338 MONTHLY WEATHER REVIEW. SEPTEMBER, 1896
sons in the north temperate and arctic regions continued from
six months before to sis months after correqonding cold
winters in the estrenie southern temperate region. The lati-
tudes visited by vessels that round Cape Horn are usually a
little less than 8. GOO, corresponding, therefore, to Behring
Sea, southern Alaska, Ca.pe Farewell, the Orkney Islands,
Christiania and St. PQterSburg, and, in fa.ct, are very little
farther south of the eqnator than the routes followed by the
steamers from Glaegow to North America are north of the
equator. The fact that tliere is so much niore ice in thiise
southerii latitudes t h n in the corresponding northern
tiides must be attributed largely to the winds of the antarctic
regions and t,he distribution of ocean and land. Any special
increase or diminution in the ice, either arctic or antarctic,
must result from a change in the winds; it may be in their
direction, or force, or t.emperature, or moisture. Whatever
the cause 'of that change in the wind, i t must affect a large
portion of the Southern Hemisphere apprecialjly. If, how-
ever, the ultimate cause consist, in some phenoiiienoii peculiar
to the equatorial regions, it may affect both the arctic and
antarctic simultaneously. In fact such conditions may pre-
vail over the whole north temperate zone as to influence the
circulation of the atinosphere in the south temperate and
antarctic regions, a.nd this inflnence may either be direct and
simultaneous, or indirect and only apparent after many
months. *
A consideration of the mechanics of the atmosphere josti-
fies the collection and intercomparison of even the popular
news items in tlie daily press describing special a11d abnormal
phenomena in distant places. As far back as 1750 Dr. E. A.
Holyoke, of Saleni (Memoirs Am. Acad., Vol. 11), made a
comparison between the weather- on opposite sides of the
Atlantic Ocean, hoping to discover some simple reason for
the contra.sts hetween tlie two regions. We, on the other
hand, are now tempted to compare together the weather of
the arctic and antarctic circles. There have opportunely
come to hand a few reports from vessels off the coa.sts of
Alaska aiid Patagonin. A letter published in the San Fran-
cisco Chronicle of October 8 from the fleet of whalers in the
Hehring Sea, dated August 14, 1895, states that-
On sailing northward in July from Uiialasks ice was met within about ICH! miles, and was always present until reaching Port Clarence, and that it had never before been seen so near t,o Uiialaska by niiy one. From Port Clarence to Point Barrow' the vessel's progress was exceedingly slow on account of the driftin ice, scarcely a mile liavin been made in the first fifteen days of lfugust. The northeast win2 t,liat usnally keeps this ice off shore has been wanting.
Dr. Sheldon Jackson, agent for the Bureau of Education,
reports a similar esperience by the revenue cutter Bew while
north of Behring Straits. The southern edge of the arctic
ice pack had remitined so far south as to prevent any passage
iiorth of Icy Cape from July 19 to August 22. Parties froni Point Barrow who had traveled down the coast for their mail,
report that the past winter, 1894-95, had not been very cold,
the lowest temperature being 3 O 0 . (See the National Geo- graphic Magazine for January, 1896.)
We are not to infer from the above that there has been
iinusual cold or an unusual quantity or thickness of ice, but
simply that the wind failed to counteract the ocean currents that drifted the ice 011 shore.
From the Loa Angeles Espress of Nov. 36, 1895, we learn
that the British ship An.gZcse:y arriving at Ran Francisco on
that date, like every other vessel that has arrived a t that city
aft.er rounding Cape Horn during t.he summer and autumn
of 1895, reports an 'unusual quantity of ice in that region,
n.nd corresponding unusual storms and freezing weather and
snow. Usually a northwest wind clrives the antarctic ice south-
ward, just as a southwest wind drives the arctic ice northward.
We may infer that there has, during the past July and Au-
gust, been a diminished tendency to northwest winds near the antarctic circle, while, at the same time, there was a dimin-
ished tendency to northeast winds in northern Alaska..
As the wilds in northern Alaska are associated with the low pressure area in Behriag Sea the abnormal wind condi-
tions indicate an abnormal condition of that low area, and,
in fact, i t was probably a t this time almost entirely obliter-
ated by ail unusual northward estension of the great high
area of the North Pacific, which area also hrought cold wea.ther
to the Pacific Coast States.
As the westerly winds at Cape Horn represent a general cir-
culation around the low area of the Antartic Continent, and as the winds a.t the Cape are usunlly northwesterly, blowing
outward from the high pressure area of the South Pacific, we
must infer that the absence of northwest and the presepce of southwest winds in this region implies a diminished activ-
ity of the high area in the South Pacific.
Now, this tendency to an increased activity of the northern
area is but an esaggeration of what ordinarily takes place in
the changes from January to July. or winter to summer, in
the Northern Hemisphere ; similarly, the tendency to a dim-
inished activity of the southern high area is an eraggeration
of what takes place in that region in the transition from July
to January (winter to summer) in the Southern Hemisphere.
Both these changes are, therefore, in harmony with those
variations in the general circulation that depend upon the
interactions of oceans and continents. The change in the
North Pacific area is that which would be produced by an ip-
creased contrast of land and water in the northern summer, and the change in the South Pacific high area is that due to
a diminished contrast between land and water in the southern
winter season. Therefore, in the north, or summer, the con- tinental air has been warmer than theoceanic air, but in the
south, or it,s winter, the continental has heen cooler than the
oceanic air in the region of high pressiire. Both of these
changes may be plausibly traced back to some one Aingle
cause, such as an increased dryness of the atmosphere, which
makes hotter suniniers aiid colder winters. Although the lat-
ter suggestion may not present the true cause in this specific
case, yet it often may be applicable to similar cases, and it
seems to enforce the general principle that widespread and
persistent seasonal variations of climate may result from a
very slight general disturbance in the quantity 0-f n~oisture
iii the air, or an excees of ascending or descending movenients
in the atmosphere.
THE LOCAL STORM OF SEPTEMBER 8 IN KBNSAS.
The niap of Sunday, September 8, a t 8 p. m., shows a gen-
eral movement of the wind from south and southeast over
Texas, Missouri, Illinois, Indiana, and thence northward to
North Dakota. This movement may be considered in either
of its aspects, viz, either as a flow of air towards a special low
pressure in Alberta, or as a flow toward the high, warm table-
land constituting tlie eastern slope of the Rocky Mountains.
A barometric pressure of 39.7 or 29.8 prevailed over the
greater part of the Plateau Region, a d a temperature of No
or 90° prevailed from Nebra.ska south and east over the Gulf States, with generally clear weather. As there was no
itrongly developed low pressure, therefore the local showers
that occurred, with thunder and lightning, during the 8th and
9th, must be considered, not as an essential part of a system
of cyclonic circula,tions, but as local incidents due more par-
ticularly to special local influences. The locations of such
storms, with reference to the center of low pressure, has but
little significance as compared with their locations relative to the winds and local topography. However, an exception
must he made in respect to the storm that. occurred on the
8th, a.. m., in Morris and Lyon counties, Kana., and moved
thence southeastward to the southeast corner of the State into
SEPTEMBER, 1895. MONTHLY WEATHER REVIEW. 339
Missouri, which was too large to becalled a local storni ; in fact., i t seems to present phenomena characteristic of the southern
end of all long troughs of low pressure. In this case the
trough may be considered as extending, on the Sth, p. ni.,
from central Kansas to southern Alberta. On the west side
the cold, drier air over the Plateau Region was undoubtedly
descending and pushing eastward, but had not yet made itself
felt as a general westerly wind, except a t Pueblo, Santa Fe, and
Amarillo. In this region, which lies a little southwest of, and
higher than, central Kaneaa, such winds prevailed, as would,
by descending to the lower level of that State, produce the
rise of 80° in teniperature that was reported from Wichita
and Dodge City, with corresponding falls in the baronie- ter. This descending stream of air, having once reached
the surface of the ground, and pushing eastward, must have
presented the usual well-known appearance of an onward
rolling or rushing wall of air, lifting up the lighter air in front, topped along its whole extent by a long and compar-
atively narrow belt of what might, to a distant ebserver,
appear as one cloud, but which, 011 closer inspection, is seen
to be a series of. isolated whirls and hailstorms, horizontal
rolls and straight-line movements, all in close jur taposi tion
and presenting a scene of violent, tn+bulent motion. A t first the front was a small region, perhaps 5 miles long in the
southeastern portion of Morris County, and northwestern
portion of Lyon County, and therefore about 100 miles we&
southwest of Kansas City, and was but the front of an enter-
ing wedge for the large mass of air to follow behind. About iioon of Sunday, the 8th, the front was not only moving east-
ward as a whole, but also lengthening by its steady growth
toward the southeast and the northwest, and by 8 p. m. had passed beyond Concordia 011 the north and Springfield (Mo.)
on the east, and was near Kansas City. A description of the
effects of the storm in southemtern Kansas, as compiled from
the WEATHER REVIEW, for that State, woulcl be about as fol-
lows :
On the east side of a line from Lyon Cl0unt.y towarll t.lie southwest corner of Kansas the corn was blown down from t,lie norblieast and on the west side from the northwest to sout.heast, showing the path to have lJeen about 25 miles wide: inside of t,liis was an area of desbmctive l i d
about 1.5 miles wide, while beyond it was the area of excessive rainfall, about Ho miles wide. The hail lay in broad streaks or paths; all liail- stornis [streaks? onthe east side of the centertraveled [trended'?] from
east. I n many of the hail streaks the leaves, fruit, and bark were stripped from the trees, and the next morning the groves were left bare, as in January; 4, 8, and 10 inches of water were nieasured in standard rain gauges, and a box standing isolated near the center of the storni path, 14 inches deep. was full of water on the morning of the %*h. This storni was succecded by hot, dry weather. The electrical display was quite rem:trkable, and was reported by ovservers in Clay County, 1% miles west of Kansas City, and Reno County, IN) miles southwest of Morris and Lyon counties.
THE EAB'I'HQUAXE OE SEFTEMBER 1.
This earthquake shook the whole of the region IJet\\'Pw
Virginia and Maine, but did not extend west of the c~ntral
portions of these Atlant.ic States. The time a.1~1 c1inractt.r
of the shocks are briefly given in the following mnimary :
li'~yinin.-Falls Church, lasted several seconds. District oj* CoZ,i~itib.irr.-~~nshingtoii, 6 hr. 8 min. 39 sec., n. ni.,
Ly the Weather Bnreau seismograph.
Jfu~~yZa,ad.-Westniinster, 6.05 ; distinct trenibling of the
house for a few seconds.
DeZuic~ure.-Wilmingtoii, slight tremor ; duration one sec-
ond, a t 6.10.
Ntw Jersey.-Perth Amboy, 6.08 ; lastecl from t h e e to ten
seconds. Beverly, 6.10; lasted a f u l l minute. Plninfield,
6.01 ; lasted ten seconds.. Englewood, 6.10. Rsiicocas, 6.0s.
New Brunswick, 6.07. Moorestown, Newark, Clinton, Bridge-
ton, Oceanic, Elizabeth, Bayonlie, Alaire, Somerville, Burliog-
ton, Palmyra, Freehold, and Toms River, recorded without
northeast to sout L west and on the west side, from northwest to south-
details. Red Bank, not noticed by people out of doors, but
distinctly felt by people inside of houses. Belmar, 6.08; ac-
curately observed hy Mr. Edward Brown. Asbury Park,
loose plaster fell from the ceilings. Sandy Hook, in the
Marine Observatory the tower ehook slightly for about eight
seconds. Jersey City, 6.10, a t the Erie Railroad depot. Eu-
glewoocl, 6.10 esactly ; oscillations lasted froni eight to ten
seconds. Port Jervis, 6.08. High Bridge, perceptible shock.
Elizabeth, 6.16. Plainfield, 6.M). Trenton, between 5.00 and
6.00. New Brunswick, 5.50. Morristown, 5.45. Orange, not
noticeable. 0 .
PeiiiisyZ.unlzicc.-Enston, chnrch bell rang shortly after 6
a. ni. Philadelphia, shortly after G a. in. ; huilclings swayed
perceptibly. Alleiitown niid I'licPnixville, lmildings swayed
and many awakened from sleep.
New I'ork.-New York, not so severe as that of August 10,
1884, August 31, 1886, or March S, 1593 ; Weather Bureau
office, vibrations for ten seconds a t 611. 11ni. a. ni. Staten
Island, a strong, single shock, lasting ten seconds, occurred
at 6 hr. 10min.a. in. Governors Island, the ordinance sergeant
was awakened by the shock. New l'ork Harbor, the ferry
boats rocked as in heqvy weather, and II rumbling sound
occurred as if the boats were scraping on the bottom. Long
Island and Brooklyn, three distiiict shocks ; the earthquake
generally travelled eastward, and was felt most decidedly on
the south side of the island. Rockaway Beach, tlie sand con-
tinued in motion about two minutes, and the siirf suddenly subsided. Bath Beach, two persons are said to have been
thrown out of bed. Jamaica and Newtown, 6.05. Flushing,
about 5.45. Northport and Mount Vernon, albout 6.00. Ton-
kers, a little after 6.00. Sing Sing and
Tarrytown, slight shock. Newburg, tliree gentle shocks felt
at Washington Heights. Bedford, 6.06.
Con ticcticut.-Greenfield Hill ( from the Bulletin of the
New Englancl Weather Service) :
Uncorrected time, 6.08 a. m.; duration not over ten seconds: direction of niovement nearly southwest to northeast. Only a few people were uut at that hour, and not many nf those f e n noticetl the shock. I n hwses it was felt generally, and neltrly all report liriving rctx4ved the i n i y i o n of the wave motion. Being in lied :it tlw time, I felt tlie
roc ing motion; the j:trring motion was f d t faintly. A iwitlrnt of the valley, about half a mile away from Iierc, s:iid his Iiunsi. was bndly sli:tkt.n. The shock passed away with a rnnring sound xinii1:ir tn that
.rent back from water wlirn a heavy gun is fired o\ (11. it.
White Plains, 6.15.
Coji twct kilt.-westport, houses swayed.
With reference to the acciiracy of the deteriiiinations of
the esnct time of the earthquake shock nt Washington, Mr. D. T. Maring,. who (in the absence of Professor PIIarviii at
Atlanta) was i n charge of the Instrument Rooiii n t that time,
wri tea follows :
The time of twtliquakr shocks is rrcordril as follows: On :t register
L' liiitlrr, iiiaking LL revolution every six honrs, tin ordinary ofice .' reg-
n!kr" clock mwks off 5-niinutts intt.rvale IJY closing II circuit tlirougli
n nisgnet of tlie registcr h n i t fuur xewnds after the ending of e:ich
tifth ininute of the hour. Whenever the needlis of the seismograph is disturlwl. it also records by closing a circuit tlirougli the same magnpt that m:tkrs t l i r dock record. ('areful interpo1:rtion between the clock nn(l seismograph records will, tlitwforr, give the time by the " regu- lator " cltJck at whicli the needle of tlir seisinogrnph was alisturbed. The accuracy of the interpolatiim lii*tween tlir. dock :tiid seisiiiopraldi records will depend upon the clc.finitcness ani1 fineness of the lines inade 1)y the recording IJ~W. In thi# particular instance the case is s~Jl11ewll~~t complicatecl because the seisniograpii record followed so closely after the clock rrconl that the two are united into one wiclt. mark, excepting at tlieir points. The niarks representing the clock rec- ords corer a space eyuimlent to :tlmut forty seconds of time, the tari:t- tion in width nnt being greater tlinn five 8econils. The conihineil width of tlie two marks, as rccordeil in this instance. is eyuiwlent to about sixty-five seconds of time on the sheet. while the tlistance between their points is equivalent to about twenty-five seconJs. The time of the vibration of the neeillc. of the seismograph here rrcorderl alJlJear8, therefore, to have been made wit.hin fire seconds of twenty-five sec- onds afkr tlie clock record was made, or at ten minutes twenty-nine sec- onds after 6 o'clock a. m., of September 1, 1895.