131 
By Measure- By Fixed Lead By Suspended 
ment. 
Weights. 
Lead Weights. 
Average. 
No. 1... . 
.. 86-246 . 
.. 86-173 
... 86-225 .. 
. 86-215 
No. 2... . 
.. 86-424 . 
.. 86-330 
... 86-435 .. 
. 86-396 
No. 3... . 
.. 85-248 
.. 85-070 
... 85-283 ... 
. 85-200 
No. 4... . 
.. 84-843 ., 
.. 84-773 
... 84-906 .. 
. 84-841 
The results show that the moments of steel bars, when 
carefully finished, may be very correctly deduced from their 
linear dimensions and weight. 
It is important that the bars should be thoroughly and uni- 
formly magnetized. This is best effected by bringing the 
ends of the bar into contact with the ends of two straight 
electro-magnets of considerable power. In this way the 
confused polarity, which more or less exists in all bars mag- 
netized by stroking, is avoided — the result being, that the 
magnetic virtue is more permanent, and at the same time, 
much more effective. 
The needle to be deflected should be of small dimensions, 
not exceeding half an inch in length. It may be conve- 
niently made of several perfectly hard pieces of common 
sewing needle mounted on perforated card. A glass index 
should be affixed to it, the ends being furnished with very 
fine bits of copper wire. These show well when traversing 
the graduated circle. 
A circle of six inches diameter will be found a convenient 
size. It should be divided to thirds of a degree. The 
deflections may be read off with sufficient accuracy by the 
aid of an eye-glass, a piece of looking-glass being placed on 
the table to ensure the correct perpendicularity of the 
eye. It is easy after a little practice, if read off the deflec- 
tions to half a minute or less. The needle should be 
protected from currents of air by being placed with its 
circle in a shallow box covered by a glass plate, in the 
centre of which there is a hole for the purpose of suspension. 
A filament of silk of six inches length, and strong enough 
