JuNE 23, 1923] 
NATURE 
849 
The E6tvés Torsion Balance and its Use in the Field. 
By Capt. H. SHaw and E. Lancaster Jones. 
| eee problem of locating from the surface mineral 
deposits in the interior of the earth presents 
numerous difficulties, and has been considered by many 
investigators who have employed methods based on the 
physical properties of these bodies, utilising electrical, 
magnetic, seismic, and density effects. Speaking gener- 
ally the useful minerals are of either very high or very 
low density, so that a method dependent upon the 
difference in density between the mineral and its 
surroundings would appear to be suitable for the loca- 
tion of such deposits, especially as these effects are 
noticeable at a considerable distance. 
The attraction due to a heavy body is superimposed 
upon the normal force of gravity at any point in its 
Vicinity, and it is by observations on this local field by 
means of a torsion balance that the disturbing mass 
may be located most readily. 
The torsion balance of Michell and Cavendish was 
redesigned and employed for this purpose in 1888 by 
Baron Roland von Eétvés, professor of physics at 
Budapest, who gave to the balance a new construction 
which is retained essentially in the modern instrument, 
although numerous improvements have since been 
introduced. 
The essential features of the balance can be seen on 
reference to Fig. 1. A light aluminium beam loaded 
with platinum weights is suspended by a fine platinum 
iridium torsion wire. 
directly to one end of the beam, but at the other end 
the weight is suspended about 60 cm. below the beam 
by means of a fine wire. 
The forces of gravity acting on the two masses are 
not wholly vertical, but have small horizontal com- 
ponents which give rise to a minute horizontal torque 
tending to rotate the balance arm. The displacement 
of the beam relative to its case is observed by means 
of a telescope and scale fixed to the case, with the aid 
of a mirror carried by the beam. 
When the position of equilibrium of the system has 
been read, the entire beam and case are rotated through 
72° and the observation repeated. Five observations 
are thus obtained in one revolution of the balance, and 
the readings in these positions are sufficient to furnish 
the information required for the station at which the 
balance is situated. 
The modern type of balance consists of twe similar 
beam systems placed side by side, with the suspended 
weights at opposite ends, and as with this type only 
three observations at angles of 120° are required for 
a complete determination, a considerable saving of 
time results. These improved instruments are in 
some cases fitted with photographic recording arrange- 
ments, the balance case being rotated automatically 
into its new azimuth after the beam has come to rest 
and the exposure made. 
The suspended system requires most careful pro- 
tection against convection currents and other disturb- 
ing influences, and consequently is enclosed in a double 
or treble-walled brass case of uniform thickness. Tem- 
perature and radiation effects are thus reduced con- 
_- siderably, while the additional precaution is taken of 
Working at night, thereby eliminating solar radiation 
NO. 2799, VOL. 111] 
One of the weights is attached | 
effects and securing a greater constancy of temperature. 
Under these conditions the instrument is found to 
give uniform and satisfactory results in the field when 
protected only by a tent. 
In order to secure the necessary degree of sensitivity 
it is essential that the period of oscillation should be 
large, and Eétvés has succeeded in obtaining a period 
of swing of 1500 to 1800 seconds by employing weights 
of 30 grams and a beam of 4o cm. length, the lower 

5. 1.—Section of balance. 
weight being suspended at a depth of about 60 cm. 
below the beam. The sensitivity of the instrument is 
partly controlled by the coefficient of torsion of the 
suspension wire, and by using a platinum wire 0-04 
mm. in diameter, alloyed with 20 per cent. iridium, it 
is possible to measure variations of gravity to within one 
1o-® C.G.S. unit per centimetre. These wires, which 
have hitherto proved the most suitable for the purpose, 
are previously subjected to a special “ baking ” treat- 
ment with the view of eliminating remanent torsion. 
Similar instruments of smaller dimensions have been 
constructed and tried by Eétvés, but were subse- 
quently abandoned as being. unsatisfactory. Quartz 
fibres have also been employed in place of the torsion 
