TRANSACTIONS OF SECTION A. 479 
throughout all space, would give rise to gravitational foree between bodies 
varying inversely as the square of the distance, and, if of sufficiently high fre- 
quency, much higher than the Réntgen rays, the absorption would be proportional 
to the mass, and so also would be the pressure. 
The velocity of propagation would be very much greater than that of light, 
and this would therefore not be open to Laplace’s objection that gravitation, 
if it were propagated at the same speed as light, would give rise to planetary 
disturbances which could easily be detected. 
The frequency would have to be very great, so that the law of force should be 
that of the inverse square for small bodies. It seems therefore in accordance 
with this theory that the wave-length is comparable with the molecular dimensions 
when the law of attraction no longer follows the inverse square law. It is note- 
worthy that the hydrogen molecules repel when those of other gases attract. Thus 
the longitudinal wave-length would appear to be comparable with the dimension 
of the hydrogen molecule. The repulsion is due to the pressure of the scattered 
radiation. Similarly with electrons. But although two hydrogen atoms or any 
two molecules of that order of magnitude may repel each other, and also two 
electrons, yet a large molecule will attract a very much smaller one, so that a 
molecule the dimensions of hydrogen atoms will attract an electron because of the 
sereening action of the larger molecules on the smaller. This appears to be in 
accordance with electrostatic attraction of unlike charges and repulsion of like 
ones. Thus the theory is promising in many ways, and is being developed further. 
8. On the Electric and Thermic Conductivities of certain Alloys of Iron. 
By Professor W. F. Barrett, F.2.S., and R. A. HApriexp. 
9. On a New Apparatus for producing Magnetic Fields of Force. 
By Professor Marcus Harrtoe. 
From a square base-plate of cast iron arise four brass columns with a screw- 
thread on their upper ends to take large flat nuts, which support a wooden plate 
with circular perforations at regular intervals. On the iron plate rest the electro- 
magnets (5), each consisting of a cylindrical coil (c) with both terminals below, and 
a cylindrical soft iron core which projects through a hole in the wooden plate. A 
source of direct current, an amperemeter, a rheostat, and a mercury commutator (e) 
(to allow of the ready alteration of the direction of the current in the individual 
coils), complete the apparatus. The material I employ is magnetite (mineral 
Fe,0,), finely powdered and levigated, or iron filings levigated in alcohol; this 
powder may be shaken on paper or suspended in glycerine, balsam dissolved or 
melted, or liquefied gelatine. As a rule a thin layer of the mixture is sufficient, 
spread on a glass or china plate giving an axial section of the field. I have with this 
apparatus produced several interesting variants of the classical figures of the 
magnetic field obtained by the agitation of paper strewn with iron filings. 
Sus-Section or AstRoNoMy AND CosmicaL Puysics. 
The following Reports and Papers were read :— 
1. Report on the Magnetic Observations at Falmouth Observatery. 
See Reports, p. 29. 
2. Report on Meteorological Observations on Ben Nevis.—See Reports, p. 55. 
