COMI'LKX NEURAL AIK'S. 97 



displ;iccme>nt. The; move>me'nl terminates abruptly, absolutely with- 

 out rebound or secondary vibration. Inspection of the curve at the 

 moment of stopping shows that the transition from the; most rapid 

 movement to complete rest occurs in about 0.002". The exposure' is 

 not absolutely noiseless. II seems to begin with a light swish and ends 

 with a light thud. Neither noise bears any resemblance to the usual 

 noisy stop of the spring or the gravity tachistoscope. (J run ting its 

 reasonable fulfillment of the main criteriaof a satisfactory exposure appa- 

 ratus, the chief advantage of this form over the camera tachistoscope of 

 Erdmann-Dodge, 1 the transparent-mirror tachistoscope of Dodge," and 

 other satisfactory instruments, is its simplicity and compactness. None 

 of these forms could have been used in our complex of instruments with- 

 out serious inconvenience to the operator or subject or both. All of them 

 are relatively bulky, and in our experimental arrangements space was a 

 valuable asset . 



VOICE-REACTION KEY. 



Considerably more difficult of construction thnn an adequate ex- 

 posure apparatus is an adequate reaction key for vocali/at ion. We 

 know of no even relatively good reaction key for recording the move- 

 ments of the vocal organs. Movements of the chin, lips, tongue, and 

 larynx may each be recorded separately, as is commonly done in 

 experimental phonetics. But then- is no one key for them all. The 

 familiar voice keys of Kraepelin, 1 * ( 'attell, 4 Erdmann-Dodge, 1 Homer, 1 ' 

 and others fnmkly surrender the effort to register the; muscle-action of 

 articulate speech in favor of the consequent air-movements. But this 

 is a questionable e;xpc;elie;nt, unless due precautions are taken to tender 

 it innocuous. Voice keys de?pe;nel on the expiration of air involves! in 

 utterance, to break an electric contact. Unfortunately, the e;hrono- 

 logical place of the expiration of air in the- total physiological process of 

 utterance is ve-ry different for different worels. Consequently, no air- 

 current key c;m ever register in ;iny reliable, way the re;d beginning of 

 the vocali/;it ion re;ie;tion. In most e'xperinient.'d investigations, how- 

 ever, this is not a materinl source of error. If one seeks the relative 

 efficie'iicy of the vocalization process under varying conditions, arid 

 if one uses a definite, unchanging series of stimulus words, such as our 

 group of words was, the precise beginning of muscular reaction is 

 relatively unimportant. For studying the effect of a drug, any one of 

 the systematically correlated movements of the reaction would be 

 equally significant in comparing the nor:n;il with the drug-reaction 

 periods. It is on these grounds and with the corresponding limitations 

 that air-movement keys are defensible in speech-reaction movements. 

 As Wirth stales in a discussion of this type; of key, "They permit 



'Erdrnann and Dodgo, Paychologuche UntersuchuM^cri iibcr das L':,-.'ii, Flallo, 1898. 

 "Dodgo, PHychol. Bull., l'J<>7, \, p. 10. "Kraepelin, Phil. Stud., 1883, 1, p. 417. 

 4 Cattcll, Phil. Stud., 1SS5, 3, p. 313. *R5mer, Kr.'irpflin'H P.syrhol. Arlx-if,., 1, p. 577. 

 'Wirlh, Psychophynik. TigerHtedt's Handbuch der physiologiHchon Mcthodik, 1912, 3, p. 400. 



