TRANSACTIONS OF SECTION G. 653. 
leaving a comparatively small surface that can be acted upon. It was found that 
the material known as Prout’s glue gave good results. Another device was to use- 
as the support a material electro-negative to lead, so that a comparatively ineffec- 
tive local couple was formed. Altogether these experiments have now resulted in 
the ability to construct a secondary battery which may be left for some time with- 
out appreciable loss, and the life of which is greatly extended beyond what was 
hitherto possible. The second part of this paper dealt with the economic genera- 
tion of steam for electric and other purposes. It described a special combination of” 
elephant boiler with the ordinary furnaces replaced by coke ovens. By these means 
the gases usually wasted from coke ovens were utilised in the production of steam. 
The coke produced was of the hard kind required for foundry and smithshop pur- 
poses, and its value was equal to or greater than that of the coal from which it is 
produced. The system had been in use considerably over two years with complete 
success, 
8. Fire Risks of Electric Lighting. By Kitumsaworta Heposs. 
There is a great difference between the electric currents which have been in 
constant use for telegraphic purposes, and those which are to be supplied by the 
undertakers under the Electric Lighting Act. The latter can only be said to be 
free from danger when the heat generated by the current is utilised in its right 
place, and not developed in the conductors or wires which lead the electricity to 
the incandescent lamps. 
The Fire Risk Committee have already issued rules for guidance of users of 
electric light ; these can hardly be said to embrace all the salient points of the new 
subject, which can only be arrived at after years of practical work. The necessity 
of proper regulations has already been recognised by the insurance offices, both in 
the United States and Germany, and some of their special rules are given in this 
paper. 
The conductors must be properly proportioned for the current they have to 
carry; whatever resistance there is in the conductor will cause a corresponding 
development of heat, which will vary with the amount of electricity passing, and! 
inversely as the sectional area. 
The material must be free from impurity, otherwise an impure section will 
increase the resistance. The extraordinary difference in the conducting power of a 
sample of ‘commercial’ Rio Tinto copper wire, as compared with the pure metal, 
was shown in an experiment by Dr. Matthiessen—the conducting power being only 
13°6 as against 99:95 for pure copper. 
The continued heating of an impure metallic conductor has a certain effect on 
its electrical resistance. With the sample just mentioned, the conducting power 
at 100° C. decreased from 13°58 to 13°558 after the wire had been heated for 
three days. It does not always follow that there will be a decrease in the con+ 
ducting power, as, with alloys, the opposite effect is produced. A copper-silver 
alloy showed an increase of ‘264, after having been heated to 100° C, for three days, 
and a tin-copper, an increase of ‘13. 
As the temperature in Dr. Matthiessen’s experiments was not increased over 
100° C., the author has made some further experiments—heating the wires by the 
electric current from a secondary battery, to within a few degrees of their melting- 
oint. 
F The materials given on the table printed on the next page were tried—the 
wires and foils having such sectional area, and so arranged that, on the current 
being increased by twenty per cent., they were immediately fused. 
The total length of each experiment was twenty-four hours, during which time 
the current passing through varied slightly, and the following is a mean of the- 
results :— 
