lation of the fixed stars when they have risen 12 or 15 

 above the horizon, give the vault of heaven a peculiar char- 

 acter of mild effulgence and repose. I have already referred 

 in many of iny delineations of tropical scenery to this charac- 

 teristic, which was also noticed by the accurate observers La 

 Condamine and Bouguer, in the Peruvian plains, and by 

 Garcin,* in Arabia, India, and on the shores of the Persian 

 Gulf (near Bender Abassi). 



As the aspect of the starry heavens, in the season of the 

 serene and cloudless nights of the tropics, specially excited 

 my admiration, I have been careful to note in my journals 

 the height above the horizon at which the scintillation of the 

 stars ceased in different hygrometric conditions. Cumana 

 and the rainless portion of the Peruvian coast of the Pacific, 

 before the season of the garua (mist) had set in, were pecul- 

 iarly suited to such observations. On an average, the fixed 

 stars appear only to scintillate when less than 10 or 12 

 above the horizon. At greater elevations, they shed a mild, 

 planetary light; but this difference is most strikingly per- 

 ceived when the same fixed stars are watched in their grad- 

 ual rising or setting, and the angles of their altitudes meas- 

 ured or calculated by the known time and latitude of the 

 place. In some serene and calm nights, the region of scin- 

 tillation extended to an elevation of 20 or even 25 ; but a 

 connection could scarcely ever be traced between the differ- 

 ences of altitude or intensity of the scintillation and the hy- 

 grometric and thermometric conditions, observable in the low- 

 er and only accessible region of the atmosphere. I have ob- 

 served, during successive nights, after considerable scintilla- 

 tion of stars, having an altitude of 60 or 70, when Saus- 

 sure's hair-hygrometer stood at 85, that the scintillation en- 

 tirely ceased when the stars were 15 above the horizon, al- 

 though the moisture of the atmosphere was so considerably 

 increased that the hygrometer had risen to 93. The intri- 

 cate compensatory phenomena of interference of the rays of 

 light are modified, not by the quantity of aqueous vapor con- 

 tained in solution in the atmosphere, but by the unequal dis- 

 tribution of vapors in the superimposed strata, and by the 

 upper currents of cold and warm air, which are not percept- 

 ible in the lower regions of the atmosphere. The scintilla- 

 tion of stars at a great altitude was also strikingly increased 

 during the thin yellowish red mist which tinges the heavens 



* Lettre de M. Garcin, Dr. en Med. a M. de RSavmur, in Hist, de 

 TAcadtmie Royale des Sciences,' Annie 1743, p. 28-32. 



