270 PROCEEDINGS OF THE AMERICAN ACADEMY. 



cumulus and alto-stratus clouds were seen floating in it at a height of 

 perhaps 4000 or 5000 meters, and from them light sprinkles of rain fell 

 occasionally. On the Peak of Teneriffe, in passing into this upper 

 current, a rise of temperature was noted, which was less than that 

 encountered above the surface-trade. 



The winds at great heights in and near the trade-wind region are 

 given in Table II. They were obtained by pilot-balloons launched 

 from the islands, excepting the one from the yacht, and the last 

 figures for each ascension show the maximum height at which the bal- 

 loon was sighted. The means of direct observations of wind-direction 

 and velocity at definite heights, obtained during the ascents and de- 

 scents of the peaks on the tropical islands of Teneriffe and Fogo, 

 and the drift of clouds passing at estimated heights above these 

 mountains, are given. Observations of the direction and velocity of 

 the wind, obtained in two kite-flights south of the trade-wind region, 

 complete the table. In that portion of the Atlantic investigated by the 

 Franco- American Expedition, the atmospheric circulation was found 

 to be as follows: (1) North of Madeira, and near the Azores, the 

 upper winds, as was already known by observations of clouds, are chiefly 

 from west and northwest, this region being generally to the north of 

 the barometric maximum over the ocean and beyond the zone of the 

 trades. (2) The winds blowing towards the equator are from north- 

 east to east in the lower region, and generally from northwest to 

 northeast above 1000 meters. (3) The return currents from the 

 equator, or anti-trades, are formed by winds having a southerly com- 

 ponent, being generally southwest in the latitude of the Canaries, and 

 southeast near the Cape Verdes, thus showing the influence of the 

 earth's rotation. The law of the vertical succession of winds, as for- 

 mulated by Abercromby, ^ namely, a shifting in the northern hemi- 

 sphere of the upper winds to the left-hand, when one's back is 

 towards the wind, is found not to hold true always, the right or 

 left-handed rotation depending upon the origin of the wind, and, 

 presumably, upon the distribution of the pressure at high levels. 



The vertical distribution of temperature and relative humidity re- 

 vealed by these observations up to a height of 4000 meters is nearly 

 the same as that found by Professor Hergesell during the cruises of 

 the " Princesse- Alice," in 1904 and 1905.^ Most of his observations 



" Nature, 36, 85. 



• Comptes Rendus de I'Academie dcs Sciences, 30 Janvier, 1005; INIeteoro- 

 logische Zeitschrift, November, 1905 ; Bulletin du Musee Oce'anograpiiique de 

 Monaco, 30 novembre, 1905. 



