March 1st to June 30th of this year. The rainfall from July 1st to
October 1st has been slightly above the normal in that section of the
country. Bangor, Me., to 9th; great suffering in farming districts;
streams drying up; stock driven long distances for water; many mills
stopped. Charlotte, Vt., severe to the 13th. Windsor, Vt., severe
at beginning of month up to 9th. Woodstock, Vt., 28th, drought of
July and August still prevails; many wells and streams dry; farmers com-
pelled in early part of month to haul water for stock. Westborough,
Mass., month warm and dry. Auburn, N. H., drought severe in early
part of month. Fort Madison, Iowa, 30th, water scarce; some wells
dry. Sandy Springs, Md., 30th, slight drought.
The percentage of mean relative humidity for the month ranges as
follows: New England, from 71 to 81; Middle Atlantic States, 66 to 87;
South Atlantic States, 69 to 88; Eastern Gulf States, 71 to 80; Western
Gulf States, 70 to 81; Ohio Valley and Tennessee, 62 to 77; Lower
Lake region, 65 to 73; Upper Lake region, 65 to 75; Upper Mississippi
Valley, 62 to 71; Missouri Valley, 60 to 67; Red River of the North
Valley, 70 to 76; Texas, 67 to 81; Middle Plateau, 19 to 39; Califor-
nia, 37 to 79; Oregon, 64 to 78. High stations report the following per-
centages not corrected for altitude: Mount Washington, 86.6; Denver,
43.9; Virginia City, 31.0; Cheyenne, 38.6.
The prevailing directions of the wind during September, 1880, are
shown by arrows, flying with the wind, on Chart. The prevailing direc-
tion in New England, the Lake region, the Ohio and Upper Missis-
sippi Valleys was southeasterly; in Florida, Tennessee, the South At-
lantic, and " Eastern Gulf States, northeasterly; in the Western Gulf
States, including all of Texas, southeasterly; the Upper Missouri and
Red River of the North Valley, northwesterly; the Lower Missouri Val-
ley and the Eastern Rocky Mountain slope, southerly. On the Pacific
coast it was northwesterly, except south at Sacramento and southwest at
Los Angeles. In the Middle Atlantic States the winds were mostly from
the northwest to southwest. In the Plateau districts they were variable.
On Mount Washington, the prevailing direction was N.W., and maxi-
mum velocity was S. 76 miles per hour on the 28th. Other maximum
velocities of 50 miles or more occurred as follows: 10th, N.E. 60; 15th,
THE \vi. vim. i;. 269
N.K. rO; 81st, N.W. 70; 22d, N.W. 60; 87th, N.W. 57. The prevail-
ing direction on Pike's Peak was S.W.; the total monthly moyemenl was
9,824 miles and maximum velocity 50 miles W., 4th. A maximum
velocity <>f 50 miles N.E. was reported From Gape Eemy on the 9th.
Local storm. -On Friday, the 3d, a tornado passed through the south
pari of Riley, 111. It came from the southwest and moved to the north-
east in a path aboul 200 feel wide. A.n observer says: It turned ;i tree,
2 feet in diameter, up by the routs, and twisted oil' two others of I feet
diameter, about 12 feel from the ground and carried the sundered parts
eight or ten rods, and scattered them in pieces over a large surface. It
broke ofl a sound hard maple of thirty years' growth, over a foot in
diameter, 10 feet from the ground, and took off several feet of the tops
of stacks of oats and scattered the bundles. It all occurred within five
minutes, and in the mean time water came down in torrents with a
whirling motion. There was considerable damage done in Harmony and
elsewhere south of Riley by the same storm, but no houses blown down.
Wafer-spouts. ā Key West, Florida, 16th, 3 p.m., about six miles dis-
tant. Buffalo, 29th, 4 water-spouts reported to have been seen on Lake
Erie, 4 miles from this city, moving from S.W. to N.E. They were said
to be cylindrical rather than conical.
Sand-storms. ā Umatilla, Oregon, 22d; Burkes, Ariz., 9th and 12th.
Thunder-storms have occurred with the greatest frequency as follows
Illinois : 3d, 18th, and 19th. Indiana : 3d, 5th, 6th, loth, 18th, 25th.
Iowa: 3d, 6th, 19th, 25th. Kansas: 3d, 18th, 25th. Michigan: 1st,
19th. Nebraska: 6th, 18th, 25th. Missouri: 19th. New York: 4th.
Ohio: 4th, 28th. Virginia: 5th and 6th. Tennessee: 4th, 5th. Texas:
3d, 4th, 6th, 7th, 12th, 22d to 26th. In other States they have occurred
with comparative infrequency. At Santa Barbara, Oal., a thunder-
storm occurred the 7th.
Atmosplieric electricity interfering with telegraph lines. ā New Mex-
ico : Silver City, on the 9th, 10th, 21st; Santa Fe, 6th, 26th; Socorro,
12th, 16th, 20th, 26th, 28th, 29th; La Mesilla, 10th, 20th.
Auroras were nightly observed, from the 27th to the 30th, inclusive,
over a wide extent of country, from Maine westward to Dakota, and as
far south as latitude 40Ā°. They were visible as far west as Fort Buford
on the 27th, 28th, 29th, and at Bismarck on the 30th. To the eastward
they were observed as far as Grafton, N. H., on 27th; Bangor, 29th; and
Gardiner, Me., 28th to 30th.
Till; WEATHER. 2 . 1
4. MONTHLY WEATHER REVIEW, OCTOBER, 1880.
The genera] distribution of barometric pressure as exhibited on the
Chart differs slightly from the mean pressure as determined from the
observations for many years. The greatest variation from the mean
occurs on the North Paciiic coast, where the pressure is seven hundredths
above the normal at Portland. The pressure has increased from three
to six hundredths in all districts east of the Mississippi, and the area of
mean high barometer which covered the Southern States in September
has increased and moved north over the Middle Atlantic States and New
Ilnrometric ranges. ā The barometric range for the several stations
increases with the latitude and is unusually great in the Upper Mississippi
Valley and Upper Lake region, located near the track of the centre of
the storm which passed over these regions on the 16th and 17th. The
range increases on the Atlantic coast from 0.30 at Key "West to 1.22 at
Portland, Me., and in the Mississippi Valley from 0.68 at New Orleans
to 1.68 at La Crosse, this being the greatest range reported. In the
Lake region, the range varies from 0.89 at Eochester to 1.51 at Duluth.
On the Pacific coast the range increases with the latitude, but is much
less than at corresponding latitudes on the Atlantic coast.
Areas of high barometer. ā Nine areas of high barometer have ap-
peared within the limits of the Signal Service stations during the month
of October, four of which ā Nos. II., V., VII., and IX. ā were traced
from the North Pacific to the Atlantic coast. No VI. apparently de-
veloped in the Southwest, and followed the general direction of the
storm of the 16th and 17th, being preceded by the most severe norther
of the month on the Texas coast.
Areas of low barometer. ā On the Chart will be found the tracks of
centres of the areas of low pressure, which have been traced from the
tri-daily weather maps of the month. The mean latitude of these tracks
is several degrees to the north of the mean latitude of low areas for cor-
responding month of previous years. Nos. IV., VIII., and XI. were
tropical storms, which developed south of latitude 25Ā°, and No. VI., the
most marked depression of the month, apparently developed in the
Southwest, probably in Northern Texas. No. XII. is the only depres-
sion that crossed the Rocky Mountains, and this disappeared before
reaching the Lake region.
.272 THE WEATHER.
TEMPERATURE OF THE AIR.
The mean temperature of the air during October, 1880, is shown by
the isothermal lines on the Chart. In the Southern .States east of the Mis-
sissippi Hirer, and in the Middle States, the temperature has been near
the normal of many years. In all the other districts east of the Rocky
Mountains it has averaged from one to five degrees below the normal,
the greatest departures being in Western Texas, and thence northward
over Kansas and Colorado. On the Pacific slope, the temperature cor-
responds with the normal, except at San Diego, where it is one degree
below, and in the southern Plateau district, where it has averaged over
two degrees below. The temperature is slightly above the normal in the
lower Saint Lawrence Valley and Gulf of Saint Lawrence.
The general distribution of rain-fall for October, 1880, is shown on
the Chart, as accurately as possible, from about five hundred reports.
East of the Mississippi Eiver, except in the Upper Mississippi Valley,
Upper Lake region, and Middle Atlantic States, where there were slight
deficiencies, there were excesses of rain-fall, ranging from 3.36 inches in
the Florida peninsula and 2.10 inches in the South Atlantic States to
0.17 in Xew England. West of the Mississippi, except in Minnesota
and the Lower Missouri Valley, where slight excesses occurred, the rain-
fall was below the average, the greatest deficiency, 1.47 inch, occurring
in the Xorth Pacific coast region.
Rainy days. ā The number of days on which rain or snow has fallen
varies as follows: Xew England, 7 to 16; Middle Atlantic States, 5 to
13; South Atlantic States, 7 to 17; Eastern Gulf States, 10 to 14;
Western Gulf States, 4 to 10; Ohio Valley and Tennessee, 10 to 16;
Lower Lake region, 14 to 22; Upper Lake region, 7 to 20; Upper Mis-
sissippi Valley, 5 to 10; Lower Missouri Valley, 5 to 7; Red River of the
Xorth Valley, 9 to 11; Rio Grande Valley, 3 to 8: Texas. 3 to 7; Rocky
Mountains, 5 to 17; Western Plateau, to 5; California, to 3; Oregon,
3 to 10; Washington Territory, 7 to 9.
Cloudy days. ā The number varies in Xew England from 5 to 14;
Middle Atlantic States, 6 to 10; South Atlantic States, 8 to 16; Eastern
Gulf States, 6 to 16; Western Gulf States, 3 to 10; Ohio Valley and
Tennessee, 8 to 13; Lower Lake region, 10 to 13; Upper Lake region, 8
to 16; Upper Mississippi Valley, 7 to 12; Lower Missouri Valley, 5 to 11;
Red River of the Xorth Valley, 12 to 14; Rio Grande Valley, 4 to 7;
Texas, 1 to 9; Rocky Mountains, 4 to 9; Western plateau, 2 to 5; Cali-
fornia, 1 to 3.
Hail. ā New York: Madison Barracks, 6th; Buffalo; 17th, 18th.
THE WKATin .1:.
I'mnsiilrania : Catawissa, Fallsington, Philadelphia, and W'c-t Chester,
28th; Green Castle, 26th. New Jersey : Freehold and Prii ton, 28th.
Illinois: Chicago, Let, 2d, 25th; Morrison and Sterling, 2d. Iowa:
Dubuque, 14th, L6th; Clinton, 3d; Cresco, 15th; Mount Vernon, 2d.
Des Moines, 15th, severesi storm experienced since opening Btation;
great damage to windows; stones varied in size from hickory nuts to
walnuts; width of track, one-quarter mile; direction from southwest to
northeast. Kansas : Eolton, 2d. Michigan: Lansing, 2d, 25th; Mar-
quette, 16th. Minnesota: Saint Vincent, 10th; New Ulm, 15th; Du-
1 ii tli. 28th. Wisconsin: Milwaukee, 25th. Dakota: Fort Bennett,
15th, 16th. Texas : San Antonio, 27th; Pilot Point, 15th. New Mex-
ico: Santa Fe, 12th. Oregon: Albany, 10th. Washington Territory:
Snow. ā New Hampshire: 19th, 24th, 26th; on summit of Mount
Washington, 1st, 2d, 5th, 6th, 7th, 8th, 12th, 13th, 18th to 23d, 25th to
31st. Vermont: 19th, 20th, 23d, 25th, 28th. Massachusetts: 19th,
20th, 24th. New York: 17th, 18th, 19th, 20th, 23d, 24th, 27th, 28th.
New Jersey: 19th, 28th. Pennsylvania: 19th, 24th, 28th. Mary-
inn, I: 19th, 22d, and 23d in mountains. West Virginia: 19th,
23d. Kentucky: 19th. Ohio: 17th, 19th, 20th, 23d, 24th. Michi-
gan: 16th to 24th, 26th, 27th. Indiana: 15th, 17th to 19th. Illinois:
16th, 17th, 19th to 23d. Missouri: 16th. Wisconsin: 16th, 17th to
19th, 23d. Iowa: 15th to 19th, 23d. Kansas : 15th, 16th. Nebraska:
14th to 18th, 21st. Minnesota : 15th to 21st, 23d, 30th. Dakota :
9th, 15th to 18th, 26th. Colorado : 9th to 13th, 15th, 26th, 30th. On
summit of Pike's Peak, 9th, 11th, 12th, 13th, 18th to 24th, 31st. New
Mexico: 10th, 11th. Nevada: 8th, 10th, 11th, 12th. Utah: 8th, 9th,
10th, 14th. Wyoming: 9th, 10th, 14th to 18th, 29th. Idaho: 14th.
Montana : 8th, 9th, 14th, 15th, 28th.
Droughts. ā Fort Madison, Iowa, 30th, water very scarce; pastures
drying up. Yates Center, Kans., 31st, streams and ponds lower than
for past six 3' , ears. Emmittsburg, Md., 30th, month very dry. Sandy
Springs, Md., 31st, great scarcity of water; streams unusually low.
Contoocookville, N. H., 1st to 30th, quite severe. Palermo, N. Y.,18th,
very dry; wells and streams low. Greencastle, Pa., last half of month
streams very low; wells and springs nearly dry; boats on Delaware and
Hudson Canal stopped running for want of water. Woodstock, Yt..
throughout month springs and wells remarkably low. "Wvtheville, Ya..
29th, drought has been very severe since the 1st; cisterns and wells gen-
erally exhausted and streams very low.
Snow from a cloudless sky. ā Springfield, 111., 18th.
Snoiv on ground at end of month. ā Virginia City, Mont., 1.50
inches; Pike's Peak, 2 feet 2 inches; Breckenridge, Minn., trace;
Mount Washington, 4.00 inches.
The percentage of mean relative humidity for the month ranges as
follows: New England, 66 to SO. Middle Atlantic States, 61 to 86
South Atlantic States, 62 to 80. Eastern Gulf States, 68 to 77. West-
ern Gulf States, 70 to 76. Ohio Valley and Tennessee, 63 to 79. Lower
Lake region, 67 to 70. Upper Lake region, 66 to 76. Upper Missis-
sippi Valley, 58 to 65. Lower Missouri Valley, 62 to 64. Red River of
the North Valley, 70 to 77. Rio Grande Valley, 63 to 71. Texas, 61
to 71. Western Plateau, 30 to 45. California, 37 to 71. Oregon, 54
to 82. High stations report the following percentages not corrected
for altitude: Mount "Washington, 80.7; Pike's Peak, 67.3; Denver,
55.9; Cheyenne, 45.0; Virginia City, 48.7; Santa Fe, 48.0.
The arrows on the Chart indicate the prevailing direction of the wind
at the several stations during the month of October. The prevailing direc-
tion was north to east in the South Atlantic and Gulf States, in the
southern portion of the area of mean high barometer for the month, and
south to west in the Lake region, Ohio Valley, northwest and interior
of Texas, the Middle States, and Xew England. Xortherly winds pre-
vailed on the Pacific coast as far south as Sacramento, and southeast to
southwest winds in the southern portion of California. At stations in
the Rocky Mountain region the prevailing direction was generally from
the higher to lower altitudes. Pike's Peak, prevailing direction, north-
west; highest velocity, 72 miles, west on 13th. Mount Washington,
prevailing direction, northwest; highest velocity, 84 miles, northwest on
the 24th and 31st.
Local Storms. ā Under this head no storms of decided severity were
reported during the month except in connection with the passage of low
area Xo. VI. during the 15th, 16th, and 17th, which are herewith ap-
pended from the reports of various stations and other sources. Morri-
son, 111., 16th, violent storm, hundreds of trees prostrated, and consid-
erable damage to fences and out-buildings. Riley, 111., 16th, great
damage to trees, fences, corn-fields, and a " wholesale destruction
of windmills;" heaviest storm ever experienced here; estimated ve-
locity of wind, 70 miles per hour. Rockford, 111., 16th, severe
gale from the southwest, causing considerable damage. Elmira, 111.,
16th, violent southwest gale nearly all day. Springfield, 111., 16th, con-
siderable damage to trees, fences, and out-buildings. Chicago, 16th, vio-
lent southwest gale with sleet from 5.15 to 9.30 p.m., buildings blown
down, and much damage to shipping; 17th, wind continued high all day.
Independence, Iowa, 16th, snow accompanied by a violent westerly gale
THE WEATIIl i:. 275
Erom L to 11 p.m. Qattenberg, Iowa, 15th, severe thunder-storm, creeks
Ln vicinity rose 11 feet above Low-water mark; 16th, hurricane from
west, accompanied by sleet. Logan, Iowa, 16th, violent westerly gale,
worst storm for twenty years. .Mount Vernon, Iowa, violent westerly
gale all day. Nora Springs, Iowa, lGth, lowetst barometer ever recorded,
Btorm of great severity, all railroad communication westward suspended.
Muscatine, Iowa, lGth, very heavy west gale. Cresco, Iowa, 16th, hi
iterly storm, lowest barometer for the year. Vail, Iowa, l<;th, storm
of unusual severity; snow-drifts, from 3 to 5 feet deep. Mitchellville, Iowa
15th, about 4 P.M. a tornado was noticed in the west coming rapidly in
direction of town, accompanied by a deep, heavy, rumbling sound. The
cloud, funnel-shaped and twisting with great rapidity, was accom-
panied by a heavy westerly gale, and sudden darknes3 overshadowed
everything as it approached the western edge of the village, where it
lifted from the earth and passed harmlessly to the northeast, when it
again seemed to descend. Dubuque, Iowa, lGth, violent southwest wind,
considerable damage to trees, fences, and out-buildings. Keokuk, Iowa,
loth, southwest 48 miles, considerable damage to property; 16th, west, 40
miles. Davenport, Iowa, lGth, great damage to fences, trees, and build-
ings, and navigation suspended. Topeka, Kans., loth, southwest, 54
miles, much damage to property. Dodge City, Kans., 14th and 15th,
north, 56 miles. Grand Rapids, Mich., 16th, 17th, violent southwest
gale. Escanaba, 16th, much damage to shipping and other property.
Port Huron, 16th, severe wind-storm, lasting thirty-six hours, much
damage to shipping and other property. Port Huron, 16th, severe wind-
storm lasting thirty-six hours, much damage to shipping. Grand Haven,
16th, southwest 48 miles, worst day ever seen at this station. Detroit,
17th, heavy southwest wind nearly all day. Genoa, Nebr., 15th, violent
wind-storm from the north, severest for many years. De Soto, Xeb.,
16th, heavy northwest gale. North Platte, 16th, violent gale from
northwest, 54 miles per hour: Wooster, Ohio, 17th, violent southwest
wind nearly all day. Flemington, TV. Va., 16th to 18th, very violent
westerly winds. Ashland, Wis., 17th, heavy northeast gale, much dam-
age to docks, warehouses, and shipping. Bloomfield, Wis., 16th, violent
southwest gale, much damage to fences and buildings. Milwaukee, 16th,
southwest 60 miles, much damage to buildings, and telegraphic commu-
nication interrupted. Madison, Wis., 16th, west, 44 miles, much dam-
age to property. Breckenridge, northwest, 56 miles, 1.30 P.M. Duluth,
16th, northeast and northwest 30 miles, very heavy sea on lake and
much damage to shipping and wharves. Saint Vincent, Minn., 16th,
violent storm from the north, highest velocity 40 miles per hour. St.
Paul, violent gale from north-northwest, considerable damage in country
and city, all telegraphic communication interrupted. Yankton, Dak.,
16th, 70 miles, northeast at 1 a.m.; roads blocked with snow and com-
munication of all kind suspended. Snow-drifts east of station reported
2TG THE WEATHER.
to be from 10 to 15 feet high. Reports from different points estimate loss
of cattle and crops in Yankton County, at $5,000. Fort Bennett, Dak.,
lGth, north, 40 miles. Memphis, Tenn., 15th, 32 miles west, much damage
to buildings in city. Little Rock, Ark., 15th, west 32 miles, consider-
able damage to bridges, fences, and buildings.
Sand-storms. ā Umatilla, Oreg., 13th; Winnemucca, Xev., 7th, 8th,
14th, 28th; Fort Yuma, Cal., 8th, 15th; Fort Garland, Colo., 22d.
Thunder-storms. ā Comparatively few have been reported during
the month, and mostly from the northern sections of the country. The
most extensive, as well as severe, was that of the 15th and 16th. Dis-
tributed among the several States, they were reported on the following
dates: Maine: 23d. Massachusetts: 12th, 23d, 25th, 30th, and 31st.
New York : 3d, 12th, and 16th. Virginia : 15th, 22d, 30th. North
Carolina: 22d. Florida: 3d, 5th, 7th, and 28th. Tennessee: 3d, 14th,
15th, 16th, and 26th. Kentucky: 14th and 15th. Ohio: 1st, 2d, 14th,
15th, 17th, and 21st. Indiana: 1st and 15th. Illinois: 1st, 2d, 3d,
14th, 15th, 21st, 24th, and 25th. Michigan: 1st, 3d, 4th, 15th, and 25th.
Wisconsin : 10th and 16th. loiva : 1st, 2d, 3d, 10th, 11th, 13th, 15th,
and 24th. Missouri: 15th and 25th. Mississippi: 27th. Louisiana:
3d, 27th, and 28th. Texas: 2d, 3d, 13th, 15th, 16th, 21st, 26th, and
27th. Indian Territory: 13th, 15th, and 25th. Kansas: 2d and 20th.
Nebraska : 2d, 3d, and 13th. Colorado : 26th. New Mexico : 24th.
Auroras. ā The most extensive occurred on the evenings of the 7th,
28th, and 30th, and the early morning of the 31st, and were reported by
various observers, both Signal Service and voluntary, from stations
reaching eastward from Dakota to Maine, and southward to the 37th
AQUEOUS VAPOR IX RELATION" TO PERPETUAL SNOW.
BY JAMES CROLL, LL.D., F.R.S.
Some twelve years ago I gave (Phil. Mag., March, 1867, "Climate
and Time," p. 548) what appears to be the true explanation of that ap-
parently paradoxical fact observed by Mr. Glaishier, that the difference
of reading between a thermometer exposed to direct sunshine and one
shaded diminishes instead of increases as we ascend in the atmosphere.
This led me to an important conclusion in regard to the influence of
aqueous vapor on the melting of snow, but recent objections to some of
my views convince me that I have not given to that conclusion the prom-
inence it deserves. I shall now state in a few words the conclusion to
which I refer.
I 11 1, v, i.A i ii I R.
The reason why snow at great elevations does not melt, but remains
permanent, is owing to the facl thai the heal received from the ran is
thrown off into stellar space so rapidly by radial ion and refled ion that tho
sun fails to raise the temperature of the mowto the melting point; the
snow evaporates bul doesnotmelt. The summits of the Himalayas, for
example, must receive more than ten times the amounl of heal 01 cessary
to melt all the mow that falls on them, notwithstanding which the snow
is not melted. And in spite of the strength of the sun and the dry]
of the air at these altitudes, evaporation is insufficient to remove the
snow. At low elevations, where the snow-fall is probably great'! - , and the
amount of heat received even less than at the summits, tho snow melts
and disappears. This I believe we must attribute to the influence of
aqueous vapor. At high elevations the air is dry and allows the heat
radiated from the snow to pass into space, but at low elevations a very
considerable amount of heat radiated from the snow is absorbed by the
aqueous vapor which it encounters in passing through the atmosphere.
A considerable portion of the heat thus radiated, being of the same qual-
ity as that which the snow itself radiates, is on this account absorbed by
the snow. Little or none of it is reflected like that received from the
sun. The consequence is that the heat thus absorbed accumulates in the
snow till melting takes place. Were the amount of aqueous vapor
possessed by the atmosphere sufficiently diminished, perpetual snow
would cover our globe down to the sea-shore. It is true that the air is
warmer at the lower level than at the higher level, and by contact with
the snow must tend to melt it more at the former than at the latter posi-
tion. But we must remember that the air is warmer, mainly in conse-
quence of the influence of aqueous vapor, and that were the quantity
of vapor reduced to the amount in question, the difference of tempera-
ture at the two positions would not be great.
But it may be urged, as a further objection to the foregoing conclu-
sion, that as a matter of fact on great mountain-chains, the line reaches
to a lower level on the side where the air is moist, than on the opposite
side where it is dry and arid. As, for example, on the southern side of
the Himalayas and on the eastern side of the Andes, where the snow-line
descends some 2,000 or 3,000 feet below that of the opposite, or dry side.
But this is owing to the fact that it is on the moist side that by far the
greatest amount of snow is precipitated. The moist winds of the S. W,
monsoon deposit their snow almost wholly on the southern side of the
Wimalayas, and the S.E. trades the snow on the east side of the Andes.
Here the conditions in every respect the same on both sides of the
mountain ranges, with the exception only that the air on one was per-
fectly dry, allowing radiations from the snow to pass without interrup-
tion into stellar space, while on the other side the air was moist ami full
of aqueous vapor, the snowline would descend to a lower level on the dry
than on the moist side. ā A mer. J'l of Science, No. 116, vol. xx., Aug. 1880.