EXPLANATION OF THE PLATES. 



Plate I.— This Plate combines in its construction the results of 1,159,353 separate 

 observations on the force and direction of the wind, and a little upwards of 100,000 

 observations on the height of the barometer at sea. The wind observations embrace 

 a period of eight hours each, or three during the twenty-four hours. Each one of 

 the barometric observations expresses the mean height of the barometer for the day ; 

 therefore each one of the 100,000 may itself be the mean of many, or it may be only 

 one. Suffice it to say, that 83,334 of them were obtained by Lieutenant Andrau 

 from the logs of Dutch ships during their voyages to and fro between the parallels 

 of 50° N. and 36° S. ; that nearly 6,000 of them were made south of the parallel of 

 36° S., and obtained from the log-books at the Observatory in Washington ; that for 

 the others at sea I am indebted to the observations of Captain Wilkes, of the Explor- 

 ing Expedition, of Sir James Clark Ross, on board the " Erebus" and " Terror" in 

 high southern latitudes, and of Dr. Kane in the Arctic Ocean. Besides these, others 

 made near the sea have been used, as those at Greenwich, St. Petersburg, Hobart 

 Town, etc., making upwards of 100,000 in all. This profile shows how unequally the 

 atmosphere is divided by the equator. 



The arrows ivithin the circle fly with the wind. They represent its mean annual 

 direction from each quarter, and by bands 5° of latitude in breadth, and according 

 to actual observation at sea. They show by their length the annual duration of the 

 wind in months. They are on a scale of one twentieth of an inch per month, except 

 the half-bearded arrows, which are on a scale twice as great, or one tenth of an inch 

 to the month. It will thus be perceived at a glance that the winds of the longest 

 duration are the S.E. trades, between the parallels of 5° and 10° south, where the 

 long-feathered arrows represent an annual average of ten months. 



The most prevalent winds in each band are represented by full-feathered arrows ; 

 the next by half-feathered, except betw^een the parallels of 30° and 35° N., where the 

 N.E, and S.W. winds, and between the parallels of 35=^ and 40, where the N.W. and 

 S.W. winds contend for the mastery as to average annual duration. 



The row^s of arrows on each side of the axis, and nearest to it, are projected with 

 the utmost care as to direction, and length or duration. 



The feathered arrows in the shading around the circle represent the crossing at 

 the calm belts, and the great equatorial and polar movements by upper and lower 

 currents of air in its general system of atmospherical circulation. 



The small featherless and curved arrows, nqrs, on the shading around the circle, 

 are intended to suggest how the trade-winds, as they cross parallels of larger and 

 larger circumference on their way to the equator, act as an undertow, and draw 

 supplies of pure air down from the counter-current above ; which supplies are 

 required to satisfy the increasing demands of these winds : for, as they near the 

 equator, they not only cross parallels of larger circumference, but, as actual observa- 

 tions show, they also greatly increase both their duration and velocity. In like 

 manner, the counter-trades, as they approach the poles, are going from latitudes 

 where the parallels are larger to latitudes where the parallels are smaller. In other 

 words, they diminish, as they approach the poles, the area of their vertical section ; 



