788 
will eventually attain a size such that they can fall 
against the direction of the upward current, instead 
of being carried up still higher. Whether these 
drops reach the earth’s surface or not will depend 
on the nature of the layers of air through which they 
must pass. If the latter are dry, and the upward 
current strong, they will have evaporated away before 
reaching the earth. If the height of the clouds is 
h,, and the height to which they are lifted is h,, then, 
before any rain can reach the earth, the ratio h,/h, 
must attain a limiting value a, which for Moscow 
has a value which varies from 2°5 to 3°5; when 
its value falls below this, irrespective of the degree 
of cloudiness, no rain can fall. The anomalously 
low rainfall of Swakopmund is explained, on this 
basis, to be due to the presence of a pronounced 
inversion caused by the sea-breeze, below the average 
cloud level, which inhibits the development of 
upward currents. Farther inland, where this factor 
does not exist, the rainfall is much greater. 
PROJECTION FOR AERONAUTICAL Maps.—In 1919 
the International Aeronautical Conference decided 
to adopt the Mercator projection for the general 
small-scale air map. In the Geophysical Supplement 
(vol. i. No. 3) of the Monthly Notices of the Royal 
Astronomical Society, Col. E. M. Jack and Capt. 
G. T. McCaw contribute a paper on the value of 
Mercator in this connexion, answering criticisms 
that have been made regarding its adoption. The 
substance of the paper was read to the British 
Association at Hull. The essential value of these 
small-scale maps is for aerial navigation; that is, 
for the direction of the course of the aeroplane from 
one point to another. In consequence it applies 
not to local large-scale maps but to maps on a scale 
of I: 2,000,000 or I: 3,000,000. In most criticisms 
of Mercator, the use of the polyconic has been 
advocated. The authors reply to these arguments 
by pointing out that the disadvantages of the poly- 
conic are threefold—the sheets do not fit together, 
the measurement of bearings is less simple and 
direct than in Mercator, and, with a single exception, 
no straight line represents anything in Nature save 
as an approximation. On the other hand, a good 
case is made for the use of Mercator, even if that 
projection, in common with all others, has some 
obvious defects. The authors answer the criticisms 
at length, but point to the essential quality that 
Mercator possesses, of representing a line of constant 
bearing as a straight line, as being of prime import- 
ance to the navigator. The varying scale does not 
trouble the aviator, but the bearing line enters into 
almost every problem with which he has to deal. 
Mercator enables him to solve these problems by 
simple graphic means, and, as a result, he prefers 
it to any other projection. At the same time the 
authors admit that, if flying beyond lat. 70° becomes 
common, some other projection will need to be used 
for polar regions, but this does not affect the problem 
at present. 
FREE-AIR PRESSURE MAPS FOR THE UNITED STATES. 
—Supplement No. 21 to the U.S. Monthly Weather 
Review is a discussion on “the preparation and 
significance of free-air pressure maps for the central 
and eastern United States,” by Mr. C. Le Roy 
Meisinger. The aim of the author is to develop 
barometric maps of free-air levels which shall have 
a direct and important bearing upon accurate fore- 
casting for aviation. The discussion opens with 
“the history and problems of American barometry,”’ 
and an explanation is given of the relations involved 
in the reduction of pressure from one level to another, 
temperature introducing a considerable difficulty. 
NO. 2797, VOL. III] 
NATURE 




[JUNE 9, 1923 
The author, referring to practical experience being 
required to demonstrate the value of free-air pressure ~ 
maps, and stating that the United States does not 
possess this experience, mentions that in Japan 
daily maps of the 3-kilometre level have been drawn 
for several years and are of great service to Japanese 
forecasters. Details are given of Dr. R. Sekiguchi’s 
experience in forecasting from thesé maps in Japan. 
Cyclonic centres in the Far East show better agree- 
ment with the trend of the isobars on the 3-kilometre 
maps than with that at sea-level. Various interesting 
examples are given, and probably valuable informa- 
tion may be gained from these for other parts of 
the globe. A large number of specimen maps are 
shown for sea level, 1-kilometre level, and 2-kilometre 
level, while details are given of each series. The 
work is of considerable importance to aviation, and 
the upper-air charts show what the winds are doing 
aloft when it is often impossible to gain the informa- 
tion from the sea-level charts. It is claimed that the 
discussion affords a glimpse of the physical processes 
at work, and may help to lift us from empiricism a 
little nearer to that ultimate goal toward which 
students of weather forecasting are striving. 






















































ENDURANCE LIMIT OF STEELS. — Engineering 
Bulletin No. 136 of the University of Illinois contains 
an account of further experiments on the fatigue of 
metals, conducted by H. F. Moore and T. M. Jasper. 
The new results, which have been obtained by 
similar methods to those described in a former 
Bulletin, confirm the existence of a true endurance 
limit for steels of different compositions. For 
moderately hard steels, this limit may be found by 
rapid tests in which the rise of temperature is 
measured. Only a few preliminary tests have been 
made by Gough’s method of determining the increase 
of deflexion as the range of stress is increased. A 
fairly close correlation is found to exist between 
the endurance limit and the ultimate tensile strength 
or the Brinell hardness, and a much less close con- 
nexion between the endurance limit and the yield 
point and limit of proportionality, there being no 
correlation with the impact or repeated impact — 
values. It should be remarked, however, that the 
range of materials studied was not very wide, and 
the authors did not examine defective materials. 
A formula for the effects of cycles of stress not — 
involving complete reversal is proposed, the results 
not being in agreement with Goodman’s diagram. 
A New Fitament ELEcTROMETER.—In the issue 
of the Physikalische Zeitschrift for April 15, Dr. C. W. 
Lutz, of the Geophysics Observatory, Munich, de- 
scribes an improved form of filament electrometer, 
which has been constructed by the firm of Edelmann, 
for use in the observatory. The filament is of 
platinum, and its lower part is attached to one end 
of a diameter of a circular loop of quartz fibre; the 
other end of the diameter of the loop being attached 
to the sliding frame by means of which the whole 
may be removed from the instrument for replace- 
ment of the filament. The deflectors between which 
the filament is placed are adjustable by means of 
screw heads outside the metal cover of the instru- 
ment. The filament is observed through a micro- 
scope magnifying 285 times, and, for distances of 
the deflectors from the filament exceeding 4 mm., 
the deflexions are proportional to the applied vol 
over the whole of the scale of the ocular of the 
microscope, which is graduated from a central zero 
to 50 on each side. The insulation throughout is 
amber. The sensitivity of the instrument may 
varied by the usual methods from o-oor to 2 volts 
per division of the scale. 
