14 MONTHLY WEATHER REVIEW.
(fig. 2b); but after a few moments these cylinders were seen to
congeal again, and change thereby into broader plates, sharp-
ened at their outer edges by two planes of a regular hexago-
nal prism. The whole crystal became thus again a hexagonal
star, but with broader and shorter rays than it hadbefore.
Other crystals, which had at the beginning such fiat and
broad rays (fig. 2c), changed these by melting into feathered
ones (fig. 3c), because on their liquefaction there remained
only the middle of each plate, like an icy needle, in the water,
until, the new cangelation ensuing, a number of needles ran
at each side out of this rib at angles of 60 degrees.
Some of the stars were feathered in the beginning, but only
at the outer half of their rays. I did not see any change take
place in them, nor did this happen with some other more com-
plicated forms. Thus I observed among others a small and
continuous hexagonal plate, with simple rays issuing like di-
agonals out of its angles; but then each adjoining pair of these
rays was still connected by a couple of needles which met at
an angle of 60 degrees (fig. 4).
But these complicated forms were comparatively rare; and
those transformed under my eyes were so predominant, and
presented a spectacle so full of motion, that at last I could
hardly help comparing them with living beings. In fact it is
only in the case of such that we are accustomed to witness
changes so mysterious without inquiring after the forces that
produce them. We got, however, a partial explanation of this
phenomenon by remarking that the outer parts of the snow-
crystal, which were the first to melt, borrowed their warmth
of liquefaction from the parts that remained solid, and thereby
cooled these below the point of congelation. The newly-
formed water could then freeze again by its collecting round
this cold ice, and by its offering at the same time a smaller
surface’ to the air whose temperature had melted the crystal.
This water then assumed in freezing a more complicated form,
because the remainder of the old crystal exerted in it a greater
variety of attraction than that which occurs in a wholly liquid
drop. Perhaps all complicated forms of snow [crystals] re-
sult from the simple one by melting and freezing again in
this way, a process which they must then undergo during
their fall thru the air; and here this hypothesis seemed some-
what confirmed by the complicated crystals being always of
less diameter than the simple ones.
Additional remark (April, 1859).-I have sometimes watched
the snow-crystals which fell at Berlin when the temperature
of the air was a little higher than the freezing-point, but till
now without seeing again the phenomenon just mentioned.
We may suppoee either that these observations were still too
rare to present some one of those neglected and apparently
t a n g circumstances that are requisite for the phenomenon
in question, or that this depended also on the spot where I
made my first observation having been at a considerable ele-
vation, and consequently not far from the atmospheric stratum
where the snow was first formed. But then, as to the expla-
nation oi the observed metamorphosis of snow, I think it
might have some connection with the equally obscure prop-
erty of some chemical precipitates, which, like carbonate of
lime, according to M. Ehrenberg, consist, when first consoli-
dated, of regularly arranged solid globules, and which are
then changed, “all of a sudden and quite wonderfully,’’ to aggregates of true crystals of microscopic size. (Cf. Ehren-
berg in Abhandlungen der Berliner Akademie, 1840.)
THE URYSTALIZATION OF UNDERC3OOLED WATER.
[Reprinted from the Physical Review, 1908, 27:509-510.1
In order to show the undercooling of water and to allow
the free development of its crystals I endeavored to introduce
into the undercooled water a piece of ice put in a finely drawn
By BORIS WEMBERQ. Dated St. Peteraburg, July, 1908.
-- ~~-____-
a VIz, the cum& eurface of a single, or of six conneoted drops.
I
out glass tube. The experiment, carried out the first time by
Michael Tvanov, gave an unexpected result. When the crys-
tallization attained the end of the tube there began to grow
at this point an ice crystal having the shape of an hexagonal
star and very similar to the characteristic snow crystals.
The greater the undercooling of the water the more numer-
ous were the ramifications and the greater the velocity of i crystallization. With water undercooled to a temperature .i between -0.3’ and -lo C. I obtained small stars with few 3 narrow ramifications, see fig. 1. Undercooling to a temperature ; between -lo and -3O C. gave rise to stars with such dense ’
ramifications that they resembled hexagonal plates, see fig. 2. ’ The plane of the stars contains the direction of the end of
the tube, and therefore when this end is vertical a sufficiently
large plate can divide the tube into two parts. An under-
cooling greater than -3” C., especially when the end of the
-3
1
tube is not narrow enough, prodbes several plates set in dif-
ferent azimuths, and the whole mass becomes at last a mass
of differently sized crystals and water, resembling the so-called
(( anchor ice.”
I
I
FIG. 1.
The crystals are often a conglomeration of several stars
which have their planes, their principal rays, and even the
ramifications of higher order parallel as in fig. 3.
If a star is broken, the pieces of it rise horizontally in the
water with slight oscillations and attain the surface. This
circumstance can explain the verticality of the o
river and lake ice.
FIG. 2.
The evolution of these artificial snow crystals can be easily
projected on a screen, if the vessel (a tumbler, an alembic, an
evaporating dish) with undercooled water is put into another . vessel with plane-parallel sides containing water at a tem-
perature somewhat higher than the thaw temperature [dew-
point?] of the surrounding air. For undercooling any water
can serve, but the refrigerating mixture (finely chppped ice
upon which is poured a strong solution of NaCl) must not be
too cold (from -4O to -6 O C.) and its level must be lower
than the level of the water which is to be undercooled.
The projection is especially beautiful when the vessel is
placed between crost nicols, as in figs. 1-3. A star on a
MONTHLY W E B
I@ound grows which gradually becomes more and more
kt and at last, when thick enough (the thickness is gen-
18 of the order of a tenth of a millimeter), shows the colors pmatic polarization. One can prove that these crystals
@tically uniaxial; if the tube is turned so that the plane ht the is at right angles to the rays of polari
of the star disappears.
II ll
ctor R. P. Stupart of the Canadian Meteorological Ser-
his letter of March 3,1909, states that during the past
he supplied barometers and a full equipment to the
g stations in extreme northern Canada:
Fort McMurray, latitude 56.40° N., longitude 111.25O W.
latitude 60.51O N., longitude 115.20° W.
latitude 64.57O N., longitude 125.0O0 W.
he observers will be paid for satisfactory service. This
has also just started two new stations in Newfoundland
Fort Norman,
THEORIES OF “HE COLOR OF THE SKY.
By EDWARD L. NICHOLS.^
Society, February 29, 1908. dential address delivered at the New York meeting of the Physical
[ABSTRAUT. ]
e author summerizes the various theories explanatory of
color of the sky, as follows:
other extraneous matter it would not, according to Ray- er, be optically empty, to use the term employed by Id be blue by virtue of reflections from the molecules
Ice of blueness, the color of the air ac-
s a blue sky by virtue of the selective
.THER REVIEW. li,
relatively to sunlight in proportion to the squaro of the wave-length. This is quite suftlcient to account for the average blueness of the sky, but not for the intenser blueness frequently observed. It cannot there-
fore be regarded as the sole or most important factor. 5. Fluorescence as a factor of blueness of the sky cannot be deflnitely considered at the present time for lack of experimental data concerning it. 6. As regards the subjective or physiological factor it may be said that were there no other muse the sky would undoubtedly appear blue; for we still see it blue where measurements with the spectrophometer indicate a composition relatively much weaker in the shorter wave-
lengths of the spectrum than the average composition of sunlight. In the present paper I shall, however, consider only the objective factors.
The problem of the color of the sky is stated as resolving
itself into a determination of the relative importance of these
various factors, the existence of all of which, with the possible
exception of fluorescence, may be regarded as experimentally
established. The phenomena of aerial polarization are believed
to indicate beyond any doubt that the turbidity of the air is
one source of the blueness of the sky. But while Rayleigh’s
masterly theoretical work-which calls for relative intensities
of the reflected ray as compared to the incident ray varying
inversely as the fourth power of the wave-lengths-has found
complete verification in the studies of artificial media, syec-
trophotometric measurements of the sky itself have led to
widely varying results. Thus, Zettwuch, who made many
measurements at Rome, calls especial attention to the varia-
bility of the ratios. Crova, at Montpellier, whose measure-
ments extend only. between wave-lengths 0.635~ and 0.510p,
found the exponent to vary from 1.61 to 6.44. The author
therefore seeks other sources than turbidity for the blue color
of the sky.
Numerous measurements of the spectrum of the sky made
by the author with a portable spectrophotometer show, in gen-
eral, f a r greater relative intensities of the longer wave-lengths
than one would expect from the theory of Rayleigh, which is
based upon the assumption of an ideal turbid medium in which
€he diameters of all the particles in the medium are small as
compared with the wave-length of light. The following are
given as obvious causes of the discrepancies between theoret-
ical and observed ratios of intensities:
(a) The presence of larger reflecting particles in the atmosphere, some- times invisible and sometimes forming masses of mist or cloud. (b) Absorption by transmission through the turbid medium itself. (c) Illumination of the atmosphere by light reflected from the surface of the earth.
Curves of ratios based on observations taken at dawn and
in the twilight after sunset, show but little variation from day
to day in fair weather, and approximate closely to the ratio
ourves ctalled for by Rayleigh’s equations. During the day,
while the sky-light taken as a whole increases greatly in in-
tensity as the sun approaches the zenith, the actual intensi-
ties of the blue and the violet me much less affected than are
the longer wave-lengths. When the moisture of the atmos-
phere condenses into cloud forms [cumulus] in the middle of
the day, there is a marked diminution in the relative intensity
of the sky-light at the violet end of the spectrum.
Evidence is found of the modifioation to a measurable ex-
tent of the character of the light of the sky by reflection from
foliage, from clouds, and from the ground.
Reference is made to Pernter’s study of the polarization of
light emitted at right angles to the incident beam by emulsions
of different colors. In general, the whiter the emulsion the less
the polarization, which is also true of the sky. For a blue
emulsion the green ray showed the greatest polarization, the
blue next, and then the red. For a white emulsion the red ray
showed the more polarization, there being a diminution toward
the violet. Pernter found this also to be true of blue and
white skies. The author found that the polarization of sky-
light was sometimes greatest in the red, sometimes in the
violet, sometimes in an intermediate color, and sometimes
uniform for all wave-lengths, probably depending upon the
size of the particles present in the atmosphere.