126 
Now how do these elements effect the conductivity of the 
steel ? Let us examine them in detail. 
First, graphite carbon must be present in the steel as a 
mechanical mixture, and so can scarcely exercise any in- 
fluence on the conductivity. Then manganese will probably 
be present in the form of an oxide, as manganese oxidises 
at the high temperatures steel is cast more readily than any 
of the other constituents of the steel. Now, oxide of man- 
ganese can hardly be present in the steel other than as a 
mechanical mixture, and thus we may disregard its influence. 
There are now left combined carbon and silicon, and in 
samples 1 and 2 a little sulphur or phosphorus. From ana- 
logy with copper, sulphur and phosphorus should increase 
the electrical resistance of iron very much. Perhaps we may 
estimate in a rough way the total effect of these elements 
on the conductivity by taking their sum as shown at foot of 
table C. If this is correct, we are led to the interesting 
result that within th e range of these experiments any increase 
in the percentage of sulphur, phosphorus, carbon, and sili- 
con present in a steel is accompanied by an increase in its 
electrical resistance, and further, an increased electrical re- 
sistance is concurrent with an increase in the resistance to 
tensile strain. 
The quantity of carbon, silicon, sulphur, and manganese 
in the samples of steel is so very small that the most careful 
analyses give results which must be regarded as approximate 
rather than definite ; hence I have much diffidence in laying 
these figures before you. But however much the ultimate 
accuracy of these figures may be called in question, I think 
we may fairly say that the electrical resistance of a piece 
of iron or steel is a measure of its resistance to tensile strain 
and of the amount of combined carbon, sulphur, silicon, and 
phosphorus it contains. 
