90 
L. Schwendler —On Electric Light Measurements. [March, 
Endeavours to cut up the electric light into a large number of small 
lights, although of great interest, must, I think, invariably result in en¬ 
gineering failure, as nobody could afford to pay for the luxury received. 
I have succeeded in working three Serrin Lamps connected up succes¬ 
sively in the same circuit of a dynamo-electric machine, but found the 
loss of light very great. 
Having satisfied myself of the difficulty and impracticability of the 
division of the electric light, I tried :— 
Diffusion, i. e., a few large lights (each light produced by one machine), 
are placed at different points of the space, and by optical means the light 
is diffused over a large area. 
This method I find perfectly practicable. 
There is naturally also a large amount of light lost (by absorption), 
but this loss will bear a constant ratio to the total light produced, nay 
probably may decrease with the intensity. 
The actual plan by which I propose to do it, and have done it 
during the trial, is to construct a Silver ed-glass reflector in which a power¬ 
ful electric light burns, throwing direct and reflected rays uj) to a white 
ceiling or any other convenient white surface. A number of such arrange¬ 
ments is to be put up .in the most convenient places, and where they have 
the greatest effect. 
The form and size of each reflector will depend on the locality where 
it is to he used. 
In future constructions of railway stations the lighting up by the 
electric light should he taken into account. 
IV.— WnAT REPAIRS ARE REQUIRED IN COURSE OP TIME, AND ARE THEY 
OP SUCH A NATURE THAT AN ORDINARY MECHANIC CAN EXECUTE 
THEM ? 
Any repairs which during this extended trial had to be made were 
entirely limited to the commutator, and the brushes of the dynamo- 
electric machines. These repairs were exceedingly simple, and any native 
mechanic would be able to execute them. 
I consider the Commutator or Dynamo-Electric Machines A and B 
the best. 
Practice will however answer this question much more effectively than 
any experimental trial can do. 
The commutators are to be often cleaned, and the brushes are to be 
well set. 
Although the electric lamps look rather formidable, they are never¬ 
theless very simple apparatus, not easily getting out of order if well 
adjusted. 
