METALLIC WIRES BY ELECTRICITY. 439 
charge, this bend merely appears as a break in the brilliant line 
of light which is reflected from a polished wire by daylight ; on 
the repetition of the same discharge, or by increasing it, the bend 
is deepened until it becomes a measurable angle. A platina wire 
(radius 0°021) was exposed to the following discharges :— 
Experiment 4.— 
Number Quantity 
of jars. of electricity. 
3 8 Sparks on the outer side of the wire— 
vapour. 
9 Violent shaking—bend. 
10 Bend deepened. 
10 . The same—fresh bends. 
This wire presents, at a considerable distance from its fixed 
points, a deeply impressed angle, and on the right-hand side of 
that several slight bends. Four other platina and two iron wires, 
which were likewise exposed to a few discharges, received in the 
same manner deep bends. Wherever the bending takes place 
without obstruction, it is obtuse-angled: in several cases I mea- 
sured it, and found it to vary little from 110°. Neither the di- 
mensions nor the material of the wire cause this to vary, as was 
shown by a thick platina wire (radius 0°0405), an iron wire and 
a copper wire. The formation of the angle is rendered difficult, 
and partly obstructed, if the wire be stretched in a straight line 
or exposed to considerable tension in an arched curve. Then, 
instead of the deep bends, only slight indentations appear in 
great number, which easily escape observation: sometimes the 
wire breaks at that part where the angle would have been formed, 
and the wire is hindered from yielding. For the same reason, 
an angle already formed hinders the production of a new one, 
and in this manner it is that the different distortions of the wire 
are caused, of which I shall speak directly. The bends occur 
on the discharge of a less quantity of electricity than is requisite 
to produce a red heat. The degree of incandescence distinctly 
visible by daylight has been taken in these experiments as the 
normal quantity, but I have satisfied myself, that. in complete 
darkness the formation of the first bend is attended with no ap- 
pearance of light (with the exception of occasional sparks at the 
connecting points of the wire). If more powerful discharges are 
allowed to pass through the wire after the first bend, new bends 
will constantly appear, which, modified by those already present, 
or distorting them, cause curves and angles in different planes, and. 
