METALLIC WIRES BY ELECTRICITY. 459 
Experiment 37.—A platina wire of the same dimensions as 
that used in Exp. 34, was melted into a number of small per- 
fectly round globules by a discharge in which s=5, g=22. 
Experiment 38.—A silver wire (radius 0°0264, length 19 lines) 
was melted into globules when s=6, g=26. 
Experiment 39.—A copper wire (radius 0:0253, length 16 
lines) became incandescent when s= 6, g = 25, and was con- 
verted into a multitude of very minute globules by a discharge 
in which g = 30. The globules were some of them so small as 
only to be perceptible with the magnifier. I made several un- 
successful attempts to obtain larger globules of copper. 
It may appear striking in these experiments, that the charge 
necessary to produce complete fusion is not much greater than 
that which causes the first appearance of incandescence in the 
wire. In oxidizable metals the temperature is increased by the 
absorption of oxygen from the air, and a chemical effect is thus 
added to the electrical. This is most remarkably shown in the 
case of iron, which often is melted by discharges, which in them- 
selves would only have produced a moderate incandescence. 
But the kind of fusion is then peculiar, as the following expe- 
riments show. 
. Experiment 40.—An iron wire of 0!""0266 radius and 17"! 
long, which when s = 3 and g = 10 would have shown the first 
signs of incandescence, became intensely red-hot on the dis- 
charge of the quantity of electricity 13. The incandescence did 
not cease immediately, as on all other occasions, but was in- 
creased to a white heat; a few globules dropped from the wire 
and sprung about on the table, emitting vivid scintillations. 
The ends of the wire, some lines in length, which remained in 
the clamps, were melted into globules, and retained the position 
which they had before the commencement of the experiment. 
The iron wire is quite differently affected by more powerful 
discharges, even when these are insufficient to produce perfect 
fusion. 
Experiment 41.—An iron wire of 0°04 line radius, 16 lines 
in length, was brought to incandescence when s=7 and g=25. 
When g=30 it was broken to pieces whilst red-hot, and was 
fused, so that three globules, which were oxidized while springing 
about, and three pieces were collected. The long wire-ends 
which remained in the clamps were forcibly bent round them ; 
and on one end a globule was melted. We know that thin iron 
2K 2 
