



EKECT AND INVERTED IMAGES OF SHIPS. 209 



Sometimes there was seen only one inverted image above 

 the real ship, and this was generally the case when the 

 real ship was full in view. But when the real ship was 

 just beginning to show its topmast above the horizon, as 

 at A, Fig. 33, two aerial images of it were seen, one at B 

 inverted, and the other in its natural position at C. In 

 this case the sea was distinctly visible between the erect 

 and inverted images, but in other cases the hull of the 

 one image was immediately in contact with the hull of 

 the other. 



Analogous phenomena were seen by Captain Scoresby 

 when navigating with the ship Baffin in the icy sea in the 

 immediate neighbourhood of West Greenland. On the 

 28th of June, 1820, he observed about eighteen sail of 

 ships at the distance of ten or fifteen miles. The sun 

 had shone during the day without the interposition of a 

 cloud, and his rays were peculiarly powerful. The inten- 

 sity of its light occasioned a painful sensation in the eyes, 

 while its heat softened the tar in the rigging of the ship, 

 and melted the snow on the surrounding ice with such 

 rapidity, that pools of fresh water were formed on almost 

 every place, and thousands of rills carried the excess into 

 the sea. There was scarcely a breath of wind : the sea 

 was as smooth as a mirror. The surrounding ice was 

 crowded together, and exhibited every variety, from the 

 smallest lumps to the most magnificent sheets. Bears 

 traversed the fields and floes in unusual numbers, and 

 many whales sported in the recesses and openings among 

 the drift ice. About six in the evening, a light breeze at 

 N.W. having sprung up, a thin stratus or "fog bank, 5 ' at 

 first considerably illuminated by the sun, appeared in the 

 same quarter, and gradually rose to the altitude of about 

 a quarter of a degree. At this time most of the ships 

 navigating at the distance of ten or fifteen miles began to 

 change their form and magnitude, and when examined by 





