I operated on them in the 

 following way. After etherizing the animals, I cut off the top 

 of the skull, exposing the brain, and the vertical semicircular 

 canals. I then seized the canals with forceps and drew them 

 out bodily with the attached sacs and their otoliths, as Kreidl 

 had done. I operated thus on four fishes, three of which recov- 

 ered. After recovery, I tested them again with the tuning fork, 

 and found that one responded to the sound about as well as 

 before the operation, and that the two others responded some- 

 what less regularly than before, though in a still perfectly 

 definite and unmistakable manner. Thus, since these fishes 

 responded like normal individuals, my results confirmed in all 

 essential respects those of Kreidl, and I came to the conclusion 

 that there must be some fundamental difference between Kreidl' s 

 methods for the elimination of the ear, and mine. The method 

 I generally used, cutting the eighth nerves, seemed to me a per- 

 fectly secure means of excluding the action of the ear. On 

 the other hand, the withdrawal of the semicircular canals with 

 the attached parts of the ear, as practiced by Kreidl, might well 

 leave behind and intact parts of that organ, and thus be inefficient 

 as a method for completely excluding the ear. To settle this 

 matter, I made careful dissections of the ears of goldfishes. 

 The ear of the goldfish is in all essential respects similar to that 

 of Cyprinus, as described and figured by Retzius ('81, p. 78)- 

 The semicircular canals are of large size; the two vertical 

 canals lie free in the brain cavity, while the horizontal canal 

 is partially imbedded in the skull. The sac into which these 

 canals open, the utriculus, is of medium size, and contains a large 

 lenticular otolith. The utriculus, with its otolith hes free in the 

 brain cavity and is the structure which is removed in connection 

 with the semicircular canals in fishes which are operated on by 

 Kreidl" s method. But ventral to these parts, and largely imbed- 

 ded in bone is another portion of the internal ear, which probably 

 represents the combined sacculus and lagena. This is not 

 removed, nor even seriously disturbed by the Kreidl operation. 

 This deeper sac extends posteriorly and ventrally until that 

 of the right ear nearly meets that of the left in the base of 

 the cranium. Each sac contains two otoliths, one long and 



