z 
FEBRUARY 15, 1884.] 
sioners, four systems of electric lighting were 
entered for competition. The Thomson-Hous- 
ton electric-lighting company submitted a sys- 
tem of are-lighting ; the Edison electric-lighting 
company, a system of incandescent lighting ; 
and the U.S. electric-lighting company offered 
a system of arc-lighting, and also one for incan- 
descent lighting. Several things conspired to 
make the tests less complete in some respects 
than was desired by those interested. The 
members of the jury were all engaged in pro- 
fessional work, and were therefore unable to 
devote their entire time to the tests. The 
dynamometers used were built after the expo- 
sition opened, and were not completed until 
after many vexatious delays. One of them, 
that upon which the most reliance was placed, 
was of a form recently devised, and the prin- 
ciples of which had never been realized in 
practice, except in an experimental model con- 
structed by its inventor. Its construction on 
a large scale necessarily involved a good deal 
of experimentation. In spite of these delays, 
the jury was enabled to begin regular work on 
the evening of Sept. 25, and to make, dur- 
ing the succeeding ten days, such tests of the 
most important features of the various systems 
as to justify them in making the awards, what- 
ever difference might have existed in reference 
to minor points, which, for lack of time, were 
not thoroughly investigated. 
The plan adopted was substantially that 
upon which nearly all similar trials have been 
conducted. The energy consumed by the dy- 
namo was measured by means of the dyna- 
mometers, and the electrical energy in the 
circuit was determined by well-known methods. 
This gave the efficiency of the machine as 
a generator. The illuminating-power of the 
lamps was compared, and at the same time 
the electrical energy which they consumed was 
measured. A combination of the results ob- 
tained by these two processes gives the rela- 
tive illuminating-power per unit of energy 
consumed by the dynamo, which represents 
the relative commercial efficiencies of the sys- 
tems. The measurements made, therefore, 
were of three kinds, — dynamometric, electric, 
and photometric; and they will be considered 
in the order mentioned. 
Dynamometric measurements. 
Two separate dynamometers were simulta- 
neously employed in measuring the mechanical 
energy expended in running the armatures of 
the four dynamos which were tested. They 
have been called, from the manner of their 
SCIENCE. 
175 
operation, the ‘belt’ and ‘ cradle’ dynamome- 
ters respectively. 
The belt dynamometer has been frequently 
employed before; and its manner of operation 
is explained by Dr. Hopkinson in Hngineering, 
vol. 27, p. 403, where he gives a figure of it, 
and the formulae used by him in determining 
the power expended in certain electric-light 
tests. These formulae tacitly assume that the 
belt is perfectly flexible and without weight: 
for otherwise terms must be introduced into 
the formulae to take account of the differences 
of tension in the belt caused by passing the 
dynamometer-pulleys, and the centrifugal force 
generated in the belt as it leaves the various 
pulleys. The velocity of the belt being great, 
it is more than probable that such terms are 
required in order to deduce accurate results 
from this form of dynamometer. 
Under these circumstances the computation 
of the power expended from the observations 
of the belt dynamometer by the theory as at 
present known was wholly unsatisfactory, giv- 
ing results, in all except the first few tests, con- 
siderably less than the truth, and in some cases 
less than the electrical power in the circuit. 
The cradle dynamometer, however, gave re- 
sults of a much more satisfactory character. 
The principle of this dynamometer is a recent 
invention of Professor Brackett of Princeton, 
N.J., and, owing to its novelty and great ac- 
curacy, merits a somewhat minute description. 
It was built at the machine-shops of Messrs. 
Lane & Bodley, Cincinnati, under the super- 
intendence of Mr. Laidlaw, from designs made 
by Mr. Eddy, to whom is due the arrangement 
of its various parts. 
It consisted of a substantial platform, cc, 
fig. 1, seven feet long by four and a half feet’ 
wide, hung at each end by iron rods, ee, from 
an axis consisting of a short piece of two-and- 
a-half-inch shafting, a, which rested upon a 
supporting-girder, gg. Fig. 1 represents the 
framework, etc., at one end of the platform, to 
facilitate raising and lowering the girder gg, 
which carried the platform cc by means of the 
jackscrews jj, upon which gg rests. The up- 
rights bb are guides passing through the open- 
ings 6b, shown in fig. 2, which is intended to 
represent the ground-plan of gg, and adjacent 
parts of the cradle. Each girder, gg, was com- 
posed of two planks, held at a distance of 
three inches apart by blocks, dd, and bolted 
firmly together. The rods ee passed between 
the planks gg, and were forged to an eye 
which fitted the axis a The axis a rested 
upon pieces of smooth boiler-plate in the 
upper surface of gq. 
