QUARTZ FIBERS. 



325 



Now supposing that the force — whether of the nature of a twist or of 

 a pull (it does not matter which) — is too small to produce an appreciable 

 twist in the wire, it is obvious that a finer wire must be employed, but 

 it is not obvious how much more easily a fine wire is twisted than a 

 coarse one. If the fine wire is one-tenth of the diameter of the coarse 

 one, we must multiply ten by itself four times over in order to find how 

 much more easily twisted it is, and thus obtain the enormous number 

 10,000; it is 10,000 times more easily twisted than the coarse one. 

 Thus there is an enormous advantage in increasing the minuteness of 

 the wire by means of which feeble twisting or pulling forces are meas- 

 ured. But if the delicacy of the research is such that even the finest 

 wire which can be made is still too stiff, then, even though with such 

 wire, which is somewhere about the thousandth of an inch in diameter, 

 forces as small as the millionth part of the weight of a single grain can 

 be detected with certaiuty, the wire is of no use ; and as wire can not be 

 made finer, some other material must be used. Spun glass is fine and 

 strong, and is still more easily twisted than the finest wire, but it 

 possesses a property somewhat analogous to putty. When it has been 

 twisted and then let go, it does not come back to its old place, so that 

 though it is much more largely twisted than wire by the application of 

 a force, it is not possible with accuracy to measure that force. There 

 is, or rather I should say there was, no material that could be used as 

 a torsion thread finer than spun glass ; and therefore physicists use in- 

 stead a fiber almost free from torsion. A single thread of silk as spun 

 by the silkworm is taken and split down the middle, for it is really 

 double, and one-half only is used. This is far finer than spun glass, 

 and being softer in texture, it is much more easily twisted. Silk is ten 

 thousand times more easily twisted than spun glass. So easily twisted 

 is silk that in the majority of instruments the stiffness of the silk is 

 either of no consequence at all, or at any rate it only produces but the 

 slightest disturbing effect. Now, if it is necessary to push the investiga- 

 tion further still by the continued increase in the delicacy of the appa- 

 ratus, silk itself begins to prevent any progress. Silk has a certain 

 stiffness, but if that were always the same it would not matter ; but 

 then it possesses that i>utty-like character of spun glass, but in a far 

 higher degree; it is affected by every variation of temperature and 

 moisture, and any really delicate measures are out of the question when 

 silk is used as the suspending fiber. 



This, I believe, is a fairly accurate account of the state of the case, three 

 years ago. At that time I was improving, or attempting to improve, a 

 certain class of apparatus of which I shall have more to say presently, 

 and I was met by the difficulty that a greater degree of delicacy was 

 required than was possible with existing torsion threads. Silk would 

 have entirely prevented me from reaching the degree of delicacy and 

 certainly in this instrument that I hope to show this evening that I have 

 attained. 



