- | 
FEBRUARY 15, 1884.] 
positions, —the normal position, and that in 
which the base of the lamp was towards the 
photometer-box ; these being regarded as the 
positions of the greatest importance. Alto- 
gether, twenty-five photometric observations 
were made in comparing the arc-lamps. The 
lamps compared were taken at random from 
those in use by the exhibiters. 
The comparison of incandescent lamps pre- 
sents questions of fur greater delicacy and diffi- 
culty. There is one element, in the economy of 
an incandescent lamp, which does not enter to 
any extent in the consideration of arc-lamps ; 
that is, the life of the lamp. Although of great 
importance, it did not seem possible, in the 
limited time which was at the disposal of the 
jury, to investigate this point. The only fair 
and impartial method of making such an investi- 
gation, involved, in the opinion of the jury, the 
continuous and prolonged burning of a large 
number of lamps belonging to the different 
competing systems. Under the circumstances, 
it was absolutely impossible to make use of 
this method. 
There exists, also, difference of opinion as to 
the proper method of comparing the efficiency 
of two incandescent lamps. They may be re- 
duced to the same illuminating-power, and the 
electrical energy consumed by each may _ be 
compared ; they may be brought to a condition 
in which they consume the same electrical en- 
ergy, and their illuminating-power compared ; 
or they may be allowed to differ in both of these 
elements, and comparisons be made in both. 
The first method has been pursued in several 
tests which have been made both in Europe 
and in this country. 
Incandescent lamps are generally made to 
be equal, nominally, to a given number of 
standard candles; but, by modifying the con- 
sumption of energy, a lamp of nominally low 
candle-power can be made to produce almost 
any degree of illumination, from nothing up 
to the equivalent of several hundred candles, 
the high illumination being, of course, at the 
expense of the life of the lamp. If this element 
is left out of consideration, the efficiency of a 
lamp increases rapidly with its degree of incan- 
descence. As it is by no means necessary that 
incandescent lamps should run at a fixed ‘ can- 
dle-power,’ it will follow that the temperature 
at which a lamp will show greatest efficiency 
(including the life element) will depend greatly 
upon its construction. 
Taking two lamps of radically different con- 
struction, however, there will be for each a 
certain set of conditions as to current strength 
and electromotive force, and including the 
SCIENCE. 
181 
element of life, under which it would show its 
highest efficiency and economy. After such 
conditions were determined for each lamp, a 
strict comparison would be possible. The re- 
duction of two such lamps to the same degree 
of illumination would probably be unfair to 
one or the other, or possibly to both, if the ele- 
ment of life is not considered. 
Suppose that a lamp in one system is at its 
best, all things considered, at fifteen candle- 
power, and that one in another reaches its 
highest degree of efficiency at sixteen candle- 
power. If they are both brought to fifteen 
candle-power, the second must suffer in the 
comparison ; and if both are brought to sixteen 
candle-power, and the element of life is not 
considered, it will again suffer, for the apparent 
efficiency of the first will be increased by its 
higher incandescence. 
As the labor of determining the most favor- 
able conditions for each lamp would be so 
great as to necessarily throw that method out 
of consideration, the jury felt constrained to 
adopt the last of the three methods mentioned 
above. The jury assumed, in fact, that the 
exhibiters of the different systems had already 
determined these favorable conditions in their 
own interest; and that in putting their lamps 
before the public for the entire period of the 
exposition, each maintaining more than two 
hundred lamps in different parts of the expo- 
sition building, they would operate them as 
nearly as possible in accordance therewith. 
In other words, it was decided to compare 
the lamps as they were used in the exhibit, 
determining the ratio of their illaminating- 
power, and measuring the electrical energy 
consumed by each. It is proper to state, that 
the lamps of both systems were spoken of 
by their respective representatives as sixteen 
candle-power lamps, although certain marks 
on the lamps which were supposed by the 
jury to refer to candle-power did not exactly 
agree. 
To secure impartiality of selection, the jury 
requested permission to have access to the 
supply of lamps kept by each company for use 
in the exhibit, which permission was freely 
granted. From each, ten or twelve were se- 
lected at random, and carried to the testing- 
room; and from these the lamps which were 
compared were taken. They were placed 
upon the photometer-bar at a distance of a 
hundred and twenty-five inches from each 
other, and a system of switches was arranged, 
so that the galvanometers could be quickly 
connected with one or the other. Measures 
of current and electromotive force were made 
