265 
1906-7.] Degree of Twist in Single Threads. 
handle of high-quality woollen goods depends upon this. Many hbre ends 
project from the surface of the thread. Therefore if, to begin with, in 
twisting, the fibres are not parallel, by untwisting we cannot arrive at the 
parallelism which it is the object of microscopic examination to observe. 
In the preceding papers * on “ Take-up in Twisted Threads " I have 
shown that the relation which exists between length and twist in a thread 
is expressible in the form 
L = l(l — kx 2 ), 
where L is the length of twisted thread. 
I „ „ untwisted thread. 
x „ turns of twist per inch. 
k is a constant. 
Evidently the thread is longest when the twist is zero. This is true 
with every description of single thread, irrespective of its process of manu- 
facture. To devise some form of apparatus which will utilise this property 
is all that remains to be done. 
The apparatus which I have designed for the purpose is shown in fig. 1. 
A represents ten inches of single thread, which is stretched between clamps 
B and C. B is carried on the end of a bar, which can be rotated on its 
axis through toothed wheel N by turning wheel M. 
The number of turns are registered on dial T, to which motion is com- 
municated by worm W carried on the rotary bar. The other clamp C is 
carried on the shorter arm of a light brass bell-crank lever pivoted at G. 
The pointer of the long arm F can move along scale S, thus indicating any 
change in the length of the thread. The method of conducting the experi- 
ment is as follows : — Take the length of thread such as will cause the pointer 
to come opposite the top of the scale ; read the twist dial. Turn wheel M 
so that clamp B rotates in the opposite direction to the twist in the thread. 
Observe the lowest point on scale S to which lever F dips ; this point will 
be reached when the thread is untwisted. Read the twist dial again ; the 
difference of the readings will give the turns of twist in the length of 
thread under test. 
It will be seen from the graph in fig. 2 that the point at which the 
thread is untwisted gives a stationary value for the length. 
There is scarcely any difference on the length for a number of turns on 
* t£ The Relation between Take-up and Degree of Twist in Twisted Threads,’ 5 Proc. 
Roy. Soc. Edin., 1906, vol. xxvi., part iii., pp. 182-206. “The Relation between Take-up 
and Yarn Number for constant Twist in Twisted Threads,” Proc. Roy. Soc. Edin., 1906, 
vol. xxvii., part ii., pp. 93-106. “ The Relation between Take-up and Degree of Twist in 
Twisted Threads with Singles of unequal Sizes,” Proc. Roy. Soc. Edin., 1906, vol. xxvii., 
part ii., pp. 107-116. 
