DECEMBER, 1903. MONTHLY WEATHER REVIEW. 573
for which we can also write,
4 A But P~~ = is the vertical velocity of ascent, corresponil- zXdaf,"oo
ing to a definite density x,,,, of the air, and to the correspond-
ing buoyancy appropriate to it. With this notittion we obtain
or ap1)roxiIuately -
I' = I %,,
$;.
Hence, the velocities of ascent are inversely as the Gth root
of the corresponding air pressures.
According to the preceding tlie maximuni height that an
elastic balloon can attain clepentls iiot a t all on tlie size of the
balloon, the nature of the gas with which it is filled, etc., lmt
only upon the capacity of tlie material for elastic expansion.
The great,er the volume of the elastic coveriiig can become
without bursting, so much the greater the height. The size
of the balloon is only to be considered in ho far a:, tliiit i t
must be sufficient, without being too greatly espamled. to gil e
the buoyancy necessary in order to raise the bnlloon and the
instruments. Tlie ballooiis met1 heretofore. a:, furnished by
the Continental GummifaLrik in Hanover. can easily stretcli to
double their diameter n ithout bursting. Therefore. the height
they will attain is about lS,OOO meters.
The following table will be of use in the einploynient of
india rubber balloons. The first coluinn sliows the density of
the air, the seconcl tlie altituc1e, the third the ratio ('.the fourth (I ,,
theratio , the fifth the ratio of the ventilation, or of the
('0,
mass of air flowing past the 1)alloon.
-~ -
Ikiisily
of the sir.
I . 2s
1.1!1
1.13
I .0 i
1.01
0. 96
u. 91
U. S i
0. s2
0. 7s
0. 74
IJ. 711
0. 66
0. 6::
0. 39
11. 56
0. 53
Blti-
tlldl.. +l,,
1. 00
I . (I'
1.04
1. I).-,
1. O i
1. IJ3
1.11
1. 1 3
1. 15
1. l i
1. 15
1.21
1.24
1. 21;
1. 2!1
I . 31
I . Y6
.l t ,' l ' ,,
1.110
1 .0 1
1. u:: 1. I):{
1. u4
1. 04
1. IJ5
1. Ill;
1.0;
1. IJS
I . I19
1 . 10
I . 11
1. I:!
1. 1:;
1. 14
1. IS
r l ,r l a ,
I . RG
1. ::!I
1.41
1 .4 4
1.49
1.5;
I . G7
1. it:
1. H i
1.91
~~~ ~
1 ,' , :',,
~-
I . 16
1. IS
1. I!l
1. 211
I . '1"
1.25
I.!!!)
I . 3::
1. :x
1 .ss l,!l9 , 1.41
2.Ili 1.44
?. IS 1.4s
2. I.-, , I . 5n
2 .z
~
1.51
?.2 S
~
1.51
0.46
0. 44
0. 43
11. 41
0. 31;
0. x:!
0. I S
10. 25
0. "I
0. 3 1
0. IS
0. I6
0. 11
0. l:j
0. 111
0. os
Table I? s h o w that very considerable heights can be
attained with closed rubber balloons i f they can espand to
more than twice their original diaiiieter. I n practise, however,
the misfortune is often noticed t,liat the enrelopes in expand-
ing develope sinall holes tlirough which the gas rapidly
escapes. In such cases i t happens that the balloons (10 not
explode although the altitudes that they can attain under
such circuinstances are very considerable, such as 12,W)O-
13.000 meters.
However, in such cases we lose one advantage whicli Assmann
iiiore especially has pointed out, namely, that a closed l~alloon
FC
11-2
ascends with increasing velocity and does not maintain any
position of equilibrium. The leaky balloon floats for some
time a t the highest altitude, so that the thermometers have
no proper ventilation, and then descends too slowly. On this
account, it would be well in all cases to give the balloon, by
a strong distension in the beginning, a more than sufficient
upward impulse, so that in the first place it will certainly
explode, ani1 in the second place be suficiently ventilated.
The fact that the velocity varies inversely as the 6th root of
the pressure, therefore, increases somewhat slowly, iiiakes a
great velocity in the beginning particularly desirable.
If me take the product of the vertical velocity by the density
of the air :IS the measure of the ventilation) then in the vicinity
o f the surface of the earth and a t 4 meters per second this will
lie 5 , h i t a t an altitude of 20,00() iiieters where the velocity,
nccvrcling to our table, has risen to G meters per second, the
ventilation will be 0.65. The latter figure is certainly no
longer snffic.ient to protect even well sheltered thermometers
axainst radiation. According to our esperieiice we must at-
tain a value of 1. This figure. however. requires an initial
ascensional velocity of 5.7 iueters per second. a velocity that can
easily be given to rubber balloons. For still greater maxiinurn
altitudes a still greater initial velocity must be given. For
these asceiicliag velocities, however, one intist use very sensi-
tive instruiueiits and iiot slug@ thermometers. In Stras-
burg, since the introduction of tlie closed rub1)er balloons, me
have with great success met1 the tubular thermometer, de-
scribed by me in the protocol of the Conference of the Iuter-
national Coniiuission for scieutific balloon ascensions a t Berlin.
This theriiiometer lias a sensitiveness more than suficient to
enable it to recortl properly during the above desired velocity
of ascension; it also possesses the lightness (weighing with the
clock mil protecting case 33) grams) necessary to make it
possible to rise with ruliber balloons of 1.50 meters diameter.
The further advantages possessed by the rubber 1JallOOllS
have been SO fully described by the inrentor, Dr. Assmann, in
the protocol to the abore-mentioned conference that I do not
need to go into any ftirtlier details. I will only close with the
wish that tliej- may be used frequently and with good results.
DETAILED CLOUD OBSERVATIONS. A PROGRESSIVE
PHASE IN WEATHER FORECASTING.
U! F:ev FRLULKI~ h I , l l r ,r \n i ~ 1 1 . 9 .T , dntr~l .Talluar~ S 11104 hIet?orulogicsl Obserrs-
tarn tbf ht. 1g11.lt1ua C O I I ~~~, I iprriau;i, o L o
Isobars have formed the stepping stones on which weather
forecasting has mounted to take its place among the sciences.
The dailj sur^ ey o f the atiiioslhere a d the publication of
weather iuaps hare enaldecl meteorologists to bring out these
ftiots: that tlie /i r t /)c w of (nir nentlier depends on the configu-
ration o f isobaric. lineh, but its r i i /+w s t /y on their gradient.
With these t n o lninciples to start with, a few decades hnve
sufticecl to de\ elop a iiietliocl of forecasting that lias met
with very encouraging rehults ant1 11:~s been of great value to
iuoht varied classes of interests. But. in spite of these suc-
cesses, tliere is a prevailing conviction aniong forecasters that
in the face o f great clifticulties progress. a t tlie present time,
is checked.
In spite of telrgrapliic bysteins aiid the imp material a t our
disposal twice each claj-, certain obstacleh bar our further
aclvance \thich iiiay l w siitniuecl up as follows:
1. We iiiih:, iuaiiy nn important detail in the iiiap.
I?. We are often left ignorant of tlie sudden formation or
dissolution of highs and lows. or of cliaiiges in their intensity.
3 . Tlie irregular progrebs of some isobaric systems can not be
detected on, IJr inferred from. even the most perfect niap The
initiated will hardly call for a proof of this statement. But
how are we to mend our condition Y -4 thousand stittioiifl
woiild linrcllg bring out the necessary cletail in a nap; and,
even if they did, the changes in atiuosplieric conditions would
574 MONTHLY WEATHER REVIEW. DECEMBER, 1903
make the map unreliable for the hours which intervene be-
tween the morning and the evening report. Moreover, a
change may occur even during the short space of time that
elapses between the observations and the compilation or util-
ization of the map. Continuous teleonieteorology might offer
an eficient remedy, but that must be considered, for the pres-
ent a t least, as a scientific utopian dream.
Much diligence and eminent talent are being brought to
bear on nieteorological problems. There is great activity
everywhere; the atmosphere is sounded by menus of kites and
balloons, clrtta are collected and collated, Init the results ap-
parently answer the *‘ how ” and the ‘6 wherefore ” o f the
theorist, rather than the pleading ‘‘ what ” and ‘‘ when ’’ o f tlie
solitary forecaster, to whom the public looks for definite iu-
formation, and who stands or falls by his personal wits.
It is fair, therefore, to say that the weather niap alone,
though an indispensable help, will not advance the science of
forecasting much beyond its present stage. We have read
everything out of the i m p which beeins contained in i t : still
we are often a t a loss, often a t fault. It will be wise, there-
fore, to look elsewhere in the hope of striking a nen trail,
which will advance 17s another step such as that taken \\hen
we began to consider the configuration of isobars.
For quite a run of years. while constantly and diligently
observing the clouils [as they come and go in their varied
forms and directions, their different colors and speecls] the ccm-
riction has firmly taken liold of me tlint they have more to
tell than we have hitherto bupposecl. I believe that a system-
atic ancl detailed study of the clouds will. to say the leabt,
advance us one more step in the science of forecasting. Poh-
sibly that step mag be iuiicli more progressive than we can c)r
dare imagine a t the present nioment.
This conviction is chiefly based on m y personal experience ;
yet it seeins that auy one who considers tlie nature of clontls
bught to come to the same conclu5ion. C~OUCZS are the iiume-
cliate product of weather conditions, not iiideecl directly over-
head, but a t some distance froin us. We have here a relation
of cause and effect. Similar causes are followed l,y siinilar
effects. The coinbination of weather elements is varied, but
any variation here will produce a corresponding change in the
effect, namely in the cloucls. I would, therefore, consider
cloucl as an index; as a hieroglyphic language of the weather,
if ,you will, written against the blue sky. I f they are unin-
telligible to 11s it is not nature’s fault; she never writes in
meaningless scrawls.
What attempt has thus far heen macle to decipher this
script? We have picked out ten or a dozen cloud forms and
observed these in a way. But is it reasonable to expect that
the ever changing weather in its endless variety can be de-
picted by means of 10 simple signs? Abercromby seeiiis to
me to contradict himself, when he writes within the space of
three pages the two following passages. After referring to his
catalogue of 10 cloud forms, he says:’
Almost all the smaller varieties are so rare or tran4ent that for ~~r a c -
timi purposes they IJP neglected. IJllt if. un the contrary, the 1u main
words are reqtrictecl to tlie forms of cloud \\e hare (I
them * * * then tlie author can
observaticin in all lon~itucle+ and i n latitud
5 5 O bouth, that gU per cent of skies in e ~e
sufficiently accurately defined liy these 10 a
Two pages further on he remarks:’
We have seen that there ih a fine-weather c.umulus, as well as cuiilulo-
nimbus, a fine-weather, as well as a dangerous cirrus, while fieecy
have not the same import in L ~n d ~i i as on the etluator. In pract good and bad forins can rarely be mistaken, hut sometimes very difficult
cases arise. Clouds. in fact, tell us by their appearance, what might l w
written in words. that more or less damp air is rising or falling under
certain conditions uf upper ancl lower wind currents: the significance
must be judged hy the siirrounding~ and antecedents, just as the zeu+e
of many words can only I J ~ judged IJY the (.ontext. ________ -
Weather, p. 118. 2 Ibld, p. 12u.
We must remark, that there is nothing so small in nature
that we mag neglect it with impunity. It may be possible
with these 10 forms to define the general appearance of 90 per
cent of skies; but 95 per cent of observers would be able to
conclude precious little about the weather conditions under
skies so vaguely described. Then the certain conditioiis, the
surroundings and antecedents, whicli are written in the clods-
these are the very things we wish to get at, and I doubt
seriously i f 10 words are sufficient to describe the great nuin-
ber of combillations furiuecl by neather factors, ~l i o s e esist-
ence must, be and is eqwebsecl in cloud lnnguage.
Cloud
language ant1 cloucl clidect seem to promise the much coveted
aid, and percliaiice this language mill be as serviceable as is
tlie incomparable barometer.
If the sky were as iutelligible to us as is t,he daily weather
niap wfth its isobars, isotlierius, wind direction, and velocity,
we could follow weather changes n itli solliething like the cer-
tainty uf a sldled pliysicinn watching the syiuptoms of his
patient toward recovery or total collapse. This knowledge,
and the facility in using it, can be acquired only with patient
and SJ steiuntic labor. We must study the letters, the words,
and phrases of clou(l language. I n the beginning. no doubt,
i t will be :L task no less arcluous than that undertaken by thobe
who fii ht decil)liere(l liieroglppliic inscriptions, but it seems
no lesh 1)roiuisiiig.
Some 1)rogress has already been iuade along these lines.
11-e have the 10 cloud forms. laid as a founclation by the in-
teriiationnl cloud cwnimittee. Blue Hill, Upsala, and several
other observatories have clone very efficient work. Still these
are isolated atteiu1)ts. and it is to l ~e feared that there are too
fern who are really interested in this iinportaiit subject, oning
to its primitive condition, niid p e r h n p to a lack of metliocl-
ical clirectiuns in its pursuit.
For these reasons I liave attelIIlJte(1 to piit the subject in the
light in which I view it, and will giie tlie clnsbification used
a t the Meteorologicd 0l)servatory of St. Igiiatius College, to-
gether with wine of our iilethocls. reseriiiig the finer tletails
for sn1,sequent articles i f the subject hliould prove interesting
to tlie readers of the RET-IEW.
PIInuy will, no doubtm, be shocked by our elaborate catalogue
of cloud varieties, but it lias nlreatly 1)een stated that the
simple, tliongh funclamental c7ivisions of the Iiiteriiational
Cloud Coininittee can do nothing to advance the science of
fiwecasting. These are a part of tlie A, 13, C! of cloud lan-
guage, and will remain so unless we progress.
Abercroin1,y lias correctly stated tlie practical object of
cloucl study when he says?
Thp fi)iinilatioii o f all modt.rn ~~l o 1 1 1 1 l i n i ~\r l i ~i l ~~ turns roniiil the rela-
tia3n o f i ~l o u d forniz to shalws o f isoLni>.
Now, there :ire more than 10 staple foriiis o f clouds, each of
nliich has its own peculiar significance. because it is the pro-
c1uc.t of a peculiar conibinatiun of weather factors, and we can
not, therefore. neglect any one of tlisiii without running the
risk of leaXing :L very seriouq gap in the s>steiuatic knowl-
eclge we are :tn.;ioub to acqnire. It will be our task to recog-
nize aiid record all staple cloud forms, together with the at-
mospheric condi tioiis that precede, acconipany, or follow them.
If I am told that cirrus preceded a certain storm, I know very
little, siiice I ha\ e many forms of cirrus in mind, and i t would
interest me to know precisely which form \vas observed in
that particular ewe, and, more than this, there are many in-
terebting, intelligible points about eT-erj- cloud that demand
recognition and interpretation. It is evident, therefore, that
cloud language ant1 cloud study can not be so siiiiple as seems
to have been taken for grantecl in a11 instructions to observers;
one will prove as coinplicatecl as the other.
Here. then. is evicleiitly an oliportunity for progress.
-_
JVeather, 11. 122.
DECEMBER, 1903. MONTHLY WEATHER REVIEW. 575
I n setting forth our system of species and varieties I have
added no definitions, as the names, for all preient purposes,
are suficiently self-esplanatory. It has been my endeavor to
make the classification as rational and exhaustive as possible.
Hence, principles of division were assumed, which mill coTer
the whole field, ancl are so simple and elastic that they inay
be applied to the most peculiar climate or sky. I n judging of
the cletail in classification, I would caution the reacler to re-
member that this is the particular dialect of cloud language as
it is found in ancl around C'leveland, Ohio. How it mould work
in other regions I know not, though it occurs to me that sev-
eral good exact monographs for fipecial localities could later
on easily be reduced to a universally practical system.
Dr. Hann remarks in a note to his chapter on cloucl classi-
fication :'
The inexpedience of atteinpts to clct4fy cloi I J n purely genetic and
physical prinuiples, i. e ., nitliout a iiiorpliok a1 lmsis, has lwen in-
sisted on repeatedly.
The reasons are plain. Our present linowlec1,rre of the causes
of difierent cloucl forms is too limited to establish a natural
classification. Even i f me understood the genesis of cloucl
forms, the huinaii eye and the cloud forms will always remain
the only factors on which the observer must rely for his incli-
cations of the coming weather.
What is desirecl, therefore, is an erhanstive, or, a t least. an
espansive system of cloud forms, even if it, be art,ificixl, arounil
the elements of which we mag group the results of our obser-
vations. Our classification of c.loucls will naturally follow tlie
mine laws and pass through the same stages as did the zoologi-
cal and botanical systeiiis of Liiinwus.
Gratitl diuision (p i i t nntrcrr or inntrrinl f!f
thP C Z O /d ).
Class I. Ice cloucls: Crystals.
Class 11. Watery clouds: Solid mater particles.
S'ditlitwioti i)ito-fiondi~.s (p r i w i ~i Z t ~ qf dirihti-thr e.rft~isioi/ (f /1 1 ~
rloutl).
Family I. (The line.) Cirrus: Fiber cluucls.
Family 11. (The surface.) Stratus: Layer cloucls.
Family 111. (The solid.)
With these five elenients, obtained 11s divicling according to
nature aiid extension, we have five genera of cloucls, a division
which is exhaustive and conies very near to that of Howard.
Cumulus: Lump clouds.
TI1 f, .t? 1.6, y ' l / P rn.
1. Cirrus.
2. Cirro-stratus.
3 . Cirro-cumulus.
4. Stratus. (a) Alto-Stratus.
5. Cumdus. (0 ) Alto-c~1~dut;.
(h ) Low stratus.
(h ) Low C U I ~U ~U S .
by dorrd.-High level, 30,000 to 27,000 feet:
(2 ) Streamer.
(3 ) Plume.
(1) Wisp.
ni i r isio it i f 0 sp"c I P S (1) r iu r Qi1P-t 11 P !j ~u t w i l o ii t I t )/ e or rsttvt s i v n ) .
C'irrns. (1) Fiher.
Cirro-stratus. (5) Baacl.
(6 ) Bar.
(7 ) Patch.
(8) Veil.
(10) Ch-o-cumulus balls.
Cirro-oumulus. ( 9) Cirro-cumulus pellets.
4Lelirbuch, p. 263. [Genu. Ed.]
ITl'ater cloud-Central level, 12,000 to 15,000 feet:
Alto-stratus. (11) Gauze.
(12) Band.
(14) Cover.
(13) Patch.
Alto-c~1mulus. (15) Alto-cumulus ba11~.
(16) Alto-cumulus plates.
Iliittv- rlvutl.-Lower level, 2000 to 6000 feet:
Stratus. (17) Sheet.
(18) Cover.
Cumulus. (19) TVoolpack.
(20) Mountain.
(21) Roll-cumulus.
(22) Strato-cu1uulus.
(23) Wrack.
(21) Fog.
From the above it is apparent that clouds may be naturally
and exhaustively cliTicled into species on the principle of gen-
eral outline. There are, however. cliflerentiations of species
constituting varieties, which, though easily and generally
nrerluoked. :we of the very greatest import, since every turn
in a staple form must have a corresponding cause in the com-
bination of weather factors that produced i t ; ant1 it is here
that we depart on new lines. What others have heretofore
clesignatecl trifles, that we now take up. suspecting them to
contain the very knowledge we are after.
I)(ffwi,ttntiolts (constituting varieties) mag be observed in:
(a). The general outline.
(11). The surface.
(c). The organization.
OUTLINE.
Hooked.
Bent,.
Branched.
Pectinate.
Screwecl.
Curved.
Tangled.
SUKFhCE. ORGANIZATION.
Undulating. Flaky.
Drifted. Granulated.
Warpecl. Watery.
Rippled: Fibrous.
Banked. Lumpy.
(a) Plain.
(b ) With interfer-
ence.
Furrowed. Marbled.
Striated:
((0 Lengthwise.
(h ) L4cross.
(c ) Obliquely.
Headed:
(I / ) Forward.
(1 )) Backward.
Wispy.
Tapering.
Feathery.
Fleecy.
Fringed.
A combination of one of these adjectives with one of the
names in the list of species makes it possible to perfectly
designate ant1 register a large number of staple cloud forms.
Certain storm clouds are so independent ancl so well organ-
ized that special terms are required to fully describe them.
They exhibit peculiar rlerelopnients. which I will call processes.
1. &lamm,u.
2. Antlers.
3. Beards.
4. Trailers.
5 . Fringe.
The well-cleveloped thunclercloud shows some or all of the
-4t /op : (1 ) Beak (forward).
Eelurr,: (1 ) Ram (front).
following processes :
(2 ) Spur (rear).
(2) Ruclder (rear).
576 MONTHLY WEATHER R;EVIEW. DECEMBER, 1903
These four, when they occur combined together, have long
Besides this classical ago suggested the name of anvil cloud.
form I may mention:
1. Cloud wedge.
2. Crescent.
3. Funnel.
4. Umbrella cloud.
Thus far I have made use of form only and have shown that
all clouds may be fully described according to a uniform prin-
ciple. But it is not the form only which has its story to t,ell;
on the contrary, form tells but half the history of the cloucl.
If we desire to derive all the information possible we must take
in the accidentals. These may be grouped under the follow-
ing heads:
1. Condition.
2. Motion.
3. Distribut,ion.
4. C!olor.
Condition: &lo tion :
3. Stable.
4. Forming.
5. Breaking up.
6. I n transforniation :
From-to.
Distribution:
1. Isolated.
2. Scattered.
:3. Flocked.
4. Aligned.
5. Radiate.
6. Continuous.
7. Coalescing.
1.
2.
3.
4.
5 .
(i .
7.
8.
Color :
1.
2.
3.
4.
5.
6.
7.
Slom.
Rapid.
Average.
Boiling.
Encl on.
Broad side on.
Aslant.
Direction w1ienc.e.
White.
Blackish.
Red.
Yellow.
Green.
Iridescent,, with. . . .
as prevailing color.
C h y .
v
The 1)rngrani of a single cloud observation will be. as followb:
We are to determine, (1) the species aiiil \ariety; (2) pro-
cesbes; (3) conditions; (4) elements of motion; (5 ) clistribu-
tion; ( 6 ) color; (7 ) configuration of isobars.
The answers to the firht six points are enterecl by means of
symboiic signs in one column of a journal, and in a coluiiln
parallel to this :we to he entered the configuration of ibobars.
together wit11 the state of the weather. Regarding the con-
figuration of isobars, it is evicleiit that most attention must be
paid to that system under whose control the ohserver happens
to be a t the time of observntion. The final alii1 crowning
work \\ill be the collating of the two columns. These mill in
the course of time show repetitions, regularly recurring mho-
ciatioils of cloucl forms and type of weather, and s~tch ~g r r l o ~
associntions n w tlir r.rlu*f-ssio,i qf n loti, (f )zntuw: the) will form
laws for the forecaster. By such method only shall we reach
any practical results in cloud study. I n way of consolation
for those who find this method too elaborate or too difticult. I
will quote a few lines froin Clement Ley.5
All said then. and dour. the laud of i.louds is not eutirel) oui' (i f fancy.
The art of distinguishing ant1 o f risrfril playing tlir ili4inction.s IIC-
tween the vaiieties of clout1 is not nra
which the amateur engages Iiim+t~lf f i r tlir
very cursory aciluaintance with tlir laliois o
if it is uot insiilious to say this o f C+er~uan and Aiiirricnu writer+-sati&r,
every one as to the fact that therr is also a 5c.iencr of nryhologg. na+iwut
though this scieuce may be. Hcw ('an practibe anil knuwle(lge lie most
successfully promotril ?
To this question, as put by RIr. Ley, I mould answer: By
systematic and detailed cloud olwervations; by attending to the
so-called fr(Pes.
5 Cloiidlantl. 1). 201
I might enter into some very interesting details myself, but
since it is my object to shorn that cloud study, as a scientific
branch of meteorology, is something more thau noting one of
10 cloud forins opposite a given date, but I will close, hoping
that some of the readers of the REVIEW will interest themselves
in one of nature's most interesting languages and lend a help-
ing hand in interpreting it ani1 teaching it to others.
'
~
__
METHODS OF FORECASTING THE WEATHER.
A Lecture ilelivererl hp Prof. Dr J. 31 Peruter to tlie A-sociation for the Advancement of svieotilic Kuunl~dge, Vienna. .Ienuar) 14. 1903. Trm~slaterl from the Vot trage dcs Verpiue* L i i r V d w i t ~i i i ~ untiiraisseir~rlinttli~ her Keuntuisse i n Wicu .13d Jebrgang,
Hf*ft 14
Allow me to-clay to adclress yon once again on the subject
of weather prophets, and this time to bring before you not
only one or two kinds of weather forecasting, but to give you
a more general survey of all methods a t present in use, be
they right or wrong, with or without results. I will keep
strictly to the title of this lecture and give the prominent
place to the iiiethocls of forecasting. I shall esplain them and
subject them to critical analysis, naming a t the same time the
advocates of each of the various methods; in the technical inves-
tigation, me have to do with the value of the methods anil not
that of the persons. I must, lionever, at once bring promi-
nently forward the fact that 11 e have a t present, unfortunate-
1y. no method bp wliich we can forecast the weather with ab-
solute certainty even for one day in advance, to say nothing
of longer periuds. Thib is already self-evident from the fact
that we are now able to speak of many methods of forecasting,
whereas if there were a sure ani1 infallihle method, then i t
woulcl be out of Illace to speak of the other methods to this
soc,iety for the ndvniicemeiit of scientific knowledge.
All methods of weather forecasting, not excepting those in
use by the central meteorological offices, are based upon ob-
serrecl weatlier conditions and have, therefore, an empirical
founilatioii. Many of tlieiii do not elen make the slightest
attempt to put their methods on a theoretical basis and content
themselves with setting u p &'weather rules. " Even the scien-
tific iiiethocls of profe onal meteorologists have not yet suc-
ceeded in cleclncing a theory capable of determining in advance
the clirtnges of the weather ah the effect of one or several
known causes. Only the advoc:ttes of the influence of the
moon hare veiit8ured solely by means of aprioristic theories to
l.calcnlate" the weather for long periocls in advance.
There are many widely different methods by which the va-
rious Glasses :ind kinds of weather prophets c y r y on the mark
of weather forecasting. There are those who iiiake use of the
behavior of animrls to foretell the weather; hunters who rec-
ognize the cliarncter of the apprnaching seahon from the ac-
tions of the wild animals; the observers of lirds. spiders,
crickets, ants. and other aniuials from whose conduct they
judge of the approaching weather. But in aclilition to tliis class
which utilizes living animals there is another opposing class
that prefers to inake use of the dead substances of the animal
or regetable kingdoms, such as hairs, strings of instruments,
roots and fibers of plants; by iueans of their espansions or
contractions, either with the aid of little weather houses and
figure6 or without them. they recoguize the coming weather.
Others prefer to consult stones ani1 walls as to the character
of the weather to be expected, ani1 turn rather to inorgmic
nature in order to learn froni the '*sweating" or Jryness of
these whether to espect rain c)r continued fine weather. Thus,
as you see, all the kingdoms uf nature are drawn upon to fur-
nish prognostics of the weather, and it may de1)encl upon tlie
occupations rind predilections of the various persons in-
terested in the coming weather mliether they give the
preference to one or the other. But I had almost for-
gotten to inention another class-perhaps the largest-
those who are not to be satisfied by any one of the three king-
cloms nor even by all three together, ani1 who rely only on