1 52 Ratinrhabk Effe^ of Refranlm on Dijlant LanJ. 



the diftances, as by the laft reduflion. There is a difliculty in making this obfervatlon when 

 the fun pafles the meridian very near the zenith, as the change in the azimutli from eaft to 

 weft is too quick to allow fufficient time : nor can it be obtained by the fextant when the 

 fun palTes tlie meridian more than 30 degrees from the zenith; for I never could adjuft the 

 back obfeivation of the Hadley's quadrant with fufHcient accuracy to be depended upon*. 



II. 



Remarhablt Effecl of Terrejlrial RefrnBkn on a Di/lant Headland. ExtraEl of a Letter 



from Andrew Ellicot to David Rittenhouse, Efq. dated at Pitijburg, Nov. 5, 

 1787, concerning Obfervatioiis made at Lake Erie\. 



yja the 1 3th of laft month, while we by on the banks of Lake Erie, we had an oppor- 

 tunity of viewing that fingular phenomenon by feamen termed looming. It was preceded 

 by a fine aurora borealis on the evening of the 12th. The T3th was cloudy, but without 

 rain. About 10 o'clock in the morning, as I was walking on the beach, I difcovered 

 fomething that had the appearance of land in the direftion of Prefqu'ifle ; about noon it 

 became more confpicuous, and when viewed by a good achromatic telefcope, the branches 

 of the trees could be plainly difcovered. From three o'clock in the afternoon till dark, the 

 ■whole peninfula was confiderably elevated above the horizon, and viewed by all our com- 

 pany with admiration. There was a fingular appearance attending this phenomenon, wliich 

 J do not remember to have feen taken notice of by any writer. The peninfula was fre- 

 quently feen double, or rather two fimilar peninfulas, one above the other, with an ap- 

 pearance of water between. The feparation and coincidence was very frequent, and not 

 unlike that obferved in fliifting the index of an adjufted Godfrey's quadrant. As fingular 

 as this may appear, it is not more fo than the double refravElion produced by the Iceland 

 cryftal. The next morning Prefqu'ifle was again invifible, and remained fo during our ftay 

 at that pofition. Prefqu'ifle was about 25 miles diftant ; its fituation very low. 



The fame evening the wind began to blow briflcly from about two points weft of north. 



• The optical phenomena which relate to the conftitution of the atmofpherc have not yet been much invcfti- 

 gated. Fiom the barometrical admeafurcment of heights, as well as from experiments with the macromcter, or 

 marine barometer of HaLley (Roy, in Phil. Tranf. Vol. I.XVII.), it is afcertained that air abounding with 

 humidity is more expanfible by heat, and Icfs denfc at like temperatures than air which is more dry. And 

 from Sir Ifaac Newton's table, the refraftive power of air, in refpe£t to its denfity, compared with that of 

 water in the fame refpcift, is as 4160 to 7S45 (Optics, Part III. Prop. 10. )• Thcfe and other f.ifts fliew that 

 the rays of light muil be varicrufty affcfled, according to the nature and circumftances of evaporation, condenfa- 

 tion, and other procelles carried 00 in the air, and particularly ne?r the furface of water. That a lighter and 

 lefs refractive (Iratum of air may thus be generated near the furface of the fca to a confiderable elevation, and re- 

 main for a time in cquilibrio without afccnding, is highly probable \ and in this cafe, whenever the quicfcent 

 (late is firom local or temporary circumnanecs deftrnyed, it fecms a natural confequcnce, that fuddcn and irre- 

 gular <urrcnts of the lower air, and descending blafts of the upper, may occafion all the ftrange tlTctfts of the 

 tyf)hon and other phenomena, not well defcribed in that interelling tropical combination of lea and land to the 

 caftward of Sumatra and Malacca. The Mediterranean Sea affords limilar appearances, particularly the afto- 

 liiOiing fpcflra in the Strait of Reggio, defcribed by Kirchcr, Minazi, and others, called Fata Morgana, of which 

 ^ purpofc fliortly to give an account, with an engraving. N. 



t American Fhilof. Soc. ill. 62. 



and 



