252 
Proceedings of the Royal Society of Edinburgh. [Sess. 
elasticity ; but there is no real limit of elasticity in testing threads, similar 
to what is met with in testing materials like steel. This difference arises 
from the constitution of a thread differing greatly from that of a continuous 
substance. A thread is composed of fibres arranged spirally or, as in carded 
woollen, lacking any well-defined arrangement. Therefore much slipping 
occurs amongst the fibres from the beginning of the pull. 
The most valuable quantity to be determined relative to the strength of 
a thread is not its ultimate breaking strength but the yield point ; because, 
if the stress exceed that corresponding to the yield point and be continued 
sufficiently long, the thread will ultimately be ruptured. This result will 
be much accelerated if the stress be periodically applied and withdrawn, as 
in the process of shedding the warp in the loom. This determination 
of the yield point provides us with a much more accurate method of 
determining the relation between the degree of twist and the strength of a 
thread. It is then possible to apply an increasing twist to the same thread 
while under tensile stress, and, if great care be exercised in the application 
of the load, quite a long series of observations of the yield point may be 
obtained on the same thread before the thread breaks. Theoretically, it 
should be possible to perform the whole range of experiments with one 
thread, because the stretch is not very great up to the yield point. In 
practice, this limit is exceeded at each observation, as it is often not well 
defined, and the accumulation of these increments soon wears the thread 
out at some weak point in its length. 
The great advantage of the second method over the first is that we are 
eliminating the effect of variation in the uniformity of the thread, and 
thus a satisfactory mean can be struck from a much smaller number of 
observations. Yarn-testing has not received the attention from manu- 
facturers which its importance demands. Many firms never test their 
yarns at all. The majority (possibly all) of those who do only ascertain 
the breaking strength and the ultimate stretch of their yarn. The ultimate 
stretch is spoken of as the elasticity of the thread. This statement is quite 
wrong ; the ultimate stretch and the elasticity are quantities not necessarily 
connected. The elasticity of a thread may be measured by its stretch for 
a given stress well within the limit of elasticity, this stretch being 
distributed throughout the length of the thread. The stretch from the 
yield point to the ultimate breaking point, on the contrary, is largely 
confined to one or more “ soft ” or slackly twisted parts in the thread. 
This local stretch is of no importance to the manufacturer, because the 
thread is useless long before it takes place. Further, the local stretch, being 
dependent on so many accidental circumstances, does not bear any definite 
