186 

during which the fire raged. It was visible from 
points near the fire—a few hundred yards—but was 
more striking from points a quarter to a half a mile 
away, from which the flames themselves could not 
be seen. It seemed to vary with some atmospheric 
condition (more or fewer snow crystals in the air ?) as 
it might be dim when the diffuse reflection of fire-light 
from the low-lying clouds would be brightest, and 
might be sharp and bright during a lull in the flames. 
It was, however, often most striking when the con- 
flagration was at its height. 
It was not due to shadow of the still-standing walls 
—the light coming from the burning interior—for it 
was first noticed when the roof had just caught fire 
and most of the light came from the burning shingles 
—a case where no wall shadow was possible. Further, 
the beam was sensibly parallel-sided. A wall-shadow 
would have given a broadly diverging beam. The 
explanation offered is that of reflection from falling 
flat snow crystals, which, of course, were not over the 
burning building but were distributed in the atmo- 
sphere between the observer and the source of light. 
The official record at the Queen’s University station 
of the Canadian Meteorological Service, taken just 
before the fire broke out and less than a mile from 
Sydenham Hospital, gave ‘“‘ Temperature 12° Fht., 
Wind N.E. 20 miles per hour, Light Snow.” In fact, 
the snow-fall was so light that the record of precipita- 
tion over the twelve hours, including the time of the 
observations on the beam, was only 0-03 inch. 
WILL C. BAKER. 
Physical Laboratory, Queen’s University, 
Kingston, Ontario, Canada. 
January 8. 

Unusual Crystals. 
THE following may be of interest to readers of 
NATURE: 
I have a bottle of pure phenol which has not been 
opened for a dozen years. During this period I have 
been interested to watch the growth of crystals from 
the sides of the empty portion of the bottle by 
sublimation. These crystals are cylinders or prisms 
many of them between two and three centimetres in 
length and as many millimetres in diameter. The 
ends are not pointed but neatly trimmed off by an 
oblique plane. 
On closer examination these crystals prove to be 
thin walled tubes. The stalk attached to the bottle 
is solid for a few millimetres. Then a fine capillary 
appears, spreading out conically until the wall is about 
a half millimetre thick’ and then continuing as a 
uniform tube. The explanation is of course that 
within the tube the air is just saturated with phenol 
vapour while outside it is slightly supersaturated. 
I do not remember meeting any published descrip- 
tion of such crystals. G. H. Martyn. 
38 Hogarth Hill, 
Hampstead Garden Suburb, N.W.11. 
January 19. 

Science and Armaments: 
Wuat Dr. Martin regrets, in his letter to NATURE 
of January 20, p. 82, is to me a consolation—to know 
that scientific men in our universities are still working 
for the safety of the realm, for across the Channel 
there are fierce black clouds and ominous rumblings 
of strife that seem almost beyond control. 
Dr. Martin says: ‘“‘ So may the temple of science 
be kept free from echoes of human quarrels,’’ and 
instances the sojourn of Davy and Faraday in Paris. 
Is the example fortunate? Davy was irresistibly 
attracted to Paris by reports of a detonator of fearful 
NO. 2780, VOL. 111 | 
NATURE 
[ FEBRUARY 10, 1923 
violence that had already deprived Dulong—its dis- 
coverer—of an eye and a finger. He spent much of 
his time there investigating another discovery of a 
manufacturer of salpetre, a substance not unknown 
to Ministers of Munitions. 
It was this very journey that occasioned the 
human quarrel that we seek to forget when con- 
templating the lives of these two great priests. of 
the temple of science. JAMES WEIR FRENCH. 
Anniesland, Glasgow, January 22. 

The Opacity of an Ionised Gas. 
In a paper read before a joint meeting of the 
American Physical Society and the American 
Astronomical Society, in December, I pointed out 
that theoretically the absorption of radiation by 
free electrons should render an ionised gas highly 
opaque. The organised vibrational energy, due to 
the radiation, of the free electrons is transformed 
by collisions into disorganised thermal energy of 
translation. A tentative application of the methods 
of the well-known free electron theory of the optical 
properties of metals to conditions in an ionised gas_ 
gives the following equation for the volume opacity 
coefficient K. The quantity K is such that in 
distance z centimetres through the gas the intensity 
of the direct beam is reduced to e~** of its initial value. 
NAD? 
T?(r +2)? 
Here \ is the wave-length of the radiation in centi- 
metres ; 7 is the ratio of the number of free electrons 
to the number of atoms and ions; # is the gas 
pressure in atmospheres, including the partial 
pressure ip/(I+7) of the free electrons; T is the 
absolute temperature, Centigrade; and A is the 
radius in centimetres of an atom or ion (assumed 
equal in size—a very rough approximation). This 
type of opacity increases as the square of the gas 
pressure, while the opacity due to general scattering 
increases only as the first power of the pressure. 
Even at fairly low pressures, however, the effect of 
absorption predominates in an ionised gas. 
The above equation follows from the following 
assumed equation of motion of a free electron in 
an ionised gas through which radiation is passing : 
du 
™m dt 
where m is the mass, and e the charge of an electron, 
and uw is its component velocity in the direction 
of the electric vector X of the radiation. The term 
2ymu represents a pseudo-frictional resistance due 
to collisions between electrons and atoms or ions ; 
y is the number of such collisions per second per 
electron. The usual assumption is made that the 
velocity of an electron after colliding with an atom 
is independent of its velocity before collision; and 
collisions between electrons are neglected. (When 
the scattering of radiation by free electrons is dealt with 
a term involving —d?u/di? is added to the left-hand 
member of the equation.) The average rate at which 
energy is absorbed from the radiation by each 
electron is the average value of feXudt; and, 
remembering that the intensity I of the radiation 
is cX,?/87, where X, is the amplitude of X, K is 
easily found. The number of collisions per second per 
electron is taken as tNA?,/3RT/m, where N is the 
number of atoms and ions per unit volume, and R is 
the gas constant per molecule. This relation assumes 
equipartition of energy between free electrons and 
the other molecules in the gas. 
The well-known experimental work of Dr. Anderson 
K= (6-7 x 1075) 
+2v¥mu=eX, 
