NOVEMBER 17, 1899. ] 
slightly fogged, either by exposing it to feeble 
diffused light or by leaving the lens of the 
camera open, the flash on development came 
out darker than the background. If, how- 
ever, the plate was fogged before the image 
of the flash was impressed, it came out 
brighter than the background, as in the 
ordinary pictures of lightning. I refer to 
the appearance in the positive print in each 
ease. This is quite different from ordinary 
reversal due to the action of a very intense 
light, for the order in which the lights are 
applied is a factor, and the phenomenon lies 
wholly in the region of under exposure. I 
repeated Mr. Clayden’s experiment, and 
obtained dark flashes without any diffi- 
culty. 
The effect cannot, however, be obtained 
by impressing an image of a filament of an 
incandescent lamp on a plate and subse- 
quently fogging the plate. Clearly there is 
something about the light of the electric 
spark which is essential to the production 
of the reversal. It is not intensity, however, 
for I found it was impossible to obtain re- 
versed images or bright sparks with the lens 
wide open. 
Fig. 1 shows a series of spark images, 
some normal, some partly reversed and 
others wholly reversed. The sparks are 
those of a large inductorium, with a good 
sized Leyden jar in circuit. The sparks 
were all of equal intensity, but after each 
discharge the iris diaphragm of the lens 
was closed a little. It willbe seen that the 
borders of the bright sparks are reversed. 
In some the image is reversed with the ex- 
ception of a narrow thread down the core. 
The images were impressed in succession on. 
the plate by moving it in the camera. A 
plate holder was dispensed with, an opening 
being made in the ground glass back by re- 
moving a strip a few centimeters wide. 
The plate was held against this opening 
and a large number of exposures made ina 
few moments. Of course, the room was in 
SCIENCE. Tag 
total darkness. After exposure the plate 
was exposed to the light of a candle for a 
second or two and then developed. 
In this series of pictures it will be seen 
that the edges of the bright images of the 
sparks are reversed, the intensity on the bor- 
der of the image being less than at the core. 
As the intensity of the spark becomes less 
and less, the bright central core dwindles 
down to a mere thread and eventually dis- 
appears, the spark’s image being feeble 
enough to reverse over its entire area. 
This explains why the dark lightning 
flashes are usually:ramifications of the main 
flash. The ramifications are less brilliant 
discharges and reverse, while the main one 
is too bright to cause the effect. 
The first thing that occurs to one is that 
it may be some peculiar radiation which the 
spark emits, which is wanting in the light 
coming from other bodies. If a small pho- 
tozgraphic plate is partly screened by a 
piece of black paper and illuminated by the 
light of a small spark ata distance of two 
or three feet, and a similar plate, screened 
in the same manner, is illuminated for a 
moment by candle light of sufficient inten- 
sity to produce the same amount of black- 
ening on development, we shall have the 
means of showing that the spark light 
differs in its action on the plate from that 
of the candle. If these two plates before 
development be half screened in a direc- 
tion at right angles to the former one, 
and exposed to the light of a candle for a 
second or two, the part of the plate which 
has been illuminated by spark light plus 
candle light, does not become as black on 
developing as the part which has only 
received candle light, whereas the part 
which has twice been exposed to candle 
light is blacker than that which has been 
only exposed once. This shows that the 
light of the spark does not act in the same 
way as the light of the candle. Wherein 
does it differ? It seemed possible that the 
