108 Major-General Sabine on the Cosmical Features 



to the form in fig. 2, which corresponds to the easterly deflec- 

 tions at Kew and Hobarton. During twelve hours, or about 

 half the diurnal period, the amount of disturbance, and conse- 

 quently the ratios, scarcely vary; there occurs then a sudden 

 and rapidly progressive increase of disturbance continuing for 

 about six hours, and forming an apex in the curve which at Kew 

 and Hobarton occurs between eleven and twelve hours of local 

 astronomical time ; and from this apex there is a symmetrica] 

 and progressive decrease of disturbance lasting also about six 

 hours. This is the general form, as I have stated, of one of 

 the curves; but the local hour at which the apex occurs, 

 although it happens to be the same at Kew and at Hobarton, 

 appears to range through all the hours at different stations 

 in both hemispheres. Another remarkable difference is found 

 at different stations. The two forms exhibited in figs. 1 and 2 

 are so distinct and dissimilar, that there is no possibility of 

 mistaking the one for the other. This dissimilarity is general 

 between the two forms of the easterly and the westerly deflec- 

 tions ; but the form exhibited in the Plate, as representing the 

 easterly deflections at Kew and at Hobarton, does not always 

 belong to the disturbances included in that category. In the 

 northern hemisphere, for example, it belongs to the easterly de- 

 flections at Kew in England, and at Toronto and other stations 

 in North America; but in Northern Asia, at Pekin, and at 

 Nertchinsk in Siberia, it belongs, on the contrary, to the west- 

 erly deflections*. We have here, therefore, apparently a marked 

 distinction in this respect between two parts of the same he- 

 misphere; and there is another no less remarkable which 

 may be conjoined with it, both appearing to point to geogra- 

 phical relations. At all the stations in North America, the 

 easterly disturbances predominate in amount, whilst in Northern 

 Asia it is the westerly which predominate f. There are there- 

 fore distinguishing characters in the effects of the sun's influence 



* Fig. 3 in Plate I. represents the ratios at the several hours of the west- 

 erly disturbances at Nertchinsk in Siberia, lat. 51° 56' N., long. 116° 33' E. 

 from Greenwich, derived from hourly observations during the years 1851 

 to 1857 inclusive. The difference from the normal constituting a disturbed 

 observation has been taken at V'96. The form of this curve has the same 

 general character as the curve of easterly deflection at Kew (fig. 2) and 

 differs altogether from the form of the westerly curve in fig. 1. 



The curve of the easterly deflections at St. Helena (south end of the 

 magnet) is drawn in the woodcut in vol. xi. p. 683, of the ' Proceedings of 

 the Royal Society,' and is another repetition of the same general form. 



t A similar difference is found in the preponderance of easterly or west- 

 erly deflections in the southern hemisphere. At Hobarton, the Cape of 

 Good Hope, and St. Helena, easterly deflections (of the south end of the 

 magnet) preponderate ; at the Falkland Islands, on the other hand, it is the 

 westerly deflections (of the same end of the magnet) which preponderate. 



