of Edinburgh, Session 1881-82. 
589 
circulation, these two among other cases are given of wonderful 
atmospherical phenomena, and an experiment is detailed exhibiting 
the marvellous powers of reflected light. A bottle half filled with 
liquid is held before a concave mirror, when lo ! the inverted image 
has the liquid in its lower half ! the mirror is able to invert the 
bottle, but cannot reach its contents. With the author of this work 
the stellar nebulae become huge concave reflectors. He exhibits 
also a ' drawing of a stately ship, her sails well filled by the breeze, 
and below, heeling over symmetrically, her inverted image, and 
below that again the topmasts of a ship hull down : he tells us 
that the unseen ship is the reality, the others only her images 
caused by atmospheric refraction. It has not entered into his 
head to consider that such a breeze would blow to the winds all 
the thereunto necessary refracting surfaces. This is an improved 
edition of the second phenomenon, the subject of the present paper. 
In the 89th volume of the Philosophical Transactions , Professor 
Yince describes a phenomenon seen by him from Ramsgate; the 
appearance was that of the mast-head and upper rigging of a ship 
hull down, in the air above which an inverted image was seen, 
having above it again an erect image of the same ship. He explains 
the conditions of the atmosphere needed to produce such appearances 
in the following manner : — 
Having drawn a straight line bz to represent the surface of the 
water, raised a perpendicular ba for the ship’s mast, and placed the 
eye e above z to indicate the observer’s height above the sea-level, 
he draws two curved lines 6sE, arE for the usual paths of light 
from the hull and from the top-mast. The directions of the 
tangents to these two curves define the apparent position b' a! of the 
vessel. Here the dominant idea is the curvature of the rays ; and 
this idea has led him to show the paths as curved while the surface 
of the sea is flat. Had he intelligently considered the matter he 
must have seen that his illustration demonstrates the impossibility 
of our seeing a ship hull-down. If the path of light were more 
curved than the surface of the ocean, we should be able to see all 
round, and, with transparent air and good glasses, to view our own 
backs, twenty-four thousand miles off. The line bz, instead of 
having been drawn straight, should have risen from b to graze the 
line arE and should then have descended to z, completely 
intercepting the ray &sE. 
