5y COSxMOS. 



they had seen stars at hroad daylight : not havmg myself 

 been a witness of this phenomenon, I did not pay much at- 

 tention to it, but the unanimous assertions of the guides left 

 me no doubt of its reality. '''' It is essential, however, that 

 t}i3 observer should be placed entirely in the shade, and that 

 he should even have a thick and massive shade above his 

 head, since the stronger light of the air would otherwise dis- 

 perse the faint image of the stars." These conditions are 

 therefore nearly the same as those presented by the cisterns 

 of the ancients, and the chimneys above referred to. I do 

 not find this remarkable statement (made on the morning of 

 the 2d of August, 1787) in any other description of the Swiss 

 mountains. Two well-informed, admirable observers, the 

 brothers Hermann and Adolph Schlagentweit, who have re- 

 cently explored the eastern Alps as far as the summit of the 

 Gross Glockner (13,016 feet), were never able to see stars 

 by daylight, nor could they hear any report of such a phe- 

 nomenon having been observed among the goatherds and 

 chamois-hunters. Although I passed many years in the 

 Cordilleras of Mexico, duito, and Peru, and frequently in 

 clear weather ascended, in company with Bonpland, to ele- 

 vations of more than fifteen or sixteen thousand feet above 

 the level of the sea, I never could distinguish stars by day- 

 light, nor was my friend Boussingault more successful in his 

 subsequent expeditions ; yet the heavens were of an azure so 

 intensely deep, that a cyanometcr (made by Paul of Geneva) 

 which had stood at 39'^ when observed by Saussure on Mont 

 Blanc, indicated 46° in the zenith under the tropics at ele- 

 vations varying between 17,000 and 19,000 feet.f Under 

 the serene etherially-pure sky of Cumana, in the plains near 

 the sea-shore, I have frequently been able, after observing an 

 eclipse of Jupiter's satellites, to find the planet again with 

 the naked eye, and have most distinctly seen it when the 

 sun's disk was from 18° to 20° above the horizon. 



The present would seem a fitting place to notice, although 

 cursorily, another optical phenomenon, which I only observed 

 once during my numerous mountain ascents. Before sunrise, 

 on the 22d of June, 1799, when at Malpays, on the decliv- 

 ^ ity of the Peak of Tenerifle, at an elevation of about 11,406 

 feet above the sea's level, I observed with the naked eye 



* Consult Saussure, Voyage dans les Alpcs (Neuchatel, 1779, 4to), 

 torn, iv., § 2007, p. 199. 



t Humboldt, Essai sur la GiograjMe des Plantcs, p. 103. Compare 

 also my Voy. aux Htgions Equinox, torn, i., p. 143, 248. 



