XXVI EXPLANATION OF THE PLATES. 



consequently there is a crowding out— a sloughing off from the lower current, and a 

 joining and a turning back with the upper current. This phenomenon is represented 

 by the small featherless and curved arrows in the periphery on the polar side of the 

 calm belts of Cancer and Capricorn. 



This dotted or shaded periphery is intended to represent a profile view of the at- 

 mosphere as suggested by the readings of the barometer at sea. This method of 

 delineating the atmosphere is resorted to in order to show the unequal distribution of 

 the atmosphere, particularly on the polar side of lat. 40*^ S. ; also the piling up over 

 the calm belts, and the depression— barometrical — over the equatorial calms and 

 cloud ring. 



The engirdling seas of the extra-tropical south suggest at once the cause of this 

 inequality in the arrangement over them of the airy covering of our planet. Except- 

 ing a small portion of South America, the belt between the parallels of 40° and 65° 

 or 70° south may be considered to consist entirely of sea. This immense area of 

 water surface keeps the atmosphere continually saturated with vapour. The specific 

 gravity of common atmospheric air being taken as unity, that of aqueous vapour is 

 about 0.6 ; consequently the atmosj)here is expelled thence by the steam, if, for the 

 sake of explanation, we may so call the vapour which is continually rising up from 

 this immense boiler. This vapour displaces a certain portion of air, occupies its 

 place, and, being one third lighter, also makes lighter the barometric column. More- 

 over, being lighter, it mounts up into the cloud region, where it is condensed either 

 into clouds or rain, and the latent heat that is set free in the process assists still 

 farther to lessen the barometric column ; for the heat thus liberated warms and ex- 

 pands the upper air, causing it to swell out above its proper level, and so flow back 

 towards the equator with the upper current of these regions. 



Thus, though the barometer stands so low as to show that there is less atmosphere 

 over high southern latitudes than there is in corresponding latitudes north, yet, if it 

 were visible and we could see it, we should discover, owing to the effect of this 

 vapour and the liberation of its latent heat, and the resulting intumescence of the 

 lighter air over the austral regions, the actual height of this invisible covering to be 

 higher there than it is in the boreal regions. 



Taking the mean height of the barometer for the northern hemisphere to be 30 

 inches, and taking the 100,000 barometric observations used as data for the construc- 

 tion of this diagram to be correct, we have facts for the assertion that in the austral 

 regions the quantity of air that this vapour permanently expels thence is from one 

 twelfth to one fifteenth of the whole quantity that belongs to corresponding latitudes 

 north — a curious, most interesting, and suggestive physical revelation. 



PiiATES II. and III. are drawings of Brooke's Deep-sea Sounding Apparatus for 

 bringing up specimens of the bottom (§ 573). 



Plate IV. (§ 723) is intended to illustrate the extreme movements of the isotherms 

 50°, 60°, 70°, etc., in the Atlantic Ocean during the year. The connection between 

 the law of this motion and the climates of the sea is exceedingly interesting. 



Plate V. (§ 781) is a section taken from one of the manuscript charts at the Ob- 

 servatory. It illustrates the method adopted there for co-ordinating for the Pilot 

 Charts the winds as reported in the abstract logs. For this purpose the ocean is di- 

 vided into convenient sections, usually five degrees of latitude by five degrees of lon- 

 gitude. These parallelograms are then subdivided into a system of engraved squares, 

 the months of the year being the ordinates, and the points of tlie compass being the 

 abscissae. As the wind is reported by a vessel that passes through any part of the 

 parallelogram, so is it assumed to have been at that time all over the parallelogram. 

 From such investigations as this the Pilot Charts are constructed. 



Plate VI. illustrates the position of the channel of the Gulf Stream (Chap. II.) for 

 summer and winter. The diagram A shows at h ermometrical profile presented by 

 cross-sections of the Gulf Stream, according to observations made by the hjdrogra- 

 phical parties of the United States Coast Survey. The elements for this diagram were 

 kindly furnished me by the superintendent of that work. They are from a paper on 

 the Gulf Stream, read by him before the American Association for the Advancement 

 of Science at its meeting in AVashington, 1S54. Imagine a vessel to sail from the 



