— 234 — 



are recapitulated here, together with their geographical 

 and their auroral latitudes. By the latter are meant the 

 complements of spherical distance from the pole of aurora 

 borealis. 



The regularity in decrease shown by auroral latitude as 

 contrasted with the irregular succession of geographical 

 latitudes is obvious; now this is not whöUy unexpected, as 

 the vector of disturbance is directed perpendicularly to 

 the isochasms, which have a course resembling that of true 

 parallels around the axis of aurora. 



And as these isochasms show certain deviations from the 

 true auroral parallels, the small irregularity in the succession 

 above mentioned shown by Bossekop and Sodankgld (resp. 

 with 66° and 63°) may be safely neglected. 



This deviation of isochasms from auroral parallels makes 

 it impossible to determine accurately the place of the aurora- 

 pole. After his observations of aurora Nohi)Knsk(>li> ^) adopted 

 81° N and 80° W, and this place correspoiuls very nearly 

 to the point where tlie meridians of disturbing- vectors, 

 mentioned above, converge. (Nachstörungsmeridiane, Cf. 

 Meteorol. Zeitschr. 1895 l.c). 



') Ann. de Chimie et de Phijsique, 1884. 



