46 REPORT — 1860. 



drift. There appear to be two, both setting to the north-east, one passing by the 

 Falkland Islands, the other having its northern terminus in the regions about the 

 Cape of Good Hope. Further south, icebergs are found all around; but in these 

 lines of drift thry are found nearest the equator. The space between the Falkland 

 drift and the Good Hope drift is an unfrequented part of the ocean. It may there- 

 fore be one broad drift, the edges of which only I have pointed out. The most active 

 currents from the south do not run with this ice. Humboldt's current is the most 

 active, but it does not get its icebergs as far north as they come by these lices. 

 This circumstance has suggested the conjecture that one part of the Antarctic Con- 

 tinent must be peculiarly well situated for the formation of glaciers and the launching 

 of icebergs. These lines of drift point to such a place. The facts stated in my 

 former letter will, I trust, when considered in connexion with these views, impress 

 you with the importance of the subject. So, trusting, and hoping that you will juin 

 with me in the cry, " Ho for the South Pole!" 



On the Climates of the Antarctic Regions, as indicated by Observations upon 



the Height of the Barometer and Direction of the Winds at Sea. By 



Captain Maury, U.S. Navy. 



In the course of my labours connected with the wind and current charts, I had 

 caused to be grouped 1,213,933 observations upon the direction of the wind at sea. 

 Each one of these observations embraces a period of eight hours, and aims to give 

 the prevailing direction of the wind during that time. Thus each individual of my 

 group is, in fact, itself the mean of many. The result of the whole is presented 

 diagrametrically in Plate I., in which the mean direction of the wind in each belt, 

 and for the four quarters, is represented by the arrows correctly, both as to mean 

 direction and average duration. 



From the labours of Lieut. Andrau and his colleagues at the Meteorological In- 

 stitute of the Netherlands*, I obtained 83,332 observations upon the barometer be- 

 tween the parallels of 50° N. and 36° S. at sea. This fine series was enriched by the 

 observations at Greenwich, St. Petersburg, and HobartTown on shore, and by Dr. 

 Kane, Sir James Clark Ross, and Lieut. Wilkes at sea, du;ing their Arctic and Ant- 

 arctic explorations. From these the barometric profile of the atmosphere (Plate I.) 

 was constructed. 



The barometric observations on shore were not found in all cases to accord with 

 those at sea. Moreover, those of Wilkes and Ross were the means of observations 

 for only a few days. 



Our ' Marine Magazine,' as the precious store of abstract logs may be called, con- 

 tained many more, and which, by their great numbers, and in consequence of their 

 having been made at all seasons of the year, would afford better mean results. In 

 ex'ension of Andrau's series, I therefore added 6915, between the parallels of 40° 

 and 6i)° south, from the log-books of this office. These, Andrau's, and Dr. Kane's 

 in the ice, form the elements of the Barometric Curve, Plate II. 



Proceeding upon the supposition that, with regard to the general movements and 

 the mean status of the atmosphere, we should have at sea the rule, on land the ex- 

 ceptions, I commenced to group these observations for discussion. 



As the North Indian Ocean, the China and West India seas, where the monsoons 

 blow, are known to present exceptional cases to the general movements of the winds 

 at sea, the observations for them were excluded from the general summing up. 



Thus premising, the winds were taken from the pilot charts and grouped in belts 

 5° of latitude broad. As a rule, the vessels that are cooperating with us seldom go 

 on the Polar side of 6CP north or south ; for our fleet of observers consists for the 

 most part of merchantmen, whom the channels of trade do not carry beyond these 

 parallels ; consequently the observations of the winds were arranged in 24 belts 

 (12 on each side of the Equator); all the observations between the Equator and 

 5° north, for example, being in one belt ; and so on for every 5° of latitude. 



Now, considering that the general movements of the atmosphere, as exhibited by 



* Maandelijksche Zeilaanwijzingen van Java naar Het Kanaal. Als Uitkomsten Weten- 

 schap en Ervaring Aangaande Winden en zeestroomingen in sommige Gedeelten van den 

 Oceaaq Uitgegeven Door Het Koninklijk Nederlandsch Meteorologische Instituut. Utrecht, 

 1859. 



