416 MONTHLY WEATHER REVIEW. SEPTEMBElt; 1906
lic intelligence, as e. g., thru the courses for workingmen; but
here also he insisted upon thoroness ancl genuineness on the
part of the teachers. No one thing was more odious in his
eyes than that popularization of science which is based upon
mere appearance of scholarship.
The severe illness which first attacked him in 1867 was an
important factor in determining the decrease of his activities
during recent years. With a slow but not uncertain step it
brought about the dissolution of the once tireless mind ani1
body which in earlier years had delighted in activity. Early
on the morning of July 13, 1877 a gentle and painless death
freed this able man from liis sufferings, and bore him from the
midst of his large circle of friends, admirers, and relatives.
EXPLANATION OF THE TABLE OF EXCESSIVE PRE- CIPITATION.
The REYIEW publishes each month, in Table IT, a statelllent
of the acciunulated precipitation during storms of vertain
intensities. The actual fall during periods of five minutes,
ten minutes, etc., is given, in the right hand columns of this
table, for each five minute interval up to 50 minutes, inclusive.
When the excessive i'ate is continuous for more than 50 min-
utes the accumulated precipitation for such longer 1)eriods
is printed in tlie following line or lineh, and the actual dura-
tion of the given amount is found by aclding to the figiires
printed at the top of the coluim an additional 50 minutes for
each additional line eniployed. The times of beginning ancl
ending given in columns 5 and 6 of this table will sliov, in
such cases, that the total period is a continuous period ant1
that the depths giren are the accurnulatetl depths from the
beginning of the excessive rate of fall. In the RETIEW for
December, 1005, for example, on page 5G6, tlie escessive rate
at Atlanta on tlie 2d was continuous from 6 4 6 p. m. to 9:13
p. in., and the depth of 1.35 inch given in the 5-iriiuute column
is not, for 5 iiiinutes, but for 55 minutes, and the depth of 2.20
in the %)-minute column of the line below is for the total
period of 1%) minutes. The case is siinilar for Jupiter on the
same page. The storm of the 2d at New Orleans, La., is
recorded on two horizontal lines, but the times are not con-
tinuous, and the amounts given on the second line correspond
to the times printed ;tt the top of the columns.
The above explanation is prompted by the fact that wine
confusion Bas resulted froill a too literal interpretation of the
esplanatory heading of these right-hand coliiiuns.--F. 0. R
THE ATMOSPHERE AND THE SOIL.
The Department of Agriculture, thru its variouH bureaus,
seeks to investigate every condition that can in any way affect
the growthof the plant and the character, quality, or quantitgof
the resultant crop. We quote the following paragraphs from
an adclress by Milton Whitney, Chief of the Bnreau of Soils,
published as Farmers' Bulletin No. 257, on '( Soil Fertility ".
The greater part of the paper is given up to the question of
manures and fertilizers, but the following paragraphs relate
to atmospheric influences.
Plants nz ust breathe.
Of course we all understand that the breathing of the plant is mainly
thru ith leaves; but the soil also may be a very important factor in the
breathing of plants. as i t is necessary to have a supply of oxygen around
tlie roots. Physiologists differ as to the office the roots have in regard
to the absorption of oxygen. Whether i t is a true breathing-the
taking of oxygen for the plant economy thru the roots as thru the leaves-
has never been decided; but i t is uuquestionably a fact that rocits of
cultivated plants require oxygen around them for their healthy growth.
We know perfectly well that cultivation of the soil is iiuportaut or nec-
essary for the best development of many crops, and we say that this is
in order to introduce osygen aud make possible the introduction of more
water into the soil.
The investigations of the Bureau of Soils seem to indicate that the
actual supply of oxygen to the roots may not be the only or even the
most important function of cultivation. It seem6 necessary not only to
introduce air into the soil, but, by stirring the soil, to permit. the escape
of noxious gases that are perhaps given off by the plants themselves, or produced by bacterial action on the remains or excreta OP plants. In a
crowded rooni a person begins to feel drowsy, languid, and his head
begins to ache. We speak of these sensations, usually, as due to de-
ficient ventilation, too little oxygen, the oxygen having been partly used
up, and to an accumulation of carbonic-acid gas; but physiologists now
believe that this is not the true explanation, but that the person suffers
because there are gaseous emanations from the lungs that are deleteri-
ous to human beiugs. On t,he streets of Washington one of the principal causes of the death of
trees is leaks in gas pipes; every year hunalreds aiid perhaps thousands
of trees have to be removed, and the usual muse is a leaking gas pipe.
The amount o f gas is so small that i t cau not be detected by the odor,
but the influence of the gas on the roots is so pronounced that the tree
suffers ant1 is likely to die. It seems probable that the ventilation of the
soil is necessary not ouly to allow air to enter, but t o allow gases formed
in tlie soil to escape.
Furthenilore, air must euter not only for tlie use of t,he root itself,
but also t.o oxidize the organic matters giwn off by the plants-to pre-
serve the proper sanitary couditious in the soil-as I shall explain later.
Ventilation to remove noxious gases might increase the yield without
atiectiug the fertility. Vent,ilation for t,he Iiurpose of osidizing organic
matter might affect fertility it.self.
The plant is exceedingly sensitive to gases.
THE CLASSIFICATION OF CLIMATES.
We call the attention o f our readerh to a iiioht inhtructive
series of articles 1)s Prof. R. DeC'. Ward, on .'The classifics-
tion of climrtes ", pu1)lisliecl in the Bulletin of the diuerican
Cfeograpliical Society, for July i m c l Augnht, 1 W G . After
e\l~lsining in detail the ~u n n \ classifications that haw been
suggested by 5 arious stutleuth, Professor Ward conclutles as
fOllo\vs :
The broad classifiration of climates into tlie three general groups of
marine, continental, and mountain, with the subordinate divisions of
desert, littoral, and moiisoon, is wmvenient for purposes of summarizing
the interaction oP the cliinatic rleinents under the controls of lanal,
water, and altitude. But in any &:tailed study stiine sclieme of classi-
fication is welled in which siiiiilar climates in diffrreut parts of the world
are grouped together, and in which their geographic distribution ieceivrs
paiticular conblderation. I t is obvious from tlie preceding paragraphs
that an almost iufinite nuniber of vlassiflutbtinns might he proposed; for we may take as the basis o f subdivision either the special conditions of
one climatic element. as, for example. the same mean annual teinpera-
ture, or mean anniial range o f temperatilie, or the same rainfall, or
rainy seasons. or humidity, ant1 so o n ; or, again. sin~ilar wnllitions of
the combination of two or more elenleiits of cliiiiate limy be made the
basis of classitication; cir we may take a Ilotauioitl or a zoo1t)gical basis.
Of the classifications which have lieen proposed, special reference is here
made to tho.e of Supan, Iiijppeu. and Hult. That of Supan, taken as a
whole, gives a rational, siii~ple, and satisfactory schenle of grouping,
whose fielluent use in climatic. tlescrii~tious woultl tenil ttJ\Vani sybtem,
simplicity, and facility of coinliaristin. It emphabizes the essentials of
each climate. and serves to impre-s these essentials iipou the mind by
means of the conipact. well-consideiwl verbal summary which is giveu
in the rase of each province ilesci ibed. Ohvit iusly, no vlassific*atioii of
climates which is a t all wmplrte can approach the fiiniplicity of the
ordinary classification of the zones.
KOiipen's admirable scheine of bubdivicling climates from the botanical
point of view is distinctly rather for the use of -,tuclmts of plaut grogra-
phy than of general climatoltigy. The l~resent limits of the different
climate.; iu Iiijppeu's ma11 will doubtless ueecl to be changed in several
cases. as more detailed botanical studies throw further light on the
geographical distribution of different plants, and iio rigiil deliinitation
of plant zones is ever satisfactory to e\ t'ryim~. Biit Ii3lipen's classifica-
tion has tlie great merit of recognizing the existing differences of climate
lietween east and west voasts, and between coasts aiid interiors. The
coordination of districts of vegetation and of climate, which this scheme
EO strikingly emlihasizes, is a noteworthy fa#% in cliii~atiilogy.
Hult's classificatiou is far 'coli detailed, if all the siualler priivinces are
taken iiito account; )Jut if only the larger kingdonis are considered, as in
Plate I1 [not reproduced], the scheme is useful. It, however, possesqes
(~v e r that of Supan. whi L takes arcouut (i f more typical
of climate. Raveustein hygrotlieriiial types rest upon
unbatisfactory data, ani1 rpqious of T ery diKerent c.limatiu (.onditions
are grouped together liecause they halipen to have the same mean
annual temperature ancl relative huinidity.
THE WEST INDIAN HURRICANES OF SEPTEMBER, 1906.
Tropical storn development was exceptionally active inilmeri-
In seeking the causes of
B) I: E. (;+RHI(DTT, Pr<Dli.*aor t ,f XIrte<~rinlv~y.
can waters during September, 1906.
SEPTEMBER, 1906. MONTHLY WEATHER REVIEW. 417
this activity we find an apparent contributory condition in the
distribution of atmospheric pressure over the region of observa-
tion. In the West Indies ancl adjacent waters barometric pres-
sure was unusually low, while in the more northern latitudes
of the Atlantic, and more especially from the Azores over the
British Isles, the barometer averaged above normal, and after
the 17th was remarkably high. This arrangement of air pres-
sure overlying the Atlantic naturally produced an unusually
strong flow of air from the inore northern latitudes toward
the Tropics, and in this accelerated movement of air currents
is found a recognized associated cause of tropical storin de-
velopment. In fact a distortion or reversal of the iisnal order
of barometric distribution invariably produces weather of ab-
normal types over consideral>le areas. A notable instance of
this kind was presented during the winter of 1901-5 when
general and excessive rains occurxetl thruout New Mexico.
Arizona, and southern C'alifornia. L4s stated bj- the C'hief o f
the Weather BUW~U a t the time tlie cause of the heavy rains
in the southwest WAS not locd, hit was associated with gen-
eral abnormal atmospheric. contlitioiis over the United States
that were in turn associated with a1)noriiixl conditions that
o1,tainecl over a large part of the Northern Hemisphere. He
stated that the association between low baroiuetric pressure
and excessive rains in the southwest and high barometric pres-
sure ancl unusual colcl in the north and east had been estab-
lished, and that during winters of excessive colcl in northern
and eastern districts of the United States t,lie winters had been
unusually wet from western Teras to southern California.
During September, 1906, tlie storiny weather in the West In-
dian region attended a distribution of baroiiietric pressure
over the Atlantic Ocean siiuilar to that observed over the
North American Continent during tlie winter storm period of
190-1-5 in the southwestern portion of the United States.
Nunierous esnmples of this observed ahsociation of apparent
baroiiietric causes with weather eRects can be cited. Those
referred to, Iiowever, will Answer the present purpose of in-
viting attention to an arraiigeinent of atmospheric pressure
that figures prominently during certain phases of abnormal
American weather.
Chart IS shows the paths of the three hurricanes of Sep-
tember, 1906, here (liscust, over the West Inclies ancl adjacent
waters.
THE WEST INDIAN STORM O F AU~TUST 31-SEPTEMBER 15, 1906.
The approximate path of the center of this storm lias been
traced from a position east of Barbados, W. I., on August 31,
to the region north of the British Isles, on September 15.
Seven clays were occupied by the storm in advancing from
the ocean east of Barbados northwestward to the nort,liern
Bahainas, ani1 two days, the Gth ancl Tth, in making a recurve
over the Bahaiiias. Noring thence northenstward the certer
past west and north of Bermuda by the morning of the !Mi
and reached the region of the Banks of Newfoundland 11s the
10th. The subsequent path of the storm was north of the
region from which observations have been received. As the
barometer over the British Isles began to fall on the l l t h and
continued to fall rapidly until the 15th, when tlie center of a
well-marked and energetic depression past north of Scotland
on an easterly course, i t map be assumed that this disturbance
was identical with the one that inoved north of east from the
Banks of Nemfonnilland during the 11th.
During September 1 and 2 the vortex of the storin p~ir-
sued a course over the ocean east of and near the Lesser
Antilles, and by the morning of the 3d had past north of
Porto Rico. No reports have been received from vemels that
encountered the storin on these dates. The following, from
islands of the Leeward group, indicate measures of protection
that were taken on advices telegraphed from Washington.
~ ~~
_______~~ ______~~~ ~
Mr. D. Hope Ross, oficial in charge of the Weather Bureau
office at Basseterre, St. Kitts, W. I., reports:
The morning of the 1st the barometer, which had been steadily falling
since 9:30 the previous night, read 29.76 at 6 a. ni. It rose slightly, t o
20.77 at tinie of observation, and then fell steadily until 5 p. m. to 29.64,
the lowest point reached, after which i t rose slowly. The wind steadily
decreased from midnight o f August 31, and at 13330 p. m. of September
1 the velocity was less than 3 miles an hour. Shortly after i t increased
slowly, shifted from north-northeast and north to southwest and south-
southwest. It lilew steadily from that quarter with increasing force, and reauliecl a maximum velocity of 60 miles at 3:35 to 3:40 a. m., of the ad,
with a11 extreme velocity for one minute of 70 miles per hour at 3:36
a. ni. Rainfalls mere heavy, aud varied from about 6 to 13 inches at different I'oints on tlie island. The telegram of advice received the
morning of the 1st WAS issued as a precautionary warning. This was
done especially by telephone, to Government offices anil other centers
in tt>wn, ani1 to outlying clistritats, and all possible information was given
to caller+ at the office. Comparatively little damage mas done by the
storin. Owing t u the prostration o f telephone liner conimunication to
outlying point.: on the island \\as teniporaiily cut off; niany trees were
uprooted. and a few small liouses were l ~l o w ~i down ; but everyone being
prepared there appears to have been 110 further damage.
The following letter, dated September 4, froiii Mr. Christo-
pher H. P:tpe, American Consul at St. Thomas, JV. I., indi-
cates the character of the disturlmnce during its passage over
the region of the Virgin Islancls:
Tour cal~legram of September 1, 1906, received, for which I beg of you
to accaept my sincere thanks ; also the business people of thih place, as
your timely warning has saved millions to +hipping in this harbor. I
sincerely hope that >ou will find i t a pleasure in advising me of all future
disturtanc~s during this dreaded seawn. I beg t o assure you this office
is at your t.ommand at all times in whatever way we can serve you in
your great work.
The Weather Bureau obserl-er at Sau Juan, P. R., reports:
The storm apparently followed the course forecast in tlie messagc
received September 1. ,411 vessels were adriswl to remain in port;
the adricw wvre heeded and no material damage has been reported.
Vieqries is the reporting station nearest to the path of the storm; no
material damage was cmisetl on that island altho 8.18 inches of rain
fell on the 4th. Tho ~iiasinrum wind Trlocity at San Juan was 35 miles
an hour from the west on the 3d.
During the 4th anil 5th the center of clisturbance iiioved
northwest\\ arc1 toward the northern Bahamas, passing north
of Turks Ihhnd on the 4th. Dnring the 6th and 7th i t gradu-
ally recurved east of the iiorthern Bahamas and was severely
felt by vessels navigating that region. At 3 a. m., Septeniber
7, the three-nmstecl schooner J d /i Rmr, John Douglas, master,
in latitude north 28" 37', longitude west 55" 4', had a ba-
rometer reading of 2!).01. This vessel was coal-laden for Key
West. ancl was held before the wind on a southwest course and
orost the path of the htorm ahead of the center. The shift
of the wind encounteretl was from northeast back to north
and northwest, but there were no cross seas until after the
center had past. By the morning of tlie 8th the storm had
coinpleted its rerurre to the northeastward. On tlie morning
of that day Beriiiucla, and Lloyds, London, were cabled that
a tropical clisturhance was approaching Bermuda from the
southwest. By the morning of the 9th the center had past
west and north of Beriilnda and the reailing of the barometer
at Hamilton wah 29.18 inches. The storm was exceptionally
severe in the trans-Atlantic steamer tracks on the 10th and
l l t h . The experience of the North Geriiian Lloyd steamship
h-ot,~iiyi/i Liiise indicates the intensity of the storm in the re-
gion of the Grand Banks. This steamship encountered the
hurricane Septeiner 10 in latitude north 3 9 O , longitude west
55". and at 8 1). in. the baronieter read 28.06 inches. The
yessel was unable to resume full speed until 5 a. m. of the l l t h ,
fourteen hours after the storm began. The subsequent course
ancl character of this storm will doubtless be determined by
vessel reports that are not now arailable.
THE ATL.4NTIC ('09sT STC)RM OF 5EPTEMBER 17.
Unsettled weather conditions over the West Indies followed
the passage of the storm of August 31-September 11 over the 'Monthly Weather Review for April, 1905, Vol. SSSIII, p. 153.
41s MONTHLY WEATHER REVIEW. SEPTEMBER, 1906
Western Atlantic, and on the 12th there was evidence of a
slight depression near Porto Rico. From this position it
moved to the neighborhood of the Windward Channel by the
13th, where there were indications of its presence on the l-ith,
after which i t appeared to pass northward over the ocean.
During the 16th falling barometer and increasing northerly
winds along the south Atlantic coast showed the presence of
a barometric disturbance off that coast, but an absence of re-
ports from the great ocean area rendered it impossible to
locate the center of the disturbance or to determine its future
course. On the morning of the 17th its close approach to the
Carolina coast was shown, and by 1 p. m. it had reached the
coast line north of Charleston, where the barometer at that
hour read 29.44 inches and the wind had reached a velocity
of 46 miles an hour from the west. A t Wilmington the maxi-
mum velocity had been 52 miles an hour from the northeast.
After crossing the coast line the storm lost strength rapidly,
and during its subsequent course to the lower Ohio Valley and
thence northeastward its energy was expended in heavy rains.
Damage on land by wind was of a minor character, and no
serious injury to property has been reported. The clamage to
shipping along the coast between Charleston and Wilmington
was, however, considerable, and crops were destroyed near
Georgetown where the storin moved inland. Indinii sta-
tions in the indicated line of its advance were notified on the
12th ani1 13th of the character of the slight barometric cle-
pression froin which this storm sprang. During the mcceecl-
ing two days its presence ofi our southern coasts was but
faintly indicated by land observations. On the 15th advices
were issued that there was evidence of a disturbance between
the Carolina coast and Bermuda that was apparently moving
northward. The storin that struck the South Carolina coast
was a small tornado-like development in the southern end, or
tail, of the disturbance referred to, and its origin ancl course
was a product of oceanic atmospheric conditions that mere not
shown by land observations.
The observer of the Weather Bureau at Charleston, S. C.,
reports as follows regarding this storm:
On the day preceding the storm no unusual phenomena were observed
at this station. The barometer hegan to fall at 11 p. m., of the 16th,
and fell steadily until i t reached a mininium of 35.44 inches at 1 p. 111.
of the l’ith, after which i t rose rapidly. Light to moderate northerly
winds continued during the 16tli, shifted to northwest at 5 a. in., l‘ith,
and hacked to south at 5 p. m., from which (luarter it continued until
midnight. From 11 a. n1: to 9 p. m. i t ldrm a moderate and at times a
fresh gale, the highest velocity, 18 iniles an hour, occurriugat 3:30 p. m.
By midnight i t had diminished to 23 miles an hour.
The damage to buildings in Charleston wassiuall, not exceeding $1000,
and was cnnfined prlwipally to sillall buildings. At Georgetown, S. C.,
a small town 60 miles nortlirast from Charleston, the damage was esti-
mated at 915,000. The wiuds being oflshore during the storni, the storm
tide did not exceed the normal high tide more than one aud a half feet,
and consequently no damage resulted from high mater in esposed por-
tions of the city.
THE CARIBBEAN SEA STORM OF SEPTEMBER 23-27, 1906.
This was the severest disturbance that has visited the Gulf
coast since the occurrence, on September 8, 1900, of the storm
that devastated Qalveston. That storm advanced from the
eastern Oaribbean Sea to the Tesas coast during September
1 to 8, 1900. The September, 1906, storm was first definitely
located over the western portion of the Caribbean Sea on the
22d and crossed the Gulf coast line west of Mobile the morn-
ing of the 27th. After leaving the Yucatan Channel, on the
24th, the storiii moved aliiiost due Bortliward over the Gulf
of Mexico. During this period there was an almost equal
chance for the center to swing to the northeastward over the
Florida Peninsula and thence along the Atlantic coast, or to
the northwestward to the Tesas coast. Daily advices and
warnings based upon careful calculations announced, how-
ever, that its course would probably be northward toward the
central Gulf coast, and beginning on the 23cl Gulf shipping
was advised to remain in port. When a storm of this type is
hundreds of miles from reporting coast stations i t is apparent
that its intensity and exact coiirse can not be accurately deter-
mined. Tlie warnings issued during the four days of its pro-
gress over the Gulf were based upon an assumption that the
disturbance possest in full degree the intensity that usually
characterizes storins that advance from the Caribbean Sea
orer the Gulf of Mexico; and clay by clay the statement was
clearly macle that conditions dangerous to shipping would
exist over the central and east Gulf. So far as known the
warnings were heeded, and vessels did not venture into the
threatened district. It is a fact, indeed, that of the millions
of dollars of clamage an infinitebimal portion was clone to
shipping in the open Gulf. The inonetary value of vessels
and cargoes thus protected, and the nuniber of human lives
safeguarded, can not now be calculated. Adequate protection
coulcl not have been macle against daiiiage and destruction by
banked up nater from the Gulf that snept over harl)ors, bays,
inlets, and low-lying coasts. and from washouts due to tor-
rential rains that attend storms of this character. The center
of the storm crost the coast line near am1 west of Mobile at
about 8 a. m., seventy-fifth meridian time, of the 27th. East
of the point where the center crost the coast an onshore
southerly g a b combined with the wa\ e from the Gulf to drag
vessels from their anchorage, and caused water damage to
shore property that was not experienced west of the storm’s
path, where northerly offShore winds pre\ ailed during the
period of the storm’s greatest intensity. After the 27th the
htrength of the storin diminished rapid), and during its subse-
quent course northward orer the lower and middle Mississippi
Valley heavy rains fell nithin its area.
The following are accounts by observers of the Weather
Bureau ancl newpapers of storm action and losses at and near
their respecti1 e cities:
Ta n i p , FZa.
The Weather Bureau followed the distiirlJance from its inception to
i t s destructi\ e violence at Mobile and Pensacola, and many congratula-
tiow hare beeu received on its good work in keeping so closely in touch
with the progress an11 development of the storni. The first information
regarding the btorm wa+ received at 11:45 a. ni., Septeinber 23, when
the disturbance was near C+raud Cayman and moving tonard the Yucatan
Channel. A t this station the approach ancl passage of the storm over
the Gulf was shown by violent sqiialls of wind, rapidly moving clouds,
and ahormally high title.;.--Jno. S. HaZen, Local Forecmler.
The Tampa Tribune, of September 30, 1906, comments
editorially on the storm as follows:
The great storm in the Gulf has given an excellent cleinonfitration of
the effectiveness, accuracy, and practical utility of the Governinent
weather forecasts. It was on la& Suuday that the Weather Bureau
oftice in Tampa flew i t s warnings of the approach of a great tropical storm which was then appearill:: in the regiou of the Yucatan Channel.
From that time on the Bureau sent claily and urgent warnings of the
progress of the disturbance. As the fitorin c a n e within the area of a
inore intinlate observation, its character, severit), and direction were
iuore closely reported The second day lielure i t struck the coast, sliecial warnings were sent t o Gulf ports fur the guitlanve of shipping,
with the most urgent advice that ve55els remain in port till after the
disturbance. It IS already known that many vessels heeded this warn-
ing and were without doubt sa.ted because of it. * * * That the storm would strike the miildle Gulf coaht with great fury mas the notification
given two days IJefore it struck. That the city of Pensacola waa in
ey)ecially grave danger, and woulil be near tlir center of the most severe
disturbance, was announced t\\ entg-four hours before the storm broke
over that city.
hTrw Ot*Iems, La.
The severe4 storm that has visited the middle Gulf coast for several
gears prevailed from September 25 to 27. The work of the Weather Bu-
reau in forecasting the progress. prolmblr path, ani1 inteusity of thib
stoim was exact in almost every paiticular. The system of distribution
of storm-warning advices is so coinplete that iiot only clid places having
telegraph and telephone service receive the warnings, but thru the
intelligent and promiit action of all employees aiid agents o f the Bureau
every fishiug cauip received timely warning. Acting on advices people
in exposed localities -ought place5 of safety, and vessels either remained
in port or sought shelter in protected places along the rivers and bayous.
SEPTEMBER, 1906. MONTHLY WEATHER REVIEW. 41i
this activity we find an apparent contributory condition in the
distribution of atmospheric pressure over the region of observa-
tion. In the West Indies and adjacent waters barometric pres-
sure was unusually low, while in the more northern latitudes
of the Atlantic, and more especially from the Azores over the
British Isles, the barometer averaged above normal, and after
the 17th was remarkably high. This arrangement of air pres-
sure overlying the Atlantic naturally produced an unusually
strong flow of air from the more northern latitudes toward
the Tropics, and in this accelerated movement of air currents
is found a recognized associated cause of tropical storm de-
velopment. I n fact a distortion or reversal of the iisual order
of barometric distribution invariably produces weather of ab-
normal types over considerable areas. A notable instance of
this kind was presented during the winter of 1901-5 when
general and escessive rains occurred thruoiit New Mexico.
Arizona, and southern C’alifornia. L4s stated by the Chief vf
the Weather Bureau a t the time ’ tlie cause of tlie heavy rains
in the southwest was not local, lnit was associated with gen-
eial abnormal atmospheric conditions over the United States
that were in turn associated with a1)norinal conditions t8h:tt
obtained over a large part of the Northern Hemisphere. He
stated that the association between low barometric pressure
and excessive rains in tlie southwest and high baroiiietric pres-
sure and unusual cold in tlie north mil east hacl been estab-
lished, and thnt during winters of excessive cold in northern
ancl eastern districts of tlie United States the winters liacl been
unusually wet from western Texas to southern C’alifornin.
During September, 1906, the storiny weather in the West In-
dian region attended a distribution of barometric pressure
over the Atlantic Ocean siiuilar to that observed over the
North American C’ontinent during the winter storm periocl of
190-1-5 in the southwestern portion of tlie United States.
Numerous examples of this observed association of apparent
barometric causes with weather effects can be cited. Those
referred to, however, will An~wer the present purpose of in-
viting attention to an arrangement of atmospheric pressure
that figures prominently (luring certain phases of abnormal
American weather.
Chart IS shows the paths of the tliree hurricanes of Sep-
tember, 1906, here discust. over the West Indies and adjacent
waters.
THE WEST INDIAN STORM OF AUtTUST 31-5EPTEMBEH 15, 1306.
The approximate path of the center of this storm lias been
traced from a position east of Barbados, TV. I., on August 31,
to the region north of the British Isles, on September 15.
Seven days were occupied by the storm in advancing from
the ocean east of Barbados northn estwarcl to the northern
Bahamas, and two days, the 6th ancl Sth, in inaking a recurve
over the Bahamas. Moving thence northeastward the cepter
past west and north of Bermutla by the morning of the !Mi
and reached the region of the Banks of Newfoundlancl by the
10th. The subsequent path of the storm was north of the
region from which observations have been received. As the
barometer over the British Isles began to fall on the l l t h and
continued to fall rapidly until the 15th, when the center of a
well-marked and energetic depression past north of Scotland
on an easterly course, it 1na-j be assumed that this disturbance
was identical with the one that moved north of east from the
Banks of Newfoundland during the l l t h .
During September 1 and 3 the vortex of the storm pur-
sued a course over the ocean east of and near the Lesser
Antilles, and by the morning of the 3d had past north of
Porto Rico. No reports have been received from vessels that
encountered the storm on these dates. The following, from
islands of the Leeward group, indicate measures of protection
that were taken on advices telegraphed from Washington.
‘Monthly Weather Review for April, 1905, Vol. SSSIII, p. 152.
~~ _______~~ ____~~ ~~~~
Mr. D. Hope Ross, official in charge of the Weather Bureau
office at Basseterre, St. Iiitts, TV. I., reports:
The morning of the 1st the barometer, which had been steadily falling
s i w e 9:30 the previous night, read 29.76 at 6 a. ni. It rose slightly, t o
29.77 at tiiiie of observation, and then fell steadily until 5 p. m. to 29.64,
the lowest point reached, after which it rose slowly. The wind steadily
decreased from midnight of August 31, and at 12:30 p. m. of September
1 the x elocity was less than 3 miles an hour. Shortly after i t increased
slon.ly, shifted from north-northeast and north t o southwest and south-
southwest. It iilew steadily from that quarter with increasing force, and
reached a maximum velocity of 60 miles at 3:35 to 3:40 a. m., of the 2d,
\\ ith an estreine velocity for one minute of 70 miles per hour at 3:36
a. in. Rainfalls were heavy, and varied from about 6 to 13 iuches at diffrreut points on the island. The telegram of advice received the
morning of the 1st WAS issued as a precautionary warning. This was
clone especially by telephone, to Government ofices and other centers
in town, ani1 to oatlyiiig districts, ancl all possible information was given
to caller+ at the office. Comparatively little damage was done by the
storin. Owing t c ~ the prostrativn of telephone lines communication to outlying points on tlie island was temporarily cut off: nmny trees were
uprooted. and a few small houses were IJlO\Yn down ; but everyone being
prepared there appears to have l~een no further damage.
The following letter, dated September 4, froiii Mr. Christo-
pher H. Payne, American Consul at St. Thomas, W. I., indi-
cates the character of the disturbance during its passage over
the region of the Virgin Islands:
Your cablegram of September 1, 3906, received, for which I beg of you
to accept my sincere tliaulis ; also the business people of thih place, as
your timely warning has saved millions to <hipping iu this harbor. I hincerely tiope that j o a will find i t a lbleasure in advising me of all future
disturlmiices during this dreaded season. I beg to assure you this office
is at our coiutuand at all times in whatever way we can serve you in
your great woili.
Tlie Weather Bureau observer at San Juan, P. R., reports:
Tlie storm apparently followed the course forecabt in the messagc
received Septeinber 1. All vessels were advised to remain in port;
the advices were heeded and no mateiinl clamage has been reported.
Vieqries is the reporting station neared tu the path of the storm; no ninterial damage wah cnusetl tliat island altho 8.49 inches of rain
frll on the 4th. The ~iiasiinuni wind \ rlocity at San Juan was 35 miles
an hour from the west on the 3d.
During the 4th and 5th the center of disturbance moved
northwestward toward the northern Bahaluas, passing north
of Turks Island on the 4th. Daring the 6th and 7th it gradu-
ally recurved east of the northern Bahamas and TORS severely
felt by vessels navigating that region. At 3 a. m., Septeiuber
7, the three-masted schooner J o i i ~i Rose, John Douglas, master,
in latitude north 28. 37’, longitude west 77’ 4‘, had a ba-
rometer reading of 2!).0l. This vessel was coal-laden for Key
West. and was held before the wind on a southwed course and
orost the path of the htorm ahead of the center. The shift
of the wind encounterecl was from northeast back to north
and northwest, but there were no cross seas until after the
center liscl past. By the morning of tlie 8th the storm had
completed its recurve to the northeastward. On the morning
of that day Bernluda, and Lloyds, Lon(lon, were cabled that
a tropical disturbance was approaching Berniucla from the
southwest. By the morning of the 9th the center hacl past
west and north of Beriiiuda and the reading of the barometer
at Hamilton was 29.18 inches. The storm was exceptionally
sei-ere in the trans-Atlantic steamer tracks on the 10th and
l l t h . The experience of the North German Lloyd steamship
Kot?,/iyin Liiise indicates the intensity of the storm in the re-
gion of the Grand Banks. This steamship encountered the
hurricane Septemer 10 in latitude north 30°, longitude west
5 5 O . and at 8 1). 111. the barometer read 28.06 inches. The
vessel was unable to resume full speed until 5 a. m. of the l l t h ,
fourteen hours after the storm began. The subsequent course
and character of this storm will cloulJtless be determined by
vessel reports that are not now available.
THE ATLANTIC C’OPST hTURhI OF qEPTEMBER 17.
Unsettled weather conditions over the West Indies followed
the passage of the storm of August 31-September 11 over the
SEPTEMBER, 1906. MONTHLY WEATHER REVIEW. 419
Every precaution was taken to reduce the damage from the approaching
storm to a minimum. On themorning of the 27th there was about three
feet of backwater in the river a t New Orleans, indicating a storm tide a t
the mouth of the river of six feet or more. Such a high tide is exceed-
ingly dangerous, and but for the timely precautions taken there would
have doubtless been much loss of life and property. The highest wind
velocity reached at New Orleans was 49 miles an hour from the north-
west at 8 a. m. of the 27th, and the lowest barometer on recortl for the
station during the summer months, 29.14 inches, was reached at 8:15
a. m. of the 27th (see fig. 1). The storm was severe east of New Orleans,
but there does not appear to have been any loss of life on the Mississippi
and Louisiana coasts. The damage to property that could not. be moved
or protected was great. A t Burwood the tide was over the wharf, wliich
.is eight feet above mean Gulf level. Railroads and crops suffered great
damage.
26 27 28
---
IOXll 2 d 6 8 IO M 2 4 6 8 IO XI1 2 4 6 8 IO M 2 4 G 8 IO XI1 2
FIG. 1.-Barogram, New Orleans, La., September 26-28, 1906.
I t is impossible to estimate the amount of damage that was
by the storm in this section, hut i t amounts to several million
Shipping remained in port and prepared for the storm, and tlie
been no great losses reported. The losses in this sectiou were excell-
tionally small for such a storm, and this is attributed by the press and
public to precautions taken on account of the timely warnings issued by
the Weather Bureau, as indicated tiy the following extracts from edi-
torials that appeared in the local press.
The Times-Democrat of September 28, 19OG:
While only meager details of the damage wrought by the st.orm on the
Gulf littoral are obtainable on account of iiijury to the telegraph service,
the reports indicate severe losses of property at several point,s on t.he
coast, much suffering and inconvenience in a number of communities,
and possibly a loss of life. No doubt the losses would have lieen much
greater but for the timely and acourat,e warning given by the United
States Weather Bureau. Notice that the disturbance was approaching
the coast, and that it was unusually severe in character, enabled ship-
ping interests to avoid the dangers of the open sea, and very greatly
minimized the damage incident to the violent visitation. Because nf the
high standard of efficiency reached by the Weather Bureau, and the es-
tension of t,he service to West Indian points, we may now escape many
disasters of t.he character hitherto experienced during these meteoro-
logical disturbances; indeed with reasonable care we should tie able to
reduce our losses on the coast, ancl in the Gulf, t o the minimum.
The Daily Picayune of September 29, 1906:
When the full record of the damage from the recent storm will be
made up i t will be found that altho the property loss has been quite os- tensive the total loss of .life has not lieen nearlysn great as in many past storms of even less magnitude. This fact was probably due to the
timely warning of the storm's approach, which enabled everybody likely
to be especially exposed to take proper precautions.
Pensrrroln, Fln.
This was the most terrific storm in the history of Pensacola, or since
the village of Pensacola, on Santa Rosa Island, was swept away 170 years
ago. The greatest loss was to the shipping interests; a large number of
ocean going vessels, tug boats, fishing smacks, launches, and craft of all
kinds were wrecked upon the beach; the wreckage is strewn from Mag-
nolia Bluff to the entrance of the harbor. During the early morning
hours the people of the city were panic stricken, many believing that a
repetition of the Galveston disaster was imminent, and large numbers
of people took refuge in the higher portion of the city, braving the high-
winds and stinging sand-filled rain in the hope of reaching a place of
safety. When the wind reached 60 miles per hour (at 1:07 a. in.), every
effort was made to protect property and life and people were driven from their homes along the water front.. Trees were uprooted; houses were
unroofed, and vessels in the harbor began to drag t.heir anchors, being slowly but surely forced upon the beach. A t the height of the storm
(between 3 and 4 a. m.) the water rose 83 feet above the normal hiph- water mark, being the highest known. The great protection afforded
by Santa Rosa Island saved Pensacola from more severe suffering. The entire water front property was inundated, the water reaching many
houses; some were either carried away completely or irreparably damaged.
Muscogee wharf was broken intwo in the middle, and the tracks on either
side of the main dock were washt away. Thirty-eight coal cars which were
on the wharf were washt away. A timber boom was broken and the tim- ber cast adrift: this timber and wreckage of every description is jammed
upon the beach in a torn and twisted mass. The water backed up in the low ground north of Intendencia street, where several cottages were in-
nundated, the water in places rea~?hing a depth of 10 feet. dbout the greatest havoc was wrought along West Main street, the south side of
which has been completely washt away. The houses along the bay shore
froin Barcelona street to Perdido are in ruins and all along t.he shore in-
numerable vessels are scat,tered. The damage t,o the fishing industry alone will probably aggregate half a million doilars and that to the tow
boat inclustry will aggregate as much more. Several wharfs are com-
pletely gone. On Palafox street from the wharf north to Wright street
there is hardly a building that has escaped damage, and in many in-
stances the losses will be heavy, as the t i n roofs were blown from many
and the driving rain damaged stocks to a great extent. The train serv-
ice in and out of the city is completely paralyzed. A train came in from the north this morning during the storm, arriving many hours late, the
train men reporting that. every few yards i t was necessary to stop and
remove trees and poles from the track. From Magnolia Bluff to Milton
the track is gone and a portion of Escambia Bridge is washt away. A
channel was washt thru Santa Rosa Island east of t.he Life Saving Sta-
tion; the station itself is gone and the entire shore is a mass of ruins.
Fort Barrancas is much damaged. Fort Pickens has suffered severely,
and Fort Mcllae is completely razed. the ruins of t.he old fort standing
alone on the shore. [Figs. 2, 3, and 4 show some effects of the storm,
and flg. 5 reproduces the barograph trace a t Pensacola.]
There were only two of t,lielocal steamers inoperat,ion on the morning
of the %th, namely, t,he U. S. S. Poe, of the Quartermasters Departnient,
net tonnage 106, and the dredge C~I.ILC~~R. of the Engineer Department, net
tonnage about 1000; these vessels, together with the Pilot, belonging to
the Pilot Association, net. tonnage 66, and a few flshiug smacks, took u p
anchorage behinil High Bluff on the peninsula between Town Point aut1
Deer Point,, which sheltered them from the full force of the easterly gale,
and they were protected from the Gulf by Sauta Rosa Island. An idea of the force of the st,orni in that locality arid westward from the navy-
yard may be gained from the following facts. The Pilot anal the Caiccica
were equipped with extra good ground tackle, but the Poe had only
common large anchors. The Pilot let down one anchor at 1 a. m. in 16
feet of water with 75 fathoms of chain out, and another anchor with 45
fathoms of chain out a t 2 a. ni.; she took up anchors in 20feet. of water,
having drifted about 150 yards; her engiues were not working. The
Caitcu~ had two anchors out and half steam ahead, and with her bius
half f u l l of water, only drifted 100 yartls. The Poe with full steain
ahead tlragged her two anchors three and one-half miles to the navy-
yard where they caught in nioorings used for coal barges, and this saved
her from going ashore.
Wharfmaster Cos stated that the tide was fully ten feet above normal
high title along the city water frcint and twelve feet a t Bayou Grande,
based upon the reckonings of five reliable men. The sea swells entering
t,he slip between Palafox ancl Baylen streets were from seven to twelve
feet high between niitlnight and 7 a. in. and were aljout one hundred
yards apart, lifting wharf timbers and boards as they rolled thru the
slip and splashed over Cedar street. Oyster boats, launches, steam
tugs, lighters. timl.iers and wreckage of all description are jamnled
together a t the corner o f Cedar and Baylen streets; the launc,h Wolver-
ine, a 1Jtiat 40 liy 3 feet, past over Cedar street and entered the lot at
the northeast corner nf Baylen and Cedar streets, followed Liy other
wreckage mentioned above. Trees fully exposed to the easterly gale
in all parts of the city were blown down, the chinaberry and sycamores
suffering most. All weak chimneys that were broadside to the east were
tUiii1Jled to the westward. Whenever the wind got under a t,in roof it
rolled i t off. The foreign merchant vessels in port on the 26th were 5 steamships.
7 full-rigged ships, and 13 barks. One bark is a total wreck, ancl 11
&her barks ant1 ships are stranded: also an American schooner is sunk
and over on her side. Before t,he hurricane there were in port 38
schooner.riggec1 fishing s m a ~k s ; now 29 are ashore and 16 of these are ,
probably tot,al wrec:ks. Of 36 lumber barges only 8 remained anoat., and of 8 tugs 2 are afloat, 3 are totally wrecked and 3 can be floated.
The navy-yard was exposed to the full sweep of tlie easterly gale and
all naval boat.s, except the Isle De Lwo?~ and two water boats, are either
sunk or ashore; this includes the wooden dry dock, and all coal barges
except one; all tugs were sunk; the Spanish steel dry dock is apparently
undamaged ]J u t aground; the entire water front a t the yanl is a mass of
wreckage. Houses along the water front in TVolsey ancl Warrington are
destroyed by either wind or water, and a great many are wwsht away.
The total known deaths are 32.
The estimate11 damage is as follows: to buildings in Pensacola by
wind, $60,000; to interior finishings and stock by rain, $100,000; to
buildings and wharfs, timber and lumber by tide from the sea, wind,
and wave, $775,000; t,o shipping, $1,iJOO,OOO; to telephone, telegraph,
-. 420. MONTHLY WEATffER R E m . SEPTEMBEB, 1906
v
FIG. 2.-The effect of the storm at Pensacola, Fla.; view looking northeastward along Baylen street, showing the jam of craft of every description,
timber, and lumber.
electric light, and trolley wires, $20,000; to foliage of the city, including
cost of clearing streets, $50,000; to railroads by tides from the sea and
washouts by heavy rains, $125,000; total estimated damage in Pensacola,
immediate anchorage, and water front as stated above, $2,120,000. In
addition to the above we have an item that can hardly be estimated; that
is, the washing away of lots along the water front caused by the water
rushing out t o seek its level in the Gulf after the storm had past from
the water to the land surface; in many places there were embankments
eight feet high where now there is a depth of water measuring five feet.
In the vicinity of Pensacola the damage is estimated to be: to navy-yard,
forts, etc., $l,lOO,OOO; by tides along the shores of East Bay, Escambia
Bay, etc., $125,000; by tides in St. Andrew Bay, 12 inches above any-
thing in past nineteen years", $5000; by tides and wind at Apalachicola,
* Fla., $12,000; by winds and rain at Goulding, Fla., $13,000; to timber
and turpentine interests in Escambia and Santa Rosa counties, $40,000.
The estimated damage by wind and wave, and height of tides at
Pensacola during the severe storms of the past twenty-seven years, is as follows:
Tide above
-I
Date. ". . ___-____
.e .
....
. _I .
.' . Feet. October 2, 1893. ................................................... $1,100 ............
Julr 7, 1896.. ....................... : ............................ 409, OOO 5.0 August 16 1901.. .................................................. 112,000 ............ Februaq"27,1902 ................................................. 1 100 ............
October 8, 1894. .............................................................. 4.5
September 26-27,1%. ............................................ 2,120: OOO 10.0
The following tidal heights from Pensacola eastward will give an idea
of the wave that attended the storm of September 26-27,1906: Pensacola
Bay, 10 feet; East Bay, 9 feet; St. Andrew Bay, 6 feet; Apalachioola Bay, 5 feet.
During the past twenty-seven years there have been eighteen days on
which the wind attained or exceeded a velocity of 50 miles an hour at Pensacola; of these the maximum velocities of 60 miles, or more, were
the following: August 20,1888, 60, SW.; October 2, 1893, 66, sw.; Octo-
ber 8, 1894, 68, ne.; July 7,1896, 72, n.; August 15,1901,70, sw.; Decem-
ber 28, 1901, 60, SW.; September 27, 1906,83, e.
The Pensacola Journal, September 30, 1906, remarked:
The efflciency of the service of the United States Weather Bureau in
giving warning of approaching storms was never better demonstrated
than during the recent hurricane, the first warnbg of which was re- ceived and sent out by the observer at Pensacola on September 22, four
days before the gale became dangerous in this vicinity.
Mobile, Ala.
storm of September 26-27 was more destructive than any other in
the meteorological history of the station, involving a greater loss of property, more numerous marine disasters, and greater destruction t o
timber. The storm approached this section without any o p t i d premonitory
signs or noticeable cloud formations. Cloudy weather with stratus clouds
began on the morning of the 25th, and there was ocwional sunshine
until about noon of the 26th; and only light rain, at times a mere drizzle,
was recorded during a D a r t of this period.
:
On the 26th the fall
FIG. 3.-The effect of the storm at Pensacola, Fla.; view northward from Muswgee wharf, showing the effect of the high seas on the embankment.
FIG. 4.-The effect of the storm near Pensacola, Fla.; fishing smack driven
ashore; damaged cement residence at Palmetto Beach; the waves
broke in the second story windows and washt the foundations from
under the veranda.
barometer pressure and the increase in wind velocity, which had been
gradual, became more rapid toward night. The wind, which was
remarkable for the severity and suddenness of its gusts, reached a maximum velocity of 95 miles an hour between 5 and 6 p. m. and
higher velocities were attained during succeeding intervals until a max-
imum of 55 miles, from the east, for five minutes, was reached at 735
a. m. of the 27th, and a single mile was made in somewhat less than a
56-3
minute; after which lower velocities were recorded and no high velocities
occurred after 11 a. m. The direction of the wind varied from north to northeast till midnight of the 26-27th, then it was northeast till 6 a. m.,
then east prevailed with momentary gusts from the northeast and southeast till 8:05 a. m., and southeast afterwards with occasional gusts
from the east. The lowest reading of the barometer, 98.84 inches, was reached at 6:30
a. m. [See fig. 6.1
The rain, which was continuous from 12:05 p. m. of the 26th, at times
became a heavy downpour. The amount which fell from l2:05 p. m. of
the 96th to 7:20 p. m. of the 27th was 6.40 inches.
During the morning of the 96th, the tide reached within six inches of the top of the wharf; it then receded and was one or two feet below the
level of the wharf till about 6 a. m. of the 97th ; a rise then began and in about half an hour the water was near the top of the whari; by 7:45 a. m.
i t had come into the third street from the river, and at 10 a. m. i t at- tained ita maximum stage, exceeding by about a foot the stage reached
during the hurricane of 1893. During that memorable storm the water
was about six feet above the wharf. I n the hurricane of 1901 the tide
did not reach so high a stage 88 in 1893. Markings at the Cotton Ex-
change Building indicate 13 inches higher water in 1906 then in 1893,
and these differ but slightly from the average of measurements taken by
Mr. Wright Smith. Mr. Smith’s figures for the high water occasioned
by the laat three storms are as follows: 1893, 8.80 feet above mean tide;
1901, 8.95 feet; 1906.9.87feet.
The following is a list of storms with wind velocities of over fifty miles
an hour which have occurred at Mobile since 1872:
August rO,1888.-Mdmum wind velocity 56 miles from the southeaet;
water 3 feet deep over wharfs; estimated damage, $10.000.
April 14, 1889.-Maximum wind velocity 60 miles from the southeast;
estimated damage, $25,000.
October 2, 189J.-Maxhum wind velocity 72 miles from the southeaet;
several vessels ashore; wholesale business district inundated, and stage of water said to be higher than at any time since 1859; seven lives lost in
Mobile County; estimated damage by tides, $100,OOO; by winds, $50,000;
storm more destructive on the Mlssissippi and Louisiana coasts.
August 15, /901.-Maximum wind velocity 60 miles from the southeast;
a number of small vessels lost; wholesale business section inundated;
no loss of life; estimated damage by tide, $75,000; by winds, $25,000.
September 2%-27, lg06.-About twenty buildings, mostly houses in the
residential section of Mobile, demolished. Nearly all buildings were more
433 MONTHLY WEATHER REVIEW. SEPTEMBER, 1906
or less damaged. Windows were blown in, chimneys felled, tin roofs
rolled up, slates and shingles ript off so that few interiors of houses
escaped damage by the rain. I n some places heavy timbers were car-
ried considerable distances. Many merchants in the wholesale district
had elevated their wares, but the tide exceeded all previous stages and
damaged the lowermost goods. All electric services were totally crip-
pled, the telegraph wires being down by 3 a. in. of the 27th. Th+i rnads
were made impassable by prostrated trees. Theie was only one life lnht
in Mobile. The wharfs were greatly daniaged and shipping suffered
considerably. The official list of American vessels wrecked, kept at the
office of the collector of customs, ib not yet conipleted. Unofficial records
of all marine disasters show that in Mobile Bay and Riler there were 11
steamships, 17 barks and schooners, and 4 steamboats ashore or sunk,
I2 tugboats sunk, ashore, or capsized, and numerous barges and bmaller
craft which met a similar fate. There ale also 11 steamships, l)arks,
and many smaller vessels dismantled or otherwise damaged. In Mobilt.
the estimated damage to buildings and electric services by the wind is $500,000 and by the tide, 850.0UU. The ilamage to merchandise IJY tlie
tide is 8175,000. and by t h r rain and wind, $250,001). The damage to
vessels owned in Mobile is $175,IJWl. The damage to railroads entering
Mobile is estimated at 8500,000.
26 27 28 -r- 10 XI1 2 4 6 8 10 M 2 4 6 8 10 XI1 2 4 6 810 M 2 4 6 8 10 XI1 2 4
FIG. 5.-Barogram, Pensacola, Fla., September 26-28, 1906.
26 27 28
---
l O X I 1 2 4 6 8 1 0 M 2 4 6 8 l O X l l 2 4 6 810 M 2 4 6 8 1 O X I I 2
FIG. (;.-Barogram, Mobile, Ala., September 26-25, 1906.
The warnings of the Bureau prevented a greater loss of life, as nearly
all vessels remained in the ricer. However, the storm was of so grt'ilt severity that ships were torn from their iworings, collisions occurred,
and the wind alone dismantled or damaged others. Till about 3 it. in. of the 27th. the office telephone was in almost continuous use; and as
the barometer pressure continued t o decrease, the information was
given that the severity of the storm would increase. Shortly before 1 a. m., when the barometer pressure began to fall rapidly, the addi-
tional information was given that high tides were probitbk altho the
northeast wind had prevented high water. The impracticability o f
securing labor during the already stormy conditioiis, the belief o f many
that only southeast winds can cause high tide, the absence of light in
the early morning hours, and the subsequent rapid rise oP the tide to a
stage exceeding past records, greatly increased the amount of clanmge
by water.
All localities in southern Alabama and southeastern Mississippi sus-
tained great damage to property, but in the southern parts of Mobile
and Baldwin counties the storm was most destructive to life ancl prop-
erty and the damage will probably reach a million dollars. On the southern coast of Mobile County about one hundred persons lost
their lives, mostly by drowning in the high tide which swept the coast.
Within a radius of 70 miles from Mobile about one half of all the
timber boxed for turpentine production and from 5 to 35 per cent of all
other timber has been blown down. It is not yet known how much may
be utilized, and i t is impractioalJle to approximate the loss as i t is
variously estiniated from ten million to one hunilred million dollars.
At Fort RIoigau an anemometer, which owing to its worn condition
registers vrlocities 5 1ber cent too low, was in teiiiporary use during the
bturm. During the eight hours ani1 forty-eight minutes ending a t
P:45 a. m., at which time the aluininuni anemometer cup5 were whirled
off the steel arms, it recorded 50l)mile<. Several times after midnight a
maximum velocity of 85 miles was recorded for a period o f five minutes;
3 mile5 were remiriled in two minuteh, which is at the rate of 90 miles
an hour; this, with a 5 per cent coriection added, is more than 94 miles
an hour.
Lieut. B. L. Brockway, of the U. S. Revenue steainer 1T'ino~n,
which was a t Scrauton, Miss., d u r i n g the storm, sends the fol-
lowing meniorancla, thru the observer at Mobile :
Inches. .................... Re) Dt P II llrer 2.5-26 29. R2
Selnteniher 26.. 8:UO :c. 111. 29. 80 ................................................ 111,1111. 2Y. 7 i
................................................ X:01) I. 111. 29.64
............................................... iiiiolu 29.45
.............................................. 4:oo PJ. 07
............................................... 4::w 29. n:!
28.96
.............................................. 6:11J 28.91)
.......................... 5:3U a. 111. 28.75
110.. ............... .................................. 6:UU 8.111. ' 2s.66
.......................... 6:XU S. IU. , 28.60
.......................... I 6:45a. 111. . 34.55
..................................................
................................................ 7:oo a. 111.
................................... ... Do ........................................................ I 1 I D(1 u.
............................................... Septeirrbrr?i-?S.. l ~liiiluiyht. 29.00
From the lieginning of the blow till 7 a. In. the wind was north-north-
east. The tide ran f l t b o i l (oliposite to the direction of the wind) until 7
a . i n ., when the barometer was a t its lowest. For about fifteen minutes
after 7 a. in. there was a lull in the wind; then i t shifted to the south-
southwest and lilew nith terrific force. The barometer began to rise,
and tlie title ran ebb with tlie bame force i t had run fliwd, the tide again
running in the direction opposite to the winil.
The facts, to ins mind, fully prove the theory that the lorn barometer
made the high tiile, and that the wind did not force the tide in, as is the
u+ual Ilelief. In each case the wind was blowing with terrific force in
the opposite direction to that in which the tide was running. Purther- more the tide ran flo1~1 as long a s the barometer dropt and when the ba-
rometer began to rise the tiile began to run ebb. Do you think that tlie low Imrometer aud the lull in the storm indicate
that we were in the center of the storm at Suranton?
The destructive tides that attend s t o r m s on our Atlantic
and Gulf consts are clue in part to w h a t are kaomn as storm
waves that are projected from the vortexes of atmospheric
clistnrbances in which baroiiietric gradients are very steep,
a i d depend p a r t l y upon the force and duration of the wind
from certain points of the compass. In the case of the storm
of September 17 on the South Chrolina coast, the duration of
the onshore winds was too brief to cause a dangerous tide
before the wind shifted to westerly. In the storiu of Septem-
ber 3ti-37 on the middle Gulf coaqt, places on and near t h e
coast west of the point where the Tortes of the s t o r m crost
the coast line experienced only northerly. or ofishore, galeb;
and the storm wave alone, as of the South C'arolina storni, was
not in itself especially destructive. To the eastward of the
path of the storm, however, low-lying coastal tracts were sub-
jected to the s t o r m wave p l u s the waves driven in from the
Gulf by a southerly, or onshore gale.
In tlie R~INTHLP WEATHER REVIEW f o r May, lSTtG,' Gen. E. P.
Alesander. of Georgetown, S. C., gives results of a s t u d y of the
destructive forces on land of tropical s t o r m s that reach the
Gulf aud Atlantic coasts during the hurricane season, and
s u g g e s t s means of protection against the winds ancl waves
that sometimes attend these storms. The following are
extracts from the article referred to:
The maxinium velocity of the hurricane wind has been known to
* Vol. S S I V , pp. 153-6.
SEPTEMBER, 1906.
Bostou, Mnss.. ....................... Saudy Hook, N. .I .................... Fad Monroe, Va ..................... so11t.h Island, S. C ... ................
MONTHLY WEATHER REVIEW.
F ~Y ~. mp.. April 16, 1S51.. ................... 5.3 1.5
Yrptrmher 10. lYS9 .... 14
hlawli IO. 1.46 ........ 12
Octdwr 19, 1893.. ................. G.8 2
423
l i s t 27. 1893.. ............... 6.4
1 1 4 ?i, lS98. ................... R. 2 Iwr 2. 1S9.3. ................... 7.0
iary !I, IBRH ............ ember ?S, lSY.5 .......... 1
exceed the rate of 120 miles an hour, but this is only in puffs of a few
seconds' duration, as the total movement of the wind for a whole hour
rarely exceeds 60 miles. Now, wind pressure is usually estimated at 2
pounds per square foot of surface when blowing perpendicular to that
surface with a velocity of 20 miles per hour, 8 pounds for 40 miles, and
18 pounds for 60 miles, the pressure increasing as the square of velocity.
[It will be observed that during the Gulf storm of September 26-27, 1906,
the wind maintained avelocityof 70 miles at Pensacola for a whole hour.]
If we assume the highest velocities and calculate the pressures by this
rule, we should expect few ordinary houses to resist them. But in the
wake of a storm, a study of the structures which fail and of those which
resist is generally calculated to surprise an observer far more by the
apparently weak ones which have resisted the winds than by the appar-
ently substantial ones which have failed. And when those which have
failed are examined, i t will be found, almost invariably, that failures are
due to unstable foundations or to lightly attached roofs. In fact,, i t may
be taken as a measure of the force of hurricane winds that the frame of
any ordinary house will resist them. But the foundations must be firm
and the roofs fairly well framed and attached. In new houses, by the use
of wooden ceilings instead of plastering, and a few angle irons ail11 bolts,
one can easily have a structure like a double box, which could be almost rolled over without injury. Old houses, badly constructed and with
poor foundations, may be easily preserved by a few stout braces or inclined props on sides opposite the wind. In short the wind of a cyclone
by itself seldom works serious injury. It is only where i t has the water
as an ally and accumulator of its forces that i t s ravages are great.
When a hurricane passes inland i t soon becomes little more than a bit
of very bad weather. I t s great instrument of destruct.ion is the so-called
tidal wave or storm tide, or, more properly, storm wave, whicli is raised
by i t and which submerges the low laud of the coast. Below the limit
to which these waves rise is the zone of danger iu a hurricane; above i t
is the zone of easily attained safety. How far this danger line may extend above ordinary high water
depends so largely upon local configuration of coasts that i t is only to be
determined for any locality by observatiun. Unfortunat,ely reliable
measurements and data upon this point are rare and ditficult ti.) obtain.
Popular accounts are always esaggerat,ed, being largely hased upon the
action of surface billows, which send water and drift far above the
general level of the storm wave. A vessel, for instance, drawing eight
feet may be carried by successive billows across a marsh submerged
only four feet beneath the general level. I have read accounts of com-
bined storm waves and high tides rifiing ten or twelve feet abore ordinary
high-water mark, but when the action of billows is eliminated and care-
ful measurements are made, the highest record o f a storm tide above
ordinary high water which I have been able to find anywhere is 8.8 feet.
This limit was reached at Fort Pulaski, Ch., in the great gale of August
27, 1893, which broke all records in the height of its waters, in the
destruction of life and property, and in the measured velocity of its
wind, which at Charleston, for a few moments, exceeded 180 miles an
hour. A s this gale is one of great interest, the reader is referred to the
records published in the MONIRLT WE.4THER REVIEW for October, 1893,
page 297. The following table shows the rise of the tide caused by this hurri-
cane, and for comparison, also, the highest storm tides ever recorded a t
several Gulf, Atlantic, and Lake ports, as shown by records of the
U. S. Coast Survey and Engineer offices.
Highest storm tides at wzrious points.
14
14
XI
fi
I
Locality. Date.
The plane of reference is ordinary high water, and the age of the
moon is given in each case to indicate whether the storm tide coincided
with the normal high tides, which occur a t all Atlantic ports about each
full or new moon. There is no tide at Lake ports, and but little in the
Gulf.
From the above we see that the serious ravages are committed 11y the
water rather than by the wind, and that they are confined to a narrow
zone seldom, if ever, reaching more than eight or niue feet a?JoVe the
plane of ordinary high water. Aliove that zone ordinarily well built
houses will easily resist the winds if the house and the roof are securely
framed together and the foundatious are stable. If there are weak
The tide of September 27, 1906, at Mobile, is reported to have been
about one foot higher than that of 1893.-E. B. Q.
points, even cheap and ordinary props or braces which can be impro-
vised rapidly, are very effective in breaking up vibrations and resisting
the pushes and shakes of the wind. Withln the zone of danger from water, the dash of the waves and the tendency of the water to lift and
float all wooden structures mubt be provided for. The limits of this article do not permit a full discussion of the magnitudes of these dangers
and the various means by which they may be met, but i t may be said briefly that pile foundations, or the equivalent, posts framed into buried
timbers, are at once cheap and efficient.
WEIGHT OF SLEET ON SUSPENDED WIRES, CABLES,
AND BRIDGES.
The breakage of telegraph lines and cables by the weight
of the accumulated sleet, ice, and snow led us some years ago
to ask that observers send to the MONTHLY WEATHER REVIEW
some observations on the weight of deet actually observed in
orclinary aud estreme cases. We nom renew the request.
Please state the size of wire, or cable, and the weight of ice
per linear foot.--l'. d.
RAINY OR SNOWY WEATHER AS FORETOLD
BY HALOS.
It is R well-known fact that rain, snow, and general storms
are frequently preceded by the appearance of halos, and es-
pecially simple circles around the sun or moon. The relation
between thebe phenomena has been carefully studied in Europe,
but I know of nothing especially hearing on this subject in
America. Would not many of our observers, both regular and
voluntary, do well to look orer their past records, and tabu-
late the dates and hours on which halos mere observed, more
especially the 22-degree and 45-degree circles around the
moon aril the sun, with a stateiiient of the weather that fol-
lowecl twenty-four hours later? Doubtless the halo will be a
ii~uch better guide in predicting the weather in some places
than in others.--(% A.
MONTHLY REVIEW OF THE PROGRESS O F CLIMA-
TOLOGY THRUOUT THE WORLD.
By ('. ~I T Z F I U I .I I Tsrnrn~, 11. CI. Weathrr Bureau.
PUBLIC'ATION O F CLIMATOLOGICAL RETURNS FOR THE BRITISH COLONIES.
I t appears from a recent report of the British Meteorological
Coiuiiiittee' that a proposition to provide for the publication,
in convenient aud accessible form, of the abundant climatolog-
ical data now accumulating in nearly a11 the British colonies
was recently considered by the Committee, and rejected on the
score of expense. Following are extracts from a correspond-
ence on this subject that past between the Colonial Ofice
and the Treasury, which latter now has control of the Mete-
orological Ofice thru the newly constituted Meteorological
Coiiimittee :
Lvtter from the Colo,iinl Ofice to l l i r Trensciry.
DOWNINC~ STREET,
5th August, 1905.
SIR: I ani dirwtetl by Mr. Secretary Lyttelton to request you to
inform the Lvrds C:ommissionets of the Treasury that as the result of an
enquiry from the Uuited States Weather Bureau for meteorological in-
formatiou with regard to W'eihaiwei, lie has had his attention drawn to
the absence of any organization for the collection and publication of
mrteorological returns from the colonies generally, and for affording
iufurniation to persons making enquiries as to climatic conditions in
~a r i o a s parts of the British Empire.
It would alipear that to a great extent in response to R circular of the
87th of July, 1895, the Meteorological Oftice receive a considerable
amount of information, as shown in the enclosed print, which could
with a little trouble be largely increased. Owing, however, to the want
of the necessary clerical assistance most of these valuable returns serve -
~ ~
~~~__~~____
1 Great Britain. Meteorological Office. First Report of the Meteor-
ological Comniittee to the Lords Comniissioners of His Majesty's Treas- -- u r i for the year ended 31st March, 1906. London, 1906.