
Sept. 21, 1871] 
NATURE 
413 

is also the variation due to the height of the observer’s 
station above the sea. This is, of course, of the first 
importance. The other fluctuations are comparatively 
unimportant, and do not blind an observer to those omi- 
nous fluctuations which precede the storm, the tornado, 
and the hurricane, The oscillations which indicate a storm 
are very marked. The tornado which recently ravaged 
St. Louis was preceded by a gradual fall of the mercury 
in the barometer, for thirty hours previous, of an entire 
inch. A: Boston, within thirty-seven years, the barometer 
has ranged from 31°125 inches to 28°47 inches, the diffe- 
rence being 2°655 inches. At London it has ranged 
through more than 3°5 inches ; but in the tropics not so 
much. 
During the passage of a cyclone the mercury oscillates 
rapidly. The most noticeable fall occurs from four to six 
hours before the passage of the storm centre. 
is often over an inch, and sometimes two inches. 
Great changes are usually shown by falls of barometer 
exceeding halt aninch, and by differences of temperature 
exceeding fifteen degrees. If the fall equals one-tenth of 
an inch an hour we may look out for aheavy storm. The 
more sudden the change the greater the danger. But it 
is too often forgotten that the fall of the mercury is a fore- 
warning of what will occur in a day or two, rather than in 
a few hours, 
A variation of an inch is certain to be followed by a 
tornado or violent cyclone. In the tropics “the glass” 
has been known to show a fall of more than an inch and 
a half in one hour! 
The following guides in predicting weather changes are 
selected from the “ Barometer Manual” of the London 
Board of Trade, and are suggestive : 
I. If the mercury standing at thirty inches rise grad- 
ually while the thermometer falls, and dampness becomes 
less, N.W., N., or N.E. wind ; less wind or less snow and 
rain may be expected. 
Il. If a fall take place with a rising thermometer and 
increasing dampness, wind and rain may be expected from 
S.E., S., or S.W.; a fall in winter with a low thermometer 
foretells snow. 
III. An impending N. wind before which the baro- 
meter often rises may be accompanied with rain, hail, or 
snow, and so forms an apparent exception to the above 
rules, for the barometer always rises with a north wind. 
JV. The barometer being at 29} inches, a rise foretell!s 
less wind or a change of it northward, or less wet. But 
if at 29 inches a fast first rise precedes strong winds or 
squalls from N.W., N., or N.E., after which a gradual rise 
with falling thermometer, a S. or S.W. wind will follow, 
especially if the rise of the thermometer has been sudden. 
V. A rapid barometric rise indicates unsettled, and a 
rapid fall stormy weather with rain or snow ; whileasteady 
barometer, with dryness, indicates continued fine weather, 
VI. The greatest barometric depressions indicate gales 
from S.E., S., or S.W.; the greatest elevations foretell 
wind from N.W., N., or N.E., or calm weather. 
VII. A sudden fall of the barometer, with a westerly 
wind, is sometimes followed with a violent storm from the 
N.W., N., or N.E. 
VIII. If the wind veer to the S. during a gale from. the 
E. to S.E., the barometer will continue to fall until the 
wind is near a marked change, when a lull may occur. The 
gale may afterward be renewed, perhaps suddenly and 
violently ; and if the wind then veer to the N.W., N., or 
N.E., the barometer will rise and the thermometer fall. 
IX. The maximum height of the barometer occurs 
during a north-east wind, and the minimum during one 
from the south-west; hence these points may be consi- 
dered the poles of the wind. The range between these 
two heights depends on the direction of the wind, which 
causes, On an average, a change of half an inch; on the 
moisture of the ‘air, which produces in extreme cases a 
change of half an inch ; and on the strength of the wind, 
This fall- 

which may influence the barometer to the extent of two 
inches. These causes, separately or conjointly with the 
temperature, produce either steady or rapid barometric 
variations, according to their force. 
PRESENT OPERATION OF THE SERVICE 
Although the Signal Service is yet in its infancy, and 
must be patiently nursed and cherished by the people for 
some years before it can expect to do and discharge its 
full mission, under General Myer’s iadefatigable care 
and skilful management it has already achieved much 
good, and more than compensated the public for the ex- 
pense of its establishment. Since it was instituted last 
summer, “the chief signal officer has,” to quote the words 
of the New York World, “thoroughly organised and 
equipped a system which now embraces in its scientific 
grasp every part of the land from Sandy Hook to the 
Golden Gate of California, and from Key West to the 
Dominion of Canada.” 
Three times every day synchronous observations are 
taken and reports made from the stations—one at eight 
A.M., one at four P.M., and the third at midnight. These 
observations are made by instruments, all of which are 
perfectly adjusted to a standard at Washington. They 
are also all taken at the same moment exactly, these 
observations and reports being also timed by the standard 
of Washington time. The reports from the stations are 
transmitted in full by telegraph. By a combination of 
telegraphic circuits, the reports of observations made at 
different points synchronously are rapidly transmitted to 
the different cities at which they are to be published. 
They are, however, all sent of course to the central office 
in Washington. These reports are limited to a fixed 
number of words, and the time of their transmission is 
also a fixed number of seconds. These reports are not 
telegraphed in figures, but in words fully spelled out. 
There are now about forty-five stations for which provision 
has been made, and which are in running order. These 
have been chosen or located at points from which reports 
of observations will be most useful as indicating the 
general barometric pressure, or the approach and force of 
storms, and from which storm warnings, as theatmospheric 
indications arise, may be forwarded with greatest despatch 
to imperilled ports. 
These stations are occupied by expert observers fur- 
nished with the best attainable instruments, which are 
every day becoming more perfect, and to which other 
instruments are being added. 
The reports of observers are as yet limited to a simple 
statement of the readings of all their instruments, and of 
any meteorological facts existing at the station when their 
tri-daily report is telegraphed to the central office in 
Washington. 
Each observer at the station writes his report on mani- 
fold paper.* One copy he preserves, another he gives to 
the telegraph operator, who telegraphs the contents to 
Washington. The preserved copy is a voucher for the 
report actually sent by the observer ; and if the operator 
is careless and makes a mistake, he cannot lay the blame 
on the observer, who has a copy of his report, which must 
be a fac-simile of the one he has handed to the operator. 
The preserved copy is afterwards forwarded by the 
observer-sergeant to the office in Washington, where it is 
filed, and finally bound up in a volume for future 
reference. 
When all the reports from the various stations have 
been received they are tabulated and handed to the officer 
(Prof. Abbé) whose duty it is to write out the synopses 
and deduce the “ probabilities,” which in a few minutes 
are to be telegraphed to the press all over the country. 
* Thin paper with black carbon paper between the sheets. The pen isa 
dry s¢y/us, and being pressed on the upper sheet, it makes a similar mark on 
the sheets beneath it. 
