30 
impure cast-iron at the present day—and whereby all 
the impurities would be removed to the surface to 
form the primal crust. The first of this primal crust 
would be the most acidic, and the last the most basic; 
and in the metallic core left, even the oxygen which 
had been the means of removing all the impurities 
would itself be undetectable. Thorium is but a higher 
member of the silicon (carbon) family as uranium is 
of the oxygen family, and the conditions which remove 
the lower members should be effective in removing 
the higher ones also. So, too, it seems only natural 
that the most acidic rocks—that is, the rocks con- 
taining the greatest proportions of silicon and of 
oxygen—should at the same time be associated with 
the greatest proportions of thorium and of uranium, 
which are but the highest members of the silicon and 
oxygen families. 
On the planetesimal hypothesis a similar distribu- 
tion of the radio-elements can hardly be imagined. 
To get a metallic core that shall be free from thorium 
and uranium, we have to imagine the planetesimals 
undergoing individually the oxidation process which 
has just been sketched, unless the planetesimals hap- 
pened to be fragments of a preformed stellar mass. 
Provided all went well, when these planetesimals were 
piling themselves together to form the earth, the 
result would be a metallic core free from thorium and 
uranium, but surrounded by a crust in which these 
elements would be uniformly distributed. As the 
acidic rocks differ only in degree from the basic rocks, 
it would be impossible for the former to rise up 
through miles of a mixture of both to form an acidic 
layer, as happens in the case of a stellar earth. Sub- 
sequent aqueous action is relatively negligible in both 
cases. 
It is probably the exigencies, of the planetesimal 
hypothesis that constrain Mr. Holmes to state that 
there is ‘clearly a marked antipathy between the 
radio-elements and native iron, for in all the terrestrial 
examples which have been examined uranium and 
thorium are barely detectable.” As a matter of fact, 
these elements alloy with iron and nickel, which are 
the constituents of native iron; and their marked 
absence is a proof that native iron had undergone an 
oxidation process at one time in its history, and so 
had its thorium and uranium removed. 
95 Bath Street, Glasgow. GEORGE CRAIG. 
THE INTERNATIONAL UNION FOR SOLAR 
RESEARCH. 
HE fifth conference of the International Union 
for Cooperation in Solar Research was held 
at Bonn, by invitation of Prof. Kayser, from 
July 31 to August 5. The attendance was about 
100, including delegates from nearly every country 
in Europe, and a large contingent from America. 
In the absence of Prof. Hale and Dr. Schuster, 
through sickness, the executive committee was 
ably represented by Prof. Turner. As on former 
oceasions, the chief business at the general meet- 
ings was to receive and discuss the reports of the 
various committees appointed for the organisation 
of observations and methods of reduction. 
An important part of the work of the union has 
been that relating to standards of wave-length. 
At the last meeting, held at Mount Wilson, Cali- 
fornia, in-rgro, it was believed that a final set of 
standards was well within sight, but further in- 
vestigations liave revealed unexpected difficulties. 
It has, in fact, been found by Goos and others that 
the wave-lengths of many of the iron lines vary 
NO. 2289, VOL. 92| 
NATURE 
| 
[SEPTEMBER 11, 1913 
slightly with the length of the arc and the portion 
of the are observed. Fortunately, most of the 
lines already adopted as secondary standards, 
from interferometer determinations, preserve their 
positions under a variety of circumstances, but 
they are not sufficient in number. Hence, it has 
become necessary to attempt the definition of a 
standard iron arc for determinations of further 
standards, and for the production of reference 
spectra to be used in the determination of wave- 
lengths by interpolation. The committee recom- 
mended the following specification for the iron 
arc: (1) Length of arc, 6 mm. (2) For lines of 
wave-length greater than 4000, current to be 
6 amperes, and for lines of shorter wave-length 
4 amperes, or possibly less. (3) Direct current, 
positive pole above negative, P.D. of 220 volts, 
electrodes being iron rods of 7 mm. diameter. 
(4) An axial portion of the arc, at its middle, about - 
2 mm. in length, to be used as the source of light. 
Cooperation in the determination of tertiary 
standards is desired from all who possess concave 
gratings, plane gratings, or prisms of sufficient 
dispersion and resolving power, and to this end 
it is recommended that additional secondary 
standards be determined with the interformeter, 
so that the interpolation method may be used with 
greater exactness. : 
At the Mount Wilson meeting a committee on 
the determination of the solar rotation by means 
of the displacements of lines was formally organ- 
ised, and a programme of research agreed upon. 
Each cooperating observer undertook observa- 
tions, at specified latitudes, in a definite region 
of the spectrum, in addition to a control region 
common to all the observers, and lines were to be 
selected so as to include elements of widely dif- 
ferent atomic weights. It now appears that dif- 
ferent observers may obtain results differing sys- 
tematically by as much as 10 per cent. from one 
another, and that serious discrepancies have also 
been found in the results obtained by different 
observers from measurements of the same photo- 
graphs. The committee accordingly recommended 
that, before proceeding further, investigations of 
these sources of error should be made by deter- 
minations of velocity at the solar equator by as 
many different methods as possible. 
Satisfactory progress in work with the spectro- 
heliograph was reported, but the Committee hoped 
that additional observatories would install instru- 
ments of high dispersion, in order that filaments 
and alignments in the upper atmosphere might be 
more completely recorded. As the result of répre- 
sentations made by Prof. Ricco and by the Royal 
Astronomical Society, the title of the spectrohelio- 
graph committee was changed to “Committee on 
Solar Atmosphere,” so as to include and unify all 
the observations on the solar atmosphere, visual 
and photographic, except those associated with 
eclipses. The organisation directed by Prof. 
Ricco, and other observers of prominences, were 
thus given the opportunity of closer connection 
with the union. After some discussion, a. sub- 
committee for visual observations of prominences 
and related phenomena was appointed, with Prof. 
