* 
~~ 
14 
frequent results ; and this is often accompanied 
by a softening to a barely coherent sand or 
clay. Erroneous conclusions are thus frequent- 
ly drawn from surface indications, as to the na- 
ture of the underlying rock. 
The subject of the decay of rocks has re- 
SCIENCE. 
> 
%, 
ees =Ne 
—-— Hf, 
Ss El 
SB 
— 
ene 
z 
‘ 
[Vou. IIL, 
Sniith, J. A. Tanner, M.D., and H. W. Eaton, 
Ph.D., Louisville — was not appointed till about | 
three weeks before the close of the exposition : 
hence thorough tests were impossible. 
As the U. S. company did not enter into — 
the contest, there was no competition on the 
r 
Fie. 7.— Broken conglomerate outcrop. 
cently been admirably treated by Dr. T. Sterry 
Hunt,’ chiefly with regard to the crystalline 
rocks ; and it deserves to be further studied, in 
the case of these more recent rocks, from its evi- 
dent importance in chemical geology, its inter- 
esting and well-known relation to topography, 
and its economic bearing. ArrHuR WINSLOW. 
Pennsylvania geological survey. 
ELECTRIC LIGHT TESTS. AT THE 
LOUISVILLE EXPOSITION. 
Tue display of electric lights at the Louis- 
ville exposition, as to number, was the great- 
est ever made in and around one building. 
The number of lights used varied somewhat, 
but the average was about as follows : — 
| Incandescent Are- 
lights. | lights. 
| 
Edison isolated lighting company . 4,600 - 
U.S. electric light company ... . | 210 29 
Fort Wayne Jenney electric light com- 
[PEAT SS Ge Oe SOO Ce tic Tea ~ 100 
Thomson Houston electric light com- 
PAE tere diver aye, Yes! ol eeliie) aitee tyaneke - 36 
The jury — consisting of Benjamin Rankin, 
Louisville; W. W. Weaver, Chicago; Charles 
1 The decay of rocks geologically considered. By T. Sterry 
he LL.D., F.R.S. American journal of science, September, 
incandescent lights. However, the following 
tests were made: connection was made with a 
circuit containing 315 lights at what was con- 
sidered an average point in the circuit ; and fif- 
teen lamps, five of them new and the balance 
selected systematically from the circuit while 
lighted, were tested in a specially constructed 
photometer-room while indicator-cards were 
being taken from the engine. 
A Bunsen photometer with a twelve-foot bar 
was used, and the horizontal intensity deter- 
mined with the carbon at an angle of 45°. 
The intensity of the (nominally) 16-candle 
lights varied from 12 to 19.66 candles, aver- 
aging 13.77 candles; and the average horse- 
power was 32.50. These figures give 9.70 
lights, or 133.57 candles, per mechanical horse- 
power. 
The action of the automatic regulator was 
*then tested with a light in the photometer, 
first 50 and then 100 lights being thrown off 
and on. In one of the six cases the variation 
was 1.23 candles, but in all the others it was 
less than .66 of a candle. Only a momentary 
flicker was noticed as the lights were thrown 
off and on. Pie 
The jury reported as follows: ‘‘ The tests 
ned 
of the Edison system are most satisfactory as 
to the efficiency of the various appliances, the 
steadiness of the light produced, and the 
