76 SECTION PHYSICS. 



connection, and here is the mode which I suggested, in which the 

 string is set apart from the reed, so that it can be used in its own 

 register as in a harmonium. When we get as far as this, and the 

 reed-tongue and string are used, there is no longer any room for 

 originality as regards either wind and string, or reed and string, but 

 there is plenty of room for improvement. When the reed-tongue is 

 put to the end of a string, you may make the string act, but the farther 

 you get from the reed the intervals get wider and wider, because the 

 amount of control exercised upon it gets less and less ; and if you go 

 as far as the sixth interval, it would be widened out immensely, but 

 you seldom even get as far as that, because the string would bring up 

 into new combinations of nodes and segments, or refuse to speak 

 altogether. After trying a whole year to make a wind violin on this 

 principle, I gave it up in despair, because I could not get enough 

 intervals, and because they were so irregular; also because the tone 

 varied. When you are close upon the reed, the reed is constrained in 

 its motion and the tone is pure, but as you get away it is less con- 

 trolled, and the tone becomes more loose, coarse, and reedy. It was 

 only within the last four months that it occurred to me to use a conical 

 form of string. The small end is applied to the reed, and as it 

 departed from it it got larger and larger in bulk, and accordingly 

 the intervals remained the same, and the tone remains the same, 

 because you encounter the firm resistance of the larger string. There 

 is another curious thing also, that whenever you timed the string the 

 intervals pulled out, because the relative intervals in the string did not 

 remain the same ; the tongue entered into its composition to such an 

 extent ; but when you tuned from the small end of this conical string, 

 instead of the diminishing bulk of the string, you are able to maintain 

 the same bulk, and accordingly this new form of string has recovered 

 to us an instrument which I once abandoned in despair. But although 

 I thought it impossible to do anything in the way of a wind violin, 

 there was still a chance of doing something with the organ. If you 

 use merely one string and one reed, you can take the best note they 

 afford, and these difficulties of the intervals do not arise. Accordingly, 

 I tried to make an organ with a separate reed and string to every 

 note, and I found that whenever the string broke up into three 

 segments, it gave a beauty of tone which I could not gain by the most 



