36 HARRY MILES JOHNSON 



be detected. Weakly blown pipes or bottle whistles are perhaps 

 the next best. If nearly pure tones cannot be obtained, the next 

 best' plan is to have the tone of the same fundamental pitch 

 sounded now on one pipe or bottle -whistle, now on another 

 which is larger or smaller than the first. Thus tones of the 

 same pitch but of quite different timbre can be produced. If 

 the animal, after being trained to a tone on one instrument, 

 shows disturbance when another instrument giving the same 

 tone is substituted, it is safe to say that he was not reacting 

 merely to the pitch of the tones, and that other characteristics 

 were more or less prominent. 



The same may be said of non-musical concomitant noise. A 

 struck fork gives a pecuHar "cluck," which is rarely the same 

 in any two ; a bowed string may have a peculiar scrape which 

 lends individuality to it ; a blown pipe or whistle always gives 

 a whisper, which may be a part of its individuality; while the 

 rattle of no two reeds is perhaps exactly the same. 



A tone comparatively pure cannot be localized correctly in a 

 closed room because of standing waves ; its apparent localization 

 shifts with the observer's position. Angell " asserts that it can- 

 not be localized even in the open air. If pure tones cannot be 

 made, the position of the stimulus instruments should be fre- 

 quently changed in control tests . High overtones and non-musical 

 concomitant noises may be localized quite easily by an animal, 

 and serve as the basis of discrimination. 



Of instruments which will fulfill these conditions, one's choice 

 is limited to tuning forks and blown pipes or bottle-whistles, 

 preference being by far on the side of the first. A reed instru- 

 ment, such as the harmonium or the mouth-harmonica, which 

 Kahscher used, does not admit of control of timbre, nor of satis- 

 factory control of intensity, and the accessory noise is consider- 

 able. Trumpets, which were used by Swift, are not certain in 

 ])itch, and the timbre \'aries greatly ^^•ith the pitch and intensity 

 of the tone sounded. 



If blown pipes or whistles are used they should be easily 

 tunable, as the daily change in temperature and density of the 

 air affects their pitch consideral^ly. Differential organ pipes 

 are fairly satisfactory. The Stern \'ariators also meet this 



" Angell, J. R. Localization of tone. University of Chicago decennial 

 put)lications, Chicago, 19()(). 



