PAST AND COMING- TRANSITS AND ARCTIC EXPLORATIONS. 271 
nomenon will only be so much the more clearly recognised ; 
and it must be quite impossible to experiment on this pheno- 
menon as on the 44 black drop,” seeing' that we cannot represent 
by a model the action of an atmosphere whose real extent and 
refractive power are unknown to us. 
Nor can photography be of any use in this matter ; for the 
more perfect the photographic arrangements, the more exactly 
will the optical difficulty be reproduced. Indeed, in the pho- 
tographic records of contact, during the recent transit, a pecu- 
liarity appears, which seems, of itself, to introduce an abso- 
lutely insuperable difficulty. It would seem that the sun, which 
photographs itself, is slightly larger than the sun we see ; in 
other words, that the gaseous matter of the sun emits light- 
waves, producing that form of chemical action on which photo- 
graphy depends, from layers extending to a greater height than 
those which emit light-waves recognisable by the eye in full 
sunlight. Janssen, at least, adopts this interpretation of the 
fact, that at ingress, as observed at Nagasaki, the planet appeared 
still attached to the solar limb, while the photographs taken, 
second by second, showed Yenus already somewhat advanced 
on the sun’s disc. It matters, in truth, very little whether this 
explanation is correct or not, seeing that the observed fact, 
however explained, indicates a discrepancy between the optical 
and the photographic records of contact which must prevent 
our placing reliance on either. 
If we abandon contact observations, but one resource seems 
to be left. It is manifest that all methods have for their real 
object the determination of the chord of transit followed by 
Yenus, as seen from different stations. When reliance was 
placed on Halley’s method, for instance, although the element 
observed was the duration of transit, the element deduced was 
the length — and with the length the position — of the chord of 
transit. When reliance was placed on Delisle’s method, the ele- 
ment observed was the epoch either of ingress or egress ; but the 
element deduced was the position of the ingress or egress end 
of the chord of transit, and therefore of that chord itself. The 
great difficulty in all other methods of determining the position 
of the chord of transit resides in the fact that the exact position 
of Yenus on the sun’s disc (not merely her distance from the 
centre, but her bearing from the centre, referred to some fixed 
line on the sun’s disc,) must be determined for a precisely- 
timed moment. So that a double difficulty is introduced ; first, 
the observations necessary to determine Yenus’s position require 
time ; secondly, the exact longitude of the station should be 
known with as great accuracy as for Delisle’s method. But the 
central part of the chord of transit — the part, namely, where the 
