DECEMBER 27, 1906] 
NAT OLE 
211 
necessary condition being the development of low baro- 
metric pressure near the southern margin of the cold-air 
belt, and the production of strong northerly currents, due 
to cyclonic circulation. To quote only one instance of the 
value of forecasts in connection with these cold waves :— 
from January 6-12, 1886, a cold wave swept the country 
east of the Rocky Mountains and produced the lowest 
temperatures noted for the last fifty years in the south- 
eastern States. Speaking of this wave, Prof. Moore has 
stated that on January 7 there was a difference of 1-1 inch in 
barometric pressure between Montana and southern Texas, 
while the isotherm in Montana was —30° and on the 
Texas coast 50°. The people of the Gulf States knew 
nothing of the danger that threatened them until warned 
by the telegraphic weather forecast; on January 8 the 
temperature in parts of Texas had fallen to zero, and, not- 
withstanding the timely notice. the estimated damage to 
crops was 3,000,000 dollars in Florida alone. 
Diurnal Range of Temperature in the Tropics.—Prof. 
J. Hann recently presented an important treatise on this 
subject to the Vienna Academy of Sciences. In continu- 
ation of a former work dealing with the district between 
lat. 15° N. and S., the one now in question relates to 
places in Africa and America (including the West Indies, 
Madagascar, and Mauritius) lying north and south of the 
above latitudes, and extending to the limits of the tropics, 
and in some cases beyond them. The work is divided 
into two parts, containing, inter alia, (1) tables of the 
daily range of temperature in the form of departures of 
the hourly from the daily means, with a general discussion 
of the results, and (2) tables of the periodical and non-peri- 
odical amplitudes, and of the epochs of the daily maxima 
and minima, in connection with cloudiness, sunshine, and 
rainfall. The mean occurrence of the minimum tempera- 
ture at all places in the tropics (mountain stations excepted) 
is approximately at 5h. 30m. a.m., both on the coast and 
inland. The time of the maximum differs; on the coast 
and in rainy districts it mostly occurs soon after midday, 
at inland and dry stations it is at 2h. or even after 
2h. 30m. p.m. At places on the West Indian coasts the 
maximum occurs about 42m. after noon; somewhat more 
inland, at Puerto Principe (Cuba), nearly an hour later, 
and at the City of Mexico about 2h. 48m. p.m. The occur- 
rence of the daily mean is retarded according to distance 
from the equator ; twenty-seven stations in the central zone 
(lat: 15° N. to 15° S.) give the mean time of 8h. 26m., 
twenty stations in the outer zones give 8h. yom. as com- 
pared with oh. 27m. at fifteen places in Austria. 
Diurnal Variation of the Barometer.—In the U.S. 
Monthly Weather Review for April, Prof. Cleveland Abbe 
directs attention to an article in Gaea for August, 1905, 
by Dr. Korselt, of Annaberg, Germany, in which he 
attempts to show how the diurnal oscillation of the baro- 
meter is an important link in the chain of phenomena due 
to the unequal warming of the atmosphere by solar 
radiation, and its unequal cooling by terrestrial radi- 
ation. One of his conclusions, which may be recommended 
to the notice of meteorological organisations the telegraphic 
reporting stations of which generally possess self-recording 
barometers, is that the minute study of the daily barometric 
oscillation may be of great value for practical weather fore- 
casting, because it ought to give information about con- 
ditions in the atmosphere at altitudes which balloons have 
not yet been able to attain. A weather chart showing the 
observed difference between the barometric ranges by day 
and by night during the preceding twenty-four hours would, 
he thinks, probably show that any temporary area of low 
pressure has a tendency to move toward the region where 
the difference of the ranges is a minimum. Prof. Abbe 
points out, however, that these ranges are so small that 
they would often be completely masked by larger non- 
periodic changes, so that misleading errors would seem to 
be inevitable. 
Influence of the Ocean upon Continental Precipitation. 
—In the same number Mr. F. O. Stetson (assistant editor) 
directs attention to a recent paper read before the Société 
helvétique des Sciences naturelles on the interchange of 
moisture between land and sea, by Prof. E. Briickner. 
The author estimates that 93 per cent. of the water 
evaporated from the ocean is returned to it in the form 
of precipitation, leaving only 7 per cent. available for 
NO. 1939, VOL. 75| 
distribution over the land; and.that of the 
tion over the land 20 per cent. is supplied 
ocean, while the remainder is due to the re-condensation 
of vapour evaporated from the continents. We cannot give 
here the data upon which Prof. Bruckner’s figures are 
based, but if they are provisionally accepted as approxi- 
mately correct, they indicate that the direct influence of 
the ocean upon rainfall over the land is less than has been 
generally supposed; but Mr. Stetson points out that the 
accurate determination of evaporation is a problem not yet 
solved, and that the rainfall over extensive tracts of land 
still remains unknown. 
Wind Currents in the Vicinity of the Canary Islands.— 
In a recent note to the Comptes rendus of the Paris 
Academy, M. Teisserenc de Bort and Mr. Rotch have con- 
firmed their opinion that the south-west winds observed on 
the Peak of Teneriffe correspond to a general phenomenon 
and are identical with those which would obtain over the 
open ocean, and consequently represent the regular anti- 
trade. This view is not in accordance with that held by 
Prof. Hergesell, to which he has again directed attention in 
Beitrdge zur Physik der freien Atmosphdre (vol. ii., 
part ii.). He maintains that his observations with kites 
in 1904, and the balloon observations of the Princess Alice 
in 1905, show that in the latitude of the Canary Islands 
during summer north-west winds prevail to the greatest 
heights, and that there can be no question of a regular 
south-west current in that part of the Atlantic, the 
occasional south-west wind observed on the Peak of 
Teneriffe being due to local effects. Prof. Hergesell in no 
wise denies the existence of the regular south-west anti- 
trade wind, but maintains that at all seasons it is only to 
be met with some degrees south of the Canaries. 
The Hong Kong Typhoon, September 18.—The Zi-ka- 
wei Observatory (near Shanghai) has sent us some interest- 
ing details relating to the progress of this most disastrous 
storm, which reached Hong Kong on September 18. The 
first signals of its approach were given by the Japanese 
observations in the islands east of Formosa on the morning 
of September 15, but owing to the distance of the stations 
from the central vortex it was not until the following 
day that it was clearly shown to be moving towards 
Formosa and China. The supplement to the Zi-ka-wei 
Daily Weather Report of September 30 contains some 
important extracts from ships’ logs, which clearly show 
the definitive track of the storm. The U.S. transport 
Caesar, bound from Cavite (Bay of Manila) for Shanghai, 
was at noon of September 15 in lat. 19° 53’, long. 
120° 20’; at 4h. p.m. she had a steady wind from N.W., 
freshening in force to 7; the usual diurnal barometric 
range was still observed, but at 8h. p.m. the barometer, 
which stood at 29-66 inches, began to fall, the wind 
freshened and veered to N.N.W., and the ship was forced 
to steer S.W. to avoid the centre of the approaching 
storm. The observations were:—at 1th. a.m. on Sep- 
tember 16, barometer 29:36 inches, wind N.W. 11; at 
2h. a.m., 29-37 inches, W.N.W. 11; at 3h. a.m., 29-40 
inches, W. 11; during this period the rain was con- 
tinuous and excessively heavy. The centre of the storm 
passed between the ship and the south Cape of Formosa 
on September 16, shortly after 1h. a.m. 
The P. and O. SS. Delhi was just entering the passes 
of Hong Kong when the typhoon burst upon the colony ; 
at noon on September 17 she was in lat. 17° 58’, long. 
111° 35’, about 420 miles from the vortex, wind S., 
force 2, and a distinct E.N.E. sea swell was noted. It was 
not until 4h. a.m. next day that the breeze veered to W., 
with occasional squalls, barometer 29-78 inches. The ship 
dropped anchor near Green Island, and the wind freshened, 
being W. by N., 8 at gh. 30m. and W. by S., 10 at 
roh. a.m., while the rain fell with blinding violence; at 
gh. 45m. the barometer reached its lowest point, 29-14 
inches. The centre of the storm passed to the north of 
the Delhi between gh. 45m. and roh. a.m. As shown by 
the observations of these two vessels and those of the 
French mail steamer Océanien, which left Hong Kong for 
Shanghai on the afternoon of September 17, the centre of 
the storm travelled from Formosa to Hong Kong, about 
380 miles, in 563 hours, at a mean rate of 6.7 miles an 
hour; the rate of translation was probably checked by 
the very high atmospheric pressure to the north, but it 
total precipita- 
directly by the 
