2 
96 
WaALORE 
[JANUARY 28, 1897 
stopcocks are wide open, any adjustment that is con- 
sidered desirable for the level of the liquid standing 
higher than the other in its glass tube, may be deliberately 
made by drawing out or letting in a little air through the 
lower air-pump stopcock. 
Either or both liquids may be thoroughly stirred at any 
time to ensure homogeneousness by alternately exhaust- 
ing and letting in air to the bottle or bottles by means of 
the air-pump and the lower air-pump stopcock, the upper 
three stopcocks being kept closed. 
Operation No. 6 of my article on the subject, in last 
week’s NATURE (p. 274), must be performed as often as is 
found necessary. Every one of the stopcocks must be 
kept closed except when it is open for operation or 
observation. 
The metal tube connecting the upper necks of the two 
bottles must be long enough, or of fine enough bore, to 
prevent diffusion of vapour to any sensible extent from 
either bottle to the other during the time of an observ- 
ation. It ought to be kept at a temperature somewhat 
higher than that of the bottles, to prevent any liquid from 
condensing as dew on its inner surface. KELVIN. 
Glasgow, January 23. 
THE GRAVITATION CONSTANT AND THE 
MEAN DENSITY OF THE EARTH. 
ips 
the year 1884, Prof. Dr. Franz Richarz and Dr. 
Otto Krigar-Menzel commenced a series of experi- 
ments having for their object an accurate determination 
of the values for the constant of gravitation and the 
earth’s mean density. The work divided itself naturally 
into two parts, and the results of the first series of 
weighings were communicated to the Berlin Academy 
of Sciences in March 1893. Since that time the second 
series has been concluded, and the main results of the 
whole investigation are summed up briefly in the same 
society's S7/zunegsberichte for November 1896, the authors 
leaving the publication of the full account, containing the 
details of the measurements, for a future period. 
The instrument with which the measurements were 
made was a kind of double-balance having two pans on 
each side of the beam, one above the other, connected in 
the vertical direction by a thin rod 226 centimetres long. 
The point of this arrangement was that the acceleration 
due to gravity on the two lower pans was greater than 
that on the upper one, in consequence of the difference 
of level. The first day’s weighing consisted in de- 
termining the difference in weight of two practically 
equal spheres, one being placed in the upper pan on 
one side of the arm of the balance, and the other in the 
lower pan on the opposite side. Gauss’ system of double 
weighing was employed throughout, the masses being 
changed from one side to the other. The measured 
difference was therefore due to two sources—to the 
inequality of the masses weighed, and to the difference 
of the force of gravity. The procedure for the second 
day was to change the positions of the masses being 
weighed ; this consisted in placing the sphere that was 
in the upper pan in the lower one on the same side of 
the arm, and in putting that on the other side of the 
beam in the upper pan: a second series of weighings 
was then made. Such a series of measurements as these 
was included in the work recorded in the first publication. 
It was found that the measured difference was not the 
same on the two days, for although the difference in 
weight of the masses always remained the same, the 
difference in the value of gravity, due to the virtual 
displacement of the masses, altered its sign. By sub- 
tracting these differences for each day’s work, the mass- 
difference was entirely eliminated, and there only re- 
mained that between the two values of gravity due to 
the two heights of the scale-pan. 
NO. 1422, VOL. 55] 
In the second part of the work, for obtaining the mean 
density of the earth, a large cubic block of lead, having 
a mass of 100,000 kilograms, was used. This was sup- 
ported firmly on massive pillars under the upper and 
above the lower scale-pans, the connecting-rods of the 
latter passing through holes in the block. The presence 
of this great attracting mass had the effect of apparently 
increasing and decreasing the value of gravity acting on 
the spheres in the pans according to their position—that 
is, according as they were in an upper ora lower pan. In 
the system of arrangement adopted—namely, that of 
placing the spheres in an upper and lower pan on op- 
posite sides of the fulcrum—the acceleration of gravity in 
the upward direction was lessened by double the amount 
of the attraction of this great mass of lead. By connect- 
ing to two days’ weighings, instead of double the dimi- 
nution of the value of gravity with the height, the result 
was lessened by four times the attraction of the block 
of lead. A combination of the results obtained, both 
with and without the leaden cube, gave the fourfold 
attraction of the leaden cube free from the variations of 
gravity. 
Very elaborate precautions were taken to eliminate 
effects of air currents, changes of temperature, &c., and 
these seem to have been fairly overcome both practically 
and theoretically. Itis also needless to state that the 
number of weighings made was very considerable. 
The value thus finally deduced for the constant of 
gravitation was 
G = (Gib85)-F 0°01) 1055 ———— 
gr. sec.* 
Using this value, the mean density of the earth obtained 
was as follows :— 
= Yee : gr. 
A = (5°505 + 0009) aa 
This value, as will be seen from a perusal of the 
following table, falls between those found by Poynting 
and Boys :— 
Mean density of 
Observer. Method. edith: 
Cavendish Torsion balance 5°45 
Reich... Re 5°49 and 5°58 
Baily © <.. eae a ” 5°67 
Cornu and Baill at oF 5°56 and 5°50 
Ph. von Jolly Long-arm balance... 5 °692 
J. Wilsing Pendulum apparatus 5°594 
The same with elimina- 
tion of known sources > 5°577 
of error cor 
J. H. Poynting... Balance «=» StaoRe 
- a {Improved torsion \ _, 
C. V. Boys ee Tance j; 5°5279 
Richarz d  Krigar- Sai 
itedeiel ane is a | Double balance 57505 
- TUBES FOR THE PRODUCTION OF RONTGEN 
wal VAS. 
ANY and various are the forms of the bulbs and 
tubes employed for the production of Rontgen 
rays, as may be seen from the designs that are repre- 
sented in the accompanying illustration from La Nadure. 
Different experimenters favour different tubes, and believe 
that the forms they use possess advantages over all others. 
Up to the present, however, it may be said that three 
processes are utilised in the production of Réntgen rays. 
There is (1) the old form of Crookes’ tube, in which the 
kathode rays impinge directly upon the glass or screen 
in front of the kathode ; (2) the form of tube in which the 
kathode rays fall upon, and are reflected by the anode ; 
and (3) the tubes in which both direct and reflected rays 
are utilised. In the accompanying illustration, the first 
