500 M I R 
between them, the water appearing very diftindtly. As 
the (hip was coming up towards the horizon, I continued 
to oblerve it, in order to difcover the variations which 
might follow ; and found, that, as the fhip approached the 
horizon, the image C gradually difappeared, and at laft jt 
vanifhed ; after that, the image v w of the fea difappeared ; 
and during this time the image B defcended; but the 
Ihip did not rife fo near to the horizon as to bring the 
main-malts together. Had I diredted my telefcope to the 
fame point of the horizon a little fooner, 1 Ihould have 
feen the two images before the Ihip iti'elf was vilible. In 
fadt, the images were vifible when the whole Ihip was 
actually below the horizon ; for, from the very fmall part 
of the malt which was at firlt vilible, that part mull then 
have been below' the horizon, and appeared above it by 
the ufual refradtion ; the altitude of «, above the horizon, 
having then been much lefs than the increafe of altitude 
which arifes from the common horizontal refradtion. 
The difcovery of Ihips in this manner might, in fome 
cafes, be of great importance; and on fuch occafions it 
might be worth while to appoint proper perfons to make 
oblervations for that purpofe. 
“ The cliffs at Calais being very vifible, I diredted my 
telefcope towards them, in order to examine w'bether 
there was any thing unufual in their appearance ; when I 
obferved an image of the cliffs, above the cliffs themfelves, 
together with an image of the fea feparating them, as is 
reprefented in fig. 6. in which, xy reprefents the horizon 
of the fea, A B the cliffs, a b their image, and v w the 
image of the fea between them : the depth of a b was much 
lefs than that of AB. It is probable, however, that viv 
might not be the image of the fea immediately adjoining 
the cliffs, but a partial elevation of the fea at fome dillance 
from them; and that the image v w might intercept fome 
part of the image a b, which would otherwile have been 
vifible; we mull not therefore conclude that the image ab, 
fo far as it appeared, was lefs than the correfponding part 
of the objedt. From the memorandums which I made at 
the time of obfervation, I do not find that I examined the 
appearance of the cliff A B, and its image a b; which, had 
there at that time been any linking marks in them, would 
have determined whether the objedt and its image were 
of the fame magnitude. The image ab was however 
eredt; the boundaries on the top of A B and a b agreeing 
together. Having examined this for fome time, and taken 
a drawing of the appearance, during which I could dif¬ 
cover no variation, I diredted my telefcope to other ob- 
jedts ; and, on turning it again to the fame cliffs, after the 
lpaee of about fix or 1’even minutes, the images, ab and vw 
■were vanilhed ; but, examining them again foon after, the 
images were again vifible, and in every refpedt the fame as 
they appeared before. A lhort time after, they difap¬ 
peared, and did not appear any more. 
“ Soon'after the above appearances, I obferved a Ihip C, 
with the hull below the horizon xy, palfing by the fame 
cliffs AB ; an inverted image D of which appeared againll 
the cliffs, as reprefented in fig. 7. The Ihip was in mo¬ 
tion,'and remained at the fame dillance on the other fide 
of the horizon : I continued my obfervations on it till it 
had paffed the cliffs for a confiderable dillance, but there 
was no change of appearance. The cliffs were illuminated 
by the fun, and appeared very diftindtly ; but there was 
no image above, as in the lalt cafe. Continuing to ob- 
ferve the fame cliffs A B, fig. 8, I foon after difcovered 
two partial elevations m, n, of the lea, by the unufual re¬ 
fradtion ; they changed their figures a little, and difap¬ 
peared in the place where they firfl appeared, and were 
equally diltindl in every part. 
“ I have here deferibed all the different phenomena 
which I obferved from the unufual refradtion, of moll of 
which I faw a great many inltances. Every fhip which 
I obferved on the other fide of the horizon of the fea, ex¬ 
hibited phenomena of the kind here delcribed, but not in 
the fame degree. Of two Ihips which, in different parts, 
were equally funk below the horizon, the inverted image 
AGE. 
of one would but juft begin to appear, while that of the 
other would reprelent nearly the whole of the Ihip. But 
this I obferved in general, that, as the fhip gradually de- 
icended below the horizon, more of the image gradually 
appeared, and it afeended ; and tire contrary when the 
Ihips were afeending. On the horizon, in different parts, 
one fhip would have a complete inverted image*; another 
would have only a partial image ; and a third would have 
no image at all. The images were in general extremely 
well defined ; and frequently appeared as clear and (harp 
as the (hips themfelves, and of the lame magnitude. Of 
the fhips on this fide of the horizon, no phenorhena of 
this kind appeared. There was no fog on our coaft ; and 
the Ihips in the Downs, and the South Foreland, exhibited 
no uncommon appearances. The ufual refraction at the 
fame time was uncommonly great; for the tide was high; 
and at the very edge of the water I could fee the cliffs at 
Calais a very confiderable height above the horizon ; 
whereas they are frequently not to be leen in clear weather 
from the high lands about the place. The French coall 
p.llo appeared both ways, to a much greater dillance than 
I ever obferved it at any other time ; particularly towards 
the ealt, on which part alfo the unufual refradtion was the 
llrongelt. 
“ During the remainder of my ftay at Ramfgate, which 
was about five weeks, I continued daily to examine all 
the Ihips in fight; but I difcovered no phenomenon limilar 
to thole above deferibed. The phenomena of the ihip 
obferved by Mr. Huddart differed altogether from thole 
above deferibed, as the inverted image which he obferved 
was below the fhip iti'elf'. An appearance of this kind I 
obferved on the 17th of Auguft, about half an hour after 
three in the afternoon, of which fig. 9 is a reprefentation. 
The real Ihip is reprefented by A, and the image by B ; 
cr, m v, the hulls; st the flag, and wx its image, juft 
touching it, with the fea xy below. Between- the two 
hulls, fome faint dark fpots and lines appeared, but I could 
not difcover what they were the reprel'entatives of. The 
vefl'el, at the time of this appearance, was not quite come 
up to the horizon ; and, as it approached it, the image 
gradually diminilhed, and totally difappeared when the 
Ihip arrived at the horizon. 
“ It remains now, that we inquire into the caufes which 
might produce the very extraordinary effedts above re¬ 
lated. From the phenomena, we are immediately led to 
the nature of the path of the rays of light to produce 
them ; and we may conceive, that the air may poliibly be 
in fuch a Hate as will account for the unufual track which 
they mull have deferibed. For, let bz, fig. 10, be the fur- 
face of the fea; ab an objedt; E the place of the eye.; 
«rE, bs E, the progrefs of two rays, by the ufual refrac¬ 
tion, from the extreme parts of the objedt to the eye ; to 
thefe curves draw the tangets E<z', E b‘, and a! b‘ will be 
the image of the objedt, as ufually formed. Now,* if we 
take the cafe reprefented in fig. 4, let a!' b" reprelent the. 
inverted image, and a " 1 b the eredt image ; join a" E, 
the a'" E, and b" E, b"‘ E, and thele lines mult refpedtively 
be the diredtions of the rays entering the eye from a and b, 
in order to produce the images «" b" and a'" b'"; hence 
thele lines mult be tangents at E, to the curves which 
are deferibed by the rays of light; let therefore an E, 
am E, bv E, biv E, be the curves deferibed. We have 
therefore to aflign a caufe which may bring rays paffmg 
above the rays a r E, b s E, to the eye at E. Now, if there 
were no variation of the refradrive power of the air, a ray 
of light paffmg through it would deferibe a ftraight line; 
therefore the curvature of a ray of light paffmg through 
the atmofphere depends on the variation of the refradtive 
power of the air. If, therefore, we l’uppofe the air lying 
above ar E, to vary quicker in its refradtive power than 
the air through which ar F. paffes, the curvature of a ray 
proceeding above that of «rE, will be greater than the 
curvature of ar E; and on this principle, we may conceive 
that a ray may delcribe the curve an E ; and, in like man¬ 
ner, if a quicker variation of refradtive power lhculd take 
place . 
