On Meteoric Showers in August. 361 
fire, and leave behind them trains of scintillations, which are some- 
times iridescent. ‘These trains not unfrequently retain their lustre 
and their position several seconds, which, considering the excessive 
tenuity of the atmosphere at the place, is a fact quite unaccountable. 
I know of nothing which renders it at all improbable, that meteorites 
and shooting stars are bodies of the same class and have a common 
origin. No one can by the eye discriminate between them. 
By well-planned simultaneous observations in distant places, 
Brandes* has proved that shooting stars come principally from that 
part of the heavens towards which the earth is tending. He has 
also proved that they do sometimes actually move upwards. ‘This 
apparently anomalous motion is easily explained. His observations 
were made commonly between 9 and 11 P. M., when of course the 
tangential region is beneath the horizon; so that a meteor coming 
from that region must, if it comes into our field of view, move up- 
wards. Observations made in this city on various days of the pres- 
ent month, between 7 and 10 P. M., show that at this season at 
least, three fourths of the shooting stars visible here at this time of 
day, move from the N. E. quarter towards the S. W. If the 
premises first stated are true, the reasons for this general tendency 
are sufficiently obvious. At a corresponding south latitude, we 
might expect this tendency to be from the S. E. 
It must ever be exceedingly difficult to determine the velocities 
of shooting stars, and the statements on this point are not satisfac- 
tory. If their rate is found to be from 18 to 36 miles a second, it 
seems necessary to suppose that they move in a direction contrary 
to that of the earth’s motion. 
Shooting stars are wonderfully numerous. Leaving out of con- 
sideration the myriads which fall during the meteoric showers, the 
average daily number for the whole globe ts at least two millions. 
It is of course impossible rigidly to determine this number ; but 
there is no difficulty in showing that this statement is probably far 
within the truth. 1. The average number visible per hour at one 
place, M. Quetelet, who has devoted much attention to the sub- 
ject, states at sexteen. ‘This calculation is based on observations 
made in Germany and in Belgium, but there is no reason for sup- 
* A valuable abstract of Brandes’s account of his observations, was published 
by Prof. Loomis, in Vol. 28 of this Journal, to which I am chiefly indebted for my 
information concerning those researches. 
