Skip to main content

Full text of "Weather charts and storm warnings"

See other formats









Weather Charts &. Storm Warnings . 

Plate I. 




Januaiy 1887. 











All rights reserved 

/I I) 

/ / 



THE second edition of this book having been long out 
of print, I have been requested by several friends to 
prepare a new edition. 

The entire text has been revised, and some new 
chapters have been added, mainly relating to changes 
and improvements in the Office weather work since the 
year 1876. 

The subjects of the new chapters are, the Distribution 
and Periodicity of Gales, Weather Forecasting, the 
Weekly Weather Report, and lastly, Cirrus Cloud Obser- 
vations. The subject of ' Suggestions for Improve- 
ment,' formerly included under 'Storm Warnings,' has 
been made a separate chapter. The chapter on the 
Weekly Weather Report contains an account of the 


iv Preface to the Third Edition. 

method of calculating ' Accumulated Temperature/ in 
use in the preparation of that Report. 

Chapter XI1L, ' On Cirrus Cloud Observations,' is of 
the nature of an Appendix to the book, but it has 
seemed best to constitute it a separate chapter. 


July 25. 1887. 


Tills little work has been put together in the endeavour 
to supply a want which has been expressed in many 
quarters : that of an explanation of the weather charts 
which appear in the newspapers, and of the remarks 
which are appended to them. 

The conceptions and principles on which the science 
of weather study is based are apparently quite new to 
the majority of ordinary readers, who still hold to the 
belief that the barometer rises or falls in direct relation 
to the weather, without any attempt to consider how or 
ic/iy it does so. 

It is hoped that the following pages may convey 
some idea, however imperfect, of the present state of 
Weather Knowledge, as distinguished from the science 
of Meteorology itself, of which this book makes no claim 
to be called a manual. 

With a very few exceptions, theories of the causes of 
storms have been left unnoticed, as the object has been 

vi Preface to tJie First Edition. 

to explain to the reader what he can learn from a careful 
study of the information published in the newspapers 
or in the Daily Weather Reports, and which is therefore 
accessible to all. 

In treating of a science now in process of rapid 
development, it can only be expected that every year 
will add to our knowledge, and that many of the prin- 
ciples stated in. these pages will be extended or modified 
by the results of subsequent experience. All that it 
is here attempted to give to the public is an account of 
the actual state of our knowledge at present. 

I must express my sincere thanks to several friends 
who have aided me with their advice during the prepa- 
ration of the book, and most notably to Mr. Frederic 
Caster, of this Office. 

The illustrations which are contained in the text 
have all been executed by the Patent Type Founding 
Company, by the same process as they employ for the 
production of charts for newspapers. 


June 10, 1876. 

















\. WEATHER REPORT. NOVEMBER 29, 1874 .... 218 

MARCH 26, 27, 1874 221 


OCTOBER 20, 21, 1874 222 


FEBRUARY i, 2, 1873 223 




ACCORDING TO CLEMENT LEY .... to face p. 65 





</ P. 46, diagram, insert " .02" before "c." 
J line 6 ab infra, for " 5 " read " 10." 

J P. 194, at head of table insert ''Figures for week ending 
April 25, 1887." 


The observations taken at these stations refer to 
Atmospherical Pressure, Temperature, Humidity (or the 
Dampness of the Air), Wind, Rain, Weather, and, at sea- 
coast stations, Sea Disturbance. 

The drawing of weather charts depends on the com- 







BEFORE proceeding to describe the charts themselves, 
or to give an account of their utility in aiding us to form 
a judgment as to probable weather, it will be well to 
explain the character of the observations and the nature 
of the service on which they are based, confining our- 
selves exclusively to the instruments employed and 
observations taken at our own telegraphic reporting 
stations (the outfit of which is necessarily less complete 
than that of a fully furnished meteorological observatory), 
and omitting technical descriptions of the instruments 

The observations taken at these stations refer to 
Atmospherical Pressure, Temperature, Humidity (or the 
Dampness of the Air), Wind, Rain, Weather, and, at sea- 
coast stations, Sea Disturbance. 

The drawing of weather charts depends on the com- 


2 Weather Charts and Storm Warnings. 

parison of various observations taken at the same time, 
at several different stations. It is therefore indispens- 
able, that all the observations to be compared should 
be freed, as far as possible, from incidental inaccuracies 
and discrepancies, by reducing the conditions at each 
place to common standards of comparison, and this is 
done by applying the several instrumental and other 

Atmospherical pressure is measured by the 
barometer, and by this is meant the mercurial baro- 
meter ; for the aneroid, however convenient it may be 
for ordinary use as a weather glass, cannot be recog- 
nised as an independent instrument, its indications 
depending entirely on the delicacy of the mechanical 
appliances employed in its construction. 

The barometer has a thermometer attached to it, in 
order to show the temperature of the instrument itself, 
and this must be read whenever an observation of the 
height of the mercurial column is taken. 

The readings of the barometer are always said to be 
'corrected and reduced to 32 and to sea-level,' and it 
is necessary to explain these expressions. 

' Corrected,' means that corrections have been applied, 

a, for the error of the scale of the instrument, which 
has been ascertained by sending it for verification to some 
recognised establishment, such as Kew Observatory. 

b, for what is called ' capillarity/ which depends on 
the bore of the tube, and 

c, for what is called 'capacity,' which depends on 

The Materials available for Weather Study. 3 

the proportion which the bore of the tube bears to the 
sectional area of the cistern. 

4 Reduced to 32,' means corrected according to the 
reading of the attached thermometer. The column of 
mercury in the barometer tube behaves like almost all 
other bodies, being lengthened by heat, and shortened 
by cold. It is therefore obvious that unless two similar 
barometers be precisely at the same temperature, they 
cannot possibly read alike, and hence is apparent the 
absolute necessity of reducing all readings to one 
common temperature ; for this the temperature of 
melting ice, 32 R, has been unanimously adopted as 
the best. 

The temperature of the column being indicated by 
the reading of the attached thermometer, we can 
discover, by means of printed tables, how much, for each 
reading of that thermometer, the column is longer, or 
shorter, than it would be at 32, and consequently what 
correction is required for the barometer reading. 

' Reduced to sea-level,' is a phrase which requires 
rather more explanation than the foregoing. The 
barometer, as its name implies, measures the weight of 
the air, and that weight, of course, depends mainly on 
the quantity of air which is above the instrument 
pressing upon the cistern, and keeping up the column 
of mercury in the tube. If, therefore, two similar 
barometers be placed one directly under the other, say, 
one in the attic, and the other on the ground-floor, it is 
evident that there will be a less quantity of air above 

13 2 

4 Weather Charts and Storm Warnings. 

the former than above the latter instrument, and, 
consequently, the barometer in the attic will read lower 
than that on the ground-floor. 

Precisely the same reasoning will apply if we are 
considering two barometers at different heights on the 
side of a mountain, or one at an inland station, and the 
other at the level of the sea ; in every instance the 
barometer at the higher station will read lower than 
that at the lower. 

Hence we find that all readings must be reduced to- 
their equivalent readings at a standard elevation, as well 
as at a standard temperature. This standard elevation is 
the Mean Level of the Sea, and the reduction is carried 
out by means of tables. 

A slight consideration of the foregoing remarks, and 
of the last-named principle in particular, will show how 
it comes about that the reading given in a Daily 
Weather Report for an inland station, like Oxford, or 
Nottingham, can never agree with a casual reading taken 
by an observer at either of those stations at the same 
time, unless the latter reading be corrected and reduced 
to 3 2, and to sea-level. It must, however, be remem- 
bered that the problem of the reduction to sea-level is 
not so very simple. Its correctness depends on the 
accuracy of the knowledge we possess of the temperature 
of the supposed column of air extending from the level 
of the upper station to that of the lower. Now if the 
height of this column be considerable, our knowledge of 
this temperature cannot be complete, and therefore for 

The Materials available for Weather Study. 5 

elevations above one thousand feet, the reduction by the 
tables is not quite exact. Accordingly, for such stations 
as that on Ben Nevis, at the height of 4000 feet, the 
reduction of the readings to sea-level by tables is not 
very generally accepted. 

Temperature. This does not require much expla- 
nation. In addition to the actual readings of the ther- 
mometers, ' in the shade ' (i.e., suspended in a properly 
constructed louvre-boarded screen at a height of four feet 
above the ground), which are taken at certain definite 
hours, the Daily Weather Reports give also two columns, 
showing, respectively, the maximum and minimum read- 
ings taken during the twenty-four hours, ending at 8 
o'clock each morning, by thermometers so constructed 
as to register, one the highest, and the other the lowest, 
temperature reached during the interval which has 
elapsed since the last setting. 

The last-named readings may also be turned to very 
useful purpose. The mean temperature of the day is 
often mentioned in connection with sanitary statistics, 
as for instance in the Registrar- General's Reports. 
This mean temperature is theoretically the average of 
twenty-four thermometer readings taken at hourly 
intervals during the day. There are well-known rules 
for determining this mean temperature out of various 
combinations of hours of observation. The simplest 
method, and, for practical purposes, nearly the most 
exact, is to take the average of the maximum and 
minimum readings given in the Daily Weather Report^ 

6 Weather Charts and Storm Warnings. 

on any day, and put that down as the mean temperature 
of the preceding day. 

Humidity. The amount of moisture in the air is 
measured by the indication of a thermometer with its 
bulb encased in muslin, and kept damp (a wet bulb), 
compared with that of an ordinary (dry bulb) thermo- 
meter, without any such mounting, observed at. the 
same moment. Tables exist from which the amount of 
moisture for each observed difference in readings between 
the two thermometers can be ascertained. 

This is not the place to enter into the theory of these 
observations. Suffice it to say, the greater the differ- 
ence between the readings of the two thermometers the 
drier is the air, and that when the two thermometers 
read alike, the atmosphere is exceedingly damp. The 
difference between the two thermometers ranges in 
this country from o to 10 or even 15, and sometimes 
a difference of upwards of 20 has been noticed. 

The chance of rain depends to a great extent on the 
degree of humidity of the air from time to time, and if 
we are dealing with reports from an extensive tract of 
country, as North America, or the Continent of Europe, 
the distribution of the moisture is an interesting subject 
of study. As regards these islands, the fact that most 
of the telegraphic stations are on the sea-coast detracts 
from the value of the reports of humidity, inasmuch as 
the amount of moisture in the air is seriously affected 
by proximity to the sea ; and so in weather study we 
are unable to use reports of vapour tension as much as 
our neighbours. 

The Materials available for Weather Study. 7 

Rain is measured by means of a rain-gauge. The 
values given represent the depth of water which would 
have accumulated on a level piece of ground, if none of 
the rain which fell had escaped by drainage, &c. Too 
much faith must not be placed on the rain returns in 
the daily reports, for two reasons. Firstly, the gauges 
are often necessarily placed in towns, where a good 
exposure is not to be had, so that the amount measured 
is not the same as might be yielded by a gauge in a 
more open situation ; and secondly, the telegraphic 
stations are so sparsely distributed over the kingdom 
that it is impossible that they can give a precise 
account of the rainfall in every county. What they 
do show is whether or not the rain is general and 

Wind. As a general rule, for the purposes of the 
Daily Weather Report, wind is not measured by an 
instrument, but is simply estimated according to the 
Beaufort scale, which is so named after the late 
Admiral Sir Francis Beaufort, who devised it for his 
own use, when in command of H.M.S. Woolwich in 
1805, The following is the scale, 1 with the approxi- 

1 Since Admiral Beaufort's time there has been a great change in the 
rig of ships by the introduction of double topsail yards. It seems 
therefore advisable to add to Beaufort's scale the amount of sail which his 
ship would have been able to carry, had she been rigged with double 
topsail yards, but under all other circumstances the same. The change 
would only affect forces 6 to 10. 

6. Topgallant sails. 

7. Topsails, jib, c. 

8. Reefed upper topsails and courses. 

9. Lower topsails and courses. 

10. Lower maintop-sail and reefed foresail. 

8 Weather Charts and Storm Warnings. 

mate equivalent velocity of the wind in miles per hour, 
for each grade as determined in the Meteorological 

per hour 

0. Calm -3 

1. Light air . Or, just sufficient to give steerage way ... 8 

I Or, that in which a well-\ 

conditioned man-of-war, i to 2 knots . . . 13 

with all sail set, and 1 3 to 4 knots . . . 18 

' clean full, ' would go in ! 5 to 6 knots . . 23 
smooth water from . J 

5. Fresh breeze . 

6. Strong breeze 

7. Moderate gale 

/Royals, &c. . . . 28 
Single-reefed topsails 

and topgallant sails . 34 
Or, that to which she Double-reefed topsails, 

could just carry in chase, ( jib, &c. . . .40 
Fresh gale . ' full and by ' . . i Triple-reefed topsails, 

I &c 48 

9. Strong gale .1 .... Close - reefed topsails 

/ v and courses . .56 

10. Whole gale . Or, that with which she could scarcely bear close- 
reefed maintop-sail and reefed foresail . . -65 

i r. Storm . . Or, that which would reduce her to storm-staysails . 75 
12. Hurricane . Or, that which no canvas could withstand . . 90 

It is obvious that as this scale refers to the rate of 
sailing of, or to the amount of sail carried by, a ship, it 
is at first sight not well suited for use at land stations ; 
but experience has shown that the reporters, guided by 
the descriptions in the second column, are never very 
far from the truth in their estimates, when once they 
have gained some practice in observing. There are 
objections to supplying instruments for the measure- 
ment of wind to telegraphic stations, owing to the great 
difficulties which would be experienced in erecting them 
in suitable positions. Anemometrical indications taken 
in the middle of a town are almost worthless, as the 

The Materials available for Weather Study. 9 

buildings produce such eddies that the true movement 
of the air cannot be ascertained from the instrument. 

As regards the numbers of the Beaufort scale, it is 
those from 6 upwards which present the most interest. 
6 is the lowest number which is taken in the Meteoro- 
logical Office to justify the issue of a warning to the 
coast, and 9 is the lowest figure which by the regulations 
of the Board of Trade can be pleaded by a captain 
as ' stress of weather,' in case of casualty to his craft. 
These velocities are not uniform, like that of an express 
train. The statement of 56 miles an hour, means 
56 miles in the hour, but during that hour the wind 
may have been gusty, and at times have had a velocity 
of nearly 100 miles an hour, while at other times its 
hourly speed may have scarcely reached 30 or 40 

It may be remarked that if we employed the pres- 
sure, instead of the velocity of the wind, to measure the 
violence of a gale, we should be able to obtain a record 
of individual gusts, and in the opinion of many persons, 
such as engineers, &c., such information would be more 
practically useful to the public, than statements as to 
velocity. The reason that pressure anemometers are 
not generally adopted is, that as yet it has not been 
found possible to reason with sufficient confidence as to 
the pressure of the wind on a structure, such as a 
factory chimney, from the indications shown by a 
pressure plate of, say, one square foot in area. Till the 
influence of the size of the pressure plate on its indica- 

i o Weather Charts and Storm Warnings. 

tions, in a wind of given strength, has been thoroughly 
determined, it seems premature to recommend the em- 
ployment of pressure anemometers. 

Weather. The various observations which are 
comprehended under this general term are those which 
do not admit of instrumental record, such as the fact of 
the occurrence of thunder, &c. These are reported 
according to the subjoined notation, which, like the 
wind scale, is due to Sir F. Beaufort. 

b Blue sky, whether with clear or 

hazy atmosphere. 
c Cloudy, but detached opening 


d Drizzling rain. r Kain ^inued 

/ Foggy. s Snow ' 

t Thunder. 


Dark gloomy weather. 

h HaiK of the weather. 

/ Lightning. 

m Misty hazy atmosphere, 

o Overcast, the whole sky being 

covered with an impervious 


p Passing temporary showers. 
q Squally. 

u ' Ugly,' threatening appearance 

Visibility,' whether the sky be 
cloudy or not. 

w Dew. 

The only thing to be remarked about the use of 
these letters in the Daily Weather Reports is that, 
when two or three are entered to the same place on 
the same day, it implies that the observations have 
been taken in the order in which they are printed. 
Thus b.c.p.r. would mean that, during the interval which 
had elapsed since the previous report, the sky was at 
first ' quite clear/ then ' detached clouds ' came over it 
* passing showers ' ensued, at last turning to persistent 
' rain ' which ' continued ' at the time of sending off the 

The Materials available for Weather Study, n 

Sea Disturbance. Lastly, the table contains a 
column for the Sea Disturbance, of which there are 
nine grades, depending on estimation, like the scale for 
wind force. 

0. Dead calm. 

1. Very smooth. 

2. Smooth. 
3- Slight. 

4. Moderate. 

5. Kather rough. 

6. Rough. 

7 . High. 

8. Very high. 

9. Tremendous. 

The height of the waves not unfrequently gives valu- 
able information as to the force of the wind in the 
offing, but we must not place too much dependence on 
this particular indication, for many of our most serious 
storms have come on without any previous disturbance 
of the sea on our coast. This is due to the fact that 
the sea disturbance is caused by the wind outside, and 
if that wind is not blowing in the direction of a coast, 
it will not impel waves towards that coast. It will 
presently appear how a wind may advance to a coast 
without blowing towards it, but for the present it may 
be said, that the fact of a wind blowing in the direction 
of these islands, and driving a heavy sea on to parts of 
our coasts, is no positive proof that the wind itself will 
ever be felt on our shores. In fact, some of the 
heaviest seas reported on our western coasts are proved 
to have been caused by Westerly gales blowing far 
out in the Atlantic, which never reached Europe 
at all. 

There are yet other remarks to be made as to the 
sea disturbance column, which is, after all, a portion of 

1 2 Weather Charts and Storm Warnings. 

our table often consulted anxiously by such landsmen 
and landswomen as may have to cross the Channel. 
In the first instance, the sea close in shore, at a place 
like Dover, is often much calmer than it is outside, as 
that harbour is sheltered from several winds, especially 
Northerly ones. Secondly, the roughness of the sea 
depends in great measure on the direction in which the 
tide is running at the time of observation. 'Wind 
against tide ' knocks up a sea at once, and frequently 
the delay of a few hours may make all the difference 
between the miseries of sea-sickness, and the enjoy- 
ment of perfect comfort, if the tide has changed in the 

These, then, are the materials with which we have to 
deal in drawing charts, discussing weather, preparing 
forecasts, and in sending warnings to the coast, and it 
is self-evident that they are far from complete. 

It is obvious that the amount of information as to 
the appearance of the weather at each station which 
can be conveyed by the three or four letters which, at 
most, are given in the proper column, is quite insuffi- 
cient to convey to our minds any clear idea of what an 
experienced person would have gathered as to the 
general character and prospects of the weather at that 
station ; and yet the general appearance of the 
weather was all that our fishermen and pilots had to 
guide them for generations, nay, for centuries, before 
the barometer was invented. Right well did they then, 

The Materials available for Weather Stiidy. 1 3 

and do their successors now, know how to profit by 
such signs ! 

In the first place, the code is necessarily so con- 
densed that it cannot give information as to the form of 
clouds, whether these be the highest ' mare's tail ' or the 
lowest rain cloud, or whether there be more than one 
stratum. The value of such particulars for the purposes 
of weather study need hardly be mentioned. The cha- 
racter of the clouds, their changes, and their amount, 
afford to the practised eyed most valuable information 
as to the condition of the atmosphere above us. 
Moreover, the code does not tell us anything about the 
motion of the clouds, so often the truthful indication 
of coming wind, though we do endeavour to obtain, by 
means of words added to the telegrams, information 
on this motion, when there is anything remarkable 
in it. 

By the researches mainly of the Rev. W. Clement 
Ley in this country, and of Professor Hildebrandsson 
in Sweden, certain principles for the interpretation of 
cloud motion have been established. A brief resume 
of these will be given in Chapter XIII. 

In these and all such particulars, the Weather Tele- 
grams, no matter how correct they may be, are only a 
poor substitute for actual personal observation. Any 
one trying to form a correct judgment of the look of the 
sky from these alone is like a physician trying to deal 
with a case without a chance of a personal interview 
with his patient. What can a resident in an inland 

1 4 Weather Charts and Storm Warnings. 

town like London know of the appearance of the 
weather on the sea-coast on any clay from any tele- 
gram, no matter how detailed ? 

The reports are therefore incomplete as to quality 
and quantity, but in many respects they arc capable of 
material improvement, especially if it should ever be ren- 
dered possible to devote more money to the service. 

As to quality, the instruments should be automatic, 
so that by looking at the record traced mechanically, we 
should be able to notice all that had passed since the 
last observation. Of late years automatic aneroids 
have been introduced at some of the most important 
stations, and have been found to be of considerable 
service. The reporters should be experienced observers, 
with outdoor occupations, so as to be constantly on the 
look-out for any change of weather which may take 
place, and therefore they should be selected from such 
a class of men as signalmen on the coast. It is obvious 
that such a weather report as can be given by a clerk 
who simply runs out to look at the sky just before filling 
up his despatch, cannot be of as much value as that of 
a man who has been in the open air watching the 
weather for an hour or so, or even for the better part 
of the day. As to quantity also, the reports are 
not sufficiently complete, both as regards time and 

As regards time, they are not nearly frequent enough. 
The Signal Office at Washington receives three reports 
every day from each of its stations ; but, as is well 

The Materials available for Weather Study. 1 5 

known, that office is most liberally supplied with funds 
by Congress. Our own Meteorological Office, however, 
can only afford one at 8 A.M. from most of our stations, 
and at best we only get additional reports at 2 P.M. 
and at 6 P.M. from a limited number of places. As 
regards Sunday mornings, the service is necessarily 
much hurried, for the telegraph offices are, as a rule, 
only open from 8 to 10 A.M., and there is no service 
at 2 P.M. at all. On Sunday evenings, however, the 
service is the same as on week-days. 

Lastly, as to space, the area over which our own net- 
work and our international exchange extends, is far too 
limited for us to gain a general idea of the conditions 
which are prevailing all around us, and it is on these 
that our weather in great measure depends. This 
might, to a certain extent, be remedied by an extension 
of our Continental communications, but, as will be 
pointed out later on, the information obtainable from 
Europe is comparatively unimportant when contrasted 
with what we can supply to our Continental neighbours. 
What we really want, to give us a better insight into 
our chances of weather from time to time, is unfor- 
tunately unattainable, and that is a system of reports 
from stations in the Atlantic, say, at a distance of 600 
miles from our coasts, for most of our storms advance 
on us from the Atlantic. It is a problem as yet 
unsolved, to moor a vessel in 1000 fathoms water, 
and to connect her with the shore by a telegraphic 
cable. The experiment of a floating telegraphic station, 

1 6 Weather Charts and Storm Warnings. 

which was tried in I 869 at the entrance of the Channel, 
in much shallower water than I ooo fathoms, was not 
encouraging, as will be explained later (p. 181). It 
has, however, been recently proved by the Telegraph 
Construction and Maintenance Co. to be possible to 
maintain telegraphic communication with a light- 

Nevertheless, in spite of all the geographical defects 
of our insular position, and the difficulties which are un- 
avoidable in the early stages of any branch of knowledge, 
satisfactory progress has been made in the science of 
Weather Telegraphy in these islands. When so much 
has been accomplished in the short space of a quarter 
of a century, since the service was first organised by 
Admiral FitzRoy in 1861, it may be hoped that with 
perseverance, and continued efforts to improve our 
methods, we shall one day arrive at some clearer know- 
ledge of the laws of our storms and our weather. 

The following is a specimen of the Daily Weather 
Report, as issued in January 1887. A chart showing 
the stations will be found as the Frontispiece. 

The matter is exhibited on four lithographed pages. 
The contents of three of these are reproduced here. 
The fourth appears as Plate III., but the actual charts 
given there do not refer to January i, 1887. 

The Remarks on the General Situation at 8 A.M., and 
on the Weather over Europe during the past twenty- 
four hours, explain themselves. 

The Forecasts will be explained in Chapter X. 

The Materials available for Weather Study. 17 

It will be seen that in addition to the absent obser- 
vations, which are marked with an asterisk, as is ex- 
plained at foot of the table, various spaces are marked 
with dots (...), indicating that the' report had not 
arrived for that day ; and others with notes of interro- 
gation, indicating that for some reason or other the 
report appeared to be doubtful, in which case it is 
sometimes inserted and queried, and at others omitted 


Weather Charts and Storm Warnings. 








24 HOURS. 

jo lunouiy 

03 - - - 

co . ... 

O 00 00 M 

CM o H o : 




o" d d d 






0^. c. 


ro (N >o 

Q - 





(6 OJ 0) -B3g 






(si o; o) 


CONN co 


w H . w - u 





wv g UIQJJ 

ro o 



+ + + + 

+ 4- + + 


+ + 1 1 1 


<S 2 1 " 5- 

^k ^ctct 




'j\rv 8 TOOJJ 


ro O O >< r^. 
N H q q q 

O O >0 00 M 

1 1 1 1 1 

SOO M 10 w 
M . M . M . M . 

r r r r r 

1 1 1 


^ c5 6 d c^ 

do odd 

*0 C? M H S" 

^ $ ^ 5. !J 





00^ .0 

,00 *~ o a 

!- lH tH 

6^,-O-n g 

o o o 

(si o; o) 


N N N co 


-1 ^-'^ 





ui 'duiajQ 

H cs ro 

^ n^ 



N M \0 

pA3T[ B3g pUB 

sin i 

o?o r^PoPo 





. ." .| . 


in* ' 

^ "^ 

1 <B.'H 


! OcoU cofe 
^^ . y^ 



Leith . 
Shields . 
Spurn Head 

I Malin Head 


Materials available for Weather Study. 19 

$m "ScTmloS W 

vp M e vo vo * ro 

vo vo * ro N vo 

m \ * * 10 (O 


^j- co ro 

mTfl ^m^-mml c? m 5 i? ? 

moo voOrx>o--^-o txO'-^-rxMO oo-* 
NNOINPIN :Mro omlMmmm mm 

c?^^ m f?mc?cr c m mm ?!% 

o t- vovo o> N 

2 SctS 3 S"8SS-J? MM" 

+ i i i 



o ro ro ^J- ro 

ro O nco oo 

vo M c> in 

fO fO W 


O O 0-2^3 O 

6 oj<~ g Low*, o La* ^^^ I 

! I i 

O O O O O O * 

tx * vo 


GO WJ C/3 C/3 

rx ^ tx M N M 
m ro m m ro * 

vo O 
UT 1 10 1 

ror^i r?> me? 



C 2 

2o Weather Charts and Storm Warnings. 











Skudesnaes . 








Sumburgh Head . 








Stornoway . 
































Malm Head 
















Belmullet . 
























Scilly . 








Dungeness . 















Yarmouth . 







4 . 

Jersey . 
Rochefort . 









WEATHER. Beaufort scale is given on p. 10. 

* An asterisk is inserted in all places for which information is not 
usually received. 

\ The evening observations are taken at 6 P.M. in France and our 
Islands, with the exception of Oxford, where they are taken at 8 P.M. In 
Sweden and Denmark, the observations are taken at 8 p. M. (Christiania 




/Highest, 30.5 inches and upwards over the South of 


J Lowest, 29.9 inches and less over the Gulf of Genoa, 

but relatively low, 30.0 inches and less, to the North- 

(^ westward of our Islands and the Westward of Norway. 

(Highest, 48 at Valencia, 47 at Belmullet, 46 at 
Roche's Point, and 45 at Stornoway and Mullagh- 
Lowest, 6 at Stockholm, 12 at Ilernosand, 14 at 
, Cambridge, 15 at Oxford, and 20 at Munich. 
Hard frost prevailed last night over Central England, 
the lowest readings being 13 at Cambridge, 17 at 
\ Loughborough, and 22 in London. 

The Materials available for Weather Study. 21 




/ Light or moderate from NE. over Germany, France, 
the Netherlands, and the English Channel. 

(Southerly elsewhere, strong in force on the West 
coasts of Ireland and Scotland. 

Dull on our Western and Northern coasts, with driz- 
zling rain at Stornoway. 
Fair in most other localities, but densely foggy over 

Central England and the Netherlands. 
(Rough at Stornoway and Valencia, but moderate to 
\ smooth elsewhere. 

The anticyclone noticed over England and the North Sea yesterday 

appears to be still moving Eastwards, and the barometer is falling generally. 

The depressions which are travelling along our Western coasts appear 

likely to advance closer to our shores, and a further increase of wind is 

probable over Ireland and Scotland. 

No important change is likely at present over England beyond a gradual 
rise of temperature. 


2ND JANUARY 1887. 



Southerly and South-westerly winds, increasing 
in force ; cloudy, some rain. 

Do. Do. Do. 

Southerly winds, moderate, fair generally. 

Same as No. 5. 

Same as No. 5. 

Easterly and South-easterly winds, light; 

foggy to fair. 

South-easterly and Southerly winds, freshen- 
ing ; cloudy, some rain. 

Do. Do. Do- 

Do. Do. Do. 

Southerly winds, strong, perhaps a gate, 

cloudy, some rain. 
*io. IRELAND, SOUTH . | Do. Do. Do. 

WARNINGS. The South cone was hoisted last evening in districts o, 6, 
9, 10, and part of I. An asterisk is inserted against those districts 
in which signals are now flying. 











2 2 Weather Charts and Storm Warnings. 


THE reports for yesterday showed that a large anticyclonic system had 
spread over the whole of Northern and North-western Europe, its centre 
being over the Eastern parts of the British Isles and the North Sea, where 
the barometer at 8 A.M. stood at a little above 30.5 inches. In the West 
and North-west of our Islands, however, the mercury had begun to fall 
steadily, while to the South-eastward of the high-pressure area readings 
decreased to 29.5 inches over the Adriatic. Thus, while South-westerly 
winds had set in over Ireland and Scotland, calms and variable airs were 
felt over England and the North Sea, and North-easterly to Northerly 
winds over the South of France and Northern Italy. The weather in the 
NW. and N. was dry, cold, and very foggy, while at the more Southern 
stations it was somewhat cloudy, with showers locally. Temperature was 
below the freezing-point at all but the most Western and Southern stations ; 
over the Western and Southern parts of the Mediterranean it was still above 
50. The highest individual readings were 53 at Algiers, 52 at Malta, 
51 at Brindisi, and 50 at Constantinople ; the lowest, 5 F. at Hernosand, 
Stockholm, and St. Petersburg, 9 at Archangel, 14 at Helsingfors, 20 at 
Loughborough, 23 at Oxford, 26 at Cracow, 27 at Lemberg. 

During the day the South-westerly winds in the West and North-west 
continued to blow steadily, but did not spread much over England, where 
dense fogs prevailed locally, and the night became very cold. 

The Wind. 23 



HAVING thus discussed the observations themselves, we 
come to the results and conclusions to be drawn from 
them, and, in order to make these intelligible, we must 
indulge in a little theorising, and refer to some things 
which, though not yet recognised as absolutely proved, 
appear at least probable, as regards the behaviour of 
the wind. 

Every one knows that the East is very different in 
its character from the West wind, the former being 
reputed to be 'good neither for man nor beast.' Any 
one with a touch of bronchitis or rheumatism can (or 
at least thinks he can) tell you, without looking at a 
weather-cock, whether or not there is Easting in the 
wind. This contrast arises from the fact that as a rule 
the temperature and dampness of the air in Western 
Europe are both lowest when the wind is about North- 
east, and highest when it is about South-west. This 
is, however, a local phenomenon peculiar to certain 
parts of the globe, for if we travel from Western 
Europe, along the same parallel of latitude as our own, 

24 Weather Charts and Storm Warnings. 

either eastwards to the Sea of Ochotsk or westwards to 
Labrador, we shall find that in these districts the 
coldest wind is near North-west, and the warmest about 
South-east In each case the coldest and driest point 
of the compass lies towards the region, in the neigh- 
bourhood of the point of observation, where the mean 
temperature is the lowest. In these islands, in winter, 
this region is northern Russia, in Labrador it is the 
Barren Grounds of the Hudson Bay Territory, and at 
the mouth of the Amur it is the district of Yakutsk. 
These two latter cold regions lie to the north-west of 
the respective coast districts to which reference has 
been made. 

The wind, then, is cold and dry when it comes from a 
cold region, warm and moist when it comes from a 
warm district, such as, in these latitudes, the sea surface 
in winter. In summer there is not so strong a contrast 
between the temperature of different parts of the earth's 
surface and between land and water as in winter, and 
so the different winds do not differ so much in their 

In former years it was the fashion to say, that all 
cold winds flowed from the Poles to the Equator, 
forming the so-called Polar Currents, and becoming the 
Trade Winds when they approached the Tropics, while 
the warm winds flowed from the Equator to the Pole, 
forming the Equatorial Currents, or Anti-trades. 

There can be no doubt that these statements were 
right in principle, as the original disturbing action 

The Wind. 25. 

which produces motion in the atmosphere is the heat 
of the sun ; and this acts most strongly in the Torrid 
zone, which ought, accordingly, to be the region 
towards which all the cold air flowed, were it not for 
the irregular arrangement of land and water on the 
globe. Recent investigations have, however, shown 
that a body of air cannot be proved to have made its 
way direct from the Equator to either Pole, inasmuch 
as it will probably have been caught up on its way by 
some of the eddies and local circulations at all times 
existing on the earth's surface. The motions of the 
atmosphere are found to be regulated by the distribu- 
tion of barometrical pressure over the globe, the 
particles moving from the regions where pressure is in 
excess the barometer is high, to those where pressure 
is in defect the barometer is low, and being modified 
in the direction of their motion by various causes, 
among which the earth's form, and its rotation on its 
axis, are the most influential. 

We have said that a body of air cannot be proved 
to flow the whole way from the Equator to the Poles 
or back again, but nevertheless it is the fact that over 
extensive areas of the earth's surface the wind does 
maintain a constant direction for a considerable period 
of time. To prove this it is only necessary to cite the 
well-known phenomena of the Trade Winds and the 
Monsoons, where the wind blows persistently from the 
same direction, in one case for the whole year, and in 
the other for months at a time. In our own latitudes- 

26 Weather Charts and Storm Warnings. 

the phenomena of the winds are not so regular, but yet 
it is found that over large tracts of Europe and the 
North Atlantic, the wind at times blows for weeks 
together in a definite direction, either Easterly or West- 
erly (speaking in general terms), and that these tracts, 
or the channels of these great currents, lie the one 
alongside the other. Occasionally very intense cold is 
experienced in winter over Central and Southern 
Europe, while these islands enjoy mild weather. It 
will be found that in such cases the winds are all 
Easterly where the weather is cold, and all Westerly 
where the weather is warm. 

It will be explained subsequently how the motion of 
the air, both in direction and velocity, is regulated by 
the distribution of atmospherical pressure at the surface 
of the earth, which is shown by the distribution of the 
readings of the barometer in weather charts ; but as 
this relation lies at the very foundation of the whole 
structure of modern weather knowledge, it will be 
.useful, at the very outset, to explain it. 

This principle, which, for convenience, is known as 
Buys Ballot's Law, from the name of the Dutch 
Professor who first stated it definitely, is contained in 
-the following statement. 

Stand with your back to the wind, and the 
rbarometer will be lower on your left hand than 
,on your right. 

These words hold good, except close to the Equator, 

The Wind. 27 

for the Northern hemisphere ; in the Southern we must 
interchange left and right. If this principle be once 
thoroughly recognised, the broad features of wind 
motion will be at once understood. 

The Force of the wind, as distinguished from its 
Direction, is related to the amount of difference of 
barometrical pressure over a given distance, and this is 
defined as the ' gradient,' a term which will be explained 
in Chapter IV. (p. 44). This force therefore in no way 
depends on the absolute height of the barometer at any 
one station, as the words printed on the scales of many 
old-fashioned barometers would seem to indicate. In 
fact, as an illustration of this statement we may anti- 
cipate the future explanation by stating that at 
Liverpool, during a severe storm on January 24, 1876, 
from 2 till 3 A.M., the velocity of the wind from SW. 
was sixty-two miles within the hour, while the barometer 
was 30.10 inches, nearly at 'Set Fair.' Conversely, at 
6 P.M. on March 9, 1876, the barometer at Wick read 
27.94 inches, far below 'Stormy/ and the force of the 
wind was only 3 (a 'gentle breeze') from NW. In 
the one case, therefore, we had a heavy gale with a 
high barometer ; in the other, a gentle breeze with a 
very low barometer. The reason of these discrepancies 
between old theories and actual facts will appear when 
we treat of gradients. 

Storms were formerly divided into two great classes 
circular storms (hurricanes and typhoons) and straight- 
line storms. The former are almost the only class of 

28 Weather Charts and Storm Warnings. 

storms which occur within the tropics, and are known- 
under the general name of cyclones. The latter class 
were formerly supposed to be the most usual type of 
storms in these latitudes, inasmuch as it is a common 
observation here that the wind will blow hard from the 
same point, and for a considerable length of time, over 
a large district. The study of observations taken at 
the same hour over an extensive tract of the earth's 
surface has, however, shown that the storms of the 
Temperate zone are without exception cyclonic, or 
partially so, in their character, although not so perfectly 
developed as those within the tropics. All cases of 
so-called straight-line storms are to be explained either 
by the persistence of the same characteristics for several 
days over the same region, or else simply by the fact 
that they are mere local phenomena, due to the contour 
of the country, like the exceptionally strong breezes 
often met with on rounding prominent bluff headlands, 
or, to use a more familiar illustration, at street corners. 
By this statement it is not meant to imply that strong 
winds from a definite point are not met with for days 
together in the region of the Trade Winds and Mon- 
soons ; but, firstly, these forces seldom reach those of an 
actual ' strong gale/ and, secondly, the conditions which 
cause them are of the same nature as those which cause 
our own storms, and it is the persistence of the con- 
ditions which determines the constancy in the force 
of the wind. 

In speaking of the general character of storms, it 

The Wind. 


should be mentioned that in most cases the vertical 
depth of the stratum of the atmosphere which is in the 
condition of storm is very small in comparison with the 
superficial extent of the area over which the storm is 
felt. Every one knows that it is a common occurrence 
to see clouds at a moderate elevation either moving 
rapidly while calm prevails below, or else at rest, or 
nearly so, while we are feeling a strong wind. On one 
occasion, August 18, 1875, when a whirlwind passed 
over a village in Sweden called Hallsberg, in the 
province of Nerike, it was expressly noticed that, while 
the branches of trees and fragments of the wreck of 
buildings were carried by the wind for miles, the clouds 
did not indicate the slightest sign of disturbance. 

We have already spoken of great currents of air 
extending over vast tracts of the earth's surface, and 
the best idea which we can gain, for practical purposes, 
of the winds which affect us in these islands is that the 
air over the North Atlantic Ocean, north of Lati- 
tude 40, is constantly flowing from west to east, like a 
gigantic river. If such a river be flowing rapidly, we 
often see on its surface small waves, each with its own 
eddies and circulations, which are carried on with the 
stream. If we could imagine it possible to look at the 
upper surface of the atmosphere, supposing it to be 
homogeneous and of equal density throughout, we should 
see much the same sort of conditions, except that 
what corresponds to the hollow of the wave would 
be a patch of defective pressure, while that which 

3O Weather Charts and Storm Warnings. 

corresponds to the crest of the wave would be an area 
of excessive pressure. We shall shortly learn how 
these areas influence the motion of the air and the 

There remains one general principle which was 
formerly brought forward as almost indisputable, and 
that is the principle known under the name of the * Law 
of Gyration,' propounded by Professor Dove, of Berlin, 
some fifty years ago. This principle is, that the wind 
changes more frequently * with the hands of a watch,' 
that is, from East, through South, to West, in the 
Northern hemisphere, than it does in the other direc- 
tion, and that this is true all round the compass. The 
enunciation of this doctrine as an universal law has 
arisen from the fact of meteorology having been first 
studied in Western Europe, where the truth of the 
principle is undeniable. When, however, we look to- 
the evidence cited by Professor Dove himself for Arctic 
stations, we find that the statements from those regions 
are not nearly so positive in favour of the law as the 
experience gained in lower latitudes, and the evidence 
from the German Arctic Expedition shows that on the 
east coast of Greenland the direction of change of 
wind was more frequently against watch hands than 
with them. 

Moreover, when we come to consider the motion of 
wind in the systems of high and low pressure, of which 
we have just been speaking, and the motion of these 
systems themselves over the earth's surface, we shall 

The Wind. 31 

find that the question of the shifting of the wind in 
accordance with the law ('veering'), or it's shifting in 
opposition to the law ('backing'), simply depends on the 
motion of the systems of circulation to which the 
winds in question belong. For instance, in the Northern 
hemisphere a wind ' veers/ or shifts, ' with watch hands ' 
at any station when an area of low pressure passes from 
west to east to the northward of that station, and it 
'backs' when the area of low pressure passes in the 
same direction to the southward of the station. See 
fig. 14 (p. 79). 

The reason that veering, especially from SE. or S. 
through SW. to W. or NW., is so common in these 
islands is that the most usual path of cyclonic systems 
of air is from SW. to NE., the track of the centres of 
the storms passing to the northward of most stations 
in this country. 

We have learnt, therefore, in this chapter, that wind 
is always connected with some disturbance of the 
pressure of the atmosphere, and it will be at once 
understood that its existence is due to the tendency of 
an elastic fluid, like the air, to regain the condition of 
equilibrium from which it has by any means been 
disturbed, while the motion of wind is regulated by 
certain fixed laws to which we have alluded, and which 
will be more fully stated presently. 

3 2 Weather Charts and Storm Wa rnings. 



WE have next to consider what is the real nature 
of these disturbances of which we have been speaking. 
They are connected with irregularities in the distribution 
of atmospherical pressure. We only say ' connected 
-with/ and do not say ' traceable to/ or use any more 
positive expression of opinion, for it is not proposed in 
this book to discuss the question of what the original 
.causes of the disturbances are. Various theories have 
been propounded to account for storms, and some will 
be mentioned in Chapter VII., but none of them have 
met with general acceptance as yet We must there- 
fore only take things as we find them, and endeavour to 
make the best of them. 

But it is necessary, at the outset, to explain the 
meaning of some of the terms which will recur most 
frequently in the following pages. 

The first of these is the word 'isobar/ which is 
.derived from two Greek words signifying ' equal weight.' 
An isobar, or isobaric line, is a line passing through 
those places where the barometrical pressure is cqiial. 

The Barometer. 33 

If we look at the chart (fig. i, p. 34) containing the 
readings from a report for a day, we see that the 
readings, all of which are shown on the chart, varied 
from 29.88 at Nice to 30.54 at Stockholm and 
Christiania. We find several lines drawn across the 
chart : all these lines are isobars, and the values of 
each are given. 

The courses of these isobars are determined with 
reference to the stations on each side of them where 
actual observations exist, the probable position of the 
line in each case being ascertained by dividing the 
distances between the stations in proportion to the 
difference in readings between them. Similarly, points 
are found for isobars crossing the North Sea by 
dividing the distance between the stations on each side 
of that sea in proportion to the difference in read- 
ings between them. In this manner the courses of 
isobars drawn over the sea, and in regions whence no 
observations are obtainable, are inferred from the read- 
ings taken at the nearest stations whence observations 
have been received. As a general rule isobars drawn 
on comparatively insufficient information are dotted 
instead of being drawn in full. 

The isobar of 30.5 forms a closed curve, passing 
through Stockholm and enclosing the southern part of 
Sweden as well as Christiania and the Scaw, that of 
30.4 ins. starts from the Gulf of Bothnia, north of 
Hernosand, passes close to Skudesnaes (30.39), York 
(30.41), and Prawle Point (30.41), through Cape 


34 Weather Charts and Storm Warnings. 

Gris Nez (30.40), and is carried out over Northern 

The lines of 30.3 and 30.2 run across Norway and 

FIG. I. January I, 1887 ; 8 A.M. Barometrical Readings 
and Isobars. 

the North Sea, on a course nearly parallel to that of 30.4, 
which has just been described. That of 30.3 passes 

The Barometer. 35 

very close to Roche's Point and Rochefort (both 30.31) 
and to Berlin (30.32). It is dotted across the Bay of 
Biscay from want of information, as also is that for 
30.2 ins. This latter passes very close to Valencia, 
Corunna, and Biarritz (30.21, 30.22, and 30.19 
respectively) ; the line for 30.1 just skirts the coast of 
Norway, passing thence between Sumburgh Head 
(30.09) and Wick (30.11), and outside the coast of 
Ireland, to reappear over Southern and Central France. 

The reading of 29.95 at Bodo necessitates the 
drawing of a line for 30.0 ins. in the extreme north of 
the map, white that of 29.85 at Nice calls for two other 
lines, for 30.0 and 29.9 respectively, over that part of 

From what has been said, it will be understood that, 
as these lines pass between stations where the observa- 
tions show higher or lower pressures respectively than 
the values assigned to the isobars themselves, it is pre- 
sumable that the reading at any other station exactly 
on any line would have precisely the value shown by 
the isobar. 

These isobaric lines are drawn on all weather charts, 
and a knowledge of their respective values in baro- 
metrical readings, and of their courses, lies at the very 
foundation of all that we know about weather. 

It will already have been seen from fig. i that the 
readings are not the same at all stations simultaneously, 
and, if we look at any chart representing the conditions 
of barometrical pressure over an extensive region for 

D 2 

3 6 Weather Charts and Storm Warnings. 

any epoch of time, we shall find that this pressure is 
not equal at all stations, being greater in some places 
than in others. If we then proceed to draw the isobars,, 
we shall find that several of these lines, like that for 30.5 
ins. on fig. I (p. 34), form closed curves round certain 
spots where the barometrical readings are either lower 
or higher than they are over the neighbouring districts. 

These two contrasting states of affairs are known by 
different names. The districts of low pressure, or of 
depression, are termed 'cyclonic,' and those of high 
pressure ' anticyclonic/ the names being derived from 
the Greek word /cu/cAow, to whirl round in a circle, 
and conveying the fact that the wind in each class of 
area has a tendency to circulate round the centre of the 
system of disturbance, but in opposite directions in the 
two cases. 

There is a very marked difference between the areas 
in which the barometer is lower, and those in which it 
is higher, respectively, than in the surrounding districts, 
for the temperature and the weather, as well as the 
circulation of the wind, differ most materially under 
the respective conditions. 

In both cases there is a calm at the centre, or over 
the region enclosed by the innermost isobar, and, as a 
general rule, in cyclonic systems the extent of this calm 
centre is less, and the isobars surrounding it lie closer 
together, than in anticyclones. We shall shortly learn 
that this indicates a material difference as regards the 
force of the wind. 

The Barometer. 37 

In order to illustrate this it will be necessary to give 
some charts as instances respectively of cyclonic and 
anticyclonic disturbances, which have existed over the 
region embraced by our telegraphic reporting system. 
It is, however, comparatively seldom that the whole of 
a system of either kind will be found developed over 
our limited district. More frequently we only find 
curves trending one way or the other, which enable us 
to conclude, from the general distribution of pressure, 
in what direction the central area of high or of low 
pressure, as the case may be, is situated. 

Fig. 2 (p. 39) gives a good example of an area of 
low pressure, or a ' depression,' or a cyclonic disturbance, 
for the terms are used almost indiscriminately, developed 
as fully as it is usual to find them in Western Europe. 
Since the mode of drawing isobars has already been ex- 
plained, it has not been thought necessary to repro- 
duce the actual barometrical readings at the different 

It will be noticed that the chart contains other 
indications, besides those afforded by the isobars, which 
require some explanation. 

The direction and force of the wind are given by 
arrows : a circle indicates a calm, as the wind has 
neither direction nor force. 

The direction is shown by the direction in which the 
arrow is flying. 

The force is indicated by differences in the symbols 
employed, which are as follows : 

38 Weather Charts and Storm Warnings. 

Forces o-i (Beaufort scale, p. 8) 
2-4 ,, 


8-10 +- 

above 10 -s > 

Thus it will be seen that there is a very heavy gale at 
Rochefort from WNW., a heavy gale at Scarborough 
from SE., a fresh breeze at Aberdeen from E., a light 
breeze at Brussels from SSE., and a calm at Toulon. 
(See Appendix A., p. 218.) 

The lowest reading (28.55 inches) is at Holyhead ; 
the highest (30.00 inches) at Corunna. 

The innermost isobar (28.6) embraces almost the 
whole of Wales. That for 28.8 is oval in shape, and 
covers nearly all England, and the east and north of 
Ireland. That for 29.0 takes in a little of France and 
Belgium, and the greater part of Scotland. The isobar 
of 29.2 envelops the whole of Scotland, but is not 
carried out over the Atlantic beyond the Orkneys 
on one side and the coast of Brittany on the other, 
and it is only dotted in, as being merely inferred, 
in the absence of observations, over the Bay of 

The other isobars are only partly shown ; in fact, the 
position of a portion of that for 30 inches is only indi- 
cated by the single observation of 30.00 at the station 
of Corunna. 

The chief feature noticeable in the northern part of 
the chart is that the isobars of 29.4 and 29.6 trend 
away to the northward, and so the curves spread out in 

The Barometer. 


a fan shape between the Shetlands and the coast of 

If \ve now turn to the wind arrows, we shall find 

FIG. 2. November 29, 1874 ; 8 A.M. Isobars and Wind, Cyclonic 

that they show a circulation round the centre of 
depression. They are : 

4Q Weather Charts and Storm Warnings. 

Westerly at Scilly and in France . . .on its southern side. 

Northerly at Valencia ) ., 

vr ii i -i-x i i r . on its western side. 

North-north-east at Donaghadee J 

East at Ardrossan ...... on its northern side. 

South-east along the whole east coast of Great \ 

Britain north of Hull 1- on its eastern side. 

South at Yarmouth and in the Straits of Dover) 

In fact, the wind sweeps round the central area of 
depression, against watch hands, and this is the invari- 
able law in all cases of cyclonic disturbances in the 
Northern Hemisphere. The wind moves in a direction 
opposite to that of the hands of a watch, and its course 
is nearly parallel to the isobars ; but to this subject, as 
well as to the relation of the force of the wind to the 
distribution of pressure, we shall return later on. 

Let us now take the converse case, an instance of an 
area of high pressure, or an anticyclone, and for this we 
have an excellent example in February 4, 1874 (fig. 3). 
On this day at 8 A.M. the absolute highest reading is 
30.67 at Nottingham, and the only English reports 
which give readings below 30.6 are Dover, Plymouth, 
and Scilly, the isobar of 30.6 enveloping almost the 
whole of England and Wales. That of 30.5 stretches 
to Holstein near Fano (30.49), passes south of Paris 
(30.52), being shown by dots, owing to deficient infor- 
mation, between these points, sweeps close to Valencia 
(30.49), is again dotted over the sea outside the coast 
of Ireland, and finally reaches Aberdeen (30.49). 

In the north and south readings decrease rapidly ; on 
the south side, the curve for 30.40 passes half-way 

The Barometer. 41 

between Rochefort and Biarritz, and 30.30 skirts the 
coast of Spain and the Pyrenees, readings being 30.28 
at Corunna, and 30.24 at Toulon. In the north 30.4 

FIG. 3. February 4, 1874; 8 A.M. Isobars and Wind, Anticyclonic 

passes below Stornoway (30.39), runs between Wick 
and Thurso and across to Jutland, where the reading 

42 Weather Charts and Storm Warnings. 

at the Scaw is 30.34. Passing further north, 30.30 
runs close to Sumburgh Head and across to the 
neighbourhood of Bergen in Norway, while above it 
still we find that of 30.20, the reading at Christiansund 
being 30.15. 

When we look at the winds in this case we find that 
their circulation is exactly opposite to that which is 
shown in fig. 2 ; it is with watch hands, being : 

North . . . .in Germany, on its eastern side. 

East . . . .in France, on its southern side. 

South . . . .in Ireland, on its western side. 

South-west to West . in Scotland, on its northern side. 

North-west . . .in Denmark, on its north-eastern side. 

The arrows also appear to draw out from the 
centre instead of drawing in towards it as in cyclonic 

The isobars are further apart than in the former 
case, and in consequence the force of the winds is much 
less. The reason of this will shortly be explained. 

We have therefore learnt in this chapter (i), that the 
distribution of atmospherical pressure on a weather 
map is made clear by the isobars ; (2), that there are 
two grand types of disturbance of the equilibrium of 
pressure, called respectively, a, cyclonic and, b, anti- 
cyclonic, according as the pressure in a given locality 
is either in defect or excess of its mean value over the 
surrounding region ; (3), that these two types are 
characterised by strong contrasts to each other, among 

The Barometer. 4$ 

which the most striking are the differences in the 
direction of motion of the wind. 

Cyclonic and anticyclonic are, however, merely relative 
terms ; the reading may be above 30 inches at the 
centre of a cyclonic area, and below 29.5 inches at the 
centre of an anticyclonic region. If at any place, or 
over any district, the barometrical reading is lower than 
at the places all round it, that place or district is the 
centre of a cyclonic area ; if, on the contrary, the 
reading at the place or district is higher than at the 
places all round it, that place or district is the centre 
of an anticyclonic area. 

44 Weather Charts and Storm Warnings. 



LET us now proceed to examine still more closely into 
the principles in regard to the motion of the wind, 
which have only been faintly hinted at before. The 
direction of motion of the wind, in relation to the dis- 
tribution of atmospherical pressure, may be easily 
perceived to be such that if you stand with your back 
to the wind the barometer will be lower on your left 
hand than on your right. This is the principle already 
cited, and generally known, as Buys Ballot's Law. It 
is simply the extension, to all cases of wind motion, of 
the Law of Storms first announced by Redfield and 
Reid for the Hurricanes of the West Indies as well as 
for the Typhoons of the China seas. The Hurricanes 
of the Southern Indian Ocean obey a similar law. The 
direction of rotation is opposite in the two hemi- 
spheres. The law as above stated refers to the 
Northern Hemisphere. The law therefore gives the 
direction of the wind, but it also enables us to judge of 
its force, for it is found that this latter depends mainly 
on the amount of difference of pressure between adja- 
cent stations. 

Gradients. 45 

Air, being a gas, is even more mobile than water, 
and as the least difference in level between two 
portions of a free surface of the latter generates motion, 
tending to produce equality of level ; so in the former 
case the slightest difference of pressure causes motion 
in the atmosphere, in the endeavour to restore equi- 
librium of pressure. 

It is evident that the greater the difference of 
pressure over a given distance, the greater will be the 
effort, and consequently the more violent and rapid 
will be the motion required to regain equilibrium. 
There is therefore a convenience in fixing a standard 
of comparison by which to measure the disturbance of 
pressure, and here meteorologists borrow an idea from 
engineers, who measure the inclination of a road or 
railroad by what is termed the 'gradient/ implying, 
when they speak of a gradient of one in sixty, that the 
slope rises one foot vertically for every sixty feet of 
horizontal measurement. 

Meteorologists speak of gradients also, but, instead 
of applying the same unit of measurement, as of feet, to 
the vertical and horizontal scales, they give the vertical 
scale in units of barometrical measurement, and the 
horizontal scale in miles of distance. These gradients 
therefore are expressed in differences of barometrical 
pressure over a given distance. The first person to 
introduce the term gradient into meteorology was the 
late Mr. Thomas Stevenson, C.E. 

The gradients adopted by the Meteorological Office 

46 Weather Charts and Storm Warnings. 

are expressed in hundredths of an Inch of mercury per 
fifteen nautical miles. 1 

In the accompanying figure (fig. 4), the horizontal 
distance between the two stations A and B is supposed 
to be fifteen nautical miles. The divisions on the 
vertical line B D are hundredths of an inch, and they 
correspond to the differences between the barometrical 
readings taken at the same hour at the two stations. 
The gradients are the ratios between the intercepts 

FIG. 4. Gradients. 

B C, B D, &c., and the line A B which is supposed to 
be = i. The gradients are given as 2 for the angle 
B A C, % for the angle BAD, corresponding to the 
several observed differences. These lines A C, A D, 
&c., are imagined to be drawn every morning between 
the most important stations given in the Daily 
Weather Report, and from their inclinations conclusions 
as to the probable direction and force of the wind for 

1 On the Continent gradients are measured as millimetres per one degree 
of sixty nautical miles. These gradients are therefore almost identical with 
ours, for one millimetre is nearly 0.04 inch. 

Gradients. 47 

the day are drawn. It is found, for instance, that the 
force of the wind will not exceed the figure 5 or 6, a 
' fresh breeze ' on Beaufort's scale, unless the gradient 
be higher than 2 (A c on the diagram). 

To reduce this statement to a practical form, we 
may put it in these words. The distance from 
Penzance to Brest is 1 1 3 nautical miles. A gradient 
of 2 between these stations represents a total difference 
in barometrical readings of o. 1 5 inch, so that, in 
accordance with what has just been said, whenever a 
Westerly gale is blowing at the entrance of the Chan- 
nel we may expect that the barometer at Penzance 
will be at least o. 1 5 inch lower than that at Brest ; 
vice versa, the readings at Brest will be proportionably 
lower than those at Penzance whenever an Easterly 
gale is felt in the district in question. 

An instance in point for the former state of affairs is 
the gale of January 8, 1870, during which the read- 
ing at Brest at 8 A.M. was 29.38, and that at Penzance 
at the same hour was 29.19. The difference between 
these readings is 0.19, and the resulting gradient 2.5. 
A very severe Westerly storm was that morning 
reported from the Channel. 

The converse conditions accompanying an Easterly 
gale were observed on May 14, 1869, when the read- 
ing at Penzance was 29.92, and that at Brest 29.68. 
The resulting gradient is 3.2, and accordingly heavy 
Easterly gales were felt on our Channel coast. 

To apply the same principle to the winds of the 

48 Weather Charts and Storm Warnings. 

British Islands generally, it may safely be asserted that 
no storm of any serious extent is ever felt over tire 
United Kingdom unless there be an absolute difference 
in barometrical readings exceeding half an inch of 
mercury between two of our stations. 1 

The difference in readings between Rochefort and 
Aberdeen on February i, 1868, when a tremendous 
Westerly gale was raging, was as much as 1.76 inch, 
the reading at Rochefort being 30.16, and that at 
Aberdeen 28.40 inches. These figures give a gradient 
of 4 over the entire distance of 673 miles, and we find 
that gales were reported from seventeen stations that 

No very precise relation has as yet been established 
between the amount of the gradient and the force of 
the wind, but as a convenient figure to be remembered 
it may be repeated that a gradient of 0.02 inch per I 5 
miles indicates the probability of as much wind as an 
ordinary yachtsman likes to meet with. 

We are now in a position to see more clearly how 
entirely this idea of gradients does away with the old 
notion that the actual height of the barometer at one 
station gives a certain indication of the probable direc- 
tion or force of the wind or of the character of the 
weather at that station, a notion which has found ex- 
pression in the worcls 'Very Dry,' ' Fair/ 'Change/ 

1 Local storms, which occasionally do great damage, may be felt when 
the barometrical disturbance is itself only local, and when the actual 
amount of difference between the extreme readings is less than half an 
inch, although the gradients for a short distance may be high. 

Gradients. 49 

' Rain,' ' Stormy/ &c. &c., which have been for so many 
years placed on barometer scales. 

The small trace of truth which the scale lettering in 
question contains is accounted for in the following 
way. On the average of a great many readings of 
the barometer, taken under various circumstances as 
regards the direction of the wind, it is found that in 
these islands the reading is highest when the wind is 
North-easterly, and, as has already been stated, this is 
when the air is coldest and driest. Consequently we 
have ' Very Dry ' put down at the top of the scale, and 
according as the mercurial level rises from ' Change,' 
29.5 inches, to its maximum height, we have 'Fair' or 
' Set Fair ' entered. Conversely, when the air is warm 
and moist and the wind South-westerly, the barometer 
is low, and so we have then the descending scale of 
' Rain,' ' Much Rain,' and ' Stormy.' It is, however, 
quite a mistake to imagine that on any given day there 
is any certainty of the weather according with the 
description of it given by the word corresponding to 
the barometer height for that day. 

There are yet other points of view from which the 
inutility, and in fact the absolute error, of these words 
may be indicated. 

Let us take for example the word ' Change ; ' this 
is placed opposite the reading 29.5 inches, which read- 
ing is naturally supposed to be taken at sea-level. If 
the barometer be removed to a station situated, say, 
500 feet above that level, the corresponding reading 


5 o Weather Charts and Storm Warnings. 

will be about 29 inches, so that the whole scale will 
be half an inch out, and the error will be the greater 
the more considerable is the height of the station. 
The lettering is therefore wrong for another reason, 
because it does not take account of the necessary 
reduction of the reading to sea-level. 

Once more, the range of the barometer is far greater 
in winter than in summer, so that the reading which 
corresponds to ' Fair ' should be much nearer to 
' Change ' in summer than in winter. The lettering 
is therefore still further in fault, as taking no account 
of this difference. The words are in fact little less 
than utter nonsense. 

It is undeniable that there is more chance of strong 
wind when the barometer is low than when it is high, 
but this arises, not from the actual height of the 
barometer, but from the circumstance that cyclonic 
areas are usually much smaller than anticyclonic, so 
that when the barometer is low there is a greater proba- 
bility of a steep gradient, from adjacent higher readings, 
existing in the neighbourhood, and causing high winds, 
than is the case when the barometer is high. 

It, however, sometimes happens that the barometer 
in these islands will remain for a day or two below 
29 inches, that is, below ' Stormy,' without any gale, 
because the area of low pressure is extensive and the 
gradients slight. An instance of these conditions has 
already been cited at p. 27 as having occurred March 9, 


The question is often asked why gradients can be 
said to be for certain winds ? The answer to this is 
very simple. Let us recur to fig. 2 (p. 39), and add 
to the curves and arrows already shown a number of 

FIG. 5. November 29, 1874; 8 A.M. Isobars and wind, with 
principal gradients. 

straight lines joining Holyhead to Valencia, Aberdeen, 
Skudesnaes, the H elder, and Brest. These lines (fig. 5) 
represent the direction of the gradients, and a glance 

E 2 

5 2 Weather Charts and Storm Warnings. 

at the chart will show that the winds, as regards their 
direction and force, bear a definite relation to these 
gradients in accordance with Buys Ballot's Law. 

It will, however, be seen that as the form of the 
central isobar is oval, not circular, and as the lowest 
reading, that at Holyhead, is at some distance from 
the centre of the oval, the accordance of the winds 
with the gradients is not so exact as would be evinced 
in the case of a more perfectly circular storm. This is 
particularly the case as regards the relation of the 
winds at Scilly and Pembroke to the gradients between 
Brest and Holyhead. The centre of the disturbance 
really lies not far from Shrewsbury, and a gradient 
from Brest to Shrewsbury would be steeper than that 
from Brest to Holyhead, and the winds would be more 
nearly perpendicular to it than is the case in the chart 
In fact, this relation of the winds to the gradients may 
be exhibited by the following table : 



of Wind 

Actual Winds. 

Valencia to Holyhead 



NW. S at Roche's Point 

Aberdeen to ,, 


E. by S. 

f E. 7 at Aberdeen 
|SE. 7 at Leith 

Skudesnaes to ,, 



SE., at Scarboro' and Shields 

The Helder to 



(S. 5 at Yarmouth 
| S. 7 at Cape Gris Nez 

Brest to ,, 



]WNW. IO at Scilly 
|W. 9 at Brest 

It is needless to multiply examples. The chart 
shows a complete cyclonic circulation, and the gradients 

Gradients. 5 3 

are taken on various sides of the centre. The entire 
condition furnishes a clear ocular proof of the state- 
ment that the wind is nearly perpendicular to the gra- 
dients, and therefore nearly parallel to the isobars. 

From the way in which the direction of the gradient 
is described, we gather the direction of the wind indi- 
cated. When we speak of a gradient from Valencia to 
Holyhead being for North-westerly winds, we imply that 
the reading at Valencia is the higher of the two readings, 
the station luith the higher reading being ahvays placed 
first, and so a man standing midway between the two 
stations, with the lower barometer on his left-hand side, 
would face SSE. and have his back to the NNW. ; the 
wind would therefore be North-North-westerly, which 
brings us back again to Buys Ballot's Law. 

We have now learnt how the idea that the reading 
of the barometer at any time gives an indication of 
probable weather is controverted by the more correct 
conception of the value of gradients, but there is another 
idea which, though not absolutely misleading, is yet an 
unsafe mode of interpretation of barometrical readings. 
This is the idea that the motion of the mercury in the 
barometer is an infallible indication of coming weather. 
It is often asserted that if the barometer falls, say, at 
the rate of one-tenth of an inch per hour, a storm is 
certain to ensue at the place. 

Depressions with slight gradients, moving quickly, 
will, however, cause the barometer to fall, at stations 

5 4 Weather Charts and Storm Warnings. 

over which they pass, as rapidly as depressions with 
steep gradients which move slowly. 

It is of course true that if an area of depression 
passes over us the barometer will fall till the centre has 
gone by, and then will rise, and that the rates of such 
fall and rise, for a given rate of motion, will bear a close 
relation to the amounts of the gradients along the path 
of the centre ; but it is not by any means certain that 
because a barometer falls rapidly there must be a gale 
at the place where this fall is observed. A very telling 
instance of the fallacy of this opinion is to be found in 
the following case, when the barometer fell with extra- 
ordinary rapidity over an extensive region, and yet 
hardly at a single station within that region was there 
any gale felt. 

November 22, 1869, a very remarkable barometrical 
depression was manifested over Western Europe. The 
actual fall of the mercury, since 8 A.M. on the 2ist, had 
exceeded 0.9 inch over the entire district extending 
from Dover to Valencia and from Lorient to Shields. 
This area is bounded by the meridians of i E. and 
11 W. and by the parallels of 48 and 55 N., and 
its superficial extent is about 200,000 nautical square 

On this occasion, although there was such an enormous 
disturbance of equilibrium, there were absolutely no 
strong winds reported on our coasts during the day r 
excepting slight gales from North-west at Cape Clear 
and from South-east at Yarmouth. 

Gradients. 5 5 

Outside our west coasts, however, on the edge of 
the disturbance, gradients were very steep, and there 
gales were felt between the 2ist and 23rd, not only 
by the Atlantic steamers Scotia and City of Brooklyn, 
which were near the Irish coast, but also by the Inver- 
ness and Foam, which were further south. In fact, the 
last-named vessel, in the latitude of Cape Finisterre, 
was driven to the southward, out of her course, and 
suffered severely. At Corunna, however, and on the 
coast of Portugal, the gale was not felt seriously at all. 
In this case, therefore, a fall of the barometer of an inch 
at several stations was not followed by a gale at those 

March 9, 1876, the instance already quoted (p. 27), 
affords quite as striking an illustration. That morning, 
at 8 A.M., the fall in the barometer within twenty-four 
hours had amounted to i.i inch at Wick, had exceeded 
an inch over the entire north and east of Scotland, and 
had been more than 0.8 inch over the rest of Great 
Britain, most of Ireland, and the north of France, yet 
the only wind of a force above 7, ' a moderate gale,' 
was at Rochefort, where a heavy Westerly gale was 
reported, and near the very centre of the disturbance, 
where the barometrical fall ivas the greatest, the force of 
the wind was the slightest, because the gradients were 

This entire depression, like that mentioned just above, 
passed off without causing any gale worth notice in the 
United Kingdom. 

5 6 Weather Charts and Storm Warnings. 

We have learnt in this chapter that the distribution 
of pressure, as defined by the gradient, is the best 
guide, by the use of the barometer alone, towards a 
knowledge of the laws of wind motion and, as we shall 
subsequently learn, of coming weather ; but it must 
be stated that it appears that the force of the wind is 
not regulated solely by the gradients, though meteor- 
ologists have not yet determined what are the other 
agencies which influence it. 

It is therefore evident that any attempt to foretell 
weather by the indications of the barometer at any one 
station, unsupported by observations of wind, clouds, 
weather, &c., must necessarily be imperfect. The 
direction and force of the wind in no way depend on 
the actual barometer reading at the time. 

We have, moreover, seen the meaning of the phrase 
'gradients for such and such winds,' which are nothing 
more than the expression in a practical form of the 
laws of wind motion. 

Cyclones and Anticyclones. 5 7 



THE consideration of gradients, and of the effects of 
the distribution of pressure, brings us to contrast the 
two classes of atmospheric systems, anticyclonic and 
cyclonic, and the weather they bring with them. 

Anticyclonic systems are marked by a very slow 
circulation of the air, or, in other words, by light winds, 
by low temperature in winter, great dryness of the air, 
at least at the centre, and consequent absence of rain, 
although fog is frequently very prevalent. These cir- 
cumstances are accounted for by the fact that the air, 
in an anticyclone, flows out from the centre, and there- 
fore must be supplied from the upper regions of the 
atmosphere by a descending current, which cannot 
possibly contain much moisture, owing to the very low 
temperature of the regions from whence it is drawn, 
and which is actually rendered drier by the fact of its 
descent. 1 

1 Air ascending is expanded, it falls in temperature in the process, and 
the moisture it contains is consequently liable to be condensed and to 
form clouds, and ultimately fall as rain. Conversely, air descending is 
compressed, it rises in temperature in the process, and consequently its 
capacity for containing moisture is increased, and it feels dry. 

5 8 Weather Charts and Storm Warnings. 

Cyclonic systems, on the contrary, are distinguished 
by the exact converse of the above conditions. The 
air circulates more rapidly, causing strong winds, and 
flows in towards the centre, so that it must naturally 
be supplied from below and ascend in the centre. 
Cyclones in these latitudes bring with them, at least 
over a considerable proportion of the area they cover, a 
comparatively high temperature, much moisture, and 
consequently heavy rain. We say, over a considerable 
proportion of the circumference, because, in certain 
parts of a cyclonic system, the wind which is felt is 
often very dry. This is in the rear of the disturbance- 
In the Tropics, there is not such a contrast between 
the different parts of a cyclone, as regards dryness. 

These statements refer to the winter ; in summer the 
conditions are exactly reversed, at least as regards the 
temperature which accompanies the respective systems 
of disturbance. 

Anticyclonic systems in summer, produce our hottest 
weather, inasmuch as the air is so dry that no heavy 
clouds can be formed, so that the sun has full oppor- 
tunity for exerting his heating power, while there 
is hardly any motion in the air to produce a cooling 

Cyclonic systems in summer, on the contrary, bring 
cloudy weather, rain, and a reduction of temperature, 
the latter being mainly due to the density of the cloud 
covering which intercepts the direct rays of the sun. 

It is easy to illustrate these statements by examples* 

Cyclones and Anticyclones. 


Let us take the instance of an anticyclone first, in fact 
the very day which we have already cited, February 4, 

FIG. 6. February 4, 1874 ; 8 A.M. Isobars and temperature, showing 
changes in previous twenty-four hours. 1 Anticyclone. 

1 The changes of temperature in figs. 6 and 7 are shown by small figures 
placed after the thermometrical readings. When these figures are above 
the readings, as at Wick (in fig. 6), they indicate a rise of temperature ; 
when they are belong as at Scarborough, they indicate a_/a//. 

o Weather Charts and Storm Warnings. 

1874 (fig. 3, p. 41), which wilt serve as a very good 
illustration of a winter anticyclone. Fig. 6 (p. 59) 
exhibits the isobars as before, but the figures on the 
chart refer to temperature. 

It will be perceived that over the central area, 
embraced by the isobar of 30.6 inches (see p. 41), the 
temperature is lower than it is anywhere else in these 
islands, and that there is a fall in temperature noticed 
at every station, varying from i at Plymouth and 
Yarmouth to more than 10 in the north of England, 
and even 16 at Scarborough. In fact, the cooling in- 
fluence of the anticyclone is noticeable over almost the 
whole map, except in Ireland and Scotland. In the 
north of the latter country, however, the rise in tem- 
perature is very marked, amounting to 8 at Aberdeen 
and 7 at Wick and Sumburgh Head. This change of 
thermal conditions locally arose from the fact that, on 
the previous day, the north of Scotland had been the 
region of an independent anticyclonic system, as will 
be gathered from the examination of fig. 24 (p. 93), but 
the barometer had fallen during the night, and the 
wind, though still light on our coasts, had shifted from 
N. to SW. both in Caithness and in the Lews. In 
fact, this rise of temperature was connected with a 
cyclonic disturbance further northward, which is shown 
in fig. 26 (p. 94). 

As regards the force of the winds, a glance at fig. 3 
(p. 41) is sufficient to show that they are very light ; in 
fact, the only British or Irish station where they exceed 

Cyclones and Anticyclones. 6 1 

the force of 5 of the Beaufort scale, a ' fresh breeze/ 
is Scilly. 

The general dryness of the air is sufficiently evinced 
by the fact that rain was not recorded for this morning 
from more than five stations out of the fifty-one men- 
tioned in the Reports, at three of them the amount 
being only o.oi inch ; the greatest quantity was at 
Scilly, and that was only 0.06 inch ; nevertheless, the 
actual difference between the dry and wet bulbs was 
but slight, showing that the atmosphere was near its 
point of saturation for the temperature. This circum- 
stance is partly owing to the fact that the observations 
were taken at 8 A.M., so that the influence of the sun's 
heat had hardly made itself felt. In fact, what 
moisture there was, was present in the form of cloud or 
fog. The reader may be interested to learn that owing 
to this latter condition the only region where the sky 
was clear was the north of Scotland, and that a 
traveller coming south would have met with an entirely 
overcast sky at Leith, while fog and mist were reported 
all along the east coast of England, as being either 
prevalent at 8 A.M. or as having been so during the 
preceding twenty-four hours. 

Frequently, though this was not the case in this 
particular instance, these fogs only last during the 
early part of the day and clear off about noon, as the 
diurnal range of temperature during the existence of 
anticyclones is often very great. 

These conditions are more or less characteristic of 

62 Weather Charts and Storm Warnings. 

the entire class of areas of high pressure, and, as we 
shall shortly proceed to show, the most striking feature 
of the weather which accompanies them is its per- 
manency, i.e., the slowness with which changes succeed 
each other. 

Let us now take the case of a cyclonic area 
(November 29, 1874), fig. 2. the very instance which 
has been already cited (p. 39), and examine into the 
features which characterised the weather which this 
system brought with it. This storm was so typical 
that the Weather Report and Charts for the day have 
been reproduced in Appendix A. (p. 2 1 8). 

In the first instance we see (fig. 7) that over almost 
the whole of the map the rise of temperature from the 
previous day (indicated by figures above the readings, 
as before) has been very striking, especially at the 
south-eastern stations, where the increase has exceeded 
10, amounting even to 23 at Cape Gris Nez, near 
Boulogne. This region lies in front of the advancing 
disturbance, and is the district of the Southerly and 
South-east winds. The motion of this particular storm 
will be traced subsequently (figs. 19-22, pp. 90-92). 

In the rear of the storm, i.e., in Ireland and Cornwall, 
where the wind has shifted to North-west, we find that 
a fall (indicated by figures below the readings) is re- 
ported at all the Irish stations, as well as at those in the 
extreme north of Scotland. That this is due to the 
direction of the wind is sufficiently clear from the fact 
that, at almost all the places where the cooling effect is 

Cyclones and Anticyclones. 63 

observable, the direction of the wind is from some point 
to the Northward of West, being NE. at Donaghadee, 
N. at Valencia, NW. at Roche's Point, and WNW. in 


FIG. 7. November 29, 1874; 8 A.M. Isobars and temperature, showing 
changes in previous twenty-four hours. Cyclone. 

the south of England, where, although no fall of tem- 
perature from the reading at 8 A.M. on the previous day 

64 Weather Charts and Storm Warnings. 

is reported, the rise at Stilly and Plymouth is only i f 
as compared with 10 at Dover and 14 in London. 

As regards clouds and rain, the facts are still more 
striking. Out of the entire list of stations, only thirty- 
seven give the amount of cloud in their reports : of 
these, twenty-three stated the sky to be * entirely over- 
cast,' thirteen others reported various proportions 
ranging from the above quantity down to that of six- 
tenths of the sky being covered, while at one solitary 
place, Hurst Castle, with a WNW. wind, the sky was 
' half clear.' 

The rainfall is equally clearly marked. During the 
twenty-four hours preceding the epoch for which the 
chart is drawn, rain fell at every station in these islands 
and France whence we have reports, except at Nairn, 
and, as to its amount, at four stations this exceeded an 
inch, and at ten others it was more than 0.7 inch. 

It may therefore be said that areas of low pressure 
are accompanied by a high temperature, especially in 
the front of the storm, a large amount of moisture in the 
air, and consequent prevalence of cloud and rainfall. 
The instance which has been given, however, although it 
illustrates very well many of the features of a cyclonic 
disturbance, does not exhibit all the circumstances 
which may be observed in a given storm, as to the 
march of the different phenomena within the system 

As regards the distribution of clouds and weather in 
such a system, the Rev. W. Clement Ley exhibited, at 

Weather Charts & Storm Warnings. 


According to the Rev .W.Clement Ley. 

Cyclones and Anticyclones. 65 

a lecture which he delivered before the Meteorological 
Society in 1878, a diagram of which Plate II. is a 
reproduction ; the original appears in ' Modern Meteor- 
ology ' (London : Stanford, 1879.) 

In this the long arrow represents the direction in 
which the system is advancing, and the curved arrows 
indicate the motion of the wind. It will be seen that 
the cloud gradually becomes denser as the centre 
approaches a station, and that as the centre passes 
there is, on the right-hand side of its track, a sudden 
clearing up, the clouds, with the NW. wind in the rear 
of the system, being broken and giving rise to showers, 
not to continued rain. 

On the left-hand side of its track there is no such 
sudden clearing up, as the cloud remains dense until the 
vane shifts beyond North, and the subsequent clearing 
is gradual. 

One of the most striking characteristics of a cyclonic 
storm is the sudden shift of wind which takes place 
between SW. and NW., accompanied frequently by a 
heavy squall and a shower, together with an almost 
instantaneous fall of temperature, and this is entirely 
absent in the case of the storm shown on fig. 7, as will 
be explained on p. 92. To take a very remarkable 
instance of this squall, we find that on February 12, 
1869, a cyclonic storm passed up the English Channel; 
the lowest barometrical reading being recorded at Fal- 
mouth at noon, when the wind flew from SW. to 
WNW., and the temperature fell 6.2 in a few minutes. 


66 Weather Charts and Storm Warnings. 

At Kew similar changes took place at 4.30 P.M., but in 
a very aggravated form : the shift of wind having been 
from WSW. to ENE., and the fall of the thermometer 
1 1.6 in about five minutes. 

The storm of Sunday, March 12, 1876, was very 
severe in the south of England. In this, the shift of 
wind at Kew was from SW. to N., and the fall of 
temperature 8. 5 in thirty minutes, accompanied by 
a heavy snow shower. It is also a remarkable fact 
that, at the moment of this sudden change of tempera- 
ture, the condition of atmospheric electricity changed 
from negative to positive. Further researches on this 
subject are very desirable. 

I have already said that one great distinction between 
cyclones and anticyclones, is that the former move, and 
the latter are usually nearly stationary, and as the 
usual motion of a cyclonic system is from west to east 
across these islands, it will be interesting to see what 
we can learn, from the continuous records at our self- 
recording observatories, of the actual changes which 
will occur at a station during the passage of a storm 
moving in that direction. 

This will naturally differ according as the station 
lies to the north or the south of the centre of the 
storm, and, as it very rarely happens that the storms 
are truly circular, it is obvious that the shifts of wind and 
changes of weather will not always be perfectly regular. 
In fact, the extraordinary changes of February 12, 
1869, just cited, afford a very good proof of the 

Cyclones and Anticyclones. 67 

unequal distribution of the gradients in such a system 
of disturbance. 

Let us first take the most common case in these 
islands, that of a station lying on the southern side of 
the path of the centre of a depression which is moving 
from west to east across central England, and that 
we are situated in London. We shall experience 
first the phenomena belonging to the front of the 
system : the appearance of cirrus clouds, ' mare's tails,' 
in the sky, then the South-easterly winds, the great 
rise of the thermometer and excessive dampness, the 
sky becoming gradually overcast and the setting in of 
mist and rain, the barometer falling persistently, while 
scud begins to drift from the Southward. As the 
system advances the barometer continues to fall, the 
wind veering through S. to SW., and rain falling 
steadily. As soon as the wind passes the SW. point 
and draws to W. or NW., the barometer begins to rise, 
often with a sudden jump, and the temperature falls, 
with a very heavy shower of rain, possibly turning to 
hail with a thunderstorm, after which the air becomes 
much drier, and the sky clears with a N W. wind, which 
as a rule soon dies down. After such a disturbance as 
this we have frequently a smart frost at night, inasmuch 
as the great dryness of the air has allowed of great 
radiation of heat from the ground. 

It is not easy to find an instance which exhibits the 
entire course of changes just described, but the table, 
Appendix B. (p. 221), will illustrate several of the points. 

F 2 

68 Weather Charts and Storm Warnings. 

The table represents the hourly readings of the 
barometer, thermometer (dry and wet), anemometer 

(direction and velocity), rain- 
gauge, and the vapour ten- 
sion from the continuous 
records at Valencia on March 
26 and 27, 1874, when the 
centre of a cyclonic storm 
passed to the northward of 
that observatory. At 2 P.M., 
March 27 (fig. 8), the centre 
of the storm lay off the 
north-west coast of Scotland, 
outside the Hebrides. The 
diagram (fig. 9) exhibits the 
course of the curves. 1 

The barometer fell steadily 
for twenty-three hours, until 

FIG. 8. March 27, 1874; 2 P.M. 
Storm passing north of Valencia. 

9 A.M. on the 27th, and then rose in twelve hours 
nearly to the same height as it had held before. Tern- 

1 These diagrams require explanation. They are copied from the plates 
in the Quarterly Weather Report, and exhibit the continuous records of: 

1. The barogram or barometer curve, marked bar. The scale for this is 
given in the lower left-hand corner. It is in inches, and the readings, 
though corrected for temperature, are not reduced to sea-level, and range 
over li inch. 

2. The thermograms, marked respectively dry and wet. The scale for 
these is in the lower right-hand corner, and ranges over 30. 

3. The curve of tension of vapour, marked vap. ten., calculated from the 
two thermograms, dry and wet. 

4. The rain in hourly amounts of fall. These are measured upwards 
from the base line of the scale which serves for vapour tension and rain, 

Cyclones and Anticyclones. 69 

perature and vapour tension remained singularly steady 
during the afternoon of the 26th, rising very slowly 
during the night, and reaching a maximum at 8 A.M. 
nearly simultaneously with the epoch of the lowest 

26th 2yth 

Noon 6 P.M. Midnt. 6 A.M. Noon 6 P.M. Midnt. 

FIG. 9. March 26 and 27, 1874. Automatic records at Valencia. 

barometer readings. The thermometers then began to 
fall briskly, and by 9 P.M. the dry bulb had sunk 5.;. 
It will be observed that, owing to the influence of the 

and is on the left-hand side, above the barometer scale ; it is in tenths of 
an inch, and extends over half an inch. 

5. The wind direction, marked direc. This is on a special scale shown 
at the left-hand side, which represents a paper wrapped round a cylinder, 
and then laid flat. N is at the top, then come successively w, s, E, and N 

6. The wind velocity, marked velocity, in hourly amounts, measured by 
the scale of miles, up to seventy, shown on the right-hand side. 

The time scale is shown at top and bottom. 

7o Weather Charts and Storm Warnings. 

storm, the temperature was higher during the early 
morning hours of the 27th than it was during the after- 
noon of either the 26th or 2/th, and yet 2 P.M. 'is 
usually about the warmest part of the day. While it 
lasted, therefore, the storm entirely disturbed the regular 
diurnal march of temperature. The secret of this dis- 
turbance of temperature lies in the direction of the 

During the whole time that the barometer was falling 
the wind remained nearly constant between SSE. and 
S., blowing a stiff gale, and a little rain fell. When the 
centre of the storm was approaching, the rain set in 
again at about 4 A.M., and became heaviest just at 
8 A.M., when the barometer was nearly at its lowest level 
and the wind was about veering towards SW. At the 
hour of the lowest barometer reading we find the first 
trace of Westing in the direction of the wind, and at 
the same hour the velocity shows a sudden increase, the 
rain still continuing. The sky then cleared, the baro- 
meter rose briskly, and the wind increased till noon, 
when it blew a heavy gale from SW.byS. It sub- 
sequently veered pretty quickly to WNW. with a slight 
shower of rain at 3 r.M., and fell very light afterwards 
at 7 P.M. 

The courses of the wet bulb thermometer curve, and 
of the curve of vapour tension, illustrate very clearly the 
contrast between the Southerly and North-westerly 
winds as regards the humidity of the air. As long as 
the wind remained to the Eastward of South the curves 

Cyclones and Anticyclones. 71 

of the two thermometers remained close to each other, 
and the vapour tension trace rose steadily, but very 
slowly. At the hour of the minimum barometer 
reading, we notice the thermometer curves going apart, 
and for the remainder of the period they pursued 
courses diverging more and more from each other, 
indicating, what is also shown by the descent of the 
vapour tension trace, that the air was becoming drier 
and drier according as the wind became more and more 

Owing to this extreme dryness of the air of the 
North-west wind, in the rear of a storm, it is believed 
by some meteorologists that this wind is a downrush 
of air from the upper regions of the atmosphere, which 
is necessarily very dry, as explained at the beginning 
of the chapter. 

The case just cited, though it exhibits some of the 
phenomena with great distinctness, is not one which can 
be made clear to the reader as an instance of a storm 
passing to the northward of a station, inasmuch as the 
system to which the gale belonged was so extensive 
that the limited area of our charts would not exhibit the 
conditions in a convincing manner. Fig. 10 (p. 72), 
however, is an indisputable case of a storm passing on the 
northern side of Aberdeen, and close to it, October 
20-21, 1874, and the course of the curves at that 
observatory is shown on fig. II (p. 73). The actual 
hourly readings will be found in Appendix C. (p. 222). 

The fall of the barometer was slow at first, but then 

72 Weather Charts and Storm Warnings. 

increased in rapidity, and in twelve hours the mercury 
sank to the extent of 1.14 inch. It then rose in the 
next twelve hours as much as 0.8 inch, and for a portion 
of the time the rate of rise was more rapid than that of 
the previous fall had been. It will be noticed that the 
epoch of the maximum rise of the barometer was also 
that of the shift of wind through three points to the 


I r ; T i/ > n i 

FIG. io. October 21, 1874; 8 A.M. Storm passing north of Aberdeen. 

Northward, as well as of the strongest wind and the 
heaviest rain. This is the heavy shower of rain, 
coincident with the shift of wind to the North-west, of 
which mention has previously been made, p. 65. Tem- 
perature (dry and wet bulb) and humidity do not follow 
a similar course respectively as they did in fig. 9, as all 

Cyclones and Anticyclones. 


the curves are far less steady in their course than on 
the former occasion. We see, however, that the diurnal 
march of temperature is again entirely disarranged, the 
thermometers remaining pretty nearly at the same tem- 
perature throughout the night of the 2Oth, and the 
sudden fall of both temperature and vapour tension 
taking place nearly simultaneously with the shift of 

Noon 6 P.M. Midnt. 6 A.M. Noon 6 P.M. Midnt. 


FIG. ii. October 20 and 21, 1874. Automatic records at Aberdeen. 

wind above noticed, and at the time when the natural 
rise of day temperature ought to be briskest, viz., about 
10 or ii A.M. 

The wind commenced at about SSW., and blew 
pretty steadily from that quarter for twenty hours, its 
velocity being at first slight, and increasing to that of a 
stiff breeze at midnight. When the barometer was at 

74 Weather Charts and Storm Warnings. 

its lowest and beginning to rise, the direction of the 
wind rapidly shifted to WNW., and in three hours its 
velocity more than doubled itself, rising from 24 to 55 
miles an hour. The velocity of a strong gale, however, 
only lasted for four hours, from 9 A.M. to I P.M. The 
absence of rain is very noticeable during the early period 
of the gale ; the reason for this absence can be seen 
from the circumstance of the great distance between 
the wet and dry thermograms ; in fact, no rain of any 
consequence fell excepting just at the period of the 
lowest barometer, already noticed. 

FIG. 12. February 2, 1873; 8 A.M. Storm passing south ofFalmouth. 

It will be seen from figs. 51 and 52 (p. 157) that 
this storm, which was a very severe one in other parts 

Cyclones and Anticyclones. 


of the country, passed just to the northward of 

Taking now the second case, that of a cyclonic 
system passing on the southern side of a station, we 
have a very good instance of the course of the phe- 
nomena in the records of Falmouth for February I 
and 2, 1873. The hourly values, as before, are given in 

ist 2nd 

Noon 6 P.M. Midnt. 6 A.M. Noon 6 r.M. Mid"t. 



29.5 - s 



FIG. 13. February I and 2, 1873. Automatic records at Falmouth. 

Appendix D. (p. 223), and the instrumental records 
are shown on fig. I 3, while the conditions of pressure 
at 8 A.M. on the 2nd are shown on fig. 12. 

Here, too, as in the previous case, we have a great 
dip in the barometrical curve, which falls very regularly 
for twenty-four hours until noon on the 2nd. It then 
rises more rapidly than it had fallen before, and at 

76 Weather Charts and Storm Warnings. 

midnight on the 2nd attains to within a quarter of an 
inch of its previous height. 

Temperature shows very remarkable changes during 
the night of the ist and 2nd, rising 10 between 10 P.M. 
and 4 A.M. It then continues high until 2 P.M. on 
the 2nd, when it falls 6 in a few minutes, a change 
nearly as sudden as that at the same observatory on 
February 12, 1869, already noticed (p. 65). 

As before, we find the explanation of these changes 
in the direction of the wind. Up to midnight on the 
1st this was steady at ESE. It then veered slightly to 
-SE., and even a couple of points further. The highest 
temperature, already noticed, coincided with the direc- 
tion S.byE. As the centre drew nearer, the vane 
shifted smartly to E.byS., E.byN., and NE., and 
the sudden fall of temperature marked a shift from 
NE.byN., to N.byE. The change of direction went 
on further, and the period closed with the wind about 

The extreme force of the wind did not precisely 
coincide with the period of the lowest barometer 
reading, as it was far greater a little before midnight on 
the 1st than at any other time. This affords a further 
proof, if such is needed, that the violence of a gale does 
not depend on the actual height of the barometer at 
the time. 

Vapour tension shows a steady curve, following that 
of temperature, until the appearance of the Northerly 
and North-westerly winds, when it sank in a marked 

Cyclones and Anticyclones. 77 

way, thus again corroborating our former statement as 
to the dryness of winds from that quarter. 

The rainfall was very remarkable ; it continued 
almost without intermission from noon on the ist till 
5 P.M. on the 2nd, being heaviest shortly before the 
period of the strongest wind. The total amount which 
was collected during the twenty-four hours, ending with 
noon February 2, was no less than 1.275 inch, and it 
will be noticed that the rain gradually lessened as the 
wind shifted through N., and ceased entirely when the 
direction was NNW. 

This persistence of rain with the Easterly and North- 
east winds in the front of a cyclonic depression is a 
very striking characteristic of these disturbances. It is 
an unmistakable proof that North-east winds are not 
always dry, and the only reason that this fact is not 
more frequently observed is, as will shortly be shown, 
that the occurrence of these winds in the front of 
cyclonic disturbances is a comparatively rare pheno- 
menon in the British Isles, for the storms usually pass 
by us on the northern side, and at first give us South- 
easterly, not North-easterly winds. 

These instances are amply sufficient to show the way 
in which we are to interpret a couplet, well known to 
sailors, which expresses Dove's Law of Gyration, referred 
to at p. 30 : 

When the wind shifts against the sun, 
Trust it not, for back it will run. 

Under certain conditions a shift of the wind against 

78 Weather Charts and Storm Warnings. 

watch hands is quite en regie, and is simply an indica- 
tion that a cyclonic disturbance is passing on the 
southern side of the observer. The actual motion of 
the disturbance of February i and 2, 1873, was along 
the north coast of Brittany, and therefore to the south 
of Falmouth. 

The rule as regards the shifting of the wind in 
connection with atmospherical disturbances may be 
thus stated. 

If the observer supposes himself at the centre of a 
cyclonic system in the Northern Hemisphere and facing 
in the direction in which it is advancing, the wind at 
all stations which he passes will shift with watch hands 
if the station be on his right-hand side, and will shift 
against watch hands if the station be on his left-hand 

This is made clear by the following diagrams (fig. i 5, 
a, b, c, d), in which I. II. III. indicate the successive 
positions of a cyclonic system, and the large arrows 
represent different directions of its motion. A is always 
on the right-hand side of the centre, B on its left-hand 
side. The successive directions of wind experienced at 
each station are shown by the figures i, 2, 3, and i', 2', 
3'. In every case it will be seen that the shift from I 
to 3 is with watch hands, and from i' to 3' is against 
watch hands, whatever the first direction of the wind 
may have been. 

The reason that a shift of wind against watch hands 
is considered dangerous is, that when the wind shifts 

Cyclones and Anticyclones. 


FIG. 14. 

So Weather Charts and Storm Warnings. 

from NW. towards SW., S. and SE. at a station A in 
fig. 14, it indicates that the SE. wind of the north- 
eastern side of another area of low pressure is approach- 
ing the station which has been experiencing the NW. 
wind of the south-western side of the previous area of 
low pressure, and which is passing away to the eastward, 
so that the wind merely backs until the station is fairly 
under the influence of the new disturbance, when it 
begins to veer again, if the centre of the depression 
passes to the northward of the station, as is usually the 
case in these islands. 

There is one point in connection with cyclonic dis- 
turbances which is gradually attracting more and more 
attention, and that is the appearance of secondary 
eddies in connection with the larger areas of baro- 
metrical depression. These smaller systems are usually 
less perfectly developed than the larger ones, and they 
manifest themselves in general on the southern sides of 
the latter. Accordingly, while they cannot exhibit any 
Easterly winds of much force (owing to the fact that 
pressure is lower on their northern than on their south- 
ern edge), they intensify the Westerly winds on the 
extreme southern edge of the original disturbance by 
increasing the gradients in that part of the system. 

It is probably owing to these circumstances that 
Easterly winds are so rare in our storms. Investiga- 
tions into the distribution of pressure over the earth's 
surface have shown that there is an almost constant 
deficiency of pressure in the neighbourhood of Iceland, 

Cyclones and Anticyclones. 


and so probably most of the depressions which cause 
our storms are simply secondary to a vastly more 
extensive permanent area of depression over the North 

FIG. 15. January 3, 1874 ; 8 A.M. Secondary depression. 

It will be interesting to give a few instances of these 
'satellite' depressions (as the French call them), to 


82 Weather Charts and Storm Warnings. 

illustrate what we have been saying, and fig. 15 (p. 81), 
for January 3, 1874, shows us on a small scale the 
conditions to which allusion has just been made as 
characterising many of our gales. We see that there is 
a small cyclonic disturbance with its centre near Holy- 
head, and that the circulation round it is very imperfectly 
developed, inasmuch as it exhibits no Easterly winds at 
all. The reason of this deficiency is at once visible 
when we look at the northern part of the chart. The 
reading of 29.23 inches at Holyhead, which is at the 
centre of the secondary depression, almost reappears 
over Caithness, in the isobar of 29.2, and to the 
northward of this we see that of 29.0 inches, so that 
evidently readings would be still lower if we went 
further north. As the reading both at Holyhead and 
at Wick is about 29.2, and as the highest reading 
between these points is 29.3 inches, it is impossible 
that there can be a gradient of any magnitude for 
Easterly winds over the intervening region, and so those 
winds do not appear. On the other hand, the circulation 
round the west, south, and east sides of the secondary 
depression is clearly shown. 

It is not often, however, that we find the secondary 
depressions so clearly marked as in fig. 15. The chart 
for October 7, 1874, gives us their more common 
character (fig. 1 6). They manifest themselves simply 
by a loop in some of the isobars, and their effect is, as 
already described, to reduce the gradients on the side 
turned towards their primary, and to increase them on 

Cyclones and Anticyclones. 

the opposite side. We could hardly have a better 
example of this than that given in fig. 1 6. The strong 
NW. wind at Stornovvay, and the SW. gales in the 
Skagerrack and in Denmark, show that the conditions 
for strong winds prevail generally, but over southern 
Scotland and central England nothing but light 

16. October 7, 1874 ; 8 A.M. 
Secondary depression. 

FlG. 17. October 22, 1874 ; 6 P.M. 
Secondary depression. 

Westerly airs are reported. On the other hand, in the 
Channel and in France the isobars are crowded 
together, and heavy gales are the result. 

Another instance is represented on fig. 17, for 
October 22, 1874, but here it is the North-westerly 
winds of the primary system that are reduced in force, 
on account of the fact of the secondary manifesting 

G 2 

84 Weather Charts and Storm Warnings. 

itself on the south-west side of its principal. Owing to 
the apparent equality of pressure over the Bay of 
Biscay there are no gradients of consequence there, 

io 5 5 I0 

FIG. 18. January 24, 1872; 8 A.M. Co-existing depressions. 

and therefore no gales in connection with the smaller 
system, but, so far as direction is concerned, the circu- 

Cyclones and Anticyclones. 8 5 . 

lation round the smaller depression is very clearly 

At times it is hard to say which of the depressions is 
the primary and which the secondary, as there is appa- 
rently not much difference in size between them. This 
was the case on January 24, 1872 (fig. 18), when we 
see one depression, with central readings as low as 28.3 
inches, near Scarborough, while another shows itself off 
Valencia. The result of the interference of these two 
systems is to produce a comparative calm between them, 
the winds being very light over St. George's Channel, 
while strong gales are blowing wherever this inter- 
ference does not exist. 

In treating of the interpretation of weather charts 
and of storm warnings, we shall show the use which 
can sometimes be made of these secondary cyclonic 

In this chapter we have considered the contrast 
between cyclonic and anticyclonic systems, or areas, as 
they are indiscriminately termed, and the character of 
the weather which accompanies each respectively. 

We have learnt that, while anticyclonic areas move 
slowly, cyclonic areas travel more or less rapidly over 
the country, and moreover that their passage past a 
station presents marked features of difference in the 
weather experienced, according as the station lies on 
the right or left side of the path of the centre. 

86 Weather Charts and Storm Warnings. 

Lastly, we have found that cyclonic systems are not 
always isolated, but occasionally have what may be 
called secondary systems in connection with them, 
which exert a material influence on the gradients, and 
consequently on the winds belonging to the original 

The Motion of Storms. 87 



BEFORE we treat of the motion of storms it must be 
admitted, as will appear in the next chapter, that me- 
teorologists are not at all agreed as to what the real 
causes of that motion are. It has not been satis- 
factorily determined to what extent the storm moves 
forward as a whole, or to what extent the advance 
observed is only due to the continued reproduction of 
the same phenomena at successive points, as is the case 
in wave motion. 

The fact that storms advance over the earth's surface, 
with a motion of translation in addition to their own 
rotatory motion, has, however, long been known. The 
earliest notice of it which we can discover is an entry 
on a map of Virginia, published in 1/47 by Lewis 
Evans, to the effect that ' all our great storms begin to 
leeward.' Franklin, in 1760, followed in the same 
strain, but it appears that his attention had been caught 
at an earlier period, in 1743, by the fact of his being 
prevented at Philadelphia, by the clouds brought by a 

88 Weather Charts and Storm Warnings. 

hurricane, from observing a lunar eclipse, while the 
eclipse was seen at Boston, which lies further to the 
north-eastward, before the storm came on. The reader 
hardly needs to be reminded that hurricanes and 
cyclonic disturbances in general are accompanied by an 
abundance of cloud. Along the eastern coast of the 
United States, therefore, the advance of storms from 
south-west towards north-east has long been a recog- 
nised fact. 

The direction of the motion is different in different 
parts of the earth. Of storm tracks which are well 
made out, those of the West India hurricanes, in the 
earlier part of their course, are, speaking in general 
terms, from east-south-east to west-north-west : in the 
Indian Ocean, the Mauritius hurricanes advance at 
first from east-north-east to west-south-west, but both 
of these classes of storms generally recurve subsequently 
at a sharp angle, and advance towards the eastward. 
The typhoons of the China seas also at first move 
towards the coast from the eastward and subsequently 

Over these islands, however, the motion of storms is 
not nearly so constant in direction as in the instances 
which have been cited, and, although as a general rule 
it may be said that our storms travel from the west- 
ward, there are many exceptions to this rule, and 
in fact storms can, and do, move from every point of 
the compass, but a motion from an easterly point is 
extremely rare. 

The Motion of Storms. 89 

It will be best to consider the usual directions of 
motion in connection with the causes which appear to 
govern them, and we shall also offer some remarks on 
the motion of anticyclones, and show what a contrast 
in this respect they present to the behaviour of cyclonic 

As regards the rate of motion, it may be premised 
that this varies very greatly, some storms travelling 
very slowly or even remaining apparently steady, while 
others are propagated with great rapidity. The rate 
of motion of the West India hurricanes ranges from ten 
to fifteen or twenty miles an hour in their early stages, 
but becomes accelerated in the latter portion of their 
path, while in these islands the motion has attained 
the prodigious speed, of fifty miles an hour, February 1 2, 
1869, an instance already cited, p. 65, and even 
seventy miles an hour, December 16, 1869, a velocity 
nearly equalled on the night of November 10 II, 1875, 
as well as on March 12, 1876. 

It must be clearly understood that the rate of trans- 
lation of the storm has no direct relation to the velocity of 
the ivind in that storm, for the West India hurricanes 
exhibit almost the strongest winds of which we have 
any knowledge, and they travel slowly, while the 
rapidly moving storms just cited were not very excep- 
tional as regards their violence. In Europe, however, 
it has been found, as a general rule, that the most 
violent storms do advance the most rapidly. 

However, one of the greatest difficulties which meets 

90 Weather Charts and Storm Warnings. 



us, in the issue of Storm Warnings in these islands, is 
our almost total ignorance of the rate at which any 
given storm is travelling, until it has already moved 
over a considerable tract of country. 

It is easy to cite instances of the motion of storms, 
and we shall select the two cases which have already 
been employed as illustrations of cyclonic and anti- 
cyclonic disturbances respectively, and trace their 
progress across the area of our weather maps, paying 

attention mainly to the 
isobars and the wind. 

Let us first take the storm 
of November 29, 1874. 
The earliest unmistakable 
signs of its approach 
(fig. 19) were at 8 A.M., 
November 28, when a rapid 
fall of the barometer at 
Valencia, with the Southerly 
wind, and the course of 
the isobar of 29.3 inches, 
showed that there must be 
an area of lower readings 
at sea, outside the coast. 

Over the greater part of 
England the direction of the wind was South-easterly ; 
a common phenomenon on the approach of a serious 
storm, owing apparently to the in- draught of air towards 
the region of diminished pressure. 


FIG. 19. November 28, 1874 ; 
5 A.M. Approaching depression. 

The Motion of Storms. 

The next chart, for 6 P.M. on the same day (fig. 20), 
shows the central isobar (of 29.0 inches) over the south 
of Ireland, while the SE. winds over England have 
veered to SW., and the isobars, previously running 
nearly north and south, show a marked curvature. 
Even in Norway pressure has given way, the isobar 

FKJ. 20. November 28, 1874; 
6 P.M. Depression advancing. 

FIG. 21. November 29, 1874 
8 A.M. Centre over Wales. 

of 29.8 inches having taken the place of that of 
29.9 inches on the previous chart. 

During the night the storm made rapid progress, 
and at 8 A.M. (fig. 21) we have the conditions already 
described at p. 39, the centre of the storm lying near 
Holyhead, and the influence of the disturbance extend- 
ing over the whole of Western Europe. 

<)2 Weather Charts and Storm Warnings. 

At 6 P.M. on the 2gth (fig. 22) we find the centre 
of the storm near Newcastle, with a reading of 28.4 

inches, while the general 
course of the isobars is that 
of an ellipse, with its longer 
axis stretching east and 
west. The Southerly winds 
have nearly entirely dis- 
appeared, and the gales are 
from East in Scotland, and 
from West over England, 
without any amount of 
Noith-west winds worth 

It" we were to pursue 
this disturbance further, we 
FIG. 22. November 29, 1874; should discover that the 

6P.M. Centre near Scarborough. Qyal gh of the isobars 

Depression passing off. 

was caused by the approach 

of a second cyclonic area towards Ireland, which checked 
the rise of the mercury and reduced the gradients in 
the rear of the former system, and thus hindered the 
development of the North-westerly winds a feature 
to which allusion has already been made at p. 65. In 
fact, the chart for the next morning exhibited two 
separate areas of depression, one over Ireland, the other 
over the North Sea, while the latter area showed evident 
signs of being filled up. 

It is not necessary to quote other instances to prove 

The Motion of Storms. 


the fact of the motion of cyclones, and we shall now 
proceed, by way of contrast, to consider the motion, or 
rather the comparative absence of motion, in anticyclones, 
and shall take the period at the beginning of February 
1874. It is not necessary to give more than the daily 

FIG. 23. February 2, 1874 ; 8 A.M. FIG. 24. February 3, 1874 ; 8 A.M. 
Anticyclone. Anticyclone stationary, but chang- 

ing in shape. 

charts for 8 A.M., as the changes from day to day are 
very slight. 

Fig. 23, for February 2, shows an area of readings 
above 30.4 inches over the Channel. In fig. 24 this 
has hardly changed its place, but a sudden rise of the 
barometer has taken place over the north of Scotland, 
so that we have two independent centres of high pressure 

94 Weather Charts and Storm Warnings. 

over our limited area. In fig. 25, for the 4th, the 
secondary anticyclone has vanished, and we have the 
conditions already cited at p. 40 as a typical anticyclone, 
and at the same time the effects of a cyclone outside 
the Shetlands are just felt on the coast of Norway, where 
the barometer has begun to give way, accompanied by 
a stiff South-west gale at Christiansund. 

FIG. 25. February^ 1874 ; 8 A M. 
Original shape regained. 

FIG. 26. February 5, 1874; 8 A.M. 
Anticyclone disappearing. 

The last chart of the series, fig. 26, shows us the 
central isobar of 30.6 inches over the Eastern Counties, 
while all over the North Sea strong Westerly winds 
have appeared, owing to the advance over Norway of 
the cyclonic disturbance just mentioned. 

In this case, therefore, which is a fair sample of the 

The Motion of Storms. 95 

behaviour of areas of high pressure, it is evident that 
the persistence of the conditions of high barometrical 
pressure over the south of England was as marked as 
the instability of the conditions of low barometrical 
readings was proved to be in the preceding instance. 
Anticyclones, however, do change their position, but 
it is needless to pursue the subject further in this place. 
Before treating of the agencies which appear to rule 
the motion of cyclones, it may be well to say a word 
or two on the influences which appear to modify 
that motion, at least in the neighbourhood of these 
islands. We find that the storm path is much affected 
by the contour of the country. It is well known that 
the west coasts of Ireland and Scotland are bold and 
rugged. It has sometimes happened that a storm has 
advanced from the Atlantic to the coast of Kerry, but 
has returned, and, passing out to sea again, has moved 
northwards along the coast until it found the opening 
of Donegal Bay, and has then crossed Ireland along 
the tract of low land stretching from Ballyshannon to 
Dundalk, to the Irish Sea, where the centre often appears 
to be arrested for a while : or else the disturbance, 
clearing Ireland altogether, and crossing Scotland south 
of the Grampians, has passed out to the North Sea. 
In fact, we have already seen, from the phenomena of 
the Hallsberg tornado, p. 29, that the stratum of air 
affected by a storm is frequently very thin as compared 
with the entire depth of the atmosphere. We can 
therefore easily understand that eddies in the stratum 

96 Weather Charts and Storm Warnings. 

of air which is in immediate contact with the surface of 
the ground may be very much affected in their character 
and motion by the irregularities of that surface, which 
have a tendency to turn them aside. 

It need hardly be pointed out that the susceptibility 
of storms to such a deflecting influence, exerted by the 
contour of the country, bears some resemblance to the 
behaviour of smoke-rings, which are deflected from their 
course by obstacles placed in their path, although they 
may not actually come in contact with the obstacles. 
Remembering that the depth of storms is insignificant 
compared with their lateral extent, we may to a certain 
extent compare cyclonic disturbances to smoke-rings, 
although the constitution of the two phenomena is 
widely different. 

As regards the actual direction of motion of cyclones, 
when we come to examine a series of charts extending 
over a wide area, like those of Captain Hoffmeyer and 
others, for Europe and the Atlantic, it is found that 
the cyclones have a tendency to move round the anti- 
cyclones, but it is almost impossible to follow out these 
motions fully when we can deal only with the limited 
area of our own charts. It is, however, from this 
mutual action of the areas of high and low pressure on 
each other that we gain some notion of the coasts 
which are likely to be visited by a storm, of the direc- 
tion which that storm will take, and of the quarter 
whence the wind in that storm will blow hardest. 

If we find the readings highest over Ireland, the 

The Motion of Storms. 


depressions will sweep down over Scandinavia or even 
over the North Sea, in a direction from the northward 
to the southward, giving us Northerly gales along the 
east coast of Scotland and England, owing to the fact 
that the steepest gradients will be on their western 
sides. These areas of low pressure will follow each 

FIG. 27. November 17, 1874; 8 A.M. Pressure highest in south-west. 
Depression disappearing over Denmark. 

other until by some means or other the excess of pres- 
sure in the west has been reduced. 

Figs. 2730 afford very good examples of the above 
conditions. We start on November 17, 1874 (fig. 27), 
with readings highest in the south-west, general North- 
west winds, and a definite depression existing over 
Denmark, marked by the isobar of 29.3 inches. At 


98 Weather Charts and Storm Warnings. 

the same time a backing wind at Stornoway and the 
SSE. wind at Christiansund show that a fresh disturb- 
ance is not far off. Fig. 2 8, for next morning, Novem- 
ber 1 8, still exhibits the traces of the cyclonic disturb- 
ance over Denmark, but the lowest readings there are 
about 29.7 inches, showing a rise of half an inch in 

FIG. 28. November 18, 1874 ; 
8 A.M. New cyclonic disturbance 
appearing in north. 

FIG. 29. November 19, 1874 ; 
8 A.M. Depression over Heligo- 
land. A third appearing in the 

pressure. All over the chart, except just in the south- 
west, the course of the isobars is changed, and the 
advancing depression is perfectly visible, and has pro- 
duced its effects on the winds, causing them all to 
' back ' to W. or SW., and even to SE. in Shetland. 

The Motion of Storms. 


The morning of the iQth (fig. 29) shows us a great 
change j the area of low pressure, which lay off Shet- 
land on the previous day, has advanced rapidly south- 
wards, and lies over the mouth of the Elbe, while 
another similar system has appeared over Shetland. 
Each of these independent disturbances has its own 
atmospheric circulation. 

The last chart of the series is fig. 30, for the 2Oth. 
In it we see the last-named 
depression situated in 
nearly the exact locality 
of its predecessor, over 
northern Germany, while 
the disturbance of the winds 
at the northern stations 

shows that the perturbations 
are not yet at an end. The 
rise, however, of the baro- 
meter over Holland indi- 
cates that the general 
distribution of pressure is 
undergoing a change. 

In this case, therefore, FIG. 30. November 20, 1874; 

We have a Succession of 8 A ' M - Depression disappearing 

over Denmark. 

cyclonic disturbances chas- 
ing each other for several days, along similar paths, 
while the region of highest pressure lies in the west. 

If the highest readings be over France, the path of 
the cyclonic disturbances will run from west to east, 

H 2 

ioo Weather Charts and Storm Warnings. 

across the British Isles, and the wind will blow hardest 
from the Westward. When the centre of the anti- 
cyclone lies over England we find that the storms 
hardly affect our stations at all, but pass outside the 
north coast of Scotland, producing strong Westerly gales 
on the Norwegian coast, above the sixtieth parallel, at 

Christiansund, Bodo, and 
northwards. It is needless 
to cite instances of these 
conditions ; they are already 
admirably illustrated by 
figs. 25 and 26 (p. 94). 

When the area of high 
pressure is located more to 
the east or south-east, over 
Denmark or Germany, the 
cyclonic systems cannot 
make good their footing 
over these islands at all, but 
pass from south to north 
FIG. 31. January 14, 1875 ; 8 A.M. outside the Irish and Scotch 
Pressure highest in south-east, coasts, producing gales from 

Depression off Hebrides. 

the Southward on our 

shores. Such storms frequently only appear in the 
west of Ireland, and are thence propagated to the 
Hebrides, Orkneys, and Shetlands, without extending 
to the English coasts. 

A very good instance of this type of weather 
occurred in the beginning of the year 1875. Fig. 31, 

The Motion of Storms. 


for January 14, shows us the barometer highest over 
Switzerland, and the isobars running NE. and SW. 
across these islands, while a depression is just visible 
off the NW. coast of Scotland. Fig. 32, for the I5th, 
shows but little change of pressure over Central Europe, 
but a fresh disturbance has appeared off the west coast 


FIG. 32. January 15, 1875 ; 8 A.M. 
New disturbance off west coast of 

FIG. 33. January 16, 1875 ; 8 A.M. 
Pressure highest in south. 
Depression off NW. coast of 
Scotland, moving eastward. 

of Ireland, causing the winds to back to S. and SE., 
and blow with the force of a gale. Fig. 33, for the 
1 6th, shows us this second disturbance in the position 
of its predecessor of the I4th, but, pressure having 
given way over Norway, the conditions are altered, and 

IO2 Weather Charts and Storm Warnings. 

the region of the highest barometrical readings is 
transferred to Spain. 

We have hitherto spoken of the cases when pressure is 
highest in the west, south, and east respectively, but the 
anticyclone must sometimes lie to the northward of us, 
and then, if ever, should the disturbances advance from 
the eastward. Such a movement is, however, exces- 
sively rare in these latitudes, though this is not the case 
with the tropical hurricanes or cyclones, all of which 
at first move from the eastward. There are, however, 
some principles which have not yet been thoroughly 
explained, and which are antagonistic to the develop- 
ment of such a motion in the storms of our part of the 
Temperate Zone. A few words will be said, hereafter, 
as to the causes which have been adduced to account 
for the motion of storms, but it will suffice, at this 
juncture, to remark that the motion of cyclones round 
anticyclones will not by any means account for all the 
motions which have been noticed in our storms. 

The typical cyclone of November 29, 1874, figs. 
1922 (pp. 90-92), to which reference has frequently 
been made, did not skirt round the region of high 
pressure on our charts, but travelled directly towards it. 
In this case it is possible that a study of the weather 
Over more extensive charts such as those of the 
North Atlantic, now being published by our Office 
might throw more light on this question of motion. 

While discussing the motion of storms, it may be in- 
teresting to trace the path of a very erratic disturbance 

The Motion of Storms. 103 

which visited these islands in April 1872, and the 
course of which has been followed out with the aid of 
the continuous records at our self-recording observa- 
tories. The following seven charts (figs. 34-40, 
pp. 104 and 105) show the successive positions of the 
central disturbance, and fig. 41 gives its track for the 
entire period. It came in over Ireland and retreated 
along its own path again. That it did not advance 
as far as the Straits of Dover is clearly shown by the 
records at Kew, and that it passed first on one side 
and then on the other of Falmouth is indicated by 
the records at that observatory, which prove by the 
direction of the shifts of wind, on the same principles 
as are explained in Chapter V., that the path of the 
centre lay first on the southern and then on the 
northern side of the station. 

This storm presents us also with the rare pheno- 
menon of an advance from the eastward. As far as we 
can at present form an opinion, this latter condition de- 
pends on circumstances of pressure far outside the area 
of these islands, or even of that embraced by our 
weather charts. It is needless to say that such a storm 
as that just described afforded a striking instance of 
failure of warnings, as will be explained in Chapter 
IX. (p. 137). 

The storms of which we have been treating have all 
exhibited motion more or less rapid, and in various 
directions, but it sometimes happens that areas of low 
pressure are stationary for two or three days together 

FIG. 34. April 20, 1872; 8 A.M. FIG. 35. April 20, 1872; 6 P.M. 
Centre of disturbance near Water- Centre of disturbance off south 
ford. coast of Ireland. 

FIG. 36. April 21, 1872; 8 A.M. FIG. 37. April 21, 1872; 6 P.M. 
Centre of disturbance near Havre. Centre of disturbance near Ports- 

FIG. 38. April 22, 1872; 8 A.M. FIG. 39. April 23, 1872; 8 A.M. 
Centre of disturbance near Pen- Centre of disturbance off south 
zance. coast of Ireland. 


FIG. 40. April 24, 1872 ; 8 A.M. FIG. 41. April 20-24, 1872. 

Disturbance passing off over Ireland. Path of disturbance. 

io6 Weather Charts anct Storm Warnings. 

nearly to the same extent as the anticyclone already 
noticed (figs. 23-26, pp. 93 and 94). It is a remark- 
able fact as regards these islands that there are certain 
localities which apparently exert an attraction on these 
systems, and so retard their motion for a time. This is- 
most strikingly the case with the region situated at the 

FIG. 42. November 30, 1872 ; FIG. 43. November 30, 1872 ; 
8 A.M.. Depression appearing off 6 P.M. Centre of depression near 
Valencia. Cork. 

entrance of the Channel, where at times, especially in 
the early autumn, cyclonic disturbances appear to remain 
for days together almost immovable. 

A very good instance of a stationary depression 
occurred at the end of November 1872. Fig. 42, for 
November 30, at 8 A.M., shows us the isobar of 

The Motion of Storms. 


28.5 inches over the south-west of Ireland. There is 
nothing else very remarkable, except that readings are 
generally very low, and that a secondary depression 
appears off the coast of Lincolnshire. Fig. 43, for 6 P.M. 
on the same day, exhibits the centre of a storm over 
Cork, but conditions otherwise do not show much 

FIG. 44. December I, 1872 ; 8 A.M. Centre of depression 
nearly in same place, but depression filling up. 

change, except that pressure has given way considerably 
over France. 1 

Fig. 44, however, for 8 A.M. December i, shows us 
the centre of the cyclonic area in nearly identically the 
same spot as it had occupied fourteen hours previously 

1 In 1872 we did not receive 6 P.M. reports from Norway, so that 
portion of the chart is blank. 

io8 Weather Charts and Storm Warnings. 

but the whole system is being filled up, for the lowest 
reading, indicated by an isobar, is 28.7 inches, while 
28.3 inches was the minimum in fig. 43. 

This disturbance subsequently took a course to the 
SSE., and passed on to the west of France. 

In this chapter, therefore, it has been shown that 
atmospherical disturbances advance over the earth's 
surface, and that cyclonic systems move more rapidly 
than anticyclonic. It has also been remarked that we 
know, very little about the rate at which a storm is 
moving until it has advanced well within the area of 
our telegraphic reporting system. We are also virtually 
ignorant of what the real cause of this motion of 
storms is whether it is due to the change of position 
of the entire mass of disturbed air, or to the continuous 
production of a depression in front of the storm while it 
is filling up in the rear. We have traced some cases of 
motion of cyclones, and have stated that this is modified 
by the contour of the country over which they have 
to pass, but that it appears to be regulated in great 
measure by the position of the anticyclonic areas in the 

The Use of Weather Charts. 109 



IF we come to consider what is the practical use, to an 
ordinary observer, of weather charts, giving, as they do, 
a representation of past conditions of weather, it will 
be necessary to recapitulate briefly the results to which 
we have been led in the several preceding chapters, 
and then to enumerate some of the general principles 
on which a judgment is formed as to coming weather, 
when the observer knows by the study of a chart what 
conditions have been prevailing elsewhere. Subse- 
quently some of the theories will be mentioned which 
have been propounded to account for the origin and 
motion of storms, and, lastly, we shall allude to the 
ideas which are entertained by some meteorologists as 
to a periodicity of rain and of storms, and as to a 
connection between disturbances in the gaseous enve- 
lope of the sun, and those which take place in our 
own atmosphere. 

It will be useful in the first place to summarise 
briefly the principles which we hope have been estab- 
lished in the preceding chapters. 

1 1 o Weather Charts and Storm Warnings. 

Chapter I. has been devoted to an account of the 
materials, in the way of observations, with which we 
have to construct our edifice. 

In Chapter II. the wind has been discussed, as the 
atmospherical phenomenon which is most directly 
related to weather. It has been shown that wind is 
always connected with some disturbances of the pres- 
sure of the atmosphere ; its existence being due to the 
tendency of an elastic body, like the air, to regain the 
condition of equilibrium from whence it has, by any 
means, been disturbed, while its motion is regulated by 
certain fixed laws. 

The subject of Chapter III. is the barometer, as 
being the instrument by means of which the amount of 
atmospherical pressure, and therefore of its disturbances 
from time to time, is measured. It is shown how the 
distribution of pressure on a weather map is indicated 
by the isobars, and how there are two grand types of 
disturbance of atmospheric equilibrium, named respec- 
tively cyclonic and anticyclonic, according as the 
pressure is locally either in defect or in excess of its 
average value over the area represented in the special 
chart in question. These two types of disturbance are 
characterised by different features, among which the 
most striking is the contrast in the direction of motion 
of the wind. 

In Chapter IV. gradients have been explained, and 
it has been shown that the distribution of pressure, as 
measured by the gradients, is the best guide we have 

The Use of Weather Charts. 1 1 1 

towards a knowledge of the laws of wind motion, and 
therefore, in some degree, towards a knowledge of 
coming weather. It has, however, been stated that 
there are apparently agencies other than the distri- 
bution of pressure which influence the force of the 
wind, but that their precise nature has not yet been 
ascertained. It has also been pointed out that the 
phrases 'gradients for such and such winds,' &c., are 
nothing more than the expression of the laws of wind 
motion in a practical form. 

Chapter V. contains a closer examination into the 
principal features of cyclonic and anticyclonic systems 
respectively, and into the character of the weather 
which accompanies each. It states that, while anti- 
cyclonic areas are usually stationary or move slowly, 
cyclonic areas travel more or less rapidly over the 
country, and shows that their passage past a station 
presents marked features of difference according as 
that station lies on the right-hand or left-hand side 
of the path of the centre. This has been illustrated 
by the continuous automatic records of some of 
the self-recording observatories in connection with 
the Meteorological Office, and it has been shown 
how the usual changes of wind are explained by this 

It has also been explained that cyclonic systems 
are not always isolated, but have occasionally what 
may be called secondary systems in connection with 
them, which exert a material influence on the gradients, 

I 12 

Weather Charts and Storm Warnings. 

and consequently on the winds belonging to the 
original disturbance. 

Finally, Chapter VI. has been devoted to the motion 
of atmospherical disturbances, and there it has been 
explained, at more length than in the preceding 
chapter, that cyclonic systems move more rapidly than 
anticyclonic. Some cases of the motion of storms 
have been traced, and it has been shown that, while we 
know very little of the rate of motion of a storm before 
it arrives, we can frequently form a conception of the 
probable direction of motion of the storm. This 
motion is affected by the contour of the country over 
which it passes, and it appears in great measure to be 
regulated by the position of the anticyclonic areas in 
the neighbourhood. 

The least consideration of the principles just enu- 
merated will abundantly show that the weather pre- 
vailing over any district on a given day is clearly 
affected by, if it may not be described as the result of, 
the conditions prevailing in the districts round about, 
so that, by observing what those conditions have been 
for some time back, we are enabled to form some 
opinion as to the weather which we are experiencing, 
and which is related to those conditions, whether it is 
likely to be permanent or transient in its character, 
and, if transient, in what way it is likely to change. 

It is hardly necessary to say that the fresher the 
information is the better, but, owing to the cost, it is 

The Use of Weather Charts. 1 1 3 

not possible in this country to follo\v the example of 
the United States, and send out by telegraph, in all 
directions, the entire mass of figures in the Daily 
Weather Reports, in order that it may be published 
simultaneously in every large town. Practically, in 
these islands the readers of newspapers and the sub- 
scribers to the Daily Weather Report cannot see their 
charts until several hours, frequently even an entire 
day, after date. 

Such information, however, can generally be turned 
to very good service, if rightly used, and we shall take 
the case of an observer resident not far from London, 
and see what he can do with it. 

It has already been explained that most of our dis- 
turbances travel from the westward, and, as they take 
a certain time to advance over the distance which 
separates the east of England from the west of Ireland, 
we can learn, by studying the chart for the previous 
day, what conditions were then existing at the western 
stations, and consequently can gain some idea of what 
is likely to be the result of the changes which we 
notice by our own local observations, instrumental and 

If, for instance, the barometer at our station is high 
and steady, with dry weather and light winds, either in 
winter or summer, we may form a general notion that 
the type of weather is anticyclonic, and, as this type is 
peculiarly permanent, we may feel sure that any change 
will give us at least several hours' notice of its coming, 


1 1 4 Weather Charts and Storm Warnings. 

by alterations in instrumental readings, occurring over 
some part of the district covered by the Reports. The 
study of the charts in this case is the more important, 
because frequently, at such a time, ' cirrus ' or ' mare's 
tail ' clouds appear in the sky, which are usually the 
precursors of coming wind ; and change of wind in such 
a case means change of weather. The charts then will 
show whether or not this change of weather had made 
its way down to the earth's surface at any place on the 
previous day within the district just referred to. 

In the case of cyclonic conditions, we have already 
learnt that it will not do to trust implicitly to the fact 
of storms advancing from the westward, but, on the 
other hand, some ideas have been advanced by means 
of which we can, at least in some cases, forecast the 
probable direction of motion of a storm. At any rate 
we are usually able to learn, from a careful study of the 
chart, whether we have to deal with disturbances on a 
large or a small scale, and, accordingly, whether the 
changes in progress indicate lasting, or merely temporary 
conditions of weather. 

These charts are therefore useful helps to the local 
observer, and will be found so by those who study them 
regularly, and combine with that study careful and sys- 
tematic observations of their own instruments, and of local 
weather ; but for the purpose of judging whether a 
particular afternoon will be wet or fine, which is all 
that the public generally care to know about weather, 
it is obvious that charts which are necessarily twenty- 

The Use of Weather Charts. 1 1 5 

four hours old in many places cannot be of much 
service. Moreover, the phenomena which we include 
under the general term ( weather ' often depend in great 
measure on the nature and conformation of the ground 
in the neighbourhood of the observer, so that one place 
will be much more liable to rain during disturbed 
weather than another, while a second may exhibit a 
greater tendency to the formation of fog at a calm 
period than an adjacent district. As, therefore, such 
exceptional tendency is confined to each special locality, 
and does not belong to the phenomena produced at all 
stations by the system of circulation prevailing at the 
time, it is necessary that the observer who endeavours 
to forecast probable weather should seek to ascertain 
under what conditions such peculiarities manifest them- 
selves, as it would be useless to apply merely general 
rules in order to interpret the meaning of phenomena 
of a purely local character. 

On the whole, it must be said that our insular and 
exposed position precludes us, in the present state of 
our knowledge, from the possibility of issuing forecasts 
of future weather with such confidence as some of our 
neighbours on the Continent of Europe, not to speak of 
the staff of the Chief Signal Office in Washington, 
which has attained a marked proficiency in weather pre- 
diction. We shall, however, in Chapter X. state what 
is being done in this direction. 

Let us now examine in what manner the knowledge 
we have gained would enable us to issue storm warnings 

I 2 

1 1 6 Weather Charts and Storju Warnings. 

to the coasts. In the first place, we can in many 
instances learn from the general conditions of wind, &c., 
in fact, of weather as a whole, what will be the character 
of the changes likely to occur, and in what direction they 
will propagate themselves. 

We then notice the conditions of pressure as shown 
by the barometer, what its changes are, and whether 
any changes which may manifest themselves at any 
spot are traceable at adjacent stations, and in what 
degree, so that we can gain some idea of the extent of 
the disturbances which are approaching. 

We next see in what degree the barometrical changes 
are borne out by the wind, both in direction and force ; 
in fact, we often get earlier intimation of approaching 
change of weather from abnormal features in the direc- 
tion of the wind, than from the barometer readings 
taken by themselves. 

Temperature, too, must be looked to, for as a general 
rule, in winter, if the temperature has been very low, 
even though the wind has been Southerly, we need not 
fear a Southerly gale until the advancing depression has 
exerted a marked influence in raising the temperature. 

We have then the reports as to weather at the different 
stations, and the deficiency of information of this nature 
has been amply explained in Chapter I. To this class 
of observations belong reports of the character and 
motion of clouds, and of the clearness or mistiness of the 
air. Each of these latter may be, in its way, a 
prognostic of storm ; the former, an unnatural clearness, 

The Use of Weather Charts. 1 1 7 

being a very bad indication before the storm has 
actually reached us and affected our winds, while the 
latter, mistiness, is often the immediate forerunner of 
coming rain, and sets in when the wind in an advancing 
cyclonic system has begun to blow from the South or 
South-east, or more especially on th^ ridge of high 
pressure or debateable ground, which lies between the 
NW. wind of a retreating, and the SE. wind of an 
advancing, area of low pressure. 

Reports of the Aurora are also very useful as cor- 
roborative evidence. The winter of 1875-6 was 
remarkably free from Westerly winds, i.e., from cyclonic 
disturbances passing to the northward of these islands, 
and was also remarkable for the rarity of reports of the 

It is, however, obvious that we, in the Meteorological 
Office, labour under prodigious disadvantages in at- 
tempting to issue warnings for the United Kingdom, 
Interest therefore attaches itself to the attempts which 
we are making from time to time to draw con- 
clusions from the experience of past weather which 
may be useful to us in forming opinions on what is 
to come. 

Attention has already been drawn to cyclonic dis- 
turbances which follow each other along the same track 
as long as the general distribution of pressure remains 
unchanged, and it has also been stated at p. 15 that 
cyclonic disturbances travel across the Atlantic for long 
distances, if not from shore to shore. This is abun- 

1 1 8 Weather Charts and Storm Warnings. 

dantly proved by the fact that steamers, when outward- 
bound to America, often meet a succession of such 
cyclonic systems on their way towards Europe, and, 
when homeward-bound to England, often run for a cer- 
tain time with such systems, so that in the former case 
the changes in the instrumental indications, as well as 
in winds and weather, are much more rapid than in the 
latter, on the simple principle that if walking in a 
crowded thoroughfare you meet many more vehicles 
than you pass or than overtake you. We shall return 
to this subject of the advance of storms from America 
at p. 183. 

If, then, these disturbances are frequently travelling 
in succession over the Atlantic, it is evident that we in 
these islands must at times encounter a series which 
bears a close resemblance to a series which has pre- 
ceded it, as to the general character of the areas of 
depression themselves, and as to the intervals which 
occur between them. This we have noticed in our 
own weather on various occasions, and the parallelism 
has been found to hold good for as many as eight 
days together. 

As to the theories which have been of late years 
propounded to account for the origin and motion of 
storms, no attempt will be made to discuss them, but 
the broad features of some of them will be indicated 
in a very cursory way. 

Professor Mohn, the Rev. W. Clement Ley, and 
others attribute the generation and motion of storms 

The Use of Weather Charts. 1 1 9 

to the condensation of moisture in the form of clouds, 
resulting in rain. This condensation takes place prin- 
cipally in front of the storm, and, so to speak, draws 
it on. 

Mr. Thomas Belt and Professor Reye account for 
the origin of the storm by the existence of a condition 
of unstable equilibrium in the atmosphere ; a cold 
stratum being situated at some height above the ground, 
if the air at the earth's surface becomes heated, it must 
eventually force its way through the superincumbent 
colder layer, and the upward current thus generated 
will be the core of the resulting cyclone. Professor 
Reye goes so far as to say that the storms of the 
largest dimensions experienced in these latitudes are 
substantially analogous to the smallest waterspouts or 
dust-whirls we can observe. 

In direct contradiction to this view, M. Faye has 
long propounded the view that cyclones are vortices 
descending from the upper regions of the atmosphere 
to the earth's surface, and that their motion is that of 
the upper current. 

Lastly, Mr. Meldrum and others consider that 
cyclones are generated in the space intervening 
between two currents, which themselves are tangents 
to the nascent cyclone. Thus, in the North Temperate 
zone, the conditions which would give rise to the 
cyclones are those of Easterly winds on the northern 
side of Westerly winds. The reader need hardly be 
reminded that every perfectly developed cyclonic 

120 Weather Charts and Storm Warnings. 

disturbance here must have East winds in the north 
and West winds in the south. 

It is clear that when 'doctors differ' to the extent 
indicated in the above sketches of the different theories, 
the world in general must be content to wait patiently 
for a complete theory of weather to be developed. 

Before we leave the subject altogether, it will per- 
haps be interesting to see the extent to which we are 
able to forecast the weather for longer periods than a 
day or two. Attempts have not unfrequently been 
made to predict the seasons for a long period in ad- 
vance, but without much success hitherto. One great 
cause for failure is that accurate meteorological records 
do not extend beyond the beginning of the present 
century at more than a few stations, and at these we 
are unable altogether to eliminate the local influences. 
Thus, it is hardly possible to say what has been the 
approximate temperature of these islands for more 
than forty years a period far too short for the 
definite recognition of a cycle. The shortest of such 
cosmical cycles which has been determined is the sun- 
spot period of I i T y years, according to Wolf, and there 
are indications of far longer periods, such as 33 years, 
or even 69^ years, according to Hornstein. 

Of late years Mr. Meldrum, of the Mauritius, has 
shown that the cyclones, for which that district of the 
Indian Ocean enjoys an unenviable notoriety, have 
been more frequent in some years than in others, and 
that these epochs of maximum frequency occur at 

The Use of Weather Charts. 1 2 1 

intervals of about eleven years, coinciding with those of 
maximum sun-spot frequency. 

This agreement is very important, and it appears to 
be corroborated by an examination of the rainfall at 
several stations which has been conducted by Mr. 
Meldrum and others. The results, for the compara- 
tively short period to which they refer, are striking, 
and are sufficient to show that a periodicity is traceable 
in the weather of the Southern Indian Ocean, which is 
suggestive of a relationship between the changes which 
take place on the sun's surface and the phenomena of 
our own atmosphere. 

It will at once be asked, Why has not this periodicity, 
if it exists, been detected long ago by an examination 
of European records, which are far more complete than 
any existing for the Indian Ocean ? The answer to 
this is twofold. 

In the first place, we are pre-eminently in the region 
of the variable winds, and our storms are not nearly so 
regular in their type as those of the Mauritius, where 
almost the sole type of storm is the true tropical 
cyclone with its concomitant rainfall. It is next to 
impossible in this country to keep a record of all the 
storms which pass over us. We have already seen 
that the existence of conjugate storms is not unfrequent, 
two or even three systems of disturbance being trace- 
able at one time within the limited area of the United 
Kingdom. Are these one single storm or several ? and 
how should they be counted in a catalogue ? Rain 

1 2 2 Weather Charts and Storm Warnings. 

also cannot be taken as a sign of the frequency of wind 
storms in a year; for although we know that warm 
winters are invariably wet and stormy, and, moreover, 
that cyclonic disturbances are accompanied by rain 
at all seasons, yet it cannot be asserted that the almost 
constant rain of 1872, and the recurrence of floods 
in 1875, were in any way related to storms of wind, 
though their connection with the presence of areas 
of depression over these islands is indisputable. 

There is, however, in the second place, a far deeper 
reason for the non-discovery of these cycles in any 
chance series of rainfall records. The sun passes 
through phases of greater and less activity, and the 
terrestrial phenomena corresponding to the epochs of 
the former character are excessive evaporation in some 
parts of the globe, and consequent excessive precipita- 
tion in others. We must therefore ascertain in what 
districts we are to look for the one and in what for the 
other of these phenomena respectively. 

It may be said at present that the precise mode 
in which the sun exercises his action on our atmo- 
sphere has not yet been explained, and, as far as the 
climate of western Europe is concerned, the warmest 
adherents of sun-spot influence must admit that obser- 
vations of the sun's surface cannot as yet be depended 
on as a sound basis for prophecy of coming weather. 

That this assertion cannot well be disputed appears 
from the figures which follow. 

The Use of Weather Charts. 




No. of 









i 879-80 














I0 5 

1884-5 4 2 -2 






The first column gives the mean temperature of the 
entire United Kingdom for the six months October- 
March, for the last eleven years, and the second gives 
the number of sun-spots observed at Kew during the 
same period. 

No approach to concordance is traceable between 
the two columns of figures. 

If we go further afield and compare the general 
temperature of the globe, at least the closest approxi- 
mation to it which is attainable, with the sun-spot curve 
for the thirty-five years ending with 1875, as has been 
done by Dr. Koppen l fig. 45, p. 124), we see that, 
though for some part of the time some of the tempera- 
ture curves appear to agree v/ith the curve of sun-spots, 
the accordance in one hemisphere is associated with a 
striking discordance in the other. 

1 Zeitschrift dcr ostcrrcichischen metcorologischcn Ccsdlschaft^ Bd. xvi. 
p. 149. 

124 Weather Charts and Storm Warnings. 

The figures which I have cited therefore support the 
statement that the precise nature of the relation be- 
tween what we may call solar and terrestrial weather 
has not as yet been demonstrated. 

As regards the whole question of prediction of the 

I 6 5 



N.Tcwp ZONG 

! 1 

FIG. 45. Diagram showing for the period of thirty-five years, 1841-75 
inclusive, the number of sun-spots and the average temperature of the 
Torrid and two Temperate Zones. 

seasons, either by sun-spots or by any other means, 
the same author, Dr. Koppen, has published several 
papers ' On Protracted Periods of Weather,' devoting 

The Use of Weather Charts. 1 2 5 

his attention especially to severe winters, and he gives 
the following summary of his results 1 : 

'The main feature of the entire investigation has 
been to prove that, for certain intervals, strongly 
marked periodical influences make their appearance and 
then vanish entirely, at times being replaced by others 
of a totally different character. No law has as yet 
been discovered for these changes, and so the outcome 
of the inquiry is on the whole negative, and indicates 
that all forecasting of the seasons is the merest 

We may therefore conclude that at the present date 
there is no immediate prospect of any one being able 
to state what the character of a future season will be, 
much less to tell a farmer in spring what crops he 
should put in with a prospect of a favourable period 
for the harvest. 

1 Zeitschrift dcr osterreichischcn meteorologischen Gesellschaft, Bd. xvi. 
p. 196. 

126 WcatJier CJiarts and Storm Warnings. 



BEFORE we treat of storm warnings and the general 
prediction of weather, it may be interesting to state 
what has been ascertained as to the frequency of gales 
on different parts of our coasts, and also as to their 
periodicity, that is, the tendency which they exhibit to 
recur at distinct seasons. 

As regards their local distribution, a table has 
recently been prepared in the Meteorological Office 
showing for each month, and for each quadrant of the 
compass, the total number of gales, and also of severe 
gales, for the different parts of the coast. The statis- 
tics refer to the fifteen years 1870-85. The coasts are 
divided into ten districts, as shown in the subjoined 

It is not necessary to give the details of the table in 
such a book as this, and accordingly the monthly 
percentages of the annual totals, and the totals them- 
selves, are all that will be reproduced, without dis- 
criminating between the different directions of the 

Distribution and Periodicity of Gales. 127 

>- > 












>, ^ 


The Shetlands 


J 3 







J 3 




Scotland, East 


J 3 




















J 3 



2 75 

Ireland, North 









J 3 


J 3 

I 9 8 

Ireland, South-west 













England, North-east 













England, East 












1 60 

England, South 












England, South-west 













The Irish Sea 












26 5 

The table bears internal evidence that the number 
of gales bears a certain proportion to the number of 
stations in each district from which reports have been 
obtained. The figures have been compiled from the 
Returns of the results of Storm Warnings which are 
prepared every year. From these Returns the north- 
west of Ireland has been excluded, inasmuch as no 
storm warnings have ever been sent to any point, except 
Galvvay, situated between the Shannon and Malin 
Head. Accordingly, the number of reports of storms 
which were collected from that region before the 
establishment of Belmullet and Malin Head as report- 
ing stations was not great. This is the main reason 
why in the column of totals the north of Ireland 
comes only seventh in point of frequency, instead of 
taking a position between Ireland SW. and Scotland 
NW., corresponding to its geographical situation. 

It will be seen that the entrance of the English 
Channel heads the list of frequency. This is mainly 





























































En 3 l 


L ^ 











( 128 ) 

owing to the extreme stormi- 
ness of Scilly. 

The eastern part of the 
Channel and the east coast of 
England are the least storm- 
swept parts of our shores. This 
is naturally to be expected, as 
the force of the storms is broken 
by their passage over the land. 

If we now exhibit graphically 
the monthly march of storm fre- 
quency in the ten districts (fig. 
46), we find that in every case 
the summer is all but absolutely 
free from storms. In almost 
every district, also, November is 
stormier than December. 

The most remarkable portion 
of the diagram is that represent- 
ing the spring time. It will be 
seen that in most of the districts 
there is a distinct maximum of 
storminess in March, which is 
especially prominent in the 
north-east of England, while it 
is entirely absent in the south- 
west of Ireland, and very nearly 
so in the Irish Sea and the 
south-west of England. 

Distribution and Periodicity of Gales. 1 29 

All these results require further study to explain 
them fully, but as far as they go they are very interest- 
ing and of much practical importance. 

In the matter of periodicity of storms there is a 
belief, which is very widespread, that equinoctial gales 
are a phenomenon recurring with almost perfect 
regularity. The remark is often made that it would 
be well for those proposing to take a voyage to 
wait until the equinoctial gales were over. In order 
to test the foundation of this general belief, I made 
a catalogue of all the serious storms which have been 
reported in the British Isles since Easter 1870, and 
communicated the results to the Royal Meteorological 
Society in i884. 1 I have recently extended the list 
up to Easter 1887, so that the figures refer to a period 
of seventeen complete years. 

The mode of inquiry has been to take fifteen-day 
periods (that is, to take seven days on each side of the 
equinox), and to divide the year into equal intervals of 
this length. Of course, we have twenty-four such 
spaces, with a remainder of five days, which have been 
omitted at the beginning of July, that being the 
calmest part of the year. 

The dates of the several storms are given in Table I. 
(p. 132), and curves are also given (fig. 47, p. 130) 
showing the march of frequency in the fifteen-day 
period, both for the crude and the smoothed figures. 

1 Quarterly Journal of the Royal Meteorological Society, vol. x. p. 236, 

K * 

1 30 Weather Charts and Storm Warnings. 

These show that the storms are all but exclusively 
confined to the winter half-year. 

A glance at Table I. (p. 132) will show how for a 
certain interval the stream of storm depressions sets 
over us and then for a time takes another path, leaving 
us at rest. In some years we have as many as four or 
five storms in a fifteen-day period, and in others we 








FIG. 47. Frequency of Storms in fifteen-day periods, according to 
Scott and Prince. 

In this diagram (representing Table III., p. 135) the vertical rules correspond approxi- 
mately to the middle days of the respective months. 

The faint full line corresponds to the original figures } 
The dark ,, ,, smoothed J 

The faint dotted . ,, ,, original ,, 

The dark ,, 



have none or only one. This happens even in the 
latter half of January, the stormiest period of all. 

The diagram shows that there is no strongly marked 
maximum of frequency at either equinox, but there are 
indications of periodicity which are very interesting. . 

Distribiition and Periodicity of Gales. 1 3 1 

Leaving the summer alone, as not worth notice, the 
frequency remains practically constant from the middle 
of August until the autumn equinox is past. The curve 
then rises rapidly, the value doubling itself and trebling 
itself in the two succeeding intervals. We then find 
a falling off at the time of the Martinmas summer in 
the first half of November, the latter half of November 
is stormy again, but the beginning of December is 
quieter. The absolute maximum of frequency is in the 
latter half of January, from which period the curve 
descends gradually, with one decided check early in 
February, to the same value which it had in August, 
and which it attains at the end of April. 

The frequency at the spring equinox is nearly double 
what it is at the corresponding interval in autumn, being 
twenty-two instead of twelve (col. i. Tab. III.). In one 
point, however, the phenomena agree : the equinoxes are 
periods of change in storm frequency. In the autumn 
this rises from twelve to twenty-nine as soon as the 
equinox is passed ; in the spring it falls from twenty- 
nine to twenty-two as the equinox approaches. 

Accordingly, persons who wait till the autumn 
equinox is over materially increase their chance of 
falling in with a storm, but in the spring it would 
apparently be wise to wait till April was well advanced 
if you wished to have calm weather at sea. 

If we look to see what evidence of the recurrence of 
storms for particular short periods is discoverable in 
our data, we find that the days most frequently so 

K 2 

1 3 2 Weather Charts and Storm Warnings. 

!!--; ;^ s ;? 5 

' - ' I - " c_"3 



? 8 M 


-IN O 

CO 00 

. ft 

; ; : i : 

: = ! : i 5 

1 1 

a -c B 

g 4- H 

Distribution and Periodicity of Gales. 1 3 3 

distinguished in Table II. are Jan. I and 19, on both of 
which storms occurred six times in the seventeen years. 





I > 








t > 



























































1 1 


































I c 


















































































































Seven days Jan. 18 and 26, Feb. 26, March 9, 
Nov. 10 and 20, and Dec. 20 show five each, and 
twenty-five days show four. 

134 Weather Charts and Storm Warnings. 

The stormiest period of all is the six days succeeding 
Jan. 22, the figures for which are 3, 4, 4, 5, 3, 4. 

The date of the Battle of Trafalgar, Oct. 21, is 
marked by two fours, but no other storm of historic 
fame is specially noticeable. 

It is of interest to find that the results above given 
for storm periodicity have recently been confirmed in 
a very satisfactory manner by the figures for storm 
periodicity in Sussex published by Mr. C. L. Prince 
in a recent number of the Quarterly Journal of the 
Royal Meteorological Society (vol. xiii. p, 79). Mr, 
Prince has carried on meteorological observations for 
nearly half a century, first at Uckfield, and latterly at 
his observatory on Crowborough Beacon, and the paper 
to which reference has been made gives the dates of all 
storms recorded by him during the forty years 1846 to 
1885 inclusive. 

As Mr. Prince's data refer to one district only, while 
mine refer to the entire United Kingdom, the total 
number of gales recorded by him in forty years is only 
371, while the larger area for seventeen years yields 459^ 
The distribution of gales in the two investigations 
agrees very closely, as will be seen by comparing the 
figures in cols. ii. and iv. of Table III. (p. 135) with 
those in cols. i. and iii. 

There is one remarkable exception that Mr. Prince 
finds a greater frequency of storms in May than my 
figures show. 

His results are shown graphically in fig. 47 (p. 130) 
by the dotted lines. 

Distribution and Periodicity of Gales. 1 3 5 

Mr. Prince concludes his paper with the very 
confident statement : ' We may now assert that gales 
at the period of the equinoxes are more the exception 
than the rule.' 







April 12 to April 27 





April 28 , May 12 





May 13 , May 27 





May 28 , June n 





June 12 , June 26 




July 2 , July 16 
July 17 , July 31 






Aug. i , Aug. 15 





Aug. 16 , Aug. 30 





Aug. 31 , Sept. 14 



ii. 5 


Sept. 15 , Sept. 29 





Sept. 30 , Oct. 14 





Oct. 15 , Oct. 29 





Oct. 30 , Nov. 13 





Nov. 14 , Nov. 28 





Nov. 29 , Dec. 13 





Dec. 14 , Dec. 28 





Dec. 29 , Jan. 12 





Jan. 13 , Jan. 27 
Jan. 28 , Feb. 11 
Feb. 12 , Feb. 26 







Feb. 27 , Mar. 13 





Mar. 14 , Mar. 28 





Mar. 29 , April 12 




8. 5 

i. Numbers according to R. H. Scott, 
ii. Ditto C. L. Prince, 

iii. The figures in column i. smoothed by the formula 

B = 



a + 2b -f c 


It may be said that this is not the first time that 
the notion that equinoctial gales are felt in northern 
Europe has been combated. Professor Dove, in the 

136 Weather Charts and Storm Warnings. 

last edition of his 'Gesetz der Stiirme ' (1873), says 
at p. 197, when speaking of storms arising outside 
the Trade- Wind Zone : 

' In western Europe these storms are rare, and pro- 
bably are never experienced in summer ; we must now 
examine into the reason of this, and must take care to 
remember the changes in atmospherical conditions 
which occur in spring and autumn. Lucretius calls 
these epochs " the wartime of the year," and in fact an 
Italian may well speak of equinoctial storms. Such an 
expression, however, sounds strange from the mouth of 
a German, for we cannot help thinking that he puts the 
equinox at Christmas, and forgets that September is 
the most usual month for the occurrence of that 

In conclusion, it may be said that in the Annalen der 
Hydrographie und Maritimen Meteor ologic for 1884, at 
p. 625, a paper appeared which showed that for the 
north coast of Germany the periods of the equinoxes 
were not remarkably stormy ; and as for the Adriatic, 
a paper by Dr. Robert Muller in Mittheilungen aus 
dem Gebiete des See^vcsens for 1887, at p. 144, shows 
that, though there exists at Pola a slightly marked 
maximum of frequency at the time of the spring 
equinox, nothing of the kind is traceable at the autumn 

Storm Warnings. 137 



STORM WARNINGS may be considered as the most 
immediate practical application of weather knowledge, 
and in fact it was the possibility of issuing such intelli- 
gence which gave the first impetus to the development 
of Weather Telegraphy. Although this branch of the 
subject has not yet been reduced to such strict rules of 
procedure as are demanded by the requirements of an 
exact science like Astronomy, meteorologists can at all 
events lay claim to having established some principles 
which are sufficient, in many cases, to throw valuable 
light on the conditions which may be expected to result 
from existing circumstances, and so to some extent to 
lift the veil which shrouds the future from our eyes. 

The first suggestion of the electric telegraph as a 
method of conveying intelligence of storms from one 
place to another is apparently due to Mr. Redfield in 
America, who shares with our own countryman, Sir 
W. Reid, the honour of having reduced the Law of 
Storms to a science. Redfield's paper is in the 
American Journal of Science and Art for 1847, 

1 3 8 Weather Charts and Storm Warnings. 

this was followed in this country, after a brief interval, 
by a communication from a man still living amongst 
us, Mr. John Ball, F.R.S., as will be seen below. 

The subject was taken up in a tentative way in the 
United States between the years 1850 and 1860, but 
further operations in this direction were brought to a 
standstill by the war in 1861. 

On this side of the Atlantic the credit of taking the 
initiative is due to Leverrier, the famous director of 
the Paris Observatory, who, as early as on February 1 7, 
1855, received the Emperor's sanction for the creation 
of an extensive organisation destined to distribute 
intelligence of weather over the countries embraced by 
it. The possibility of rendering such an organisation 
really useful had been previously shown by the study of 
the Balaclava storm of November 14, 1854, which had 
wrought immense damage to the allied fleets in the 
Black Sea, and by whose effects hardly a country in 
southern Europe had been left unscathed. It was evi- 
dent, from a mere cursory study of the facts of this 
storm, especially of its rate of progress, that timely 
notice of its approach might have been issued from the 
coasts which were first visited to those which subse- 
quently felt its effects. 

The plan gradually assumed a definite form, and in 
the spring of 1860 M. Leverrier addressed a letter to 
Sir George Airy, inviting the co-operation of this 
country in his scheme. This letter contains certain 
expressions which have such an important bearing on 

Storm Warnings. 139 

the subsequent history of storm warnings that they are 
worth translation and quotation. They are as follows : 

' The ultimate result of the organisation which we 
are establishing should be to announce a storm as soon 
as it appears at any point in Europe, to follow it on its 
course by means of the telegraph, and to give timely 
notice of it to the coasts which it may reach.' 

We shall presently learn to what extent this pro- 
gramme has been carried out, and to what extent it 
has been possible to go beyond it. 

In this country the idea of storm warnings had been 
broached before 1860, for at the meeting of the British 
Association at Swansea in 1848, Mr. John Ball, F.R.S^ 
read a paper entitled ' On rendering the Electric 
Telegraph subservient to Meteorological Research.' 
This communication is so interesting and brief, that we 
reproduce it in a note. 1 Eleven years subsequently, at 

1 ' What is popularly termed the weather is a general expression for 
the physical condition of the atmosphere with reference to heat, 
pressure, moisture, and the velocity and direction of its motion. Two 
classes of causes determine these conditions at any given point of the 
earth's surface. The first class may, for short periods of time, be con- 
sidered as constants, depending on the position of the point of observation 
on the globe, and the physical conformation of the adjoining district. The 
second class, upon which the proverbial uncertainty of the weather 
depends, arises from the influence exerted by each portion of the atmo- 
sphere upon those surrounding it, by virtue of which a disturbance of 
equilibrium at any one point is rapidly propagated in all directions. In 
common language, this is expressed by saying that the direction of the 
wind is at once the cause and the indication of changes of the weather. 
However far we may be from a general solution of the problem of atmo- 
spherical disturbances, meteorologists have made considerable progress in 
tracing the connection between successive states of the weather, owing to- 
the mutual influence of contiguous portions of the atmosphere. The^e 

1 4Q Weather Charts and Storm Warnings. 

the Aberdeen meeting of the same Association in 1859, 
.a formal resolution was adopted in favour of the organ- 
isation of such a system. A month after the meeting 
the loss of the Royal Charter on the coast of Anglesea 
arrested universal attention, and as there seemed, at 
first sight, a chance that that ship might possibly have 
been warned of her danger, the hasty conclusion was 
drawn that no storm could ever come on without giving 
timely indication of its own approach. Subsequent 
experience soon proved how very unsafe such a gene- 
ralisation was. 

Early in 1861 the first tentative warnings were sent 
out, and by the beginning of 1862 Admiral FitzRoy's 

cases have been studied d posteriori, comparing the known results with 
observations extending over considerable areas. Now that we have the 
means of receiving information in an indefinitely short space of time by the 
Electric Telegraph, these problems, under favourable circumstances, may 
be studied d priori. In London we may receive instantaneous intelligence 
of the condition of the atmosphere, as to the five ab.ove-mentioned 
elements, from nearly all the extremities of Great Britain. With a delay 
of about four hours we can have similar intelligence from the western part 
of Ireland, and with a still shorter delay, our communications may extend 
to the centre of France, the banks of the Rhine, and even to the frontiers 
of Hungary and Poland. 

' I do not pretend to say that with such elements for calculation we should 
at once be enabled to predict changes in the weather with absolute 
certainty. It would require some time to eliminate the action of accidental 
and local causes at particular stations ; but there is no reason to doubt that 
in a short time the determinations thus arrived at would possess a high 
degree of probability. The ordinary rate at which atmospherical disturb- 
ances are propagated does not seem to exceed twenty miles per hour; 
so that, with a circle of stations extending about five hundred miles in each 
direction, we should in almost all cases be enabled to calculate on the state 
of the weather for twenty-four hours in advance.' Report British Association, 
1848 (' Notices and Abstracts,' p. 12). 

Storm Warnings. 141 

system was definitely established. The principles on 
which our present warning system is based are mainly 
the same as the principles of that devised by the 
Admiral, but exhibit some contrasts to them. 

Admiral FitzRoy, when he issued a signal, intended 
by it to imply that the storm of which he gave warning 
would be likely to occur within the next seventy-two 
hours. He therefore had the signals kept up until the 
evening of the day on which they were hoisted, and 
then lowered, while by the present system they are not 
lowered until all danger of the gale appears to be past 
It is evident from this that if, as often happens, threaten- 
ing symptoms disappeared without the actual occur- 
rence of a gale, Admiral FitzRoy had no means of 
announcing to the ports, on the day after the issue of a 
warning, the fact of the improvement in the weather 
prospects. This arose from the fact that once his 
signal had been lowered there was no public indication 
that it had even been hoisted at all, and yet the warn- 
ing had been intended to cover three whole days. By 
the present system the fact that no signal is exhibited 
is, to a certain extent, a proof that no danger is appre- 
hended at the Office in London. 

It is hardly necessary to describe the signals, as their 
appearance is familiar to all who have been at seaside 
places, but for the sake of those who may not have been 
on the coast in winter, at which season the signals are 
most frequently visible, the following official explana- 
tion of their meaning may be reproduced. 

142 Weather Charts and Storm Warnings. 


The Meteorological Office issues (free of charge) to ports and fishing 
stations approved of by the Board of Trade, notices of atmospherical dis- 
turbances on or near the coasts of the British Islands. 


The fact that such a notice has been received at any station is made 
known by hoisting a Cone, which is three feet high and three feet wide 
at base, and appears as a triangle when hoisted. 

Southerly Gale. 

The Cone point downwards means that gales, or strong winds, are 
probable, at first from the Southward (from S.E. round by S. to N.W.) 

Should it appear likely that a gale beginning from between N. and N.W. 
is likely to veer towards N. or N.E., the North Cone is hoisted in preference 
to the South Cone. 

Northerly Gale. 

The Cone point upwards means that gales, or strong winds, are pro- 
bable, at first from the Northward (from N.W. round by N. to S.E.) 

Should it appear likely that a gale beginning from between E. and S.E. 
is likely to veer towards S. or S. W., the South Cone is hoisted in preference 
to the North Cone. 

Sudden Shifts of Wind. 

It must be remembered, that a Southerly wind is much more likely to veer 
quickly to a point North of West than a Northerly wind is to veer to a 
point South of East, and that a gale from East is more likely to back to the 
Northward than to veer to the Southward. 

Accordingly, when the South Cone is hoisted and the anchorage of 
harbour is exposed to the North- West, it is advisable to bear in mind that 
a gale commencing at South- West may shift suddenly to North-West* 

S tor in }Varni\ 



Day Signals. 


Gale probably Gale probably 

from the from the 

Southward. Northward. 

Night Signals (instead of the above) Lights in triangle. 

Three lanterns and one yard, 4 ft. long, will be sufficient. 

FIG. 48. 

144 Weather Charts and Storm Warnings. 
On Hoisting the Signals. 

The signal is to be kept flying until dusk, and then lowered to avoid 
unnecessary wear and. tear, but to be hoisted again at daylight next morning ; 
and so until the end of 48 hours from the time at which the message was 
issued from London, unless orders to lower the signal are received pre- 
viously. The time of issue is always marked on the telegram. 

At dusk, whenever a signal ought to be flying if it were daylight, a night 
signal may be hoisted in place of the Cone, consisting of three lanterns 
hung on a triangular frame, point downwards (for South Cone), or point 
upwards (for North Cone) as the case may be. They should be kept 
burning until late in the evening, say 9 or 10 o'clock, or longer if the loca^ 
authorities should think it important. 

These signals may be made with any lanterns, showing white or any 
colour, but the lanterns should be all alike. Red is best. Lamps are 
better than candles. The haulyards should be of good rope, and should be 
protected from chafing. The lanterns should hang at least three feet apart. 

Meaning of Signal. 

The hoisting of any of these signals is intended as a sign that in the 
neighbourhood (say, within a distance of 50 miles) of the place where the 
signal is hoisted, there is an atmospherical disturbance in existence which 
will probably cause a gale from the quarter indicated by the signal used, 
and the knowledge of which is likely to be of use to ihe sailors and fisher- 
men on that part of the coast. Its meaning is simply, "Look out ! It is 
probable that bad weather of such and such a character is approaching 

Hitherto it has been found that these signals have been justified in four 
cases out of five as follows : 

(l.) In three cases out of five by Fresh, Strong, or Heavy Gales 

(*".<?., Force eight or more). 

(2.) In one case out of five by Strong Winds or Moderate Gales 
(i.e., Force six or seven). 

In every case some of the principal reasons which have led to the hoist- 
ing of the signal are briefly explained in the telegram, which should always 
be kept posted up for public inspection while the signal is flying. 

The simplicity of the idea of using a cone, which 

Storm Warnings. 145 

must necessarily show the same shape as a triangle, no 
matter how it is blown about by the wind, is the great 
merit of FitzRoy's plan, as the sailor need not burthen 
his memory with a complex code of signals. In fact, 
when the Meteorological Office in 1868 made an 
attempt to introduce a more complex apparatus, like a 
railway semaphore, in order to convey intelligence to 
ships in the offing, the experiment fell perfectly flat, 
owing to the difficulty of instructing our coasting sea- 
men, for whose benefit storm warnings are originally 
intended, in the use of a totally novel system. 

The question of signals is not an easy one to set 
finally at rest. There is no doubt that the information 
conveyed by them would be infinitely more valuable if 
they could be shown from headlands which shipping 
usually make, such as the Lizard, than if they were 
simply exhibited in harbours. Here then we see the 
necessity for a special signal code. By our present 
system we can only give a limited number of possible 
signals, so that we cannot convey detailed information 
to ships at any distance from the shore ; say, even 
riding at anchor in the Downs. 

Our system is to hoist the signal as an indication 
that some information of a certain kind has been sent 
to the signal station, which it is of importance to the 
seaman to know : but he must land and read the tele- 
gram himself if he wants to learn more precisely what 
the facts are. We are as yet a long way off from the 
condition which may be described as the acme of 


146 Weather Charts and Storm Warnings. 

weather-signalling, and which has often been suggested, 
that of having a weather signal book containing charts 
of different types of weather, much like the illustrations 
in these pages, duly numbered, so that by simply hoist- 
ing a number on the signal staff, we could tell the 
seaman to what page in the book he had to turn, and 
what weather he had to expect. Thus, such a chart 
as that for November 29, 1874 (fig. 2), might be de- 
scribed as No. 371, and so on. It will, however, be 
many a long year before our knowledge of weather will 
justify our issuing such a collection of typical weather 
maps as would be likely to be of service for such a 
purpose, but possibly in the future something of this 
kind may be done. 

If it be asked what the real practice of storm warnings 
is at present, the reply is that it is a moderate advance 
on Leverrier's anticipations in 1860 (p. 139). It will 
appear from the previous pages that we have some sort 
of knowledge of the circumstances under which the 
various types of storms approach the different parts of 
our coasts, but these are not sufficient to tell us in 
every case when a storm is imminent. 

Some conditions give us more warning than others, 
but the cases in which we have definite premonitions 
are not of very frequent occurrence, and in general it is 
not until the storm is quite close at hand, and the 
barometer has begun to fall briskly at the outposts, 
that we feel ourselves justified in issuing a definite 
warning. The appearance of one of the small secondary 

Storm Warnings. 147 

depressions, however, not very unfrequently precedes 
that of a more serious storm ; accordingly, if we notice 
one of these lesser disturbances, we are at once put on 
our guard. 

A rapid and unexpected rise of the barometer is 
often the precursor of a coming depression, so that 
whenever we see a sudden rise we may expect an 
equally sudden fall, and must be on the look-out for 
the slightest tendency of pressure to give way. 

It should, however, be remembered that, as regards 
barometrical indications, we are guided not only by the 
actual fall of the mercurial column on a certain line of 
our own coasts, but also by the behaviour of the baro- 
meter at far distant Continental stations. Thus, in the 
case of the storm already so often quoted, November 29, 
1874, a warning had been issued to the coasts at 
3.30 P.M. on the 28th, when the most alarming sign 
was not so much the fall of the barometer at our south- 
western stations, taken by itself, as the fact that this 
was accompanied by a slight rise, of o.oi inch, since 
8 A.M. at Rochefort, readings in the south-west of France 
being already seven-tenths of an inch higher than in 
the south of Ireland. This showed that pressure was 
banking up to the southward, and that gradients were 
consequently becoming steeper along the Channel. If, 
when the barometer falls at our south-western stations, 
the fall affects those in the south of France, we know 
that we have less to fear from Westerly winds, as the 
tendency of such a change is to reduce the gradients 

L 2 

148 Weather Charts and Storm Warnings. 

for winds from that point, while a rise of pressure over 
France when it is already relatively high there, puts us 
at once on our guard for a blow from the Westward. 

Similar reasoning will apply to barometrical indica- 
tions in other parts, but the changes in direction of the 
wind are of almost equal value to them, and in some 
respects of even higher importance, as giving earlier 

The proverbial danger of a backing wind has already 
been alluded to, p. 78 ; this, however, is chiefly true 
as concerns the * backing ' of the wind from NW. to 
SW. and S., which indicates the approach of a fresh 
disturbance, for it has been explained that if one 
cyclonic system is following another in quick 
succession, as the wind on the southern side of each 
runs gradually from SE. through S. and W. to NW., 
the approach of the second depression must cause the 
vane to shift back from NW. towards SE. This back- 
ing is almost the most dangerous sign of all. 

It has, however, been stated already that it is not 
safe to depend on the observations from our western 
stations as to the direction and force of the wind, 
owing to the hilly nature of the coasts, so that we 
cannot implicitly trust the reports on this subject 
which we receive. 

Another very important sign as regards the wind, is 
the indraught of air towards the storm area. This is 
perfectly intelligible on the supposition that the air 
really flows in towards the centre and rises there. On 

Storm Warnings. 149 

the approach of a cyclone to the west coast of Ireland, 
we frequently find the winds South-westerly on the 
west coast of France, as on November 28, 1874 (p. 90), 
while we can hardly say that the barometer on that 
coast, at such a distance from the centre of the disturb- 
ance, had yet felt its influence. 

The reports of sea disturbance are even more un- 
certain guides to us than the observations of wind, as 
is explained in Chapter I., p. n. 

Temperature is also a very useful aid in gaining a 
knowledge of coming disturbance, for it will be remem- 
bered how the rise of the thermometer in front of a 
storm is a marked characteristic of areas of low pres- 
sure, but the indications of temperature have, for various 
reasons, not yet been reduced to strict rules like those 
of the barometer. 

We know, however, that a great contrast of tempera- 
ture between adjacent stations, being an indication of 
serious atmospherical disturbance, is the precursor or 
concomitant of a serious storm, so that if we find a great 
difference between the reports of temperature, we are 
at once warned of approaching disturbance. Some 
striking instances of this fact may be cited. On 
November 13, 1875, at 8 A - M -> tne thermometric report 
from Scilly was 57, and from Wick 21. The result- 
ing difference is 36. The gale of Sunday, November 
14, with its accompanying high tide and consequent 
inundations on the south coast and in London, was 
long remembered by the residents in that part of 

1 5 o Weather Charts and Storm Warnings. 

England. Again, on December 8, 1886, an unusually 
severe storm crossed the north of Ireland and England. 
This had been preceded by remarkable unsteadiness of 
temperature. On the 4th of the month the change in 
24 hours had been 21 at Hurst Castle and Dungeness, 
1 8 in London, and 17 at Loughborough. On the 
6th the change had been 22 at Loughborough, 21 in 
London, 20 at Oxford, and 19 at Cambridge and 
Dungeness. Accordingly, the atmosphere had been in 
a state of disturbed equilibrium for four days before the 

Besides all the symptoms of the approach of a storm 
which have been enumerated, there are the innumerable 
local signs which enable the fisherman and shepherd, 
not to speak of the inferior animals, to judge of coming 
bad weather. Among all these, almost the most 
important are the character and motion of the upper 
clouds. Another great sign is the clearness of the 
atmosphere, the unusual visibility of distant objects, 
which is well known as a sign of a coming gale. 

It must always be remembered that, in order to issue 
perfectly correct storm warnings, we should require to 
know the size, shape, position, and motion, in direction 
and rate, of an advancing depression, and also whether 
it is becoming deeper or the contrary, and that there is 
not one of these conditions of which we have a really 
sufficient knowledge at present, while of most of them we 
can have no knowledge at all till the storm has burst 
upon us. The problem which is put before the 

Storm Warnings. 1 5 i 

Meteorological Office daily, is similar to one which 
astronomers would at once recognise as impossible of 
solution, and that is to determine the elements of a 
comet's path from a single observation taken, say, in 
a brief clear interval on a cloudy night. The first 
glimpse we get of a storm must suffice for us to issue 
our warnings. 

It is therefore evident that for our own exposed 
western and northern coasts we can but rarely issue 
timely warnings, but fortunately these iron-bound shores 
are not frequented by an amount of coasting craft at 
all to be compared with that navigating the compara- 
tively calmer wafers of the two Channels and the North 

The results of the warnings to our oivn coasts have 
been printed in the Reports of the Office for several 
years back. The following are the figures for the year 
1885, being the last which have been published : 

1 5 2 Weather Charts and Storm Warnings. 




'C O " 

2L C 



< & 


.^ W 

vo" -t-J : "i : : 




^ J W 





od'in'S Sb 4^ 




c a. ta< c tj 'S 




^<J <J t^O fe 



Cd * 

S3Jt B 5S T W 


S oo 
H ~ 

jooag; ui sSuiuau^ 

W "-< 

> Q 

sjojaq uon^jg SUQ JB 

hH ; M H4 M N ; ; ; 






P w 

p9LJDK9J O 9Z)JO^ 94"[ 


s ^ 

AnBqjEd sSuiujByW 

c to 


5 p^ 

O &3 

gjojsq SUOIJBJS 




<J C! 

5! a 



o a 


juanbasqng Aq pay 
-psnf jou sSuiuaB^Y 




g eu GJ 

1/3 *o 6 


L pus 

o s 

H H 

9 ssoaoj ' S P U 'AV. 
Suojjg juanbasqng 
Xq payijsnf sSuiuaB^ 

*O r^ ro ^"OO O oo vO O 



** Q) 2 



* o % 

W < 

-dn puB g aoaoj 






S3[BQ juanbasqng 
Xq paynsnf sSuiuae^ 

rj-rOrj-M co^j-Tl-W N 



'Si 1 1 3 






8 I I 


-y^ pus jsioq oj sjap 




r-( ff M 



%> % 

> a> 










^xT'O T: 13" T3 r a'r3 w 




Storm Warnings. 


If these figures be compared with those for the 
preceding eight years, we see that the results in per- 
centages are fairly constant. 

Warnings justified 

Warnings not 

\ EARS. 

By Subsequent 

By Subsequent 
Strong Winds. 

Total Success. 






I6. 4 

I8 79 







20. 6 













21. 1 




56.2 21.6 







12. 1 











Accordingly, it appears that more than half the 
signals of approaching storms (force upwards of 8, 
Beaufort scale, ' a fresh gale ') were fully justified, and 
the same is the case for four out of five signals of 
approaching strong winds (force upwards of 6, Beaufort 
scale, ' a strong breeze '). 

It is therefore evident that the system of storm 
warnings renders it possible to give useful intelligence 
to the coasts, but it need hardly be said that the system 
for these islands is rather of the ' from hand to mouth ' 
type, and that our warnings would be far more useful if 
they could be issued sooner, so as to render it possible 
for captains to get intimation of a coming gale before 
leaving port This, however, can hardly be hoped for 
in the present state of weather telegraphy, and a few 

154 Weather Charts and Storm Warnings. 

instances will suffice to show some of the principal 
causes of failure of warnings, in addition to those 
arising from the imperfections of our arrangements, 
which have been already described. 

We shall commence with a case in which even a 
practised seaman's experience was at fault, so that we 
can hardly blame our telegraphic reporters, who are 
mostly landsmen, for their failure to notice, from the ap- 
pearance of the sky and sea, that a storm was imminent. 
Nearly fifteen years ago, on November 22, 1872, there 
was a tremendous gale in the Channel, when the Royal 
Adelaide was lost on Chesil Bank. On that day at 
noon the telegraphic reports had shown an apparent 
improvement on the weather of the previous day, so that 
the signals then flying were lowered on the south coast. 
At night the storm came on, and of course the comments 
on the Office were not favourable. On this day one 
of our best sea-observers, the late Captain Thomas 
Donkin, of the Inverness (one of the three ships that 
rode out the Madras cyclone of May 2, 1872), was out 
in the gale, and was blown back by it, hove to, from the 
Lizard to the Casquets, off Alderney. We wrote to him 
as soon as we heard that he was safe at anchor at 
Portland, to ask him whether he, being at sea, had 
anticipated the storm from the look of the sky and sea ; 
and his answer was : ' With respect to the weather on 
November 22, I may say that at noon I was standing in 
towards the land, between Falmouth and Plymouth, 
and a pilot cutter came alongside, and if I had had the 

Storm Warnings. 


least apprehension of such a gale as followed being 
near at hand, I should have taken a pilot and gone into 
Plymouth. The appearance of the weather at the time 
was fine, though the glass was falling, though not low at 
the time for SW. wind and unsettled weather.' 

In this case, therefore, the signals, once hoisted, had 
been lowered too soon. 

FIG. 49. October 20, 1874; 8 A.M. Depression approaching 
coast of Norway. 

The next case, which will be illustrated by charts, 
affords specimens of a double failure of warning. In 
the first instance the signals had been hoisted on the 
north-west coasts on the morning of the 2Oth, and had 
been lowered again in the afternoon, while subsequently 
they were hoisted unnecessarily and had to be lowered. 

156 Weather Charts and Storm Warnings. 

The gale is that of October 21, 1874, which has already 
been noticed, p. 72. 

Taking the first chart (fig. 49, p. 155) for 8 A.M. 
October 20, we find pressure highest over the entrance 
of the Channel, and lowest in the far north, where 
SW. gales are reported. These gales, with a falling 

FIG. 50. October 20, 1874; 6 P.M. Depression off Christiansund. 

barometer in Scotland, caused some alarm, but this was 
allayed by our finding conditions unchanged at 2 P.M. 
By 6 P.M., however (fig. 50), the depression existing in 
the north had advanced to the coast of Norway, where 
the barometer had fallen 0.6 inch, but the signs of 
danger to ourselves were the bend southwards of the 
isobar of 29.4 round the Butt of Lewis, and the backing 

Storm Warnings. 


of the wind, though still light in force, to South-west 
at Stornoway and over the Moray Firth, and, though 
last not least, the steady rise of the barometer at the 
entrance of the Channel. 

During the night the storm came on, and at 8 A.M. 
October 21, its centre lay half-way between Peterhead 
and the Shetlands (fig. 5 I ), while a terrific Westerly 

FIG. 51. October 21, 1874; 8 A.M. 
New depression off north of 

FIG. 52. October 21, 1874; 6 P.M. 
Depression has advanced to the 
coast of Norway. 

gale was blowing over the whole of the United King- 
dom, and the area of high pressure had been displaced, 
and lay over the extreme south-west of France. 

The next chart (fig. 5 2), for 6 P.M. on the 2 ist, shows 
us that the disturbance had made rapid progress east- 

1 5 8 Weather Charts and Storm Warnings. 

wards, and now lay over western Norway, while the 

winds in these islands had moderated in violence, and 

had veered towards North-west 

If we now turn back to fig. 17 (p. 83), we shall see 

the cause of the unnecessary warnings to which allusion 

has just been made. 

At 8 A.M. on the 22nd, a brisk fall of the barometer 

took place in the west of 
Ireland, with a ' backing ' 
wind, and warnings were at 
once issued, but, as will be 
seen by fig. 17, this was 
only a false alarm, for the 
b fresh disturbance was only 
secondary to the more 
serious one which had pre- 
ceded it, and its only effect 
was to equalise pressure in 
Ireland, by reducing the 
gradients and moderating 
the force of the wind. 

FIG. 53-January 21, 1875; SA.M. The storm which has 

Small depression over Belgium, been taken as a typical cy- 
Mand diStUlbanCe appr achi " e clone, that of November 29, 

1 874, whose course we have 

traced (figs. 1922, pp. 90-92), is a good instance of a 
partial success and partial failure. On the 28th, 
warnings were issued to the west, north, and east of 
England, but not to the Channel coast or to the 

Storm Warnings. 


French stations. It will be seen from fig. 21 (p. 91) 
that while the warning was quite successful on all the 
coasts to which it was issued, its omission was a signal 
failure at the southern stations. 

It will be interesting to examine yet another case 
of failure of warnings, and one which has arisen from 
the unexpected direction of 
motion of the storm. 

Fig. 53 (p. 158), for 8 A.M. 
January 21, 1875, shows us I" 
no less than three cyclonic 
disturbances over the area 
of our chart. One large one 
is disappearing over Nor- 
way. A very small second- 
ary one is lying near Dover, 
while the third the one 
we are about to trace is 
just showing itself off the 
west coast of Ireland. 

Warnings were issued to FIG. 54. January 21, 1875; 2 P.M. 

New disturbance near Pembroke. 

all our own coasts except- 
ing Scotland and the north-west of England, from 
Fleetwood northwards. Fig. 54, for 2 P.M. on the 
same day, shows us the disturbance with its centre 
near Lundy Island, while pressure over the south of 
France has barely changed since the morning. 

At this time, however, the depression was becoming 
deeper, and was extending rapidly to the southward. 

160 Weather Charts and Storm Warnings. 

55> f r 6 P - M - on the same day, shows us the 
centre of the storm lying over Dover, and a very heavy 
gale from North-west is blowing all along the west 
coast of France, while our shores had nothing but 
moderate breezes. 

These warnings, therefore, were quite unnecessary for 

FIG. 55. January 21, 1875 ; 6 P.M. New disturbance near 
Dover, in same position as that at 8 A.M. 

several of the English stations to which they were 
issued, but they ought to have been transmitted to 
France, where they would have been of good service. 

Lastly, a storm like that of April 2024, 1872 
(p. 103), which doubles back on its own course, naturally 
causes much difficulty in the issue of warnings. 

Storm Warnings. 161 

The instances cited are sufficient to show that there 
is great uncertainty in the practice of storm warnings, 
owing to the deficiency of our information as regards 
quality and quantity, and that even with all the in- 
formation we can procure we are often to a great 
extent in the dark as to the character and motion, in 
direction and rate, of the storms which approach our 
coasts, until they have wrought more or less damage. 

With all these drawbacks, however, it has been 
shown at p. 153 that we are able to maintain a 
general success of about 80 per cent, for our warn- 
ings, a result on which we may fairly congratulate 


1 62 Weather Charts and Storm Warnings. 



THE recognised value of storm warnings to the sea- 
faring community soon led to a general demand, on 
the part of the public, for predictions of the weather 
day by day. These were issued by Admiral FitzRoy 
almost at the first institution of storm warnings in 
1 86 1, but the forecast service was suspended at the 
end of 1866, and was not resumed for the space of 
more than ten years. 

Meanwhile, the Meteorological System of the United 
States was organised, and it at once commenced the 
issue of its ' Probabilities.' This service speedily 
attained great popularity, owing to the very general 
correctness of the predictions. In our own country, 
however, forecasting was not taken up again until the 
re-organisation of the Office in 1877. The country was 
partitioned into eleven districts, the same division as 
has been adopted for the Weekly Weather Report, and 
which has reference to the predominant character of 
the agriculture in the different parts of the United 
Kingdom. The districts are shown in the accom- 



panying map (fig. 56). The reason that Scotland 
North is called o, is that when the division was first 
planned it was considered that the Highlands of Scot- 
land were not strictly agricultural. 

The principles of preparation of the forecasts are on 
the whole similar to those of the preparation of storm 
warnings, but the task is evidently much more difficult ; 
and besides the duty is constant, and not merely occa- 















FIG. 56. Chart shewing Districts. 

sional. The great difficulty consists in the fact that the 
weather is not, and cannot be, the same over the whole 
of each of the districts, for the variations in the contour 
of the country ; the hills and valleys ; the lie of the land, 
whether it faces north, south, east or west ; and its 
position with regard to the sea, all exert a great 
influence on its weather. More especially is this the 
case with regard to ranges of hills. Supposing such a 

M 2 

1 64 Weather Charts and Storm Warnings. 

range to lie along a coast-line, the wind coming from 
the sea will be forced to rise over it. As it rises it is 
expanded by the reduction of pressure. In the process 
of expansion the air is chilled, and a proportion of the 
moisture in it is condensed and falls as rain. When 
the air reaches the ridge of the hills it begins to 
descend. It is compressed in the descent and warmed 
thereby, so that its tendency is not to produce rain but 
to absorb moisture. The air is accordingly damp on 
the weather side of the hill and dry on its lee side. 

To give a practical instance of what is meant, I may 
say that in the month of August 1886 I was living 
on Speyside, with the mountains of Perthshire and 
Inverness-shire to the south and west. The weather 
was uniformly dry, while at the same period the 
residents in Argyllshire were complaining of constant 

To prove this by figures is not difficult. The three 
stations selected are Laudale on Loch Sunart, Argyll- 
shire, among high hills, close to the coast ; Glasgow 
Observatory, lying well to the south of the Grampians, 
but at some distance from hills ; and Nairn, lying to 
the north-east of the hills of Inverness-shire. The 
following are the recorded rainfalls, in inches, in the 
months of August and September 1886 : 

August. September. 

Laudale . . .3.78 ... 9.50 

Glasgow . . .1.81 ... 4.96 

Nairn . . . .1.19 ... 2.47 

It is not possible to frame forecasts which shall 

Forecasting. 1 6 5 

exactly suit these different conditions existing in one 
district, but it is equally impossible for the forecaster in 
London to subdivide in his mind the districts with 
reference to all the local peculiarities of climate which 
exist in each, and to frame several hundred special 

Notwithstanding all these disadvantages, we are able 
to point to a very good proportion of success for 
our forecasts, as the figures shortly to be cited will 

The forecasts are prepared three times a day, on 
the receipt of the several telegraphic reports, for 8 A.M., 
2 P.M., and 6 P.M. respectively, as explained at p. 15. 
Each forecast is intended to cover the interval of 
twenty-four hours, and no more, from the time of its 

The 1 1 A.M. forecasts are intended for the early 
editions of the evening papers, for the clubs, and for 
public exhibition in various parts of London. 

The 3.30 P.M. forecasts are not generally published, 
but in the summer-time they are issued to a limited 
number of gentlemen to aid them in the saving of the 
hay harvest. In this case the recipients engage to 
send in regularly to the Office reports as to the 
accuracy of the forecasts. 

The 8.30 P.M. forecasts are supplied gratis to any 
newspaper which will send for them to the Office, as 
explained in the Annual Reports. 

The forecasts are also issued to subscribers at a 

1 66 Weather Charts and Storm Warnings. 

small charge, and in addition the Office is always 
ready to answer, to the best of its ability, inquiries as 
to weather, for not more than twenty-four hours in 
advance, on terms which are published in the Postal 

In order to check the accuracy of the forecasts, they 
are regularly compared with the weather reported in 
the various districts on the days to which they referred, 
and the results of this checking of the forecasts issued 
at 8.30 P.M. during the year 1885 is as follows : 






Complete Partial* 
Success. ; Success. 



Scotland, N. . 

55 3i 

9 5 


Scotland, E. . 

5o 35 

10 5 


England, N.E. 

53 34 




England, E. . 

53 3 1 

ii 5 


Midland Counties 

51 34 n 4 


England, S. 

57 3i 

9 3 


Scotland, W. . 

48 30 




England, N.W. 

48 32 




England, S.W. 

50 32 




Ireland, N. , 

46 35 




Ireland, S. 

46 33 




Summary . 

5i 32 




* Note, " partial" implies "more than half." 

The forecasts have been issued and checked regu- 
larly since the beginning of the year 1879, and the 
following are the figures of the summary for the entire 
period of seven years, under the same headings as 
those just given : 

































I88 3 











5 1 





The Office, however, does not rest content with its 
own estimate of its own work, and accordingly the 
reports of the Hay Harvest Forecasts, which are 
furnished in the manner explained above, are tabulated 
and reduced to percentages. These figures therefore 
are quite independent of any possible colouring by the 
Office. The tables correspond exactly to those just 
given, except that they come down a year later, as the 
results for the season of 1886 have already appeared. 






.Is " 

1 Partial 
1 Failure. 


of Succes: 

Scotland, N. 
Scotland, E. 

Golspie and Munlochy 
Longniddry, Aberfeldy, 
Braco, Glamis 


4 1 





England, N.E. . 

Ulceby and Chatton, North- 


S 2 


14 * 85 

England, E. 
Midland Counties 

Rothamsted and Thorpe 
Circncester and East Ret- 





ford .... 






England, S. . 

Horsham, Maidstone, and 







Scotland, W. 

Dumbarton, Stranraer, and 

Islay .... 



X 3 



England, N.W. . 

Ley bum, Knutsford, and 
Prescot .... 






England, S.W. . 

Bridgend (Glamorgan), Clif- 

ton. Glastonbury, and 

Falfield .... 






Ireland, N. . 
Ireland, S. . 

Hollymount and Moynalty 
Moneygall and Kilkenny . 






Mean for all districts . 






68 Weather Charts and Storm Warnings. 







of Success. 






7 6 






7 6 






7 6 



























3 1 

J 3 



At the risk of wearying the reader with a recapitu- 
lation of dry figures, this summary of results may be 
concluded with quotations of the figures obtained by 
independent private observers during the last year 
or two. 

In the Times of February 7, 1887, a letter appeared 
from the Rev. G. T. Ryves, of near Stoke-on-Trent, 
giving his experience of the forecasts for the Midland 
Counties in 1886. The writer says: 

'The forecasts were tested under six heads viz., 
I. Prevalence of rain, snow, or sleet. 2. State of the 
sky (whether clear, cloudy, overcast, or hazy). 3 and 
4. Direction and force of wind. 5. Temperature 
(whether warmer or colder). 6. Electrical disturbances 
(lightning and hail). The success or failure of the 
forecast as to each head referred to in it was entered 
in its appropriate column, and the result was then 
entered in one of these columns (according as the 
actual weather agreed or disagreed with the forecast), 



cither as an absolute success, an absolute failure, or a 
partial or doubtful success. 

'As no forecasts are issued on Sundays and Good 
Friday, and there were t\vo days in December when it 
was impossible to issue any, owing to the breakdown 
of telegraphic communication, the number of forecasts 
sent out by the Office during the year was 310. Of 
these, 309 were tested in the manner above described, 
with the following result : Absolute successes, 247 ; 
absolute failures, 26 ; partial or doubtful successes, 36. 
That is to say, we get, omitting small fractions, 80 per 
cent, of successes, SJ per cent, of failures, and 1 1 \ per 
cent, of doubtful cases.' 

In the same connection it may not be without 
interest to give the following figures showing the 
results of the forecasts for England, SW., as tested 
at the Observatory Rousdon, Lyme Regis, by Mr. 
C. E. Peek, F.R.A.S., for the last three years. At this 
observatory the forecasts, arriving by post, are delivered 
about noon. The last entry of the weather for the 
period covered by the forecast is made at 9 A.M., and 
these facts are compared with the forecast on its 
arrival : 


Wind and 







l g l 



I88 4 

Per cent. 


Per cent. 
6 9 

Per cent. 

Per cent. 

Per cent. 



Per cent. 

Per cent. 

1 70 Weather Charts and Storm Warnings. 

The results of our work may be compared very fairly 
with those published by the other European Offices 
which issue forecasts. It is undeniable that occasion- 


FIG. 57. 

ally complete failures of the forecasts occur, and these 
are due very much to the same causes as the failures in 
storm warnings already described. A notable instance 



of this may be cited, being that to which the Bishop of 
Carlisle drew notice in Nature for November 2, 1882. 


FIG. 58. 

The date of the occurrence was October 24 in the same 

The two charts (figs. 5 7 and 5 8) will show the com- 
pleteness of the failure of the forecast. On the evening 

172 Weather Charts and Storm Warnings. 

of the 23rd (fig. 57), there was nothing in the reports 
from any of the stations to indicate the approach of any 
disturbance ; accordingly, none was forecasted and no 
warnings were issued. 

Yet, at 8 A.M. next morning we find a well-marked 
secondary depression with its centre near Southampton 
(fig- 58), which has brought with it to the south of 
England torrents of rain, subsequently turning to snow, 
and to the French coast and the south-east of England 
a storm of wind. Fortunately the wind on our south 
coast was not at its worst until about noon, for at 
8 A.M. not one of the south coast stations reported 
a gale except Hurst Castle, and that only reported 
force 8. 

The following are the reports of the wind force 
collected afterwards by the Office : 

Jersey, 8 A.M WNW. 9*. 

Hurst Castle .... WSW. 10. 

Owers (Lightship), 10-11 A.M. . SW.byW. 12*. 

Varne (do.), noon . . . SW.byW. 10. 

Dover, 2 P.M WSW. 8. 

London . . . . . SSW. 10. 

Kentish Knock (Lightship), 3 P.M. W.byS. 10. 

Yarmouth, ii A.M. . . . SE.byE., 62 miles per hour. 

Leman & Ower (Lightship), 6 P.M. W. 10. 

Spurn Head .... WNW. 9. 

Where no hour is stated none is given in the report. 
At the two stations marked, the force of 9 was reached 
before the warnings, which were issued at 9 A.M., could 
have reached the place, so that strictly speaking the 
warnings were issued before the gale had reached 

Forecasting. 1 7 3 

a dangerous height, except in the Channel Islands 
and at the Owers Lightship, off Selsea Bill. 

Now as to the possibility of foreseeing this gale, 
a careful examination of the photographic curves at 
the observatories shows that at Falmouth, the first to 
feel the influence of the disturbance, the barometer did 
not begin to fall till 10 P.M. 

In a paper by Mr. H. Harries, 'On the Typhoon 
Origin of the Weather over the British Isles during the 
Second Half of October 1882,' which appeared in the 
Quarterly Journal of the Royal Meteorological Society, 
vol. xii. p. 10, this particular disturbance comes under 
notice, and the author says : ' I have examined the 
records of many ships over a large area west and 
south-west of the British Isles, but find nothing to 
justify our supposing that the storm had any existence 
at all until late on the night of the 23rd.' 

After quoting facts to justify this assertion, Mr. 
Harries goes on to say : ' At the earliest, therefore, no 
prediction of an approaching storm could have been 
issued from the Meteorological Office before 3 A.M. of 
the 24th, and shortly after this hour the tempest was 
raging over the Channel and the French coast.' 

This failure of forecasts attracted much attention, and 
various suggestions have been made to obviate the 
possibility of its recurrence. If we come to analyse 
these, we find they are two in number, and both of 
them have been more than once mentioned in these 
pages (pp. 15, 1 6). They are: I. More frequent 

1 74 Weather CJiarts and Storm Warnings. 

telegrams ; and II. More complete and accurate reports 
of the state of the sky. 

More frequent telegrams demand a greater ex- 
penditure of money; the telegraph offices close before 
nightfall, and to keep them open would cost a very 
large sum. In fact, it has well been said ' meteor- 
ology is money,' and at every turn we are met by the 
same difficulty of providing funds for any improve- 
ment. The second suggestion, that if a lunar halo, 
which was noticed in London at 10 P.M., had been 
reported two hours earlier, the Office might have had 
some intimation of impending change, is more practical ; 
but here, too, the fact is that the halo, if it existed at 
6 P.M., was not reported from a single station, and 
the storm therefore took the Office at unawares. 

It is needless to cite more instances of incorrectness 
of the forecasts ; they are all more or less similar in 
character to that which has just been quoted. Their 
correctness, on the contrary, is constantly being dis- 
cussed in the newspapers, and in addition a large 
number of anonymous letters and postcards relating to 
the subject arrive at the Meteorological Office. 

Many of these are very plain-spoken. One corre- 
spondent, whose handwriting is perfectly familiar to 
us, and who has been sending us postcards for years, 
says : 

'Where is your violent cyclone to burst on British 
coasts between nights of 1 4 and 1 6 ? The deadest 
of calms, of course!' This individual holds the Office 

Forecasting. 1 7 5 

responsible for the announcements of the New York 

Another gentleman is straightforward enough to sign 
his name, and he says : 

' If you cannot give better information in your 
weather forecast you had better shut up your shop. 
You have been quite wrong every day this week. Go 
away, false prophet. I shall have a petition prepared 
to abolish your Office.' 

Some of the letters are of a different tenor, and call 
for special forecasts. One arrived in June 1886, 
bearing no less than ten postmarks. It was addressed 
' Weather Office, Strand, London.' Its contents were : 

' Three next days order to be fine.' 
Another, of similar import, was addressed : 

1 Clark of the Weather, 

'No. i, Rainy Quarter.' 

The Post Office added, 'Try Meteorological Office.' 
This letter was dated and reached us by first post. 
Another was addressed : 

' Right Hon. Clerk of Weather, 
1 9 Downing Street, 

' London, W.C.' 
It ran as follows : 

' My Lord Clerk, 

1 May it please your Lordship, you will greatly 
oblige your humble servant by writing or sending me 

176 Weather Charts and Storm Warnings. 

a telegraph whether it will be fine or not on the 5th of 
November 1867. I have the honour to remain, 
' Your Lordship's most obedient servant, 


It need scarcely be said that this child received an 
answer, the letter being evidently in good faith. 

The last to be mentioned is the prettiest of all. It 
reached me at Rome at the time of the meeting of the 
International Congress in 1879, and created great 
amusement among the savants assembled there. It was 
addressed : 

' On Her Majesty's Service. 

1 To the Clerk of the Weather, 

1 Government Offices, 

'Whitehall, London. 

1 To the Cleark of the Weather [sic]. 
' Please Mr. Clerk of the Weather, tell the rain, snow, 
and hail to stop for the afternoon and rain in the 

"Good bye, my dear old Clark [sic]. 

1 Go on next page. 

1 My name is Blanche 

' Parsonage .' 

Miss Blanche was duly informed that she was not 
the only person who cherished such wishes, but that 
her desires could not be fulfilled. 

The number of weather prophets that write to us is 
very large indeed. Here is a verbatim copy of one letter : 

Forecasting. 177 

' April 28, 1879. Colliery. 

'Notice Sir this to warn all shipowners and 
captans and all seefarenmen that there wil be see 
storm, on the 1 6 or 1 8 of May from North East with 
rain lighting to get there ships trimed squalls of 
wind.' This was duly signed. 

The weather was perfectly calm between the I5th 
and iQth of May 1879. 

One gentleman issued on postcards monthly forecasts 
for more than twelve months in 1882 and 1883. To 
each forecast was appended the following note : 

'These forecasts being founded on a universal law, 
allowance must be made for local differences and 
peculiarities. They are applicable to all places in the 
Northern Temperate Zone.' 

The reader will have gathered from the foregoing 
pages that the weather is seldom quite the same on 
any one day all over the limited area of the British Isles. 
How then can we expect a forecast to be generally 
applicable when the North Temperate Zone embraces 
climates so different from ours as those of St. Peters- 
burg and of Salt Lake City ! 

One thing is at all events admitted by all who have 
experience of forecasting, and that is that as yet no 
scientific principles have been established which enable 
us in Western Europe to forecast the weather for more 
than a few hours. 


178 Weather Charts and Storm Warnings. 



IT will be seen, from what has been said in the fore- 
going pages, that there is room for improvement in the 
entire service, but improvement requires money, and 
money is limited. There is no lack of criticisms, 
friendly or the contrary, of our work, or of sugges- 
tions of new sources of information. 

In some quarters the Meteorological Office is often 
expected to be far more venturesome in its announce- 
ments of weather than it is at all prepared to be. 
Once a gentleman called on us with the request that we 
would prepare for a newspaper, which he had recently 
started, a forecast of the probable weather for a week 
in advance. On our declining the task, he remarked 
that Admiral FitzRoy had issued forecasts for three 
days in advance, and that we ought to be able to do so 
for double that period. He quietly ignored the fact 
that Admiral FitzRoy's three-day forecasts were not 
successful enough for them to be continued, and that 
accordingly they had not been kept up. 

Other men call on us to be as precise in our 

Suggestions for Improvement. i / 9 

announcements of * probabilities ' as the Chief Signal 
Office at Washington, in simple oblivion of the circum- 
stance that while that Office has an immense continent 
to deal with, and is itself situated on the eastern side of 
its area of observation, so that the gradual march of 
weather changes over the western stations can be 
observed and reported long before they reach the more 
thickly populated portions of the States, we have to 
deal with an extended oceanic coast, and can by no 
possibility have a series of stations on our western side, 
so that, unless by the appearance of the clouds and sky, 
we have no chance of learning what is going on out of 
sight of land. 

Our reporting system is shown on the frontispiece, 
and a glance at it will prove how badly we are off for 
stations on the west coast of Scotland. This is a most 
important district for gaining early intelligence of 
weather changes, but in the first instance telegraphic 
communication is only very scantily developed in such 
wild regions, and in the second place the habitable spots 
are almost all in sheltered bays, where the true force of 
the wind is hardly ever felt, owing to the fact that few 
ordinary houses could bear the brunt of a winter 
Atlantic gale, blowing, as it sometimes does, with a 
velocity of over 80 miles in a single hour, and reaching 
the speed of at least 100 or 120 in gusts. The 
observers therefore cannot possibly send up perfectly 
true reports of the direction and force of wind, as will 
be at once admitted when the position of the stations in 

N 2 

i So Weather Charts and Storm Warnings. 

western Ireland and Scotland is considered. Valencia 
lies on the shore of a narrow sound, with high hills 
about it. Belmullet is at the head of Blacksod Bay, 
and is therefore quite sheltered from the sea, though its 
wind reports are good enough, as the land is low in 
the neighbourhood. Malin Head is well exposed. 
Ardrossan is a valuable station, but it is far up the Firth 
of Clyde, and at least fifty miles, as the crow flies, from 
the extreme western point of Lslay. We have no other 
station till we reach Stornoway in the Island of 

The most obvious idea is to obtain information from 
the westward, either from moored signal ships, from 
islands such as the Azores, Faroes, Iceland, and New- 
foundland, or lastly from the mainland of America. 

Let us take these different suggestions in order. 

The idea of the employment of ships moored at a dis- 
tance from the coast, and connected by a telegraphic cable 
with the shore, has already been noticed in Chapter I. 
This plan has over and over again been suggested as 
a most valuable mode of obtaining information, not 
only of the position of vessels in distress or wind-bound, 
but of the meteorological conditions prevailing out at 
sea. Some scientific men, as the late Mr. Morse, have 
even gone so far as to propose that these signal ships 
need not even be manned, but might be simple buoys 
provided with electrically self-registering apparatus, 
which might record its indications automatically at the 
nearest station on land ! 

Suggestions for Improvement. 1 8 1 

The plain state of the case about any such visionary 
projects is that, even if we were to overcome the diffi- 
culty of mooring a ship in 1000 fathom water, we 
should find it very difficult to maintain the telegraphic 
cable in working condition. Some years ago the 
Admiralty lent H.M.S. Brisk to be moored as a trial 
signal ship at the entrance of the Channel. The experi- 
ment was given up in six weeks, mainly on the ground 
of the difficulty of maintaining the telegraphic connec- 
tion in a perfect state, and it was stated that the venture 
cost the promoters about as many thousand pounds as 
the Brisk was days at sea. 

In the year 1886, however, the Telegraph Construc- 
tion and Maintenance Co. succeeded in keeping up 
connection with the Sunk Lightship, off Harwich, for a 
year, so that the possibility of working such a scheme 
has been proved. 

As regards reports from islands situated in mid- 
Atlantic, there is but little prospect of our ever making 
much progress in that direction. It has often been 
suggested that a cable should be laid from Portugal to 
the Azores, and that this country should bear a consider- 
able share in the cost of daily telegrams frcm those 
islands, as it was assumed that, as the Azores are situated 
in the Atlantic, they must be, so to speak, pickets, to 
give us early news of danger approaching from that 

All this is very plausible, but when we come to test 
the simultaneous reports taken daily in those islands 

1 82 Weather Charts and Storm Warnings. 

and at Valencia, as has been done at the Meteorological 
Office for two years and a half (Quarterly Journal of 
the Meteorological Society, vol. i. p. 183), we find 
that there is very little connection between the pheno- 
mena occurring at the two stations, so that we could get 
no direct warnings of approaching storms from the 
Azores, though of course a daily knowledge of the 
conditions of weather over the part of the Atlantic 
where those islands are situated could not fail to be of 
value to us, as is evident from what has been said in 
Chapter VI. In fact, the Azores lie to the southward 
of the usual track of Atlantic storms, and it would 
apparently be a far better investment of money for this 
country to spend it in obtaining reports from the Faroes, 
if ever a cable were laid to those islands, than in paying 
for telegrams from the Azores, as many of the most 
destructive storms on our north-east coast come down 
on us from the northern part of the Atlantic Ocean and 
pass over the Faroe Islands. The value of St. Kilda 
as an outpost could hardly be overrated, but there is no 
present hope of telegraphic connection with that island, 
and in all cases it must be remembered that unless a 
cable will pay commercially it is hopeless to look for 
the establishment of telegraphic communication for 
scientific purposes. The telegraphic connection of the 
Azores with Europe has been talked about for a quarter 
of a century, and repeatedly it has been announced that 
a concession for laying a cable has been granted, but 
up to the present date (April 1887) capitalists have not 

Suggestions for Improvement. 1 8 3 

been induced to put down the necessary sum. To give 
some idea of the expense which would have to be met 
if the cable were laid and maintained for scientific 
purposes alone, it is sufficient to say that such a scheme, 
for communication with the Faroes and Iceland, was 
proposed to the International Meteorological Committee 
at their meeting at Copenhagen in 1882, and that in 
that scheme the annual contribution of this country, 
ranged naturally in the first class for maritime import- 
ance, was 3285 for three daily reports, being more 
than we pay at present for all our home and European 
telegrams put together ! The idea of maintaining cables 
for meteorological service alone may therefore be dis- 
missed as visionary. 

As regards reports from the other side of the 
Atlantic, in the year 1868 a direct service of reports 
from Heart's Content, Newfoundland, was organised. 
These messages were transmitted gratis, thanks to the 
great liberality of the Anglo-American Telegraph Co. 
It was found, however, that we could not turn them to 
practical account, mainly owing to the fact that the 
situation of the station of Heart's Content was in a 
sheltered nook, where no storm could hurt the shore 
end of the cable, and accordingly the wind reported by 
the observers bore little relation to that blowing at sea 
outside. The service was therefore discontinued in the 
year 1871. 

Speaking in general terms, we may say, as regards 
storm warnings from America, that though storms do 

1 84 Weather Charts and Storm Warnings. 

sometimes cross the Atlantic from shore to shore, they 
change their character en route, some increasing and 
others dying out, so that Ave cannot predict which 
storm, out of several starting from the States, will 
reach us. Professer Loomis, a very high authority on 
American weather, very strongly supports the view 
just stated in the following words, taken from the 
American Journal of Science and Art for January 

* When storms from the American continent enter 
upon the Atlantic Ocean they generally undergo im- 
portant changes in a few days, and are frequently 
merged in other storms which appear to originate over 
the ocean, so that we can seldom identify a storm in its 
course entirely across the Atlantic.' 

To give an instance of a storm failing to advance 
over any great distance, I shall take one which is very 
striking. On November 30, 1874, the steamer La 
Plata, with a telegraph cable on board, foundered in a 
heavy gale near the Channel Islands. This storm 
entirely died out and disappeared before it crossed the 
North Sea. 

These statements show how rash is the opinion 
which is held by Professor Daniel Draper, of New 
York, and announced in the Reports of the Central 
Park Observatory for 1872 and 1873, to the effect that 
out of eighty-six storms which he had investigated, as 
having started from New York, only three had failed to 
show themselves either at Valencia or Falmouth. 

Suggestions for Improvement. 1 8 5 

We now come to the practical results of Atlantic 
Telegraphy, as it has been conducted of late years. In 
February 1877, owing to the great public spirit of 
Mr. J. G. Bennett, the Weather Bureau of the New 
York Herald, under the late Mr. J. J. Collins, com- 
menced the transmission by telegraph of Weather 
Warnings to Europe. These messages have been 
continued ever since ; they appear in the public press, 
and so have attained an extensive circulation. 

When these warnings were first issued, their general 
purport was to the effect that a disturbance was on its 
way across the Atlantic, and might be expected to 
cause gales on the western coasts of Europe at a given 
date. The following is a specimen of these messages 
it arrived July 4, 1877 : 

' Storm centre will arrive British, French coasts, 
probably 5th or 6th ; NE. to SW. gales, rains and 
lightning. Warn outward-bound ships.' 

The practical utility of these warnings has been 
carefully tested in our own Meteorological Office, for 
the first year of their continuance, and of late by 
Professor Cleveland Abbe, of the Chief Signal Office, 
Washington : the results in both cases are much the 
same. About one-quarter of the warnings are fol- 
lowed by gales of any consequence on the British 

The actual percentages which my discussion, published 
in the Nautical Magazine for March 1878, yielded, were 
as follows : 

1 86 Weather Charts and Storm Warnings. 

Absolute success . . 17.5 

Partial ,, . . 25.0 

Very slight ,, . .15.0 

Absolute failure . . 42.5 


The paper goes on to say : ' These figures show that 
not 45 per cent, of the warnings (in 1877) can be 
considered to be really successful. What is meant by 
" really successful " is that the information conveyed by 
them was of real value to seamen in British ports.' 

Professor Cleveland Abbe, of the Chief Signal Office, 
Washington, states in a letter addressed to myself: 

' A full report on the verifications of the New York 
Herald predictions was submitted by me to the late 
General Myer, in January 1878 : this covered all cases of 
published predictions up to November 29, 1877, and in 
it I allowed the Herald 42 per cent, in all i.e., about 
25 per cent, for well-verified predictions and 17 per 
cent, for partial successes, and for the remaining 50 per 
cent, [sic] of failures. This was for Great Britain, but for 
France and Spain the percentages were decidedly less. 
These were of course apparent verifications, and took no 
account of the fact that sometimes storms arrived in 
Great Britain and served to verify a prediction, whereas 
they really were different and distinct from the storms 
to which the predictions referred.' 

It will be seen that Professor Abbe's investigation 
refers to a portion of the same period to which my in- 
quiry referred. It is very satisfactory to see that the 
two discussions, totally independent of each other, con- 
firm each other so completely. 

Suggestions for Improvement. 187 

Of late years the Herald warnings have been gene- 
rally couched somewhat in the following words : 
4 A disturbance is central near Newfoundland, and may 
be expected to disturb the weather of the British Isles, 
&c., at such and such a date.' 

These telegrams have been much more successful 
than the positive predictions of gales, as the phrase 
' disturbance of weather ' allows considerable latitude, 
and it is a fact that if a storm even approaches our 
coasts, it cannot fail to disturb our weather to a certain 
extent, at least in Ireland or Scotland. 

To give an idea of the extent to which storms 
striking our coasts could really be predicted by means 
of reports from America, the late Captain HofTmeyer, 
the head of the Danish Meteorological Service, who 
himself had drawn and published daily weather maps 
of the Atlantic for the space of three and a quarter 
years, investigated the course of Atlantic storms in a 
pamphlet which he published in 1880. He discussed 
285 depressions in all, covering an interval of twenty- 
one months, and found the following percentage propor- 
tions of the storms prevailing over the ocean source of 
depression : 

Source of Depression. S^-W. 

Arctic Regions .... 8 
North America . . . -44 
Probably Tropical ... 9 
Secondary, arising at sea . . 37 
Untraceable 2 

It will be seen from this that not one-half of the 

1 8 8 Weather Charts and Storm Warnings. 

storms which existed in 40 W. came from the United 
States. Now of the 285 storms only 145, or,* in round 
numbers, one-half, passed the meridian of 10 W., and 
these were thus divided in percentages: 

Source of Depression. . 

Arctic Regions . . . .12 
North America . . . 47 
Probably Tropical ... 5 
Secondary, formed at sea . . 33 
Untraceable .... 3 


We see therefore that of the storms which reach us, 
not quite one-half could possibly have been predicted 
by telegrams from America. The figures given above 
show that the actual successes attained by prediction 
were much less than this. 

For the last few years the Meteorological Office, in 
conjunction with the Bureau Central Meteorologique in 
Paris, has been in regular receipt of telegrams for- 
warded by the Chief Signal Office, Washington, and 
conveying intelligence of storms met with by steamers 
arriving at American ports. It was considered that if 
a steamer encountered a storm to the westward of the 
meridian of 45 W., she could reach an American port 
before the storm could reach Europe. Accordingly, 
timely intelligence from her log might be transmitted 
by telegraph to Europe. It has not, however, been 
found practicable as yet to issue warnings based solely 
on these telegrams, but their receipt puts us on our 

Suggestions for Improvement. 1 8 9 

We learn therefore from all this, that we in the 
United Kingdom have to rely on our resources and our 
own stations, for gaining information for storm-warning 
purposes, and so we must only do our best to improve 
our own arrangements. 

It is evident that, if we had an extensive continent 
like North America to deal with, we too, in the Meteoro- 
logical Office, could probably make good our claim to 
a proportion of successes for our warnings, rivalling that 
with which the Chief Signal Office is justly credited. 
That this is not an idle boast is shown by the following 
figures taken from the Report of the Meteorological 
Office (the Norddeutsche Seewarte) at Hamburg for 
1874. A system of warning for Hamburg from our 
Office had then been in operation since 1867, and the 
general results were, that 301 warning messages were 
issued from London in the course of the seven years. 
72 per cent, of these warnings were followed by gales, 
while in only three cases did the storm outrun the mes- 
sage. Naturally in these figures no account is taken of 
the gales missed owing to the Sunday interruption of 
our service which prevailed up to 1877. 

The reason of this success in issuing to Hamburg 
timely warning of all storms, a success which is far 
more brilliant than we can ever show for our own 
coasts, is that our network of stations, as will be seen 
from the frontispiece, entirely surrounds the North Sea, 
and that no storm can reach Hamburg without passing 
over our outposts, unless it comes from the Continental 

Weather Charts and Storm Warnings. 

side, a very unusual direction for its advance. As a 
proof of this statement I may say that in the year 1869 
twenty-three storms were felt in Hamburg, and twenty- 
two of these had previously passed over some part of 
the United Kingdom. 

Finally, we come to the subject of high-level stations, 
such as that on Ben Nevis. A considerable number of 
such stations has been organised. Among these we 
may mention, in the United States, Pike's Peak and 
Mount Washington ; in Portugal, Sierra da Estrella ; 
in France, the Pic du Midi, Puy de Dome, and Mont 
Ventoux ; in Switzerland, the Santis ; in Italy, Mont 
Cimone arrd others ; in Austria, the Sonnblick and 
Hoch Obir ; in Germany, the Brocken and Schnee- 
koppe ; and there are several others. 

The discussion of the results they yield has led to 
most interesting conclusions in relation to the diurnal 
march of temperature, barometrical pressure, and the 
force of wind at different levels, and more especially as 
to the changes in the dampness and the temperature of 
the air produced by its passage over mountain chains. 

The popular idea of the value of these stations 
is, however, that they are mainly of importance as 
weather-reporting points. On this head I recently 
inquired in Europe and the United States, of the chiefs 
of the meteorological services which possess such 
stations, as to what practical benefit their forecasting 
service derived from these reports. 

The outcome is very scanty and quite incommen- 

Suggestions for Improvement. 1 9 1 

suratc with the cost and labour of maintaining the 
stations, supposing tJicsc to be kept up solely for forecasting 

The late General Hazen replied that no telegraphic 
reports were received from Pike's Peak, but that the 
information from Mount Washington was of consider- 
able value in weather prediction for the adjacent New 
England States. 

From Europe the replies were less decided. Professor 
Mascart said that the reports from the Pic du Midi were 
often of value in giving early warning of floods in the 
low country, owing to the melting of snow. He also 
stated that the reports from that station and from the 
Puy de Dome indicated shifts of wind before any signs 
of their approach were noticeable at the lower level. 

The other directors gave no positive reply. On one 
point all authorities are in accord, that the barometrical 
readings cannot well be used in forecasting. This 
arises mainly from difficulties connected with the reduc- 
tion of the readings to sea-level. It is very interesting 
to learn that the barometrical oscillations at the summit 
of Ben Nevis are particularly sudden and extensive, 
but no indication has yet been given of the mode of 
using such information in forecasting. 

It must always be remembered that the preparation 
of forecasts and storm warnings is necessarily a rapid 
proceeding, and it must be carried out with intelligence, 
often defective, especially in stormy weather, when 
completeness of information is most desirable. The 

192 Weather Charts and Storm Warnings. 

calculations therewith connected must above all be of 
a simple nature, and such as can be carried out by 
assistants of ordinary education. 

In conclusion, I can only say that the foregoing 
pages contain what is, necessarily, a very condensed 
and brief abstract of the present state of our knowledge 
of weather, in so far as this knowledge depends on the 
system of Weather Telegraphy, a branch of investiga- 
tion which is confessedly of great intricacy. 

To do the subject justice, double the space, and at 
least treble the number of charts and illustrations, would 
have been required. If, however, the contents of these 
pages shall have awakened in any of my readers a real 
conception of what the weather work of the Meteoro- 
logical Office is, and an interest in the careful study of 
weather, one of the most enticing of all inquiries not 
to mention its practical importance to every one the 
trouble of putting together these notes will have been 
far more than repaid. 

The Weekly Weather Report. 193 



IN addition to the Daily Weather Reports, which have 
been the theme of the preceding pages, the Office has 
for the last nine years issued weekly returns contain- 
ing information intended to be used in agricultural 
and sanitary investigations. The subjects on which 
data are given in this publication are Temperature, 
Rainfall, and the Duration of Sunshine. 

In the year 1884 the character of this Report was 
much improved, and its size was considerably enlarged, 
and a specimen of the summary of results which now 
appears weekly will be found on pp. 194, 195. 

This table appears in a somewhat abridged form in 
the London newspapers on Thursday mornings, and in 
several agricultural papers at the end of the week. 

The nature of the information as to Rainfall and 
Sunshine is easily intelligible : it consists of statements 
of the amount of each recorded for the week, and of the 
quantity which has accumulated since the beginning of 
the year. In each case the difference from the average 
is given. 


194 Weather Charts and Storm Warnings. 


ire3j\[ uiojj 



ro H 5 vo M" CO ^ 



Q -|5 l&ll a 

00 O CO O PI O 

O co co * M tv. 


1 1 


Q"+ + + i i i 

1 1 1 + 1 1 



||* !&??! 

CN M ft CO * ft 



S^ b/J M ^ ro ^- ro (N 

1 1 + + \ 1 






*^Mro rovo^O OiO 

Q-o M M 

ro ro t^ <*- -t- O 






T T TT f T 



V>- w 

|| S; q. 5 ?^ 

O w ON fO CO N 




> ii ^ ^ > 

^o _o rt ^o ^o JD 

.2 r2 r2 -2 ^ 

M ro TJ- o ro Pi 




O M 00 Tj- M 00 

5 i ^. 5: =: 

o co M o >o q 
10 10 vd vd ^d oo 
* * >* <! * * 



ro N O O oo N 

i i ^ i i ? 



o oo' M ^- t^ vd vd 
-4- 10 10 >o in ir> 

>0 ft 10 10 IO 



o a R a a a a 

R & -8 'S % 


o vg ^^ 8. 

CO t-x xo it- * O 
10 10 \O \O VO IO 


H , ' S s 

ta H . -, H r 
o 2 W " ^ S 

^ W > c w 
3 S a" a" Q a" 

1 K ^ s 1 

^>:il - 

g g P a 


1 8 8^.8 

w oi W W g H 

O w PI ro 4- o 

X 2 2 

c^ W H ^ 
vd t> oo ON o | 




77/6- Weekly Weather Report. 





i + + + + + 

+ + + + + + 










ajqissoj jo 

vo N ro t>. M M 
N ro n-j ri ro fO 

M M M H O f^ 

rr m ^ ro * TJ- 




unsj^ uiojj 


O\ to o f-. O 
N fO M t^ 10 vo 

VO N ro *> O oo 

O O rn <* w N 







N >O O O O\ VO 

N ro ro * ro fO 

r> \O ro M 1**. VO 

^ % S> 5= 



s .; 




aiqi'ssoj jo 


N M o n i^ O 
M N ro rr <)- 

^ 8* $ *8 % S 



N N O N h~ OO 

tx Ov <* vo OO * 


ucaj^ uiojj 


S <* N N fl ro ro 
- + 1 1 1 1 1 

O O CO Ov <> 
c^. CO (A. M ro -4- 


x O ** f> >0 O 

CO I-* N vo O> vo 



II B J I B 1X 

^ ro vd * 4- ^ 4 

1-1 vo vo OO OO 10 





1O t>- VO * N 

1 1 1 1 1 1 

Ov 2" t-. vo Ov Ov 
1 1 1 1 1 1 






jo jaquiri^ 

^ 1$ 1 5- 5- 

^ $ $ S 5 ? 



n 33 A\. H 


aqj u^qj 

SS3-7 JO 3JOJ^ 

s 2 a a 

1 J jg 

< M H M M 

H Si 2 JJ JJ H 

o o o o o o 


O w OO 10 * 



? .0 0^ * ,r ro >0 


'II B J I B 1X 

l M 

g ^ - J? M * 

S^BQ XuiKy; 

jo aaqum^ 



S g a" a" a a" 
5 5 z 2 z z 

1 < < < < 

r r j j j j 

o o g g S g 

rj t > 2 SS .~ Z 

CA dx M m X H 

ENGLAND, S.W. . . 


C 11 N CO -4- 10 

vo t- 00 0% | 






1 96 Weather Charts and Storm Warnings. 

The statistics as to Accumulated Temperature, 
however, are not so easily intelligible, as at the head of 
the column we find the entry 'day-degrees/ which 
calls for some explanation. 

At one of the Conferences in connection with the 
Health Exhibition in 1884, I read a paper on Cumu- 
lative Temperature, which is here reproduced, as giving 
an explanation which it is hoped will be sufficient for 
the reader. 

'As to the growth of crops, Sir J. B. Lawes, Bart, 
F.R.S., and Dr. Gilbert, F.R.S., have for many years 
conducted with scrupulous accuracy extensive investiga- 
tions into this subject at the famous experimental farm 
of Rothamsted. Let us therefore hear what they have 
to say : 

' " On this point we may remark that we have found 
the really most luxuriant and heavy crops, both of hay 
and of wheat, to have been very materially influenced 
by the characters of the winter and early spring periods 
quantity in both cases depending very greatly on the 
early development of the plant. With favourable 
conditions in this respect some of our very heaviest 
crops, both of hay and wheat, have been obtained under 
by no means specially favourable meteorological 
conditions during the period of most active above- 
ground growth. On the other hand, high proportion of 
corn to straw, which sometimes gives really high yield 
per acre, depends more upon the conditions of the 
periods of active above-ground growth and maturation. 

The Weekly Weather Report. 197 

In the same way, great weight of hay sometimes 
depends upon the conditions of maturation rather than 
of luxuriance. 

4 " It is obvious, therefore, that in discussing the 
relations of meteorological condition and agricultural 
production, it is essential to be able to arrange the 
records for any selected periods of the year." 

* And again : 

( " It is obvious that different seasons will differ 
almost infinitely at each succeeding period of their 
advance, and that, with each variation, the character of 
development of the plant will also vary, tending to 
luxuriance, or to maturation, that is, to quantity, or to 
quality, as the case may be. Hence, only a very 
detailed consideration of climatic statistics taken to- 
gether with careful periodic observations in the field, 
can afford a really clear perception of the connection 
between the ever-fluctuating characters of season and the 
equally fluctuating characters of growth and produce. 
It is, in fact, the distribution of the various elements 
making up the season, their mutual adaptations, and 
their adaptation to the stage of growth of the plant, 
which throughout influence the tendency to produce 
quantity or quality. It not (infrequently happens, too, 
that some passing conditions, not indicated by a 
summary of the meteorological register, may affect the 
crop very strikingly, and thus the cause will be over- 
looked, unless careful observations be also made and 
the stage of progress, and tendencies of growth, of 

198 Weather Charts and Storm Warnings. 

the crop itself at the time, be likewise taken into 

1 Again : 

' " Those characters of season which are very un- 
favourable for land in poor condition, may be favourable 
to land in high condition, and vice versa" 

1 As regards the three elements represented on the 
diagrams, 1 it is evident that the curves for rainfall and 
sunshine require but little explanation, as the successive 
steps of the curves show the successive weekly totals of 
rain or sunshine. 

' The case as regards temperature is, however, 
different, and those who examine the diagrams will see 
that there are two curves, one ascending, the other 
descending. A certain arbitrary base line is assumed, 
and the values are measured above or below that line. 

' It is proved, almost beyond a doubt, that each plant, 
say each individual cereal, requires a definite amount 
of heat to bring it to maturity. Thus, maize requires 
more than wheat, and wheat again more than barley or 

1 We all know that maize is not an English crop at 
all, and that wheat is not grown to much profit in 
Ireland or the west of Scotland, being replaced by oats 
or barley. The ordinary reason assigned for these 
differences as to crops is that in each case the 
temperature is not high enough for the particular crcp 

1 Diagrams were exhibited in the Meteorological Annexe. These are 
not reproduced here. 

The U'cckly Weather Report. 199 

which has failed. If the inquirer pursues the subject 
further he generally finds out that his informant has no 
very precise idea as to how the temperature acts. 

' Now various investigators, and notably Boussingault 
and Professor Alphonse de Candolle, of Geneva, have 
devoted much attention to this subject, and the latter 
writer, in his " Geographic Botanique," has come to the 
conclusion that a certain total amount of temperature 
above a definite base line is necessary for plant growth, 
and that this amount, or, as he calls it, this " sum of 
temperature," varies for each crop. 

' He finds that plants, as a rule, do not begin to give 
indications of active vegetation until the temperature 
rose above 6 Cent. This temperature, 6 Cent., or in 
round numbers, 42 Fahr., that is ten degrees above 
the freezing-point, is taken as the base for all the 

' Although Professor de Candolle propounded his 
views some years ago, yet as recently as the year 1874, 
at the Agricultural Conference at Vienna, meteorolo- 
gists were quite at sea as to how these sums of tem- 
perature were to be calculated. 

' The credit of solving this problem belongs to 
Lieut.-General Richard Strachey, the Chairman of the 
Meteorological Council. He proposes to adopt a 
certain unit of temperature to supply a standard for 
calculation, the unit being one degree continued for 
the unit of time, either one hour or one day, as the 
case may be. Such a unit may be conveniently called 

2oo Weather Charts and Storm Warnings. 

an hour-degree, or a day-degree. The unit of time 
adopted for the calculations to which I am about to 
refer is a day, and the unit of what may be termed the 
effective temperature is therefore a day-degree. A day- 
degree therefore signifies one degree Fahr. of excess or 
defect of temperature above or below 42 Fahr., con- 
tinued for twenty-four hours, or any other number of 
degrees for an inversely proportional number of hours. 

' Now, the first idea I want you to take in about these 
day-degrees is that, when we speak generally of the 
mean or average temperature for a day, or month, or 
year, we imply that the resulting temperature is the 
same as would be observed if the thermometer indicated 
this mean temperature throughout the entire period for 
which the mean is taken. Thus, if we were dealing 
with daily means, an average daily temperature of 62 
Fahr., which is an ordinary temperature for a warm 
summer's day, would indicate twenty day-degrees of 
temperature for that day, starting from the assumed 
base line of 42 Fahr., which has already been 

'The first step therefore towards determining this 
effective temperature in day-degrees resolves itself into 
determining as speedily and simply as possible the 
average temperature for the period under consideration. 

'We have, fortunately, to our hands, a very simple 
mode of arriving at the mean temperature with accuracy 
sufficient for our purposes. Almost all observers record 
the maximum and minimum temperatures once in the 

The Weekly Weather Report. 201 

twenty-four hours. It is found that the half sum of 
these readings, the mean between them, is nearly, but 
not exactly, the average for the day. It must, of 
course, be understood that the instruments must be 
read regularly and at the same hour every day. 

* The next points which require attention are : 
whether the maximum and minimum are both above 
42, which occurs in summer, or both below that point, 
which occurs in winter ; or, finally, whether one is 
above, and the other below. 

1 In the first case all the accumulated temperature is 
to the good ; it is all on the positive side. In the 
second case it is all on the negative side. The third 
case is the only one which presents difficulty, for when 
the extreme temperatures are on either side of the line 
of 42, one portion of the effective temperature for the 
day is positive and the other negative. 

1 Now, General Strachey carried out a long series of 
calculations, based on the observed hourly temperatures 
at Kew Observatory, and at other stations in the 
United Kingdom, in order to ascertain the magnitude of 
the co-efficient by which the difference between either 
of these extreme temperatures and the base temperature 
(42 Fahr.) should be multiplied in order to obtain the 
values of the temperatures in excess or defect of 42 
Fahr. expressed in day-degrees, and he found that this, 
for a weekly period, was o.4. 

' Accordingly we get the following rules : 

' If the mean of the day is above 42 Fahr., we 

202 Weather Charts and Storm Warnings. 

multiply the difference between the minimum and 42 
by o.4 (four- tenths), and call this the negative effective 

' To find the positive effective temperature we subtract 
from the difference between the mean for the day 
and 42, the negative effective temperature just deter- 

' If the mean of the day is below 42 Fahr. the 
proceeding is similar ; but we first ascertain the positive 
effective temperature, and subtract that from the differ- 
ence between 42 Fahr. and the mean, thus obtaining 
the negative effective temperature. 

4 The method of determining the effective tempera- 
ture, which may briefly be called the accumulated 
temperature, is fully explained in a paper by General 
Strachey, which was published in the volume of the 
Quarterly Weather Report for 1878. 

' The practical application of the data thus obtained 
as standards of comparison for the growth and ripening 
of various agricultural products must, of course, be left 
to the agriculturists, and it will be interesting to learn 
how far a correspondence between the character of the 
several crops and the accumulated temperature of the 
year can be established. 

* The measure of temperature afforded by this system 
of computation appears to be as well suited to supply 
a standard of comparison of climates for hygienic 
purposes as for agriculture, and the diagrams indicate 
in a forcible manner the characteristic differences of 

T/LC Weekly Weather Report* 203 

climate, in respect of temperature, of the portions of our 
islands to which they refer.' 






Mia land Cauntiej 


Channel fa 
/relanJ S. 

V/laJ Count,, 

FIG. 59. 

Diagram showing the curves of Accumulated Temperature, above and 
below 42, for the period 1878-1885, in various districts. 

204 Weather Charts and Storm Warnings. 

By way of illustration of the foregoing, the curves of 
accumulated temperature for four of the districts have 
been reproduced in fig. 59 (p. 203). The districts selected 
have been I. The Channel Islands ; II. Ireland South ; 
III. The Midland Counties ; and IV. Scotland North. 
The period is the interval of eight years, 1878-85. It 
will be seen that the total amount of accumulated heat 
above 42 Fahr. is highest, 3547, in the Channel 
Islands ; Ireland South comes next, with 3064 ; then 
follow the Midland Counties, with 2931 ; and last of all 
comes Scotland North, with 1890, only one-half the 
amount which falls to the share of the Channel Islands. 

If we now look at the lower part of the diagram, 
Showing the amount of accumulated heat below 42 
Fahr., it will be seen what advantages the Channel 
Islands and Ireland South derive, mainly from their 
exposure to the sea and its warm currents. On the 
other hand, the Midland Counties and Scotland North 
run closely together in the amount of what may be 
called negative accumulated heat, the inland position of 
the former having the same effect on its winter climate 
as the higher latitude of the Scotch Highlands. 

All of these curves cf accumulated temperature 
below 42 run parallel to the base during the summer, 
inasmuch as at that season the temperature does not 
fall below 42 ; but while in the Channel Islands the 
point at which this condition is reached is 93, it is 
239 in Ireland South, 504 in the Midland Counties, 
and 510 in Scotland North. 

The Weekly Weather Report. 


It will also be noticed how very closely the curve for 
Ireland South follows that for the Channel Islands at 
the beginning of the year, while during the autumn 
it practically coincides with that for the Midland 

As regards the use which may be made of these 
data, a paper on the Connection between the Accu- 
mulated Temperature and Agricultural Production was 
read by Professor J. H. Gilbert, F.R.S., before the 
Societe Helvetique des Sciences Naturelles, and pub- 
lished in the ' Bibliotheque Universelle/ vol. xvi. ($rd 
Period), and of this some extracts may be interesting. 

Dr. Gilbert gives the following table : 

Fixed Dates, or from Commencement of Active Above-ground Growth , 
to Date of Harvest at Rothamsted. 


Jan. i. 

Feb. i. 

March i. 

April i. 

From date of 
Excess nearly 
Every Week. 

From Date of 
Weekly Excess 
after a Break. 






































I88 3 




























Dr. Gilbert says : * The results show that for wheat, 
from the Meteorological Office figures, the total number 
of aggregate degrees of heat is, for the last two columns. 
1900 and 1836 respectively. 

206 Weather Charts and Storm Warnings. 

' Hervc Mangon's results, for north-west France, 
give 1854, which is not far from the Rothamsted 

' Risler's figures, for Nyon in Switzerland, correspond 
to 2192. 

' Boussingault and Herve Mangon have both investi- 
gated the question for barley, and their results give 
respectively, for spring-sown barley 2005 an ^ 1989, 
nearly approaching the Rothamsted values for wheat. 

' Boussingault's results for maize correspond to 
35 1 5.' 

The foregoing extract is sufficient to show one 
direction in which the figures representing accumulated 
temperature may be turned to account, but the number 
of years for which this information is available is as yet 
too small to allow of many important deductions being 
drawn from them. 

On Cirrus Cloud Observations. 207 



THE following Instructions, devised by the Rev. W. 
Clement Ley, were adopted by the Meteorological 
Council in 1885, and have been issued to such of the 
observers connected with the Office as are willing to 
undertake the reporting of cirrus movements : 

The attention of the observer is especially directed 
to clouds of the 'cirrus' type, including what are heie 
defined as true cirrus, sheet-cirrus, high cirro-cumulus, 
and cirrus-haze. They are the highest kinds of cloud, 
and it is to these types only that the following Instruc- 
tions refer. 

True Cirrus. When in moderate quantities it is 
commonly white, though when seen through haze it is 
usually somewhat cream-coloured, and when the sun's 
rays have reached it through a long stratum of hazy 
atmosphere, is often of either an orange or rosy tint. 
In all cases it has a very delicate appearance. It is 
sometimes arranged like bunches of fine hair ; and such 
tufts of cirrus are often called ' mare's tails.' At other 
times it resembles small curled feathers. Quite as 

208 Weather Charts and Storm Warnings. 

commonly, however, it lies in thin light strands, like 
pale gossamer threads. 

Sheet- Cirrus. When cirrus overspreads as a con- 
tinuous sheet a large portion of the sky it becomes 
what is here termed 'sheet-cirrus! In this state, 
when not too thick, it produces 'halos' or large rings 
round the sun and moon. The sheet-cirrus sometimes 
appears fibrous, sometimes reticulated. When the veil 
of this cloud becomes thick it assumes some neutral or 
muddy tint, except in those cases when it is so disposed 
that the rays of the rising or setting sun are reflected to 
us from its under surface, which then appears of either 
an orange or rosy colour. 

1 High cirro-cumulus' differs from simple cirrus in 
consisting of small detached masses somewhat rounded 
in form, a great flotilla of which is often seen in the sky, 
especially in fine summer weather. Clouds at lower levels 
are frequently disposed in the same manner, but the 
observer must be very careful to distinguish between 
these and the high cirro-cumulus, which latter is either 
white, or changes its colour under the same circumstances 
as true cirrus, no part of any of the cloudlets ever seeming 
to be decidedly throivn into shadoiv by another part. 
The higJi cirro-cumulus also possesses the same faintness 
and delicacy of outline which we observe in cirrus. 

Cirrus-Haze. Sometimes cirrus is only visible as 
either a milky or an oily-looking haze, which is here 
termed ' cirrus-Jiaze? 

On Cirrus Cloud Observations. 209 


Amount. Whenever on looking round the sky the 
observer notices clouds of the kinds above described, he 
should in the first place endeavour to ascertain their 
amount. In doing this he should observe whether they 
cover, (i) only a trifling portion ; (2) about a quarter ; 
(3) about a half; (4) about three-quarters; or finally, 
the whole of the sky. The extent, not the density, of 
the cloud is here referred to. If there be any consider- 
able quantity of lower cloud in the sky, so that the 
amount of upper cloud cannot be accurately ascertained, 
it should be invariably entered as doubtful, by a ' ? ' 


Direction of Movement. The next thing to be 
attended to is the direction from which the upper 
clouds move. Considerable difficulty will here be en- 
countered. If some of the clouds happen to be nearly 
overhead, attention should at first be confined to such ; 
and if possible the observer should so place himself as 
to have the projecting corner of a roof or chimney, 
the summit of a steeple, flagstaff, or other stationary 
object, very close to the line between his eye and the 
portion of cloud which he is about to watch. 

A convenient plan is to set up a pole reaching 5 feet 
7J inches above the level of the observer's eye ; a 
mark or ring should be placed round the pole at the 


2 1 o Weather Charts and Storm Warnings. 

latter level. The summit of the pole must carry two 
thin rods, fixed crosswise, and set truly to the four 
cardinal points. Through or near the extremities 
of these rods should pass a thin circular iron ring of 
3 feet diameter, the use of which will be mentioned 
presently. The observer should, when opportunity 
offers, so station himself that some recognisable part of 
a cloud appears to move vertically either upwards from 
the top of the pole, or vertically downwards towards it. 
The direction of the pole from the observer's position, 
which may be judged of by the cross rods, is then, in 
the first case, the direction of the upper current ; in the 
second, its opposite. In watching the movements of 
the clouds, the observer may conveniently use a staff, 
one end of which is sharpened so as to stick in the 
ground, while the other is fitted with a crescent-shaped 
top in which to rest the arm, in order to support and 
steady the head. If the clouds be observed from a 
window it is some assistance to have a few wires some 
horizontal, others vertical fastened across the window 
to the wall on the outside. 

When, however, no part of the cloud is moving 
directly either towards or away from the observer, the 
effects of the perspective render it difficult to estimate 
the direction of movement exactly. To perfect oneself 
in the art, it is advisable as often as possible to 
endeavour to estimate the direction of movement of 
some portion of the cloud which is inconveniently 
placed, and then to find the actual direction of 

On Cirrus Cloud Observations. 2 1 1 

movement of another portion of the same cloud, or 
same sort of cloud, which happens at the time to be 
conveniently placed. By patiently practising the eye 
in this process the observer eventually learns to make, 
automatically, tolerably exact allowance for the errors 
arising from perspective. In no case (should he have 
the least doubt as to the direction of movement) must 
any conjecture be telegraphed, but the figure for 
' motion doubtful ' should be inserted in the telegram. 

It is always well, when observing the movements of 
the upper clouds, either to rest the head against a wall 
or tree, or to support it on a staff, or by some other 
means, so as to be quite motionless while taking an 


Apparent Velocity. The next particular to be 
attended to is the apparent velocity of motion in the 
clouds i.e., the force of the current which carries them. 
The observer is requested to notice whether the cloud 
be actually motionless, or whether its motion is very 
slight, moderate, rapid, or very rapid. Clouds of the 
cirrus type are seldom really motionless, but owing to 
their great distance they commonly, though not always, 
appear to move more slowly than the lower clouds. To 
estimate, even according to the very rough scale pro- 
posed, the apparent velocity of movements, we have to 
encounter precisely the same difficulties which have 
presented themselves in finding the direction, and the 

P 2 

2 1 2 Weather Charts and Storm Warnings. 

training of the eye requisite for deciding this is identical. 
A cloud which is near the horizon of course presents 
much less movement to the eye than one which is near 
the zenith, because it is at a much greater distance ; and 
further, the apparent movement of clouds which are 
travelling either across, or more or less obliquely to, 
the line of sight, is different from that of clouds which 
are travelling directly to or from the zenith. To test 
the accuracy of his estimate of the apparent velocity of 
clouds moving crossways and obliquely, the observer 
should, when occasion offers, compare it with his 
estimation of the apparent velocity of clouds in the 
same stratum passing overhead at the time. To test 
the accuracy of his latter estimation he will do well 
occasionally to notice how many seconds of time a 
recognisable portion of upper cloud takes in travelling 
15 on the sky, to or from the zenith. The pole 
arrangement already described can be employed for 
this purpose. In the last case a recognisable portion of 
cloud must be first caught in the zenith, the observer 
looking up the pole, applying his eye as closely as he 
can to the ring or mark above mentioned. He should 
then note the time, and again at the moment when the 
selected portion of cloud, as viewed from this spot, 
reaches the circular wire. The time taken by the cloud 
in passing over this distance from the zenith, which 
is (approximately) 15, should be then noted. When 
clouds of the cirrus type take a longer time than 
600 seconds (or ten minutes) in traversing 15 to 

On Cirrus Cloud Observations. 2 1 3 

or from the zenith, their motion should be regarded as 
' very slight/ When a less time than this, but a longer 
time than 300 seconds, or five minutes, as 'moderate ;' 
when less than 300 seconds but more than 60, as 
' rapid ; ' and lastly, whenever they take a less time 
than 60 seconds, as ' very rapid.' 

It may be mentioned that in simply looking at the 
sky without the aid of any appliance, inexperienced 
observers commonly estimate the position of the zenith 
wrongly because they do not throw the head far enough 


' R.-point? Clouds of the cirrus kind very com- 
monly lie in streaks, lines, or bands. Attention must 
next be given to the direction of these lines i.e., their 
position with respect to the points of the compass 
altogether irrespective of the motion of the clouds. When 
one of these bands happens to pass through the 
zenith, the point at which it reaches, or, if continued, 
would reach, the horizon, is to be noted. This point 
is called, for convenience' sake, the R.-point, or ' point 
of radiation,' because the cirrus bands, while really 
parallel, appear in perspective to radiate from this 
point, or rather from one beyond it, and beneath the 
horizon. It is, however, comparatively seldom that a 
cirrus band happens to lie immediately over the 
observer's head : the bands usually appearing in per- 
spective as arches, or portions of arches, lying in any 

2 1 4 Weather Charts and Storm Warnings. 

part of the sky, and variously inclined to the horizon, 
according to their distance from the observer. It is 
easy in these cases to find the R.-point, if we imagine 
ourselves able to travel over the earth's surface so as to 
get immediately under one of these bands. The best 
plan, however, is to note, first, the point on the horizon 
which lies beneath the summit of any visible complete 
arch of cirrus. If there be no complete arch, but only 
a short thread visible, the observer should extend his 
arm so as to point with the forefinger towards this 
thread, and then pass the arm briskly two or three 
times backwards and forwards from one part of the 
horizon to another, continuing as closely as he can the 
curve of the cirrus thread. The point on the horizon 
which is immediately beneath the highest part of the 
curve so formed should then be noted, and in every 
case the R.-point will be a point at right angles to this 
point on the horizon. 

It is obvious that there must always be two R.-points 
opposite to each other. Although it is in itself im- 
material upon which of these the observer fixes, he is 
requested as a matter of convenience, when he has 
already found a point from which the upper clouds move, 
to report the R.-point which is nearest to that direction, 
if one of the two be nearer than the other. For 
example, if cirrus travels from WNW. and lies in strands 
which are disposed lengthwise NW. and SE., the 
R.-point NW. should be reported in preference to the 
R.-point SE. 

On Cirrus Cloud Observations. 2 1 5 

The * selvedge ' or margin of a sheet of cirrus 
usually gives the R.-point in the same way as a streak 
of detached cirrus. And when sheet-cirrus or cirrus- 
haze completely covers the sky, a warp may frequently 
be seen in it which supplies the same datum. 

Cross rays or tangled threads may very often be 
noticed, frequently transverse or nearly so to the cirrus 
bands. These are especially discernible in some 
varieties of sheet-cirrus, when they form what we may 
call the 'woof of the cloud sheet, and give to the 
latter the reticulated appearance already mentioned. 
No notice of these threads is required in the telegrams. 
In rather common cases, however, the cross threads are 
so marked that it is impossible to decide whether a 
particular point of the compass, or one at about right 
angles to it, is rightly to be regarded as the R.-point. 


Lastly, whenever a bank of cloud of this type appears 
on any horizon, or whenever being overhead, or nearly 
so, such a bank appears thicker in one part of the sky 
than in the other parts, the observer is to report the 
position of bank (z>., the direction in which the bank 
lies or in which it is thickest) and its density. The 
direction will not necessarily be at right angles to the 
R.-point, though it most commonly is so. 

As to results from these observations, our experience 
of them is too short to enable us to say much, but the 

2 1 6 Weather Charts and Storm Warnings. 

following is a brief summary of the most obvious appli- 
cations of cirrus observations in weather prediction : 

During bright weather and Westerly (NW. to SW.) 
airs or light winds, the appearance of very high clouds 
of the mare's-tail type, moving from the North-west- 
ward, is usually an indication of a backing of the wind 
to the Southward, and its increase in force, probably to 
a fresh or strong gale. This movement of the very 
high clouds under such conditions is especially dan- 
gerous, if at the same time a batch of such clouds be 
rising in the west, and the barometer, after rising, is 
inclined to fall again. 

Again, when the wind is Westerly or North-westerly 
of moderate strength, if high hair-like or thready clouds 
appear moving from N. or NNE., they very commonly 
portend a great increase of wind from the North- 
westward, with snow, sleet, or soft hail in winter. 

Again, if the wind be Easterly, and high clouds 
appear, similar to those just mentioned, but moving 
steadily from SSW., they point to an increase in the 
force of the Easterly wind ; and during sultry summer 
weather, to the early approach of thunderstorms, 
followed probably by a shift of wind to the South- 

In all these cases, however, the direction of the wind 
at the surface must be carefully noted ; and it may be 
added that the value of the sign is increased when it 
occurs after a spell of exceptional bright weather and 
a sudden rise of the barometer. 





K ^ 

5 ^ 





6 oj o 

sanojj tz ISTJJ m ipyure^; 

Aq 'j 

moo * oo * 

.ONhHQfO vOf^Orf- 
OO'-O-MO ^-00 

" o o o o 'dodo 

>i oj o 'jpnoQ jo lunouiy ********* 

00000 000 

(31 01 0) 

(zi 01 o) 


O N 

1 ^- 1 ^- "^OO v> m t^ VOOO CO O CO 

t^ t> O 

>" re 

W W W 

W W '-> 

VO TJ-00 r<1 H-I M M TJ-M r^- r~ I-N. l^ O rj-v3 ON OJ 

r i i i r r i i i i i i i r i i i i 

-" >-" O OOOOOn-. 

C 3 

For explanation of the columns see p. 220. 

Appendices. 2 1 9 

i-Ht>r^i-ir< 1 5r^N vO ^ O rovO O 

oo l ~ l Q *> O\vO t^.# rOTt-N^J-i-ifcdfcfcc} 

d d dd"~ddpd'-<dc)6'- : 06666 6 6 6 d_P __ 6 

O ""> t^oo OOO* OOt^OO*** O* O 

O M VO N vO \O 

MVO w O 

O * 
Tj-^* rl" 

** "-" 1OO fOiOt^^-^JvO x^)00 10 


odcxjooooooooooo \o 06 


220 Appendices. 


















Thurso . 



































Valencia . 









Scilly . 









London . 


















The barometer continues to fall, excepting at Rochefort, most on our 
SW. coasts, and the wind at Valencia has backed to E. The sea is 
increasing at the mouth of the Channel, and rain falls along our western 
coasts. Warnings issued to our W. and N. coasts. 


BAROMETER. The letters r (rising), f (falling), and s (steady), indicate 
the motion of the mercury in last 14 hours. EXTREME WIND is the 
strongest wind experienced in past 24 hours. WEATHER : see p. 10. 

* An Asterisk is inserted in all places for which information is not 
usually received. The ' extreme ' wind for the Swedish, Norwegian, and 
Danish stations is an observation taken at 8 P.M. 


For Sunday \ November 29, 1874, at 8 a.m. 

During the past twenty-four hours a serious fall of the barometer has 
taken place over the United Kingdom, amounting to more than an inch at 
Nottingham ; and the centre of a very deep and well-defined depression 
has advanced to Wales and the western central parts of England. From 
Corunna to Pembroke there is a gradient of 0.13 inch per 50 miles for 
W. winds ; while from Skudesnaes to Holyhead there is one of inch 
for SE. winds. 

Temperature has risen very considerably at all but the most western 
stations. Over central England the change amounts to 14 in the twenty- 
four hours. 

Strong SE. gales have set in on our NE. coast, while very strong W. to 
NW. gales are felt in the SW. and Bay of Biscay, and S. winds in the SE. 
of England. The sea runs very high, especially in the N. and NE. 

Very heavy rain has fallen over the whole of the United Kingdom, 
accompanied by hail and snow in some places ; the fall continues at most 
stations. The weather is extremely unsettled. 

NOTE. The chart which faces this page is reduced from the Daily 
Weather Chart, but is printed in the form in use in 1887, not in that which 
actually appeared in 1874. 

Weather Charts & Storm Warnings. 

Plate ffl. 

( Explanation of Map ) 

ecu*, truth if an. i/..* 

Thf, fly wuft t*, 

ivuui <"<^ tlw barkf and* tnvBitrs rtftr tft t'vt'tx. 

. - fonm, abort TO. > -fijrcM 8 to , 

.-fonM. * to 7, fvw > to 4, 

at ^$rjs'fi *&o. 

_^T-.* T|OM *- 

DonghHie . 
Holjhe^L. _ 

Booba'i Pomt 



PAST 2fi- WOUflS. 


A dotted, luu ttpantUf te dietnct* in. wftiA rauv 

kxu falim rtwn riw Arr <Ar hat been, nan* 
FatU of Sin* or mart, arr </u,</ ii^figurrs. 

, of July. 


64 8 
53 9 
68 7 

upcraftira mjaw < ftr * 
arc nuvd^ <hor Ibr ttu- tC vt 

WEATHER CHARTS, Nov.29.1874-.8AJ4. 

Reduced from the Daily Weather Report. 



o o o >"> o 

. . .rONfNi-ivQ ** 

: : : q q q "i q : : : : :q 

o o" o" d d d 

B jo uoisuax * 

i O) ro r^OO CO~O OO ONt^CS OoO t>it^iOC< CN~O 

< * to^^n^^o^^o^)^ -> ^^ ^^^ ^ S i ^ 

. c OO O * fi N O ^" t^ r^vo O OO O~N t^ O~O~ON^T>vO^d 

,ir i^NC^NWC^c^NC^c^t-Ht *o ON ON HH ONOO 00 t^* t^ 

vO ^OO N O ^J~OO ^"O vO O ^ ^t"OO ^*OO O ^"^O vO ^}" 

g ^ 

N fo^j-vnvo t^oo OvO * x * ci co^j-mvo fxoo ON o -" 'g 

" "S- 

vom o O 
QQ... ">"-!.... 


* d d d d 

XaiDop A | : : : : : : : : : : ON *>. N_ o vp -^vp \p o o oq o~oo 


1 ::::::::: co co co co en co ^co ^^cn ^co co 


^ "" _,*. ON uic N\ot^r^ON<r 

jnoa A : : :Nf)Ti-vou-)T^Tj-Tfvo TJ-VO o 

JJH jo uoTSuaj^ : rprOfOfproro"" 

ed *^ i N * O O O r>.00 00 ON O t>. ON 

5 o. J ON ON d d d ON ON ON ON o" d d i-; 

JJ g Tj-Tj-U-JU-lLOTf-^^-T^-inVOL/llOVO 

^ I ONOO O oo r^vO t^ r^vo ON Tfvo O ON 



O 00 Tj- TJ- rj- O TfvO Tj-oo N vO W 
ON ON Ox?Oo'oO^OO -- ^ ^^ " 

ON O -> '5 




i-i cooo ro TJ- u*>\o r> 1-1 00 u~> <-> rj- t^ ON O O <N ON >-O ON r^ > N 
p$ -jnodBA t^o O I-H M roo ON TJ- ^ "-> "3- O O>oo ONCX) r>. t^oo O O * O 

W JO UOTSU3JL NcorornrororOWNNWNWdi-jhNi-. 'J'-J^^MN^ 

g I -^-^"'--"'" 

O ONOO vo LO rj- en ro ro 'JvO t^OO ON ON O O 

odoo'cx5o6o6o6o6ododcx5o6 ONO^ 


t^CX) ON O -" '5 



I:::::::::::::::: q ::::: 

XjpopA | :::::::::: ^"^^ ^i?^ ONQN>-< N ^"^^ 

c/5 c/5 coc/i 


fi <n w ww 

JIIOUUA vo* r-^ rC r^OO^CO ONOO ON ON O * O^ 

S jo uoisuax 

I~ ri-O 1 -oONTj-i-iO - ^'O l Of r 5 l O'-" 

ci, > ::::::::!:: *ovd <o vd vd <o MD vovd vo r^.oo' oo" 
p :::::::::::odoNod ONOO t^. ^ 1^.00" od d ~ 

r^vO M vO VO ON 01 ON^-OO H~rp<s 

* j ; j *. : : : : ;iou^^oLoi-niotoTj-rj-T^-roNt-i 

W ro ^ i/VO t-CX3 

r^OO ON O -" 'g 








ents at Falmouth Obse 

lf -recording 

Hourly Readings of t 

I TJ- ro I-H "H N .ro.rO*i>-iwr'NOQQ > 

ooqqqq : q :qqqqqqqq 
do'ddo'd d dddddddo" 


JO U01SU3J, 

1co O tXi C^ O *O PO POvO * OO t^ vo ro O ^O O ^ O ^" 
oo oo t^ ^ t^vq vq vq vq t>. 1^.00 ON q ** ** w cs i ro ro 

00 oo' 00 OO oo' CO OO OO' 00' OO CO OO OO OO OO OO 00' ON ON ON ON ON ON ON 

i-iO^HMf^Tl- VO>O t^OO C\ O n ' 



AJp 0l 3 A j 

- N xo ro 







"O~O "4- CO 

oo'oo'od iAr>.r>.t^.r^. i^.vo vo vo oo" 

r^. ON 

ro to 

N \O N Tt - 
N O O 

O O ON ON ON ON ON Ovoo' vO l^. r^. ON 

\O xoiOTj-n-cocON 

t^QO ON O - ' 



A BBE, Cleveland, on results of 
New York Herald warnings, 

Aberdeen Observatory, curves for 
October 21, 1874, 73 ; readings 
for same period, 222. 

Accumulated temperature, calcula- 
tion of, 196 ; curves of, for 
United Kingdom, 203. 

America, value of telegraphic re- 
ports from, 184 ; present service 
of telegrams from, 188. 

Anticyclones defined, 36, 40 ; illus- 
tration of, 40 ; characteristics of, 
57 ; typical, 59 ; motion of, 93. 

Ascending motion, effect of, on air, 
57, 164. 

Atlantic Ocean, advance of storms 
over, 15, 117, 183; cost of tele- 
graphy across the, 183. 

Atlantic weather, Hoffmeyer on, 

Aurora as sign of storm, 1 1 7. 

Azores, possible value of telegraphic 
reports from, 181. 

1 g ACKING ' of wind explained, 
31 ; reasons of, 78 ; why 
considered dangerous, 78, 148. 
Balaclava storm, experience of, 138. 


Ball, John, suggestion of storm 
warnings, 138, 139. 

Barometer, corrections of, 2 ; for 
capillarity, 2 ; for capacity, 2 ; 
reduction of, to 32 F., 3 ; to 
sea-level, 3 ; a single reading of, 
at any time, no indication of wind 
or weather, 27 ; scales, wording 
on, misleading, 49 ; rise and fall 
of, not infallible sign of weather, 
53 ; instances of serious fall of 
with calm weather, 54. 

Beaufort, Sir F., scale for wind, 8 ; 
notation of weather, 10. 

Belt, Thomas, theory of origin of 

cyclones, 119. 
| Bennett, J. G., Neiv York Herald 

warnings, 185. 

i Ben Nevis, difficulty of using re- 
ports from, 191. 

Boussingault, M., on accumulated 
temperature, 199, 206. 

Brisk, II. M.S., as signal-ship, 

Buys Ballot, Prof., law of motion of 
wind, 26, 44. 

CARLISLE, Bishop of, on failure 

of forecasts, 171. 
Charts, weather, explained, 37. 




Circulation of wind with cyclones, 
39 ; with anticyclones, 42. 

Cirrus -cloud observations, instruc- 
tions for, 207 ; some results from, 

Cirrus-clouds, value of, as signs of 
change of weather, 1 14. 

Clearness of atmosphere as sign of 
weather, 116. 

Clouds, deficiency of information 
respecting, in the Daily Weather 
Reports, 13 ; accompanying a cy- 
clonic storm, 64. 

Co-existing depressions, 84. 

Collins, J. J., New York Herald 
warnings, 185. 

Crops, dependence of, on tempera- 
ture, 196. 

Currents, equatorial and polar, 24. 

Curves from self-recording instru- 
ments at times of storm, 68, 


Cyclone, typical, diagram of clouds 
and weather in, 65. 

Cyclones defined, 36, 58 ; illustra- 
tions of, 37, 58 ; typical, 62 ; 
passage of, on the northern side 
of a station, 68, 72 ; passage of, 
on the southern side of a station, 
75 ; advance of, 65 et seq. ; 
probable rate of, unknown, 89 ; 
affected by surface of ground, 
95 ; probable direction of, known 
to some extent, 95 ; affected by 
conditions of pressure, 96. 

T~\E CANDOLLE on accumu- 
lated temperature, 199. 

Depressions defined, 36 ; very ex- 
tensive, instance of, 54; secondary, 
80 ; co-existing, 84. See Cyclones. 


Descending motion, effect of, on 
air, 57, 164. 

Districts of British Isles, arrange- 
ment of, 163. 

Draper, Prof. D., on storms crossing 
the Atlantic, 184. 

EQUINOCTIAL gales, supposed 

recurrence of, disproved, Scott, 
129 ; Prince, 134 ; in North 
Germany, 136 ; in Adriatic, 

Evans, Lewis, first published state- 
ment of motion of storms, 87. 

P ALMOUTH Observatory, curves 
for February 2, 1873, 75 ; read- 
ings for same period, 223. 

Faroe Islands, importance of infor- 
mation from, 182. 

Faye, M., theory of origin of storms, 

FitzRoy, Admiral, institution of 
storm warnings, 140 ; storm-sig- 
nal system, 119. 

Fogs with anticyclones, 61. 

Forecast?, FitzRoy, 162 ; Signal 
Office, Washington, 162 ; resump- 
tion of, in England, 162 ; prepa- 
ration of, 165 ; hay harvest, 165 ; 
failure of, Oct. 24, 1882, 171 ; 
criticisms of, 174. 

Forecasting, difficulty of, 163 ; re- 
sults of, 1 68; Ryves, 168; 
Peek, 169. 

Franklin, Benjamin, first recognition 
of motion of storms, 87. 

Q.ALE of Oct. 24, 1882, 172. 

Gales, equinoctial, their recur- 
rence disproved, 129. 



Germany, results of warnings from 
London to, 189. 

Gilbert, Dr. J. H., on accumulated 
temperature, 196, 205 

Gradients explained, 45, 53; in- 
stances of, 47, 52; for certain 
winds, 53 ; relation of, to winds, 


Gyration of wind, law of, stated, 
30 ; explained, 77. 

T-TAMBURG, warnings from 
London to, 189. 

Harries, H., on storm of Oct. 24, 
1882, 173. 

Heart's Content, telegraphic reports 
from, value of, 183. 

Iloffmeyer, on Atlantic weather, 

Hoffmeyer's charts, 96. 

Humidity, importance of, as indi- 
cation of weather, 6 ; changes of, 
during progress of storms, 70, 73, 

Humidity of air, action of ascend- 
ing and descending motion on, 
57, 164. 

Hurricanes, 44, 88. 

ICELAND, cost of proposed tele- 
grams from, 183. 

Isobars defined, 33 ; mode of draw- 
ing 35- 

T7"OEPPEX, Dr., on sun-spots 
and temperature, 123; on 
periodicity of weather, 124. 

JAW, Buys Ballot's, 26, 44 ; of 
Gyration, 30, 77 ; of Storms, 
44, 137- 


Lawes, Sir J. B., on accumulated 
temperature, 196. 

Leverrier, M., first proposal of 
storm warnings in Europe, 138 ; 
ideas as to object of storm warn- 
ings, I39 ! 46. 

Ley, Rev. W. Clement, theory of 
motion of storms, 118; diagram 
of typical cyclone, 65 ; cirrus- 
cloud observations, 207. 

Loomis, Prof., on storms crossing 
the Atlantic, 184. 

A/TELDRUM, C, theory of origin 
of storms, 119; periodicity 
of rainfall, 120. 

Mistiness of atmosphere a sign of 
weather, 117. 

Mohn, Prof., theory of origin of 
storms, 1 1 8. 

Monsoons, 25. 

Motion of cyclones in different di- 
rections, effects of, 78 ; probable 
rate of, comparatively unknown, 
89; in relation to mountains, 95; 
in relation to distribution of pres- 
sure, 95 et scq. ; erratic, instance 
of, 103. 

Mountains, effects of, on motion of 
cyclones, 95 ; effects of, on winds, 
179 ; effects of, on weather, 164. 

Mountain stations, 190; value of, 
Gen. Hazen, 191 ; Prof. Mas- 
cart, 191. 

Miiller, Dr. R., on equinoctial gale?, 

phic reports from, value of, 183. 
New York Herald warnings, 185 ; 
results of, 186. 



pEEK, C. E., on results of fore- 
casts, 169. 
Periodicity of weather, 124 ; of 

storms, 129. 
Pressure, relation of, to wind, 47 ; 

to motion of cyclones, 99. 
Prince, C. L., on equinoctial gales, 

"D AIN, not properly indicated by 
telegraphic weather reports, 

7 ; with cyclones, 67 ; with 

easterly winds, 77; produced by 

ascending currents of air, 57, 164. 
Rainfall, periodicity of, 120. 
Rate of motion of storm', 89. 
Redfield, discoverer of the Law of 

Storms, 137. 

Reid, Law of Storms, 137. 
Reports, telegraphic, defects of, 12. 
Results of Storm Warnings for the I 

British Isles, 152; of Forecast?, I 

Reye, Prof., theory of origin of I 

storms, 119. 
Ryves, Rev. G. T., on results of j 

forecasts, 168. 

CT. KILDA, value of, as a pos- 
sible reporting station, 182. 

Sea disturbance, scale for, n ; 
value of, as indication of weather, 

Sea-level, reduction of barometrical 
readings to, 3. 

Secondary depressions, instances of, 
80 ; effects of, on their primaries, 

Shower, with shift of wind to north- 
west, 65, 72. 

Signal-ships, proposal of, 15, 181 ; 
proved to be possible, 181. 

Signals, storm, explanation of, 142. 

Signs of storm, 146. 

Storms, definition of, 8; with high 
barometer, 27 ; vertical depth of, 
slight, 29 ; advance of, 87 ; rate 
of advance of, unknown, 89 ; 
probable direction of advance of, 
95 et see/.; advance of, over the 
Atlantic, 15, 118, 184; distri- 
bution of, in the British Isles, 

Storm warnings, institution of, 138; 
difficulty of issuing correct, 148; 
results of, 152; suggestion of, 
Ball, 138 ; Leverrier, 138. 

Strachey, Gen. R., on calculation 
of accumulated temperature, 199. 

Straight-line gales, explanation of, 

Summary of principles recognised 
in weather study, no. 

Sunday interruption to weather re- 
ports, 15. 

Sunk light - ship connected with 
Harwich, 181. 

Sun-spots, relation of, to rainfall, 
121-5; an d temperature in 
England, 123; over the globe, 

TEMPERATURE, correction of 

barometer for, 3 ; information 
respecting, given in Daily Weather 
Reports, 5; mean, definition of, 5; 
mode of ascertaining, 5 ; 
instances of sudden falls of, 65, 
76 ; regular daily course of, dis- 
turbed by storms, 70, 73 ; differ- 




ences of, as sign of storm, 149 ; 

accumulated, calculation of, 

196 ; effect of, on crops, 196. 
Thermometer, attached, 2 ; wet and 

dry bulb, 6. 

Tornado in Sweden, 29. 
Trade winds, 24. 
Typhoons, 44, 88. 

"yALENCIA Observatory, curves 
for March 27, 1874, 69 ; read- 
ings for same period, 221. 

Veering of wind explained, 31 ; 
reasons of, 78. 

storm, first pro- 
posal of, in America, 136; 
first proposal of, in Europe, 138 ; 
first proposal of, in England, 139; 
results of, for Hamburg, 189 ; 
results of, for the United King- 
dom, 152; instances of failure of, 
154, 161. 

Washington, Chief Signal Office, 
14, 179, 188. 

Weather, Beaufort notation for, 10 ; 
value of reports of general ap- 


pearstnce of, 12 ; charts, explana- 
tion of, 33 ; with anticyclones, 
5$i 59 J with cyclones, 58, 62 ; 
charts, mode of using, 113 ; local 
characteristics of, 115; periodi- 
city of, Koppen, 124. 

Weather Reports in 1887, specimen 
of, 18; in 1874, specimen of, 

Weekly Weather Report, institution 
of the, 193. 

Wind, Beaufort scale for, 7 ; cannot 
well be measured instrumentally 
at telegraph stations, 8 ; measure- 
ment of, by velocity instead of 
pressure, 9 ; characteristics of 
main currents of, 23 ; currents, 
polar and equatorial, 24 ; Trade, 
25 ; Buys Ballot's law for, 26 ; 
motion of, in cyclones, 40 ; mo- 
tion of, in anticyclones, 42 ; 
cause of, 45 ; relation of, to gra- 
dients, 52 ; reports of, from tele- 
graphic stations sometimes de- 
ceptive, 148. 

Wording on barometer scales, mis- 
leading, 49. 




HOW to PLAY WHIST : with the LAWS and ETIQUETTE of 
WHIST : Whist Whittlings, and 40 fully annotated Games. By 
"FIVE OF CLUBS " (Richard A. Proctor). Crown 8vo, 5s. 

HOME WHIST : an Easy Guide to Correct Play, according to 
the latest Developments. By " FIVE OP CLUBS" (Richard A. 
Proctor). IGmo, Is. 

The POETRY of ASTRONOMY. A Series of Familiar Essays 
on the Heavenly Bodies. By RICHARD A. PROCTOR. Crown Svo, 6s. 

NATURE STUDIES. Eeprinted from Knowledge. By 
CLODD, and RICHARD A. PROCTOR. Crown 8vo, 6s. 

LEISUKE READINGS. Reprinted from Knowledge. By 
RUNYARD, and RICHARD A. PROCTOR. Crown 8vo, 6s. 

The STARS in their SEASONS. An Easy Guide to a Know- 
ledge of the Star Groups, in 12 Large Maps. By RICHARD A. 
PROCTOR. Imperial 8vo, 5s. 

The STAR PRIMER. Showing the Starry Sky Week by 
Week in 24 Hourly Maps. Crown 4to, 2s. 6d. 

The SEASONS PICTURED, in 48 Snn- Views of the Earth, 
and 24 Zodiacal Maps and other Drawings. By RICHARD A. 
PROCTOR. Demy 4to, 5s. 

With 9 Illustrations. Crown 8vo, 5s. 

HOUGH WAYS MADE SMOOTH: a Series of Familiar 
Essays on Scientific Subjects. By RICHALD A. PROCTOR. Crown 
8vo, 6s. 

contrasting our Little Abode in Space and Time with the Infinities 
around us. By RICHARD A. PROCTOR. Crown 8vo, 5s. 

The EXPANSE of HEAVEN: a Series of Essays on the 
Wonders of the Firmament. By RICHARD A. PROCTOR. Crown 
Svo, 5s. 

Crown Svo, 6s. 

PROCTOB. Crown Svo, 6s. 



The SUN; Kuler, Light, Fire, and Life of the Planetary 
System. By R, A. PBOCTOB, B.A. With Plates and Woodcuts. 
Crown Svo, 14s. 

The OEBS ABOUND US ; a Series of Essays on the Moon 
and Planets, Meteors and Comets, the Sun and Coloured Pairs 
of Suns. With Chart and Diagrams. Crown 8vo, 5s. 

OTHER WORLDS THAN OURS; the Plurality of Worlds 
Studied under the Light of Recent Scientific Researches. With 14 
Illustrations. Crown Svo, 5s. 

The MOON: her Motions, Aspects, Scenery, and Physical 
Condition. With Plates, Charts, Woodcuts, and Lunar Photo- 
graphs. Crown Svo, 10s. 6d. 

UNIVERSE of STARS ; Presenting Researches into and New 
Views respecting the Constitution of the Heavens. With 22 Charts 
and 22 Diagrams. Svo, 10s. 6d. 

on Scientific Subjects, Natural Phenomena, &c. 3 vols. crown 
Svo, 5s. each. 

CHANCE and LUCK: a Discussion of the Laws of Luck, 
Coincidences, Wagers, Lotteries, and the Fallacies of Gambling ; 
with Notes on Poker and Martingales (or Sure (?) Gambling 
Systems). Crown Svo, 5s. 

LARGER STAR ATLAS for the Library, in 12 Circular 
Maps, with Introduction and 2 Index Plates. Folio, 15s. ; or Maps 
only, 12s. 6d. 

NEW STAR ATLAS for the Library, the School, and the 
Observatory, in 12 Circular Maps, with 2 Index Plates. Crown 
Svo, 5s. 

TRANSITS of YENUS; a Popular Account of Past and 
Coming Transits, from the First Observed by Horrocks in 1639 to 
the Transit of 2012. With 20 Lithographic Plates (12 Coloured) 
and 38 Illustrations engraved on Wood. Svo, 8s. 6d. 

STUDIES of YENUS-TRANSITS ; an Investigation of the 
Circumstances of the Transits of Venus in 1874 and 1882. With 
7 Diagrams and 10 Plates. Svo, 5s. 


198 Main Stacks 

Home Use 







Renewals and Recharges may be made 4 days prior to the due date. 
Books may be renewed by calling 642-3405. 


NOV 2 2001 

50M 6-00 Berkeley, California 94720-6000