Fesruary 28, 1907] 
NATURE 
427 
(2) It has been recently shown by Messrs. Stanton and 
Bairstow (Proceedings: of the Institution of Civil Engineers, 
1906) that the primitive yield-point of a rolled or forged 
steel is usually an artificial figure, and is due to a stiffen- 
ing caused by this mechanical treatment. Such is not the 
case with these alloys. Their primitive yield-point is the 
true one. 
(3) The duectilities (considered as a product of the per- 
centage elongation and reduction of area) of alloys con- 
taining from o-1 per cent. to 7-35 per cent. of aluminium 
are very high and practically constant, even although the 
tenacity increases markedly with rise of aluminium. 
(4) The tenacity and ductility of the widely-known 
“aluminium bronze ”’ or ‘‘ gold,’’ containing 10 per cent. 
of aluminium, have been found to be as good in the 
form of small chill castings as in the rolled bar, where 
an 80 per cent. reduction of area of the original ingot has 
been effected. So far as the authors have been able to 
learn, this result has no parallel. At their request, there- 
fore, independent tests were instituted at the Broughton 
Copper Works, and these have confirmed the above result, 
which may have important practical consequences. 
(d) The research has brought to light several striking 
instances of the profound influence of a small quantity of 
aluminium upon copper, notably in the tension tests, but 
especially in the torsion and electrical conductivity experi- 
ments. QOne-tenth of 1 per cent. raises the angle of twist 
of copper in torsion 90 per cent.; it lowers the electrical 
conductivity 23 per cent. 
(e) The behaviour in torsional stress of the alloys con- 
taining from o-r per cent. to 7-35 per cent. of aluminium 
is one of the outstanding features of the report. 
(f) The alloys containing from 5 per cent. to 10 per 
cent. of aluminium have come well out of the dynamic 
stress tests. The particular merit of alloys Nos. 9 and 13 
when tested in alternating stress is the close approximation 
of the maximum stress under which they will bear an 
unlimited number of reversals to the stress at the elastic 
limit as determined in a tensile test. In this respect they 
are markedly superior to the iron and steel specimens 
hitherto investigated. 
Alloys Nos. 6, 9, and 13 stood up well when repeatedly 
stressed beyond the yield-point in Arnold’s test. In fact, 
Prof. Arnold has informed the authors that ‘‘ Alloy No. 9 
constitutes a record in its capacity of resisting alterna- 
tions. ’ ‘ 
(g) At about 15 per cent. of aluminium the 
entitled to rank with quenched steels in hardness. 
the hardness number of No. 
cast 
0-45 
‘ 
alloys are 
Thus 
17 (15:38 per cent.) in the 
state (untreated) is’ 539, which is about that of a 
per cent. carbon steel quenched in water at 20° C. 
(68° F.), and is only slightly lower than that of a 0-66 per 
cent. carbon steel similarly treated. 
(h) In the corrosion tests, which were purposely made 
as severe as possible, alloys containing from 1 per cent. to 
10 per cent. of aluminium have shown themselves to be 
practically incorrodible by sea-water, whether alone or 
bolted to a plate of mild steel: In these tests they 
showed themselves superior both to Muntz metal and 
naval brass, which corroded appreciably. In tap water 
of medium temporary hardness the positions were exactly 
reversed. 
(k) In view of the discussion in the previous report as 
to the trustworthiness of temperatures measured with a 
protected thermo-junction, the exact influence of the jacket 
(a fire-clay tube 1/16th inch in thickness) between tempera- 
tures of about 110o® ©. and sso° C. (2012° F. and 
1022° F.) has been determined. It has been found to 
cause a lowering of not more than 3° C. (5°-4 F.) at the 
higher, and 9° C. (16°-2 F.) at the lower temperature, and 
above 800° C. (1472° F.) comes within the experimental 
errors and uncertainties of the method. 
(1) Finally, a special comment must be made on the 
truly extraordinary similarity in physical and mechanical 
qualities between alloy No. 13, which consists of 90-06 per 
cent. of copper +9-90 per cent. of aluminium, and Swedish 
Bessemer rolled steel of about 0-35 per cent. of carbon 
and thirty-eight tons per square inch ultimate tensile 
stress. 
NO. 1948, VOL. 75] 
A PROPOSED INTERNATIONAL 
THE SIDEREAL 
ATTACK 
PROBLEM. 
ie a brochure’ written by Prof. J. C. 
Groningen Astronomical 
ON 
Kapteyn, of the 
g Laboratory, the author out- 
lines the chief points of a very comprehensive attack, 
which he proposes should be made as soon as possible, 
on the main problems concerning the structure of the 
sidereal universe. 
Whilst the “‘ Carte du Ciel,’’ parts of which are now 
approaching completion, gives us the relative projected 
positions of all the stars down to the eleventh magnitude, 
and will, by duplication after a number of years, afford 
material for the accurate determination of proper motions, 
it leaves untouched the extremely important question as 
to the distribution of different stellar types in actual space. 
Prof. Kapteyn proposes to supplement this enormous work 
by the preparation of a Durchmusterung which shall con- 
tain all the necessary data for a preliminary discussion of 
the structure of the universe. In fact, he proposes that 
in the same way that the geological has supplemented the 
geographical study of the earth, so shall an astrological 
supplement our astrographical study of the heavens; but 
it is obvious that to attempt a scheme like this for the 
whole of the heavens at once would be to court failure. 
The plan would probably die of senile decay ere it showed 
sufficient results to have justified its existence. For this 
reason, and acting on the advice of eminent astronomers 
who favour the idea of such a survey, Prof. Kapteyn 
limits his proposals to a number of selected ar of the 
sky. This would reduce the work immensely, and would 
probably lead to a first approximation of the truths which 
it is hoped to educe. 
The general scheme is based on the method of “ gaug- 
ing "’ as carried out by the Herschels, only that now, in- 
stead of considering simply. the numbers of stars, every 
ascertainable fact in regard to the objects studied must 
be considered. The chief data to be obtained, as enumer- 
ated by the proposer of the scheme, are visual magni- 
tudes, photographic magnitudes, spectral types, astro- 
nomical proper motions, radial velocities, and parallaxes, 
to which list he adds the determination of the amount of 
light received from different parts of the sky, as being a 
subject of great importance to the problem under con- 
sideration. 
As Prof. Kapteyn points out, there are already sufficient 
data for the brighter stars, partially excepting parallax 
and photographic magnitudes, to allow of a fairly thorough 
statistical treatment, but much of this data needs a great 
amount of arrangement and classification ere it can be 
included in a homogeneous attack. The great need in 
such an inquiry as that proposed is the international study 
of the fainter stars. Work already completed, or now in 
hand, will take us down to the seventh or eighth magni- 
tude for most of the elements named, but it is self-evident 
that, in any attempt to solve the riddle of sidereal struc- 
ture, the Milky Way is an all-important feature, and, 
therefore, far fainter magnitudes than this must be in- 
cluded. 
Put into its briefest form, the scope of Prof. Kapteyn’s 
proposals is :—‘ For 206 areas regularly distributed over 
the sky, and for another less extensive series of particu- 
larly interesting regions, to obtain astronomical data of 
every kind for stars down to such faintness as it will be 
possible to get in a reasonable time.’’ The 206 areas first 
named come under the designation of “‘ the systematic 
plan,’’ and are again divided into two classes, the first of 
which would comprise 118, and the second eighty-eight 
areas. These are so arranged that the first class might be 
completed independently of the second, and would furnish 
sufficient data for a first approximation. Then, if there 
were evidence that this could be executed in reasonable 
time, the second class might be intercalated without inter- 
fering with the other, except to provide further data which 
would, in all probability, enhance the value of the final 
1 “Plan of Selected Areas.’’ 
By Prof. J. C. Kapteyn. 
1906. 
(Groninge, 
