Unufual Jppearances by lerrejirial Rtfra£fion> 143 



mafts became vifible; and continuing my obfervations, in order to difcover whether any or 

 what variations might take place, I found, that as the fliip defcended, the images, B, C, afcend- 

 ed ; but as the fhip did not fink below the horizon, 1 had not an opportunity of obferving at 

 what time, and in what order the image would have vanifhed, if the {hip had fo difappeared. 



Being defirous of feeing whether the fame efFe£l was produced upon the other (hips which 

 were vifible, I direfted my telefcope to another ftiip, A (fig. 2), whofe hulk was jufl in the 

 horizon, *■>•; when I obferved a complete inverted image, B, the main-mafl of which jufl 

 touched that of the (hip itfelf. In this cafe, there was no fecond image as before. The fhip, 

 A, moving upon the horizon, B continued to move with it without any variation in its ap- 

 pearance. 



The next fhip which I directed my telefcope to, was fo far on the other fide of the horizon, 

 «•;?, as juft to prevent its hulk from being feen, as is reprefcnted by A (fig. 3). And here I 

 obferved only an inverted image of part of the fliip ; the image, >r, of the topfail, with the mail 

 joining that of the fhip, the image at of the top, a of the other mafl, and the image z of the end c 

 of the bowfprit only appearing at that tim.e. Thefe images would fuddenly appear and difap- 

 pear very quickly after each other ; firfl appearing below, and running up very rapidly, fhewing 

 more and lefs of the mafts at different times as they broke out ; refembling in the fwiftnefs of 

 their breaking out, the {hooting out of a beam of the aurora borealis. As the fhip was 

 defcending on the other fide the horizon, I continued my obfervations upon it, in order to dif^ 

 cover what changes might take place ; when I found, that as it continued to defcend more of the 

 image gradually appeared, till at lafl the image of the whole fhip was completed, with their 

 main-mafls touching each other; and upon the {hip defcending lower, the image and the fhip 

 leparated ; but I obferved no fecond image as in the firft cafe : a fecond image, however, 

 might probably have appeared if the {hip had continued to defcend. 



Upon moving my telefcope along the horizon, in order to examine any other fliips which 

 might be in fight, I obferved, juft at the horizon, x y (fig. 4), the top, «, of the maft of a fhip; 

 and here an efFecl was obferved which had not been difcovered ; for there was an inverted 

 image, B, vertical to, a-, an ereft image, C, both of them very perfe(£t and well defined ; and 

 an image, v iv, of the fea between them, the vrater appearing very difliniflly. As the fhip was 

 coming up towards the horizon, I continued to obferve it, in order to difcover the variations 

 which might follow, and found that, as the {hip approached the horizon, the image C 

 gradually difappeared, and at lafl it vanifhed ; and after that, the image v w o( the fea difap- 

 peared ; and during this time the image B defcended ; but the fhip did not rife fo near to the 

 horizon as to bring the main-mafls together. Had 1 directed my telefcope to the fame point 

 of the horizon a little fooner, I fhould have feen the two images before the {hip itfelf was 

 vifible ; in fail, the images were vifible when the whole fhip was aiSlually below the horizon; 

 for from the very fmall part of the mafl which was at firfl vifible, that part mu{l then have been 

 below the horizon, and appeared above it by the ufual refraftion ; the altitude of a above the 

 horizon having then been much lefs than the increafe of altitude which arifes from the common 

 liorizental refraftions. The difcovery of fhips in this manner, might in fome cafes be of great ' 



V 2 importancei 



