587 
of Edinburgh, Session 1881-82. 
the ordinary state of the air, the light would pass in the usual way 
below the curve Ba^E ; and even now, since the extraordinary 
condition does not extend below that line, it will still so pass. 
Wherefore, according to this explanation, T should have been seen 
double. Moreover the picture* tells us that light from between B 
and A reaches the eye in directions intermediate between vE and 
wE, and now we have this most wonderful fact that the air situate 
between the extreme rays has two indices of refraction, one for 
Dover Light, one for Hill Light. But more absurd still ; — the rays 
AyE, BxE are the usual paths of light, whence it necessarily follows 
that the point B is usually visible from E, and thus the phenomenon 
made clear by this explanation is : — “ that the Castle being usually 
seen clear above the top of the hill, was, on this occasion, seemingly 
imbedded in it, the ridge having been apparently raised,” exactly the 
opposite of what Vince imagined himself to be explaining. 
We may get a clear view of the matter by considering the 
broken branch in the foreground of the picture. This branch is, 
in reality, beyond yon far-off hills B ; and we have to explain how 
it can come to be seen as represented. 
One, indeed the obvious, supposition is, that the matter forming 
the hill had become translucent for Dover Light, which hypothesis 
will account also for the feebleness of the Hill Light. Or we may 
suppose that the rays of light from Dover had skimmed over the 
high ground B, had descended on the northern side until they found 
out, each the true continuation of its path, and then had come 
straight on to the eye. Of the two hypotheses, the former appears 
to me to be the simpler and the more comprehensible. 
Having thus far discussed the information given to us by Pro- 
fessor Vince, we may examine what is to be got from other sources. 
On a small map of Dover published by Messrs. Johnston, the 
Castle is shown as consisting of several detached buildings, none of 
which can be reconciled with the form and position inferred from 
the perspective drawing ; but, in the course of three quarters of a 
century, the buildings may have been changed. To the north of 
the Castle, in the direction of Eamsgate, there is marked St. Mary’s 
church in ruins, which surely should have been seen in the fore- 
ground of the Castle. 
* The proprietors of the Encyclopaedia have kindly lent the preceding 
woodcuts. 
