SEPTEMBER, 1907. MONTHLY WEA!CEER REVIEW. 403
crystals, especially of type WBB, cease t o grow and are
succeeded by type WFC. I n many cases the latter type adopts
the former as a nucleus and grows outward therefrom.
We have selected 18 examples of this most elegant type of
window-frost crystal for use as illustrations, as follows: Nos.
30, 31, 32, 69, 73, 115, 117, 130, 131, 132, 146, 147, 148, 177,
179, 105, 196 A, 196 B. Nos. 130,131, and 196 A are particu-
larly lovely and interesting examples. As will be noted, some
of them reseinble sea moss. Tiny crystals-of type WBB form
the nuclei of specimens Nos. 180, 131, 132, and 195.
Sometimes a very slight change (increase) in the indoor
humidity, accompanied by a fall of but a few clegrees in out-
door temperature, will cause crystals of this type to form and
grow between or around crystds of type TS'BB, as shown in
the very interesting photograph No. 196 B.
($4) T!yp WMD. JfiotiderijiT wj)idow frost.
a single straight or curving spike or quill, and in others of a
central quill adorned with a few or with inany secoudary tuft-
like branches. Such tuft-like secondary branches as imy
form upon and around the primary quill or spike usually ap-
pear clustered together into tufts thereon. Singularly en( nigh
many of the secondary branches grow outward from the cent,ral
quill perpendicularly rather than a t an angle of GO".
These odd crystals are essentially very cold weather types
and form only within cold, unheated rooms during below-zero
weather. They form freely only when the indoor tempera-
tures range from lFi0 F. ~lownwarcl. They form freely in a very
cold atmosphere, when the relative humidity of the air ranges
from 77 to 86 per cent, and less freely within warmer, air at
humidities of from 55 to 6.5 per cent. Sometimes the longer
varieties of this type of crystal heem to possess a structure
nearly continuous, as in Nos. 175,17G, and 1'78, while in other
cases they seem to be .built outwardly in sections, as in Nos.
58 A, 58 E, 59, 154 A, 154 B, and 188.
(35) T!ype TV'E. S f e l l ~f ~r n i I y s t a l a .
Crystals of this type are of a somewhat close structure.' I n
many cases a simple hexagonal star forins the nucleus, the
rays of which terminate in tiny solid hexagonal plates.
They cominonly form around and upon tiny bits of ice, or
tiny frost crystals, ancl usunlly develop parallel to, but
slightly raised from, the window g l a s ~, except when their
nucleus i R itself attached to tlie glass. They evidently grow
011tward in intermittent, rather than continuous, orcler, wlien-
ever the temperature suddenly falls, or a winclowpane is sud-
denly cooled far below its previous temperature, as the result
of a cold gust of wind suddenly springing up ancl striking it.
They are slow-growing, cold weather t,ypes, and occur only in
cold rooms and during below-zero weather. Indoor tem-
peratures must be as low as 20" F. for them tn form
freely. They form most freely within cold rooms a t humid-
ities ranging from 71 to 90 per cent, less freely a t milder
temperatures, when the relative humidity ranges from 55 to
ti5 per cent.
They usually fail to attain perfectly symmetrical propor-
tions, for the reason that they commonly form in such a posi-
tion that their growth does not progress equally in all direc-
tions. I n rare instances their position ancl environment is
such as to allow development to proceed in all directions, and
in such cases they assume cluite symmetrical fornis. See
Nos. ltil and 1.54 B (it is to be noted that one point of No. 161
is broken partly of€).
Crystals of this same general character sometimes form
around, or develop annex-fashion from, other frost types.
No. 145 A is a very beautiful and unique example, in which
the stelliform crystal formed around aud grew outward as an
innes to the branching type of window frost, type WBB.
These singular frost crystals consist in some czs L b es of but
It is to be noted that this type of frost, altho for con-
venience grouped under the title (' Window frost ", develops
largely outward from, and independent of, the surface of the
glass, and hence is in reality a type of indoor hoarfrost.
(See type HTF, tahular hoarfrost, section 16.)
[To be confintted.]
INTERNATIONAL METEOROLOGY.
The following is an extract from an address by Prof. Arthur
Schuster, of Manchester, delivered before the Royal Institution
of Great Britain Friday, May 18, 1006, and is reprinted from
Nature ancl from the Annual Report of the Smithsonian Insti-
tution for 19OK
I n an address delivered to the British Association a t its
Belfast meetiug, in 1902. I espressed the opinion that nneteor-
ology might be advanced more rapidly i f all routine observa-
tions mere stopped for a period of five years, the energy of
observers being concentrated on the discussion of the results
already obtained. I aiu glad to say that meteorologists have
taken seriously a remark the echoes of which still reach me
from distant parts of the earth. They disagree with me, but
their clisagreeiueiit is of the apologetic kind. I clo not wish
to retract or to weaken my previous statement, but merely
now qualify i t to the extent that i t is only to be applied to
two-tiimensional meteorology. There is a three-dimensional
meteorology as far reiiiorecl from the one that coniines itself
to the surface of the earth as three-iliniensional space is from
a fiat area. Three-dimensional meteorology is a new science,
which a t present requires the establishment of new facts be-
fore their discussion cnu properly begin. The extension of
our range of observations by kites and balloons is of compara-
tirelj recent origin. Mr. drchilxtld in this country was one
of the Ibinneers of meteorological iuvestigatioii by means of
instrmnents attached to kites. I n tlie United States Mr.
Rotch, having establishecl a separate observatory, sncceecled
in convinciug scientific inen of the great value of the results
which coulcl be obtaiuecl. RIr. L. Teisserenc de Bort, who
estaldishecl and maintained an observatory for dynamic mete-
orology a t Trappes, near Paris, rendered similar services with
regard to '< pilot " or unmanned balloons carrying autographi-
cnl instruments. The aeronautical department of the Royal
Prussian Meteorological Institute, with Doctor Assmann at its
head, under tlie tlirection o f Prufessur voii Bezolcl, also made
a numlJer of iinportant contributions in the early stages of the
work. Professor Hergesell, of Strasliurg, similarly made nu-
merous experiments; and chiefly through the efforts of those
whose naiiies have been mentioned, and more especially Profes-
sor Hergesell. an international agreement has been secured by
means of which kite and balloon ascents are made in several
countries on the first Thursday in each mouth and on three
consecutive days during two months of the year. A large
station for aeronautical work was recently established a t Lin-
clenl,urg, near Berlin, where kites or balloons are sent up
daily for the purpose of securing meteorological reoorcls. The
greatest height yet reached was during the ascent of the 25th
of November, 1905, when by mems of several kites sent one
after another on the same wire the upper one rose to an alti-
tutle of 6430 meters, alniost exactly four miles. Owing to want
of funds this country cuuld until recently only participate in
this work through the indiviclual efforts of Mr. Dines, who re-
ceived. however, some assistance from the British Associ a t' ion
and the Royal Meteorological Society.
The reconstruction of the meteorological office has made it
pnssible now for Mr. Dines's work to be continued as part of
the regular work of that office, ancl further stations are being
establihxl. MI.. Cave carries out regular ascents at his own
expense a t Ditcham Park, and through the cooperation of the
404 MONTHLY WEBTHER REVIEW. SEPTE~IBER, 1907
Royal Meteorological Society and the University of Manches-
ter, assisted by a contribution for apparatus from the Royal
Society government grant fund, a regular kite station is being
established on the Derbyshire moors.
The international committee which collates the observations
is a commission appointed by a union voluntarily formed be-
tween the directors of meteorological observatories and insti-
tutes of countries in which regular observations are taken.
The meeting of directors discusses schemes of observations
and encourages uniformity.
If I mention a few of the difficulties which stand in the way
of a homogeneous system estending over Europe, I do so in
the hope that it may perhaps ultimately assist in removing
some of them. It is obviously desirable that the charts, which
are intended to show the distribution of pressure and tem-
perature, should be derived from observations made at the
same hour. Germany observes a t 8 o’clock of Central Euro-
pean time, and France observes simultaneously (or nearly so)
by choosing 7 o’clock, Paris time, for its readings. We observe
at 8 o’clock, Greenwich time, which is an hour later. It is the
great desire of continental meteorologists that our standard
hour should be 7 o’clock, and what prevents i t from being so?
Chiefly and absolutely the additional cost which the post-office
must claim for the transmission of telegrams; because mes-
sages transmitted before S o’clock are subject to an additional
charge of one shilling, which may be claimed by the postmas-
ter, the claim being possibly increased to two shillings when
the postmaster and telegraphist are different persons. This
is prohibitive, but i t does not exhaust the inconvenience of the
additional charge. For the purpose of weather forecasting
it is clearly necessary that telegrams should be received as
early as possible by the meteorological oftice. But the S o’clock
rule delays telegrams from some Irish stations, because 8
o’clock by Dublin time is 8 9 5 by Greenwich time, and there-
fore Irish telegrams may have to wait until nearly half past 8
if they are to be transmitted without extra charge.
While the international organization of meteorology is well
on its way, though difficulties such as those I have mentioned
may temporarily retard it, another question not altogether
disconnected with i t has been raised by Sir John Eliot. This
is the establishment of an institution devoted to the collective
study of meteorological problems affecting all parts of the
British Dominions. It is t,rue, not only in this but also in
other matters, that in order to take our proper position in
international work it is necessary that we should set our own
house in order, and we must give Sir John Eliot’s proposals
our hearty support. I f I do not enter further into this clues-
tion, it is because I am now dealing more especially with prob-
lems which go beyond the limits of the Empire. I assume the
existence of a national organization, but lay stress on the
insuficiency of this limitation.
The importance of the subject, however, may be my justifi-
cation, if I direct attention for a moment to the meteorologi-
cal question as it presents itself in India. We all know and
realize the vital importance of the rainy season, and the benefit
which the native population would derive if i t were possible
to predict, even i f only imperfectly, the setting in of the mon-
soon. It appears that Doctor Walker, the present director of
observatories in India, recently obtained Fery encouraging
results in this respect. According to his investigations, a
forecast of the monsoon may be derived from a knowledge of
the weather during preceding months in different parts of the
world. Thus a heavy rainfall in Zanzibar in May is followecl
by a weak monsoon, while a pressure deficiency in Siberia
during the month of March indicates a probable deficiency of
rain in India during the following August. I need not insist
on the importance of these results, which at present are purely
empirical and require further confirmation, but i t is quite clear
that for the successful prosecution of these inquiries political
boundaries must be disregarded and a system of intercommu-
nication organized between the countries chiefly concerned.
Doctor Walker informs me that he has successfully arranged
for telegraphic reports to be sent to him at the beginning of
June from six different stations in Siberia. It is hoped that
this cooperation, which was unavoidably discontinued during
the late war, may now be reestablished.
The course of international organizations does not always
run smoothly. The efforts made toward cooperation in earth-
quake records have unfortunately led to differences of opinion,
which have hitherto prevented a truly international system
being formed; and if I give a short historical account of the
circumstances which have led up,to these differences, i t is only
in the hope that this may help to remove them. The scientific
investigation of earthquakes may be said to have begun when
British professors of physics, engineering, and geology were
appointed at the Imperial College of Engineering in Tokyo.
Some of them on returning home succeeded in interesting the
British Association in the subject. Ever since 1880 that asso-
ciation has been an active supporter of seismic investigations.
The much disturbed region of the Japanese islands was natu-
rally the first to be studied, but in 1895 Professor Milne, as
one of the secretaries of the committee, issued a circular call-
ing attention to the desirability of observing waves which
have traveled great distances, and some months later Dr. E.
v. Rebeur-Pasch witz, of Strasburg, drew up suggestions for
the establishment of an international system of earthquake
stations. To this scheme Professor Milne and other members
of the British Association committee gave their approval.
The cooperation which thus seemed SO happily inaugurated
was broken by the unfortunate death of its originator. Cir-
cumstances then arose which compelled the British Association
committee to go its own way. Under its direction a system
was established which now includes about forty stations dis-
tributed all over the world. But the needs of different coun-
tries are not, and. were not. meant to be satisfied by this
organization.
There is always a certain number of earthquakes having
purely local importance and requiring cliscussion from a purely
local point of view. For the purpose of such discussion
relating to the disturbances which chiefly affect central Europe,
the union, so-called ‘‘ kttrtell ”, of the academies of Vienna, Mu-
nich, Leipzig, and Gcttingen formed a committee and did ex-
cellent work. I n the meantime Professor Gerland, who had
succeeded Doctor Rebeur-Paschwitz a t Strasburg, had per-
sonally invited a number of friends interested in the subject
to a conference at Strasburg, with the object of forming an
international association. This was followecl in 1903 by a
formal conference called by the German Government, at which
Great Britain was represented by Sir George Darwin and
Professor Milne. This conference drew up a scheme for an
international association, and a large number of countries,
including Russia and Japan, joined. Strasburg was selected
as the seat of the central bureau. The matter came up for
discussion at the meeting of the International Association of
Academies, which was held in London in 1904, and a commit-
tee was appointed for the purpose of suggesting such modifi-
cations in the constitution of the seismic organization as might
bring i t into harmony with the views of the associated acad-
emies. This committee, over which I had the honor to pre-
side, met at Frankfort and recommended a number of impor-
tant changes, which were unanimously accepted by the second
seismic conference, held last summer in Berlin. In conse-
quence of this acceptance, it appears that Italy and the United
States joined the seismic association, while England declared
its willingness to join under certain conditions, of which the
simultaneous adhesion of France was one. The following sum-
mary of the States which have joined and their population is
copied from the official report of the last meeting at Berlin:
SEPTEMBER, 1907.
German Empire.. .. Belgium.. .......... Bulgaria. ........... Chile.. ............. Kongo State.. ...... Spain ............... United States.. ..... Greece ............. Hungary.. ......... Japan .............
405
In a paper of later date’, Angstrom discusses the probable
origin of the auroral line, and gives it as his opinion that the
spectrum consists of two portions, the chief line being due to
phosphorescence (or fluorescence) and the other lines to
rarified air, the conditions of temperature and pressure,
so different in the upper regions of the atmosphere from
those in the laboratory, sufficing to explain all divergencies.
He says:
The one spectrum results from the monochromatic yellow light which
ih always seen even in the feeblest traces of the aurora, and which, on clear winter nights, is perceived from all parts of the sky. The other
spectrum con~ists of extremely feeble lines or bands, and it is only in
the strongest auroras that it is possible to measure their position even approsimately. Piazzi Smyth 4 observes that the chief line of the aurora
falls between the second and third lines of the greenish yellow hydro-
carbon group; ant1 Vogels remarks that it is coincident with a line in the
spectrum of rarifled air. This latter observation is correct, but in my
opinion this is the result of the purest accident. The red lines-flrst
observed by Zijllner-seldom occur. There is no doubt that the spec-
trum varies at different times. The discharge is sometimes disruptive,
aud sometimes IJS conduction; it occiirs sometimes in the extreme limits
of the atmosphere, and sometime.; at short distances from the earth’s
surface.
,&ngstr;im observed (on April 14, 1873) lines at 5567, 521-,
501-. 487-, 472-. He quotes the observations (of presumably
the same lines) by Barker6 as 431-, 4705, by Vogel’ as 4694,
5233, and Lemstriim8 (November 19 and 23, 1872) as 5568,
5235, 1694, 4262, and thinks that the means of these measure-
ments agree fairly well with the lines 5327, 4707 and 4272
observed in the violet light at the negative pole in rarified air.
Vogel observed the auroral spectrum on several occasions
in 1470 and 1471. The instrument he used was the star
spectroscope belonging to the 11-inch equatorial of the Both-
katnp Observatory, consisting of a set of direct vision prisms,
with slit, collimator, and observing telescope with a magnify-
ing power of four. The measuring apparatus consisted of a
micrometer screw, the head of which was divided into 100
parts. This moved a fine steel point in the field of view, and
the point could be illuminated by a small lamp if required.
The readings of the micronleter were reduced to wave lengths
by means of readings of about 100 Fraunhofer lines found in
hgstrbm‘s Normal &lap of the Solar Spectrum. After each
observation of the auroral spectrum readings of the sodium line
or of the hydrogen lines were taken by way of control.
A very bright aurora was observed on October 25, 1870.
The spectrum showed the green line and several fainter lines
toward the blue on a dimly lighted background, extending
over E and 6, and half-way from b to F. No measurements
were taken on this occasion.
On February 11, 1871, an aurora was seen; the average of
six readings of the brightest line was 5573. No lines in the red
were seen. On February 12 an average of six readings gave
5572; Doctor Lohse also took six readinga and obtained 5569.
The appearance of the spectrum on this occasion was essen-
tially different from that of. February 11. The green contin-
uous spectrum was present from 5572 to b and traversed by
some bright lines between h and F, one line beyond F, and
just before G a faint broad band. Later a red line between C
and P, but nearer C: than D, made its appearance.
On April 9, 1S71, a very bright aurora was seen, showing a
red sheath rising nearly to the zenith. The spectrum was
similar to that of February 12, but much more intense, show-
ing five lines in the green, and a broad band in the blue. The
red ray allowed him to recognize seven lines, for which the
~~ ~~ ~ ~~
3 .ingstriim, Pogg. Ann., Jubelbaiid. p. 124.
4 Piazzi Smyth, C . R. [Comptes Rendus], LXSIV, p. 597.
5 Vogel, Pogg. Ann., CSLVI, p. 583.
6 Barker, American .Journal of Science, (3), 11, 465; V, 81.
7 Vogel, Pogg. Ann., CSLVI, p. 573.
8 Lenistriim, Polarljuset och Polarljusspelitruin. Helsingfors, 1873.
Vogel, Pugg. Ann., CSLVI.
60,000,000
7.000.000 3,700, 000
3,000,000
19 000 000
19, 000,000
76’0W’OMI
2: 500: 000
19,250.000
48,000, OW
2160
40
20 ‘20
80
80
160
20 30
160
-1 1 I
Italy.. ............ Mexico ........... Norwy. .:. ....... The eo onies of the Netherlands .... Portugal. ......... Roumania ........ Ru~sia ............ Switzerland.. .....
33 000 000 13: 600: 000
2,500,000
6,500,000
6,400,000
6, WJ, 000 129. ooo, 000
3,300,000
5164
80
20
40
40
40
160
20
It was decided a t the Berlin meeting that Professor K h e s -
ligethy, of Budapest, should be secretary and Professor Pa-
lazzo, of Rome, the vice-president of the International Seismic
Association. Professor Gerland had already previously been
designated as director of the central bureau. The ofice of
president of the association was left vacant until the final cle-
cision of Great Britain as to its adhesion had been settled.
There the matter stands for the present.
The disastrous results of recent earthquakes and volcanic
eruptions have directed increased attention to the subject.
Its thorough investigation is indeed likely to yield important
information on the interior constitution of the earth. A heartmy
cooperation to obtain and circulate the material for a detailed
discussion can not fail to bear fruit, and even though there may
be legitimate grounds for dissatisfaction a t the inanner in
which a particular scheme has been organizecl, I must express
my own opinion that at the present moment the permanent
interests of this country would be best secured by our joining
the association and helping to direct its work in a manner
which would assist rather than hamper the present organizn-
tion of the British Association.
THE SPECTRUM OF THE AURORA BOREALIS.
By Dr. W. hIARsHaLL WATT*. Dated Sydenham, Eoglswl, SeptemlDer 14, l9J7
Among the rare and striking phenomena of nature, the
aurora borealis, or ‘‘ northern lights ”, has always excited
much interest. The earlier explorers of the Arctic regions
macle frequent observatious of it, and speak of it in glowing
terms, describing it as either a long tranquil arc of silver
light, stretching along the northern horizon, with its highest
point in or near to the magnetic meridian; or as broken cloud-
like masses of a glowing ruddy light, with streamers and
golden rays, compared to aerial spears, shooting from the arc
and sometimes reaching to the zenith, forming an appearance
resembling a crown-the “ corona borealis ”. Upon the invention of the spectroscope scientific observers
were eager to apply the new method of investigation to the
aurora. Among the earliest observers of the spectrum was
,&ngstrom, who, in the winter of 1867-68, observed’ on sev-
eral occasions the spectrum of the luminous arc bordering the
dark segment, which is always to be seen in the aurora even
when faint. Its light proved to be almost monochromatic,
and consisted of a single brilliant greenish line to the left of
a group of known lines of calcium.
Measuring the distance of the auroral h e from this group,
Angstrom determined its wave length to be 5 ~t i ‘i .~, Besicles
this line, of which the intensity is relatively great, Angstrsni
observed also, by widening the slit, traces of three faint
bands extending nearly to F. angstrom remarks that this
chief auroral line does not coincide with any known line in
the spectra of simpie or compound gases. Otto Struve, of
Pulkowa, to whom Angstrom communicated his result, also
made measurements, which gave for the position of the line
1259 of Rirchhoff’s scale, or wave length 5553. -___ ~-
hgstrijm:
That is, 5567 “hgstrijm units” or hundred-milliontlis of a centimeter,
or 5567 ten-millionths of a millimeter, or 556.7 micro-millimeters or
556.7 pp, or 0.5567 microns, or 0.5567 EDITOR.
Recherches sur le Spectre Solaire ”.
55-3