HABIT FORMATION IN THE DOG 37 



requirement. The vernier scales and the card -board index 

 glued on the dials of the variators, however, are useless and 

 should not be relied on. There is wide difference between the 

 marks on the scale and the actual pitches under variable atmos- 

 pheric conditions; and the play of the cogs in the device for 

 raising and lowering the piston, is in some cases greater than 

 three points on the scale. 



If pipes or whistles are blown from an ordinary Stern or 

 Whipple tank they are not accurate in pitch within a limit of 

 about 3"% in either direction, as the air pressure is not con- 

 stant. For preliminary or rough work the 3^ are often useful 

 when blown in this way. but should not be depended on for 

 finer work. The Stern tank is much too small. It will not blow 

 a large whistle or a c-256 d.v. pipe. The air is so quickly ex- 

 hausted that if it is in the same room with the animal, the noise 

 of raising it is often disturbing. If chords are blown on varia- 

 tors blown from it, it requires filling after nearly every stim- 

 ulus. For finer work Watson -^ has devised and installed 

 in the psychological laboratory at the Johns Hopkins Univer- 

 sity an air-system consisting of a tank filled by a motor-driven 

 positive pressure blower. The tank supplies air-streams from 

 a distant room. The pressure is remarkabl}^ constant, and 

 makes possible an accuracy of pitch in blown pipes or whistles 

 far beyond that hitherto attainable. The apparatus is rather 

 expensive. 



The apparatus adopted for finer discrimination work of this 

 kind is a system of "tandem-driven" tuning forks similar to 

 that recommended and used by Helmholtz -^ In the present 

 tests the apparatus was an-anged as follows: A c-64 d.\/. fork 

 is mounted in a' room 100 feet distant from the experimental 

 room for electric actuation. A diagram of this fork and its 

 ecjuipment is shown in Fig. 2. The electrical connection is as 

 follows : From the positive pole of a six volt two ampere stor- 

 age cell to a rheostat, thence to the pole a, through the fork 

 and platinum contact p with the mercury cup c, to the magnet 

 m, to the pole b, to the negative pole of the storage cell. Thus 

 with each vibration of the fork the current is made and broken 

 at the contact p with the mercur}^ cup c. The mercury in c 



^'Watsox, Johv B. Article as yet unpublished. 



^^ Helmholtz, H. v. Sensations of Tone (Ellis' translation). 



