452 
thus wafted by the winds. Under certain meteoric conditions, this minutely 
subdivided matter returns again to the earth. J ust as, in ordinary conditions 
of the atmosphere, the moisture descends to the earth as rain or snow, we 
know that under certain more extraordinary conditions, large masses of ice are 
formed, supported during that formation contrary to the laws of gravity, and 
then hurled to the earth. There appears some analogy between this latter 
phenomenon and that of the formation of a meteorite. The peculiar noise, 
like a discharge of small arms, which heralds the fall of a meteorite from a 
cloud, in a somewhat modified form, accompanies the formation of blocks of 
ice in the air. We know, as in the beautiful test for arsenic discovered by 
Marsh, that a solid metal arsenic may be combined with an invisible gas, 
hydrogen, and form together an invisible gas. When this is combined with 
oxygen, a spark of electricity is sufficient to combine the oxygen and hydrogen 
into water, and precipitate the arsenic in a pure metallic state. May there 
not be some analogy between this fact and the formation of meteoric iron ? 
I may add, as it appears to militate against the electrical origin of falling star 
storms, that I have ascertained that no disturbance was observed in the deli- 
cate magnetic needles of Greenwich Observatory during the late November 
display. On the other hand, I believe that it is recorded that the most 
delicate electrometers have not been in the slightest degree affected during a 
magnificent display of aurora borealis. 
The Rev. Dr. Irons. — I think Mr. Warington’s point was this. There is ad- 
mitted to be a certain moment of periodicity respecting the wonderful displays 
of the meteors called falling stars, and we know that the last shower was pre- 
dicted. Could you affirm anything about displays of the aurora borealis being 
predicted ? because, if not, it would seem hard to connect the two things 
together — the one being predictable, and the other casual. 
The Rev. Walter Mitchell. — I may say in reply to this question that I 
do not know of any display of aurora borealis being predicted with the same 
degree of precision as to any particular day as these exhibitions of falling 
stars, but I have called attention to the fact that the maximum appearance 
on any day does not follow any period of years ; that for a number of years 
in succession so few stars have been seen to fall on the 13th or 14th of 
November, or from the 10th to the 12th of August ; that certain of the French 
observers, one of whom devoted attention night after night to counting them, 
gave up altogether the theory of periodicity. I might perhaps say that the 
prediction of the meteoric display of 1866 was a philosophical “fluke” — it 
was a fair guess from probabilities founded upon the years 1766, 1799, 1833. 
It was a good guess to say 1866 ; but, as I have before pointed out, there have 
been displays between those years which have not and could not have been 
predicted ; that of 1766 was included by mistake ; nor can we predict whether 
in November, 1867, we shall have a more abundant shower of falling stars 
than we had last year, or in 1864 or 1865. 
The Chairman. — I quite agree with Mr. Mitchell that there is no perio- 
dicity with regard to meteoric showers, for I have been a great deal in 
tropical latitudes, where falling stars are constantly seen night after night, 
