OCTOBER, 1897 440 MONTHLY WEATHER REVIEW.
excellent in the afternoon may weaken materially during thc
night. I found it advisable a t Clark University, some year1
ago, to keep an extra set of dry, clean Mascart table insula
tors always in readiness. Almost every morning, after tests
the insulators were exchanged with those under the collectoi
which stood in a wooden shelter, projecting from a second,
story window and exposed to the weather. Indeed, a verj
good hygroscope might be devised upon this variation of in. sulation with humidity. The accompanying diagram show
the potential curve obtained with a photographic Mascari
electrometer and outfit. It is interesting to note the fall ir
the potential with the occurrence of fog. This may have
been a natural fall, for in the diurnal curve a minimun:
occurs about the time of sunrise; or i t may have been, and
in all probability was, brought about by the decreasing insu- lation. There are, however, dieerent kinds of fog; so also.
do we find different values of the potential a t times of fog.
and to some degree the electrometer takes note of the differ.
ence. Generally with fog, haze, and smoke there is a rapid
fall from high positive to low positive or negative values.
Dust whirls, because of frictional effects, give marked dis-
turbances. But fogs, after fine drizzling raiu, as a rule, give
high positive values. On October 4, 1886, a t the top of the
Washington Monument during a light and seemingly dry fog,
I obtained potentials of over 700 volts, and had no difficulty
in drawing small sparks. Simultaneous observations were
made at a station 460 feet lower, and potentials of 140 volts,
on the average, were obtained. These were high values for
this elevation and exposure. On the day following, about the same iu general character except for the fog, the potential
values were but one-half the former at both stations.
The great problem then in atmospheric electricity has been
to preserve a high insulation in the apparatus. There is an-
other question, viz, the breaking down of the air itself. For
example, after the passage of an electric current, as in the
case of a flash of lightning, the particular path is defective
thereafter and allows easy escape of high surface charges.
Then again, as Hallwachs, Elater, and Geitel, and others have
shown, a negatively charged surface will discharge into air
under the influence of ultraviolet light. What is of sigui6-
cance to the meteorologist, in these investigatione, is the appar- ent relation, as Schuster points out,’ “between the ultra-
violet radiations and the amount of aqueous vapor present in the air.”
Much could be said concerningthese potential falls and the
accompanying phenomena; but the object of the present
paper is rather to trace the dependence of the industrial
applications of electricity upon atmospheric conditions. More particularly the point is that slectrical development is
reaching a limit because of unsuitable insulation. I n other
words, so rapid has been the advance in the use of high poten-
tials that the electrical engineer to-day is very nearly in the
position of the experimenter in atmospheric electricity. He
has to deal with potentials of enormously high voltage, and
realize8 that the successful transmission of the same will be
‘Lecture before Royal Institute, February 22,1895.
determined largely by the insulation, and this in turn is de-
termined by the weather.
Electrical joiirnals are largely filled a t present with ac-
counts of the transmission of electrical energy from moun-
tain strenine to t o m s and cities many miles distant. For esaniple, a power company with a plant in the mount,ains,
coutracts to deliver a t a city 25 miles away, a three-phase
alternating current of 2,500 volts pressure. Through the agency of step-up transformers the potential is sent up to
11,000 volts and transmitted thus,niile after mile in the open
country. Reaching the city, it is transformed don7n, first to
2,000 or 2,500 and then again to 100 or less. In reading the description of such a system, one is attracted first by the
step-up trausformers. These cast iron, oil-filled, water-jack-
eted boxes are more or less a t the mercy of the weather. Sec- ondly, the large copper wires carrying the high potentials,
must be very well insulated. As we read of triple-petticoated
insulators, we recall the earlier struggles of the investigator
of atmospheric electricity.
In further applications of high potential electricity then,
there will be limitations because of atmospheric conditions.
Will the meteorologist, with hisknowledge of the physicsof the
air, come to the rescue and point out the proper directions for experiment with the aim of overcoming present difficulties 1
That the views set forth above are not exaggerated, the fol-
lowing editorial from the Journal of Electricity of May, 1897,
will prove:
* * Science has well in hand the control of lightning, of static effects, arid of the mechanical features that give permanence, safety, and reliability to overhead transmission lines, nor is difficult expe- rienced in the handlin of line voltages at present used, but tKe brr- rier-thus far insu erafle-to the em loyment of higher potentials is, as stated, embodies in the insulator arone. Climate exerts a potent influence on the reliability of insulators, and so great is this influence that to it rather than to the insulator itself ins be attributed the existing limitations to the commercial use of hi$er voltages. In addition to being a practically perfect noncon- ductor, the insulator should be wind, rain, snow, sleet, dust, and insect proof. Its reliability as P sure preventive against all possible sources of trouble should be absolute, for if susceptible to breakdowns, ita weakness necessitates the duplication of the traiismission lines that no interruption to service may occur in making repairs. The saving in co - g.r which will result from the invention of r perfect insulator wouh e twofold in that it would enable the use of higher voltages and re- quire only a single transmission line. Truly mav it besaid that unlim- ited reward awaits the inventor of a perfect high-tension insulator.
ATMOSPHERIC ELECTRICITY : ITS ORIGIN, VARIATIONS,
AND PERTURBATIONS.
By Prof. MARCEL BRILLOIJIX, of the kook Normale Sup6rieure of Paris.
[Translated from the Revue G6nCrale des Sciences pures et AppliquCes, August 30, 1897 ; and from Ciel et Terre, October 1, 1S97.1
The origin of atmospheric electricity still remains quite
tinknown ; the suggested theories all rest upon properties
that are hypothetical or even contradictory to experience.
It has long seemed to me that the action of the ultraviolet
radiations from electrified bodies furnishes nn entirely satis-
factory explanation, assuniing only that ice acts in the same
way as metals; as to this latter assumption I was able to
aatisfy iiiyself during this past winter.
The following lines give a brief synopsis of this physical
theory of the electric phenomena of the atmosphere :
1.-In 1887 Herz discovered that the electric discharge
Dccurs more easily under the action of ultraviolet light than
in the dark. In 1888 Wiedemann and Ebert showed that
this action takes place a t the ‘Lcathode” or negative elec-
trode; that its effect is a maximum when the air is under a
pressure of about 300 niillimeters of mercury; according to
Arrhenius the maximum occurs when the air is under a pres-
3ure of 6 millimeters, but according to Stoletow, a t a pressure
that varies very nearly in proportion to the intensity of the
alectric field.
-0-
OCTOBER, 1897. MONTHLY WEATHER REVIEW. 441
A n attentive study of this action has shown that all metal
surfaces ehargcd with negative electricity lose this electricity when
they are exposed to ultraviolet radiations, however feeble the
negative charge may be. The action upon positive electricity is null.
M. Rhigi and M. Stoletow have even been able to make use
of this action for the purpose of nieasuring differences of
potential a t the very point of contact of two electrodes. 2.-M. Buisson who has tested this most delicate action of
the iiltraviolet light has, a t niy request, made a series of ex-
periments upon ice .as conipared with zinc. A beam of ultraviolet light (electric arc between aluniiniitii
electrodes) traverses a perforated brass plate that is positively
electrified and then falls upon a block which forms the nega- tive armature of tha condenser. This block of ice rests upon
a metallic plate having insulating feet and is connected with
an electrometer. A t first the ice and the electronieter are
put in electric conimunication with the ground, whereby
they are brought to the same negative potential and then this connection is broken. As soon as the beam of ultraviolet
light is thrown upon the ice the needle of the electrometer
nioves and shows that the block of ice loses its negative electricity until the potential of the ice and of the brass plate
are equal.
The action upon a block of dry ice, when just taken from a
refrigerating mixture, is very intense (of the tenth to the twentieth order relative to the action on zinc). As soon as
the surface of the block begins to melt tlie action of the ultraviolet light decreases very much ; finally, when t.he
water resulting from melting covers the whole illuminated
surface of the block the loss of negative electricity becomes
inappreciable. Such are the results that were obtained during this winter
(1896-97) by M. Buisson in the physical laboratory of the
b o l e Norniale at Paris. Ice is very sensitive to the dtraviolet rccdiutio~e; wdcr is
nonsensitive. 3.-Wheii we consider the unquestionable influence of the
dimiuutioii of pressure upon the action of light and on the absorption of the ultraviolet solar light by our atmosphere,
these laboratory results transforni my hypothesis as to the
origin of atmospheric electricity iiito an experimental theory
worthy of publication.
If, at any time whatever, an electric field exists in the
atmosphere, the ice needles of the cirrus clouds become elec- trified by induction, positively a t one end, negatively a t the
other. If now, the negative extremities of these ice needles
should receive ultraviolet solar radiations, the needles t h u s
illuminated would lose all their negative charge and remain
electrified positively. Th.e qaeutral or negative state of th.e cirri i s tciastabla; every cir-
ms cloud ihn~incited by the sun becomes posdtit,e.
4.-Furtherniore, experience has shown that the a.ir thus
illuminated by sunlight remains a good insulator (contrary
to what occurs with the Roentgen rays). In the laboratory
experiments, where the positive conductor is very near to the
negative conductor, the transportation of electricity by the
movement of the air is rapid. In the atnlosphere it will be
quite otherwise. If the negative electricity lost by the ice needles is de-
posited in the surrounding air, then the cloud, as a whole,
appears positive when the ice needles become separated from
the surrounding air. The neutral state of the air is unstable. The air which
traverses a region where illuminated cirri are formed becomes
negative. The neutral air into which a positive cirrus cloud
has evaporated into invisibility has become positive.
In the forniation of cirri by mixture there frequent.ly occur
independent movements of contiguous maases of air-some
REV^
cloudy and others clear. The negative air will then separate from the positive cirrus.'
If the ninss of negative air descends, and if continuing
negative (for the electricity can not be destroyed) it reaches
the cultivated soil, the innumerable points of the blades of
grass and leaves will facilitate the discharge of electricity
between the earth and the air. Hence, the soil of a continent i s charged neyatively by exeha?i.ge with th,e air.
A t the surface of the ocean there occurs nothing similar
to this; the descending air remains negative; it becomes
saturated with vapor, but when this vapor ascends and by expansion is condensed into fine droplete these little drops,
like fine points, borrow their electric charge from the air.
Hence, th.e czimdi .formed by eqjansion over the oceanic wgions
are aegatim
At the surface of the ground no direct action of the ultra-
violet radiations is perceivgd, both because these radiations
scarcely reach it, and because there is no appreciable amount
of mater there, and because the pressure of the air iR high.
5.-It seems useless to dwell upon the characteristics of
the diurtial variation of atniospheric electricity and compli-
cations that the transporta,tion of electrified air may pro- duce. The influence upon thunderstorms is evident; the
same gugt of wind that brings rain and showers a t night
brings thunderstornis a t the close of the daylight, when the
solar action has electrified the cirrus clouds, and when coii-
vection has curried off t.he negative air. The slowness of this convect.ion also explains the two or three days of threat-
ening weather which, in our climate (of France), generally
precede the storm itself.'
In regions or seasons where the air is nearly calm (as at the boundary of the cone of the circumpolar shadow of the
earth during the winter season) there is eo little convection
that the cirrus, electrified positively throughout its en tire niass during the daylight,, remains surrounded by negative
air. As soon as the d n r k ~l ~s of nighttime prevails the stable
condition changes : the electric discharge between the nega- tively electrified air and the positively electrified needles of
ice permeates the whole mass of the cloud. This explana-
tion harmonizes perfectly with all tlie details of the aurora
borealis; i t applies also to t.he luminous clouds sometimes observed in Europe and to the diffllse light observed during
~uniiiier evenings arid called heat lightni,jig.J
6. Finally, bhe iiiecha.nisni of the action of solar disturb-
nnces, 8s seen froin this point of view, becotiies very simple.
Every variation of brilliancy in t.he nl traviolet light from the
sun has an ininiediate action upon auroras and atmospheric electricity wherever cirrus clouds exist; but its action on
thunderstornis may he rebarded for several days wherever
the cumuli existing below the cirri are neutral or nearly EO.
The necessity of having cirriis clouds, eithm preexisting or in
process of forma.tion, and of liavit~g cumulus clouds localizes
this action of ult,raviolet rays upon thunderstorms in a man-
ner that varies with the general meteorological conditions.
The iniportance of the perturbations that are produced in
the earth's atmosphere has no relation to the apparent im-
portance of the solar spots and faculre, but depends exclu-
'See memoir on Winds and Clouds by M. Brillouin, to be published in the Annals of the Central Meteorological Bureau of France for 1898;
see the summary on page 437 of this number of the MONTHLY WRATH= REVIEW. This idea that positive cirrus and negative air will separate from each other needs to be substantiated by some very convincing experiments.-C. A.
* In the United States these days of threatening weather represent niore properly the movement of the whole disturbance eastward; at other times they represent the gradual accumulation and spread east- sard of thickening stratus haze.-C. A. There is no longer m y doubt that the so-called Real ZigMning is gen- erally produced by distant thund~rstorms whose lightning illuminates the sky.-A. Lancnster. [The visibility of auroral beams in daylight is not explained by Bril1ouin.-C. A.]
442
January ........................................ February.. ...................................... March ............................................ A ril ............................................. ay., ............................................ June.. ............................................ July .............................................. August .......................................... September.. ..................................... October.. ........................................ November ....................................... December
2
........................................
MONTHLY WEATHER REVIEW.
IMIUS.
6. T3 9.64
8. m 6. ft2 4.19
1.31 6.78 6. U 6.09 6.89
4. n
7. ia
OCTOBER, 1897
Imhsr. 8.86 6.99 4.40
6.69 6.49 8.69 1.89 8.16 6 .S 8.88
5.37 6.34
~ ~ ~
sively upon the intensity of the ultraviolet radiation that is
transmissible through our atniosphere. Thus, the faculs, and
especially the spots seen with the naked eye, are only imper- fect indicators, and it is greatly to be desired that observa-
tions such as those of M. Deslandres' should be organized
and published systematically.
7.-Other phenomena, such as the breaking up into drop-
lets of water falling upon any obstacle, have for some years
past been mentioned as having to do with the production of
atmospheric electricity. I believe that they play only a sec-
ondary r81e as a disturbing action, and that the fundamen- tal r61e is that which I have described above. I n general:
Atniosphere'c electricity i s maintained hy the action qf ultra- violet solar radiations wpo?a the ice needle8 of the c i m u clouds.
To the same cause is dzie the ~e c e s s a y initial electric je2d that i s inevitably produced by the relative displacenwnts qf the
upper ntasses of the atmosphere in relation to the magnetized ter-
restrial globe.
THE! AREA OF HEAVY RAINFALL IN THE SOUTHERN APPALAUHIANS.
By BAnBT C. HAWKINS. Volunta Observer. Weather Bureau. Station Horse Cove; Post Omce, Hlghlands,%acon Co., N. C. (dated November 8, I h ).
For several years it has been well known by meteorologists
that there exists in the southern Blue Ridge Mountains a
region where the annual total of raiufall is ahnormally large, a t least 70 inches, or more. The literature of the sub-
ject is about as follows: Numerous brief references have
appeared, the most extended being an article in the Ameri-
can Meteorological Journal (May, 1894, Vol. XI, pp; 6-10) by Mr. A. J. Henry, Chief of the Records Division, W eather
Bureau. A brief general statement regarding this area and
its probable cause is to be found in Prof. M. W. Harring-
ton's ideal and exhaustive report on the rainfall of the
United States, published as Bulletin C by the Weather Bu-
reau. Other briefer references are as follows: A statenlent
concerning the portion of the area in North Carolina in the
most excellent report on the Climatology of North Carolina,
published by the North Carolina State Weather Service and
prepared by Mr. C. F. von Herrniann, now director of the
North Carolina Section of the Climate and Crop Service,
Weather Bureau ; also, in Georgia, A Hand Book, issued by the Department of Agriculture of the State of Georgia. It is
not my intention to attempt to give reasons for this area of
heavy rainfall, but merely to present a comprehensive sum-
mary of what is known concerning the geographical distribu-
tion of the rainfall, and also the general statistics, including
the distribution in the different months.
The area included is the extreme southern Blue Ridge, and covers portions of extreme northeastern Georgia, extreme
western South Carolina, and southwestern North Carolina.
The counties included are as follows, it being understood that
only certain portions in the case of nearly every county are
referred to, the records being too few to define the limits
more minutely. I n the case of Macon County, N. C., it is certain that only the eastern part has an excessive rainfall.
The counties marked by a dagger (t ) are doubtful, it being
only inferred that they are within the area.
North Carolina: Macon, Polk, Clay,t Jackson, Transylva-
nia,+ and Henderson.+ Georgia: Rabun, Habersham,t and Towns.#
South Carolina: Oconee, Pickens,t and northern Spartan-
burg.t
Brillouin evidently refers to the numerous memoirs of Deslandres relative to the ultraviolet radiations from the sun, beginning with his first memoir on the ultraviolet spectrum of aqueous vapor and ita rela- tion to the dark lines in the solar spectrum, published in the Paris Comptes-Rendus foy 1885, Vol. .C, p. 854, and concluding with recent papers recording daily the condition of the sun tu to the intensity of ita ultraviolet radiations.-C. A.
-0-
Inch#. 7.B 6.34 7.60 6 .a 4.88 4.69 4.69
S.Q8 4.711
' 6.88 6.11 6.78
~ ~ ~ ~~ ___
The stations where records have been kept are as follows,
those discontinued being marked with a star (*):
North Carolina: Macon County, Highlands, N. 36O 5', W. 83" 11', elevation, 3,817 feet; Horse Cove, N. 36O 0', W. 83O 6',
elevation (estimated), 2,800 feet. Jackson County, Cashiers,+
N. 35O 4', W. 83O 5, elevation, 3,813 feet. Polk County, Co- lunibuR* (now Skyuka), N. 35O 14', w. 8 2 O ll', elevation,
3,000 feet.
Georgia: Rabiin County, Rabun Gap," N. 34O 55', W. 83O 2O',
elevation, 2,168 feet; Clayton, N. 84O 5O', W. 83O 30', eleva-
tion, 2,100 feet. The following stations in North Carolina may possibly be
in the area: Henderson County, Hendersonville,+ N. 35O 17',
W. 82O 27', elevation, 2,167 feet; Flat Rock, N. 35O 15', W.
82O 25', elevation, 8,214 feet. Transylvania County, Brevard,
N. 35O 15', W. 82O 45', elevation, 2,500 feet.
Over all the rainy region the annual total is 68 inches, or
more. The following are the annual average total rainfalls
(in inches) a t stations where the data are a t hand: Rabun
Gap, 68.35 (nine years) ; Highlands, 76.29 (nine years) ;
Horse Cove, 74.99 (five years) ; Cashiers, 78.60 (one year). The distribution of the rainfall by months is very interest-
ing. There is a well marked double maximum and minimum ;
the mrtsimiim is in Fehruary or March and the minimuni in
April or May for the first half of the year; the second maxi-
niuni is in July or August and the second minimum in October
or November.
The maxinia at Highlauds are: 9.64, in February, and a
lesser maximum, 6.04, in September. The minima are:
May, 4.19, and a secoud one, in October, 5.69. I t is seen that the total for the wettest nioiith is double that for the driest.
If other stations had long records for the 8ame term of years,
they would probably have similar features.
The record a t Rabun Gap does not closely correspond with
that at Highlands. The monthly and annual averages for
three stations are given helow, but there are several recent
years of observations a t Highlands which are not included:
d n t h s .
Annual .................................... 1 78.soI 74.991 68.36
The average number of rainy days has been computed for
the station a t Horse Cove from a nine years' record. Some
very interestiug facts are shown. The number of rainy days
annually is 134. This shows clearly that rain does not OCCUT
any qfteirer than i71 regioiis where the fall is 50 inches or less,
but that raiiis are heavier when they do come. The nionthly
maximum is eighteen days in July ; minimum, six days in
October. Surrounding the region of 68 inches, or more, there is a
large area having a fall of 60 inches, or more, annually, and
this area extends to the Gulf Coast ; but this area of 60 inches
does not come within the consideration of this paper. It ie
not believed that the area is continuous, as mentioned above,
but that islands of heavy rainfall exist here and there on the
uplands, with small areas of lighter rainfall on the lower lands, corresponding with the topography. I do not believe
that any portions of the region having elevations of 1,500
feet or less are included in the area of 68 inches, or more, of