ON THE REDUCTION OF METEOROLOGICAL OBSERVATIONS. 89 
epochs and directions of the diurnal phases. Middletown and Flushing 
are intermediate in character. The Baltimore curve is very regular and 
shows the diurnal phases very distinctly, though not extensive. At Ba- 
hama the diurnal oscillations are obliterated at the beginning and end of 
the series, a uniform and perfect level being maintained from 14 hours to 24 
hours. 
(B.) This group, with the exception of the Bahamas, presents a series of 
curves evidently forming a system of oscillations, the area extending over 
21' of longitude and 7° 31! of latitude ; this area, on which most of the Ame- 
rican stations are situated, appears from the succeeding observations to have 
presented the most complete systems of oscillations. 
(B.) In proceeding from north to south there appears to be some irregu- 
larity in the increase, but by arranging the stations as in the Table*, it will 
be seen that there is a gradual increase on each side of the meridian of Mon- 
treal. 
September 1836.—Montreal, Burlington, Albany, Gardiner, William’s 
College, New York, Flushing, Middletown, New Haven, Cincinnati, Bahama, 
Ohreala, Guayaquil. 
(H.) All the curves of the first nine stations are on the whole nearly level, 
with undecided fluctuations, referable chiefly to diurnal oscillation, until about 
the 20th or Z1st hour of September 21, when they begin to dip downwards. 
The Cincinnati curve runs level, in three distinct stages of level, beginning 
and terminating at about 0:07 inch higher than its middle portion; this, 
however, is merely an effect of the periodical oscillations, in which the nightly 
maximum and morning minimum are blended and obliterated. 
The smoothest. curve in this term is that of Gardiner. Middletown and 
New Haven are also smooth. Albany is most broken (chiefly in the evening), 
which seems to be its general character. Of the West Indian curves I have 
already spoken. 
(B.) This sheet exhibits a well-marked and very interesting group, occu- 
pying nearly the same geographical area already noticed, namely from New 
York on the west to Gardiner, Maine on the east, the northern and southern 
boundaries being respectively Montreal and Flushing. The general similarity 
of these curves is very apparent also, the double curve indicating a diurnal 
oscillation at each of the stations; the increase of oscillation is also distinctly 
marked proceeding towards the north. 
(B.) The intermediate elevations and depressions offer very interesting 
matter for remark, the greatest development of them is at Flushing. At 
Montreal, the station of greatest oscillation, they have nearly disappeared. 
The intermediate curves are generally apparent when the oscillation is less 
than ‘100. At Cincinnati, to which the system did not extend, the extent of 
oscillation was ‘031 for 27 and *141 for 37 observations; here the curve was 
simple. 
(B.) The falls at the latter portions of the curves are much steeper than 
those at the anterior portions; the falls at New York and Flushing are less 
steep than the others. 
December 1836.—Quebec, Montreal, Albany, Flushing, New Haven, New 
_ York, Gardiner, Bahama. 
(H.) This term exhibits finely the barometric features and local progress 
of astorm. The barometric ranges will be found in the general table. The 
greatest intensity of the storm seems to have prevailed at Quebec, M ontreal, 
* Vide the Synoptic Table at the commencement of the American group. 
