238 
IOWA ACADEMY OF SCIENCE Vol. XXIV, 1917 
If the radii are plotted as abscissas and the specific resistances 
as ordinates the curves in figure '39 result. (The radii of the 
circles equal the probable errors of the respective specific re- 
sistances.) From the curves it is evident that the two sets of 
wires do not vary in the same manner. The curve for Set 2 
seems to approach some limiting value of the specific resistance. 
This is what we would expect if the change due to drawing is 
a surface effect but the number of wires in a set is too small 
to draw any definite conclusions. In general, however, the 
specific resistance increases as the radius decreases, both sets 
agreeing in this respect. 
2 6 10 14 18 22 26 
Radius times lO^cm. 
Figure 39 
In each of the density determinations a known length of the 
wire was necessary in order that its volume might be calculated. 
The wire was held taut between two clamps and the distance 
between the faces of the clamps was determined as stated before. 
The wire was then cut off as near the faces as possible with a 
pair of diagonal cutters and the length of the ^tub-ends was 
measured with the micrometer-microscope. The length used in 
the computation was the measured length minus the length of 
the stub-ends. The mass of this known length of wire was then 
determined on a chemical balance which had a sensibility of 
about four scale divisions per milligram, weighings being made 
on each pan so as to have a check on the results. The zero of 
the balance and the sensibility were determined for each weighing 
for each wire. The percentage probable errors of the masses and 
