974 
references in the reports of the British Associa- 
tion Committee on Meteorological Photography 
and other places. 
So long ago as August, 1889, I had shown 
conclusive proofs that the phenomenon was 
not due to any difference in the refrangi- 
bility of the light of the spark and that of the 
reversing light. I showed that the light of the 
sparks themselves could effect reversal of the 
images of others. Perhaps I may be allowed 
to quote from the paper. 
‘‘A plate was then exposed in the camera to 
a series of sparks, then to the direct light of 
more sparks without the interposition of the 
lens, and finally to a second set of sparks. The 
images of the first set show reversal while those 
of the second are direct. 
‘Next a plate was exposed to one set of 
sparksand without removing it from the camera 
the light of some more was diffused by holding 
a sheet of ground glass in front of the lens. 
Finally a second set of sparks was photographed. 
The results were similar.’’ 
These two experiments enabled me to repro- 
duce the phenomenon of a bright flash crossing 
a dark one, and the reversal of one flash by the 
diffused glare of another. 
In the second place plates were exposed to a 
number of spark images and then to a tolerably 
pure spectrum. The result was reversal in all 
parts, and by varying the length of exposure to 
the spectrum it was shown that the reversing 
power was simply proportional to the direct 
actinic power, maximum reversal occurring 
when the direct actinic effect of the reversing 
light was equal to that of the spark images. I 
was, therefore, entitled to sum up thus: 
‘(Differences of refrangibility, therefore, do 
not seem to lie at the root of the matter. 
Neither can a difference of intensity be the 
cause ofthe reversal, for the less intense the 
light of the spark the more easily is its image 
inverted. It seems to me that the extreme 
shortness of the exposure to the electric spark 
may be the explanation.’’ 
A similar conclusion was indicated by the 
fact that the image of a spark very much out 
of focus did not lose the property of reversi- 
bility. 
But how were we to account for the experi- 
SCIENCE. 
[N. 8. Vou. X. No. 261. 
\ 
ments showing that the spark images could be 
reversed by the light from other sparks? Was 
it possible that objects illuminated by these re- 
versing sparks (card, objects in the room or 
ground glass) reflected or modified the light 
sufficiently to change its action on the photo- 
graphic film ? 
I also tried to imitate the phenomena by brief 
exposure to other luminous objects trying in 
turn slits illuminated by gas, lime-light, mag- 
nesium and sunlight. I had no are lamp avail- 
able then. Here Professor Wood has done 
better, my results were nil and I congratulate 
him on his success. 
However I should like to suggest that it is 
just possible that light from a source whose ex- 
citement is electrical may differ from other 
kinds of light in some manner at present un- 
known and that it is not safe to regard it as 
proved that the time element is the only one in- 
volved until the phenomena have been repeated 
without employing electricity at all. 
For ten years the facts have been before the 
world. They were partly verified eight or nine 
years ago by Mr. Shelford Bidwell and it is 
highly satisfactory to find them verified again 
in so many particulars, by another physicist 
who has reached the same conclusions by means 
of somewhat different experiments. 
ARTHUR W. CLAYDEN. 
RoyaL ALBERT MEMORIAL COLLEGE, 
EXETER, 
SOCIOLOGY AND PSYCHOLOGY. 
To THE EDITOR OF SCIENCE: The relation of 
sociology to psychology suggested by Professor 
F. H. Giddings in his article, ‘Exact Methods 
in Sociology’ (Popular Science Monthly, Decem- 
ber, 1899), is so misleading that it demands a 
word of protest from the psychologist. We 
must regard it as a capital mistake when any 
sociologist tries to make his science a means of 
measuring psychological quantity. Thus, when 
Professor Giddings (p. 155) would measure the 
‘intelligence’ of societies by comparative sta- 
tistics of literacy, for instance, he overlooks 
such facts as these: that mere reading, like 
talking, signifies little—the main point being 
what is read, whether Hegel or the yellow 
journal—and that how much is understood 
