formation. In 1870, the aëronaut Rollier, escaping from besieged Paris in a balloon, landed in Norway, and he claims to have heard a persistent sound the whole time he was in a certain cloud, accompanied by a strong odor like ozone, very irritating to the bronchial tubes. A fine aurora was observed at precisely that time. Bergmann compared this odor to that of sulphur, and Trevelyan to that of electricity (?).
It is not necessary, at this juncture, to enter into a discussion of the extent, position, or periodicity of auroras, nor their relation to sun spots. But some new facts may be added that are valuable. There is a close resemblance between certain effects of the aurora and cirro-cumulus and cirro-stratus clouds, which are the highest clouds; so that it is difficult to tell whether a given effect is due to real auroral light or to these high clouds lit up by reflected light. Sometimes these bands of clouds stretch out in long parallel lines; again, they are due north and south and lie parallel to the needle of the compass, and are then called "polar bands." It has been discovered that when these polar bands occur during the day an aurora follows that night.
As long ago as 1580 and 1590, Tycho Brahe observed that the appearance of these polar bands and halos coincided with the presence of spots on the sun. Klein, in our own day, comparing twenty-five years of observations made with the greatest care at Cologne by Dr. Garthe, has confirmed the fact that these high cirrus clouds and polar bands follow as to their frequency the same laws as the spots on the sun; they succeed each other or even coexist. Many observers of the aurora are of the opinion that the appearance of the aurora depends on the presence of these clouds in the sky, which are in turn due to magnetic disturbances of the earth's photosphere originating in the sun spots. There are other evidences that the aurora is intimately connected with material particles in the atmosphere, like clouds and fog; for instance, when two or more rays in an aurora cross, the light is augmented, or when an aurora makes a fold on itself as in drapery forms, showing that there are two material "thicknesses" of the substance; moreover, the wind also acts on the aurora, which is torn after a tempest, showing that the wind has acted upon luminous clouds which are part of the aurora; finally, the presence of clouds seems to favor the formation and development of auroras. So there is an intimate relation between them.
II.
The rare phenomenon, St. Elmo's fire (balls of fire alighting on ship masts and spars), is frequent during auroras, but this is about the only form of electric disturbance of the atmosphere occurring
