QUALITATIVE ANALYSIS. 53 



methods. The descriptions of speech sounds contained in the usual works 

 on phonetics are careful and clever, but they go no further than unaided 

 observation; the use of the method of registration and tracing is like 

 introducing a phonetic microscope which from the start opens up a new 

 field. 



The qualitative analysis as already described in this section has 

 been made without any experimental aids. There is a future, however, 

 for the introduction of a system of "experimental analysis"; if such a 

 system can be developed its results will certainly be as startling as the 

 series of discoveries in chemistry resulting from the introduction of simi- 

 lar methods. 



The experimental analysis of the action of the vocal organs in speech 

 is already well developed and forms almost a science by itself. A first 

 attempt at something different, namely, an experimental analysis of the 

 sounds heard by the ear, will here be described. 



As already repeatedly pointed out, no two waves of a vowel are alike; 

 the differences are often so great that we may be sure that one part sounds 

 utterly different from another, although the ear apparently gets only a 

 single general impression. A method must be found whereby the ear 

 can be enabled to hear the sound of each wave separately; for this pur- 

 pose the apparatus illustrated in plate xi has been constructed. 



A "zinc strip" with the etched curve — which may be a photographic 

 copy of any speech curve or an arbitrary curve — is moved horizontally 

 between guides l)y a "felt wheel." Resting in the etched groove on the 

 strip there is a sharp steel "motor point" at the end of the "reducing 

 lever"; this point follows the vibrations of the curve as the strip is moved. 

 The movement imparted to the reducing lever is registered by a second 

 steel point — the "recording point" — near the fulcrum. The registration 

 is done on a rotating zinc "recording disc" coated with asphalt varnish 

 or with wax, or on a wax disc. The relations of amplitude, that is, the 

 degree of reduction, are arranged by adjustment of a little shde carrying 

 the recording point. The horizontal axle in the reducing lever is to allow 

 for differences in thickness of the recording disc. 



To establish the necessar^^ relations between the movement of the 

 zinc strip and the rotation of the disc they are both moved from the same 

 source. The motor belt turns an axle carrying a "spur driving gear"; 

 this turns another "spur gear" which is attached to the "driving axle" 

 for the two parts of the apparatus. This driving axle carries two worms. 

 One " worm" rotates a " worm gear" on the axle with the felt wheel. The 

 other "worm" rotates a "worm gear" on an axle with two pulleys; of 



