336 
volume charts are given for the several months of 
ocean currents and winds, also isobars and isotherms 
of air and sea, with routes, trajectories of cyclones, 
limits of fog, ice, and of the trade-winds and mon- 
soons. For the charts the observations are in some 
cases grouped into 5° squares, and in others into 
2° squares. The results are of a very high order 
owing to the large number of observations available 
over the greater part of the Indian Ocean, and the 
method of dealing with the observations is about all 
that could be wished. The data are limited solely to 
the sea, but probably in some cases land data might 
have been combined with advantage. In the pressure 
charts isobars are given for each 2:5 millimetres, 
There are numerous islands over the Indian Ocean at 
which long series of observations have been main- 
tained; the entry and use of these stations would 
have enhanced the value of the work. 
UnpbeER the joint auspices of the New York State 
Commission on Ventilation, the American Museum of 
Natural History, and the American Museum of Safety, 
an investigation of three of the best methods of deter- 
mining the amount of dust in air has-been carried out | 
by Messrs. G. T. Palmer, L. V. Coleman, and H. C. 
Ward, and the results obtained are given in vol. vi. 
of the American Journal of Public Health. In the 
first of the methods investigated the dust-laden air is 
forced against a surface smeared with glycerine, and 
the number of dust particles caught by the glycerine on 
selected areas of the surface counted under the micro- 
scope. In the second method the air is drawn through 
syrup, which is afterwards diluted with water, a-drop 
of which is placed in a shallow cell under the micro- 
scope, and a count made in the same way. In the third 
method the air is drawn through a spray or curtain of 
water, which is then examined under the microscope as 
in the second method. In the second and third methods 
a rough determination may be made by a comparison 
of the turbidity of the water with that of a series of 
prepared samples. The authors conclude that the 
water-spray method is the best and most convenient 
to use under normal conditions, and they give a num- 
ber of directions as to the best way of carrying out 
the observations. 
AccorDING to the Chemical Trade Journal of Decem- 
ber 9, the Secretary of the U.S.A, Department of Agri- 
culture has decided to erect on the coast of southern 
California a Government plant for the production of 
potash from kelp. Unlike the private companies 
which are already manufacturing potash from this 
source, the Government is determined not only to 
produce the potash at the minimum cost, but also to 
conserve the nitrogen, iodine, and other by-products. 
In the Government plant the kelp will first be dried 
in a series of rotary driers. It will then be distilled 
in a modified coke-oven in such a way as to prevent 
the loss of the nitrogen, iodine, and other by-products. 
The potash salts will be dissolved with water out of 
the resulting charcoal, which may afterwards be used 
as fuel. The combustible gas obtained by the distilla- 
tion may also be used as fuel. The U.S.A. Govern- 
ment experts hold that by such economical methods 
the process can’ be made to pay in ordinary times. 
We have received from Messrs. Dulau and Co., 
Ltd., their December catalogue of botanical and 
natural history works. It is classified under botany 
and horticulture; orchids and orchid culture; mam- 
mals, birds, and reptiles; entomology; mollusca, and 
periodical publications. Many rare and out-of-print 
books are included. We notice that a set of NaTuRE 
from 1869 to 1907 is listed at a low price. 
NO. 2461, vor. 98] 
NATURE 
| DECEMBER 28, 1916 
OUR ASTRONOMICAL COLUMN. ' 
Mercury aN Eveninc Star.—Mercury will be at its 
greatest elongation east on the morning of January 3, 
and visible on several evenings before and after that — 
date over the W.S.W. horizon about 3h. or th. after 
sunset. The planet will be visible to the unaided eye 
should the atmosphere be very clear at the important 
time, though this apparition will not be so: favourable 
as the one which will occur on April 24 next. — Be- 
tween about December 28 and January 10 next Mer- 
cury will set about 1} hours after the sun, and, with 
good weather conditions, the ruddy, scintillating light 
it emits ought to be detected without difficulty, though 
it will be at a very low altitude. 
Torat Eciirse or THE Moon,—This phenomenon 
will take place on the morning of January 8, and the 
following are the astronomical times of its various 
stages :— 
Janeesie 
. mM. 
Moon enters penumbra 16 36 
Moon enters shadow 17 50 
Beginning of total eclipse 19 0 
Middle of eclipse 19 45 
End of total eclipse 20 29 
Moon leaves shadow 21 39 
Moon leaves penumbra 22 53 
Magnitude of eclipse=1-369 (moon’s diameter=1-0). 
At Greenwich the moon sets at 20h. 1om., while the 
moon is totally eclipsed. The various phases of the 
event may therefore be watched until after the middle 
of the eclipse. It will be interesting to observe whether 
the earth’s shadow proves a very dark or light one- 
on this occasion. At some lunar eclipses the disc of 
our satellite has remained quite bright, at others it 
has been scarcely visible at all. It usually happens 
that, at the time of a lunar eclipse, the atmosphere 
is sufficiently favourable to allow successful observa- 
tion. : 
LonG-PERIOD VARIABLE StTars.—. \ valuable contribu- 
tion to the investigation of variable stars was recently 
made by the Rey. T. E. R. Phillips, in a presidential 
address to the British Astronomical Association (Journ. 
B.A.A., vol. xxvii., p. 2). Observational data for 
such stars have been much extended by the carefully 
determined light-curves of twenty-one long-period stars 
observed by the members of the Variable Star Section 
of the association, and by sixty-seven light-curves pub- 
lished in the Harvard Annals. Mr. Phillips thas 
analysed the curves of eighty stars, and from a com- 
parison of the phases of the second and third har- 
monics he finds that the long-period variables fall into 
two distinct groups. This grouping is supported by 
the evidence of the ranges of variation, and, in a 
somewhat less degree, by the coefficients, brightness, 
and periods. The general characteristics of the two 
classes are also shown in the light-curves. In 
Group I. the intervals from maximum to minimum, 
and from minimum to maximum, are not very unequal, 
and there is a tendency for the curve to show a pause 
on the rise. 
maximum, followed by a long period of decline. As 
a rule it is possible to determine the group to which 
a star belongs by mere inspection of the light-curve, 
but harmonic analysis is necessary to fix its place in 
the group. The stars of Group I. are brighter than — 
those of Group IT., and it is possible that they may be Bs: 
found to include “‘ giants"’ and ‘‘dwarfs” respectively. 
The inequalities in the intervals between successive — 
maxima, and in the brightness at different epochs, call — 
for much further investigation. 
In Group II. there is a sharp rise to a 
