I9I4.] AN ARTIFICIAL HISS. 329 



We were unable to determine the pitch of the continuative hiss by the 

 Kundt dust-method; the lycopodium powder obstinately refused to move. 



The accuracy of the Edelmann tables has been questioned by C. S. Myers 

 (" On the Pitch of Galton Whistles," Journ. of Physiol, XXVIII., 1902, 417 

 fif.). Edelmann does not give the m. v. of his scale readings; but it is pos- 

 sible that his technique is so accurate that the variation is minimal, or even 

 that a single count suffices. Neither does he tell us how he compresses his 

 bulb; it is probable that he uses some mechanical device which ensures a 

 constant compression. We have ourselves made the following determinations 

 with whistle No. 423 (temperature read as the mean of four thermometers) : 

 (i) Ordinary vigorous squeeze, such as is employed in terminal determina- 

 tions: 



Found from 5 trials with the dust-method 8,897 — 18 v. d. 



By Edelmann's table 8,775 



(2) Weaker squeeze, used in our experiments: 



Found by dust-method 8,594 ± 63 v. d. to 8,522 ± 27 v. d. 



By Edelmann's tables + judgments of coincidence of 

 tones 8,430 



(3) Very violent squeeze, 10 trials by dust-method 9,046 ± 71 v. d. 



It is clear that Edelmann uses a " normal " compression, of the same 



order as that which an experimenter naturally employs for terminal tests. 

 Random determinations of our two whistles at different points of the scale, 

 with the ordinary vigorous squeeze, agree within about 100 v. d. with the 

 Edelmann tables. 



We have had but little experience with continuous tones under change 

 of water-pressure ; but we find, so far, that the pitch of our whistles does 

 not alter appreciably within the limits of 90 to 160 mm. of pressure. 



While, then, we do not question the accuracy of Myers' determinations, 

 we think that there is no need either to doubt the reliability of the Edelmann 

 whistle under " normal " conditions. 



