251 
1906-7.] The Strength of Twisted Threads. 
make observations on which to found a general law. The law, if true at 
all, could only be so within the narrow limits of observations made. 
The twist on the thread should have been gradually increased in a series 
of threads from zero to a high degree, and the corresponding breaking 
strengths observed. It is evident that the law cannot he true much beyond 
the higher limit, because the statement involves a constantly increasing 
strength of thread as the twist increases. That a thread cannot go on 
getting stronger indefinitely with twist is so obvious as to require no 
demonstration ; there must be a change in the character of the relation 
as the torsion proceeds. 
In justice, it ought to be stated that the defects noted in no way 
invalidate the main results of the above interesting paper. The care 
manifested in working out the details of the chief subject of inquiry 
contrasts strangely with the treatment accorded to the twist problem — so 
much so, that one is led to the conclusion that the authors did not give it 
their best attention. This question of twist is not one in which an 
investigator can indulge in a little by-play as a prelude to what he may 
consider more serious business. Twisted threads really present a group 
of the most complicated problems which a physicist can set himself 
to solve. 
There are two methods of experimental investigation which appear 
commendable : — 
(1) The direct method, which is as follows. Take out all the twist from 
a thread and test its strength. This can only be done satisfactorily on a 
machine which admits of the twist being varied, and also of the test for 
strength being made. Then take out all the twist from another thread of 
the same hank and put on one turn per inch, and test the strength of the 
thread in this condition. Perform the same operation for two, three, and 
more turns per inch of twist. Then repeat the whole sequence of operations 
a sufficiently large number of times to ensure a satisfactory mean being 
determined. 
(2) The indirect method, in which the thread is loaded to some 
convenient point short of breaking. When tensile stress is applied to a 
thread, the thread stretches, but the elongation is never quite proportional 
to the stress. The increments of length corresponding to any fixed 
increment of stress gradually increase as the stress rises. But there is no 
sudden change until the stress is considerable when the yield point is 
reached, after which the increments of length for any given increment of 
stress become very much larger, and go on increasing as the stress increases 
up to the breaking point. The yield point may be termed the limit of 
