IN THE MONTH OF FEBRUARY, 1842. 179 
been seen. ‘There would seem to be here very little cause for any oscillation of the baro- 
meter, and it is probable that if the entire surface of the globe were in the same situation, 
the variations of pressure would not exceed a small fraction of an inch. The range of 
the barometer observed at Melville island, December, 1819, was 1.65 inch; J anuary. 1820, 
1.18 inch; February, .83 inch; range for the winter months, 1.67 inches, which is equal 
to the extreme range which has been observed in Hudson in a period of five years. 
There can be little doubt that these oscillations are mainly the effect of storms prevailing 
in lower latitudes. 
The storm of February 4 was quite extraordinary—for its extent, amount of rain, 
and long continued elevated temperature; nevertheless it is presumed, in its main fea- 
tures, to resemble our ordinary January thaws. 
There was another phenomenon incidentally connected with the storm of February 16, 
which deserves a passing notice. Halos were observed at several places, during the 
period in question. The following are some of the observations. 
Washington city, February 15, from 6 hours 15 minutes to 8 hours P. m., a perfect 
lunar halo. 
Detroit, February 15, 10 a. m., halo round the sun. 
Detroit, February 16, 10 a. m., halo round the sun. 
Detroit, February 17, 8 to 10 p. m., halo round the moon. 
Toronto, February 16, 8 p. M., imperfect halo round the moon, diameter about 35°. 
Fort Gratiot, February 17, halo round the moon. 
Hudson, February 17, evening, halo observed. 
Fort Constitution, February 17, p. m., parhelion. 
The halos observed on the fifteenth, at Detroit and Washington, were formed in that 
cloud which is represented on chart 2, which at first was a mere haze, and gradually 
thickened. February 16, another halo was formed at Detroit and Toronto, in the cloud 
which followed the storm; and on the seventeenth, a halo was noticed at three places 
together with a parhelion. If any doubt previously existed as to the cause of this phe- 
nomenon, the facts which have already been presented may serve to dissipate it. The 
argument may be thus stated. A halo of 22° radius would be formed by the refraction 
of the light of the sun or moon through prisms of ice, having angles of 60° and floating 
irregularly in the atmosphere. The halo is proved to be formed by refraction. Snow 
consists chiefly of spicule having angles of 60°. Flakes of snow were floating in the 
atmosphere, February 15, over Detroit and Washington. The halos observed were, there- 
fore, formed by light refracted through these prisms. 
GENERALIZATIONS. 
I. DIRECTION OF THE WIND. 
The question which has, within a few years, excited most interest in Meteorology, is, 
whether storms are rotary or centripetal, and it has been supposed that either diagram, 
No. 1 or No. 2, must represent the motion of the wind. 
