Dr. It. Konig on Manometric Flames. 11 



overtone 4, which coincided with the proper note of the sound- 

 ing-chest. On turning the plate more rapidly, the flame- 

 picture became simpler, until at / it became one single flame, 

 so that the overtone 3 must be quite wanting in this sound 

 of the siren. The ascending scale had scarcely passed / when 

 there appeared between each two large flames a small but 

 sharply defined flame, which quickly increased in size, and 

 towards c reached nearly the height of the chief flames, where 

 the effect of the resonance-box confirmed the fact of its being 

 the overtone 2 of the sound of the siren. Above c the 

 smaller flame leant always more towards the larger one, until at 

 a it completely disappeared in it : after this again only single 

 flames appeared (PL I. fig. 7). 



In order to make the sound in these experiments act strongly 

 on the capsule, I provided the resonance-box with a tube, and 

 placed its interior in direct connexion with the flame-apparatus. 

 These experiments, in which the air-impulses of the siren are 

 prevented from passing immediately into the atmosphere, being 

 compelled to pass through a resonator which remains unchanged 

 for all the fundamental tones of sound, give a visible picture of 

 the process of the formation of vocal sounds ; for it is known 

 that the air contained in the cavity of the mouth, when speak- 

 ing or singing the same vowel in different tones, is always tuned 

 to the same note, so that the mouth must act on the air-waves 

 produced in the larynx in the same way as the sounding-chest 

 on the air-impulses of the siren. Nevertheless the series of 

 flame-pictures of the same vowel, sung in the tones of two oc- 

 taves, does not show such sudden changes as might have been 

 expected without closer research. 



In order to produce the pictures of the vowels, I sing them 

 into a small funnel-shaped mouthpiece which is connected with 

 the cavity before the membrane by a short india-rubber tube ; 

 thus they reach the capsule with great intensity (fig. 6). 



I had already in 1867 sketched and had painted the pictures 

 of the vowels u } o, a, e, i sung to the notes of the two octaves from 

 C to c. I proceeded in the following manner. In order to 

 be sure that I had not changed the character of the vowel in the 

 transition from one tone to another, I first verified the proper 

 note of the mouth with the tuning-fork ; then, while I sang 

 into the apparatus, an artist drew the picture which he saw in the 

 mirror. I also drew the same picture independently : and if both 

 our drawings were identical they were looked upon as correct ; 

 if, however, there were discrepancies, I repeated the experiment 

 until the error was discovered. 



The five finished drawings (PI. II. fig. 8) were unfortunately 



