METEOROLOGY. 371 



of sun-spots, but the next followinpf auroral minimum coincides with 

 the sun-spot maximum. 



5. Some observations by Bloch, 1841 to 1S4G, show that the number 

 of auroras is very large at the minimum of sun-spots, but decreases as 

 the sun-spots increase. 



G. By separating the evening and morning auroras he finds the former 

 twice as numerous as the latter, and that it is probable tliat the evening 

 auroras give us the best idea of the true condition of the atmosphere 

 with reference to these phenomena. 



7. The morning and evening auroras both show a maximum in the 

 polar region at the time of the winter solstice. 



8. Kleinschmidt distinguishes eight types of auroras, three stationary, 

 and five of them movable. Of these the most interesting is his type 

 number 7, namely, rays arranged in a band stretching from the north- 

 east to the southwest and passing through the zenith. This phenomenon 

 rises in the east lilve a column of fire from some point in the interior of 

 Greenland. The column is narrowest and brightest over the land, while 

 its southwestern extremity, namely, over the sea, is slightly broader 

 and fainter, so that in general it does not seem to touch the southwest 

 horizon. Those arches almost always take the same position. fThis 

 form of arch, which is comparatively rare in Europe, is much more fre- 

 quent in the United States and Canada. An attempt to locate several 

 such arches and to explain their structure will be found in the report 

 of the Chief Signal Officer for 187G.] 



9. The aurora very rarely appears to the north of Godthaab, the cen- 

 ter of the phenomenon is most frequently between the south and south- 

 southeast, and this holds good for the whole of the west coast of Green- 

 land. 



10. The color of an aurora is almost invariably white, except only a 

 little red or green during the most rapid movements ; not the slightest 

 noise has ever been heard. 



11. With reference to the oscillation north and south of the zone of 

 greatest auroral frequency, Tromholt finds that the relative frequency 

 of auroras which occur in the zenith or to the north of the observer has 

 an annual period showing two minima at the equinoxes and a maximum 

 at the winter solstice. 



12. In every month he finds that the morning auroras are more fre- 

 quent in the zenith and the north than in the south, and that the re- 

 verse holds good for the evening auroras. 



13. As regards the auroras in the daytime, his figures confirm Wey- 

 precht's conclusion that, in the neighborhood of the autumn equinox, 

 the auroral zone stretches toward the south and then turns toward the 

 north to reach its northernmost position at the winter solstice, it then 

 returns southward and aasits southernmost position at the spring equi- 

 nox. 



14. In the course of twentv-four hours the auroral zone moves north- 



