Marcu 18, 1897 | 
NATURE 
Aan, 
SOCIETIES AND ACADEMIES. 
Lonpon. 
Royal Society, March 4.—‘‘ Experiments on the Absence of 
Mechanical Connection between Ether and Matter.” By Oliver 
Lodge, D.Sc., F.R.S., Professor of Physics, University College, 
Liverpool. Received January 19. 
The author gives an abbreviated account of a long series of 
experiments conducted by himself and his assistant, Mr. Davies, 
in continuation of those related in PA2/. Zrans., 1893 (Aberration 
Problems, &c.). The method consists in bifurcating a beam of 
light, and sending each half in opposite directions round a closed 
periphery very near a rapidly rotating mass of matter, and then 
observing by means of interference fringes whether the velocity 
of light is affected in the slightest degree by this neighbourhood 
of moving matter. The steel disks have been now whirled to 
higher speeds, chiefly at 3000 revolutions a minute ; the steadi- 
ness of the machine and the definition of the bands have been 
improved, other minor improvements have been made, and a 
long series of micrometric readings have been taken, both at 
increasing and at decreasing speeds. 
Further, the steel disks have been replaced by a much more 
massive lump of iron, weighing } ton, with a narrower channel 
for the light to travel in; and the bands have been observed 
close up to the moving surface, and even when reflected in it. 
The rotation was also continued for some hours to see if by 
chance ¢2ze had any influence. 
Moreover, the iron mass was strongly magnetised by a steady 
current, so that the light travelled across a moving magnetic 
field; and lastly the steel disks were replaced, with an insulated 
third disk between them, and strongly electrified, so that the 
beam of light travelled across a moving electrostatic field. After 
a number of spurious disturbances had been gradually eliminated, 
the author finds that in none of these ways is the velocity of 
light at all appreciably affected, and accordingly concludes that 
there is no viscous connection between the ether and matter of 
observable magnitude ; z.e. that whatever motion moving matter 
may confer upon the ether must be of an irrotational kind. It 
was demonstrated theoretically in the previous memoir that no 
optical experiments could be competent to detect motion of this 
latter character, and accordingly no attempt has been made to 
look for any kind of motion except such as would be caused by 
something akin to viscosity. 
Incidentally the author points out that by rotating the whole 
optical apparatus and observer, instead of the disks, at a very 
moderate speed, a shift of the bands should be seen; and even 
that the earth’s rotation would with a large enough frame 
produce an effect, which latter, however, it appears difficult or 
impossible to observe, not on account of its smallness, but on 
account of its constancy. 
The effect to be expected on Fresnel-Fizeau principles from 
whirling az”, was unfortunately just too small for the author to 
safely observe. The residual disturbing causes just masked it, 
but it is probably not beyond the reach of another attempt with 
a still more thoroughly steady machine, if any one feels inclined 
to persevere so far. At the same timeif it be supposed that any 
microscopic trace of true ether effect still possibly exists (which 
the author wholly disbelieves), and if a further attempt be here- 
after made to observe it, a number of slight residual disturbing 
causes would be got rid of (and probably other difficulties intro- 
duced), by rotating the machine in a vacuum. 
Physical Society, March 12.—Mr. Shelford Bidwell, 
President, in the chair.—Mr. William Barlow read a paper on 
a mechanical cause of homogeneity of structure and symmetry, 
geometrically investigated, with special application to crystals 
and to chemical combination, illustrated by models. The author 
has previously established that every homogeneous structure 
displays one or other of the thirty-two kinds of crystalline sym- 
metry. He now shows that homogeneous structures possessing 
most, if not all, of these kinds of symmetry may be produced 
mechanically, as the equilibrium arrangements of assemblages of 
mutually-repellent particles; and also that these mechanical 
systems of particles exhibit characteristics entirely analogous to 
certain crystalline and other properties of matter. The funda- 
mental concept may be summarised thus: A number of different 
kinds of mutually-repellent particles dispersed through space ; 
the amount of this repulsion being some inverse function of the 
distance between the particles concerned; the particles are 
destitute of polarity, and the difference in kind consists in a 
difference in the degree of mutual repulsion which two particles 
NO. 1429, VOL. 55] 
exercise, according to the kinds taken. It is further premised 
that the assemblage is agitated so as to render unstable all but 
the final equilibrium arrangement, and a means is provided for 
linking the particles symmetrically, and unlinking them, under 
certain circumstances, so as to modify the repulsion between the 
particles affected. The data thus summarised may be regarded 
as merely provisional, because the making of the equilibrium 
arrangement one in which ‘‘ closest packing” prevails is the 
object primarily aimed at ; and these concepts are mere devices 
for attaining this end. By the employment of particles of 
different kinds, a large amount of variety is provided for. The 
first step taken is to deduce the law of ‘‘ closest packing,” which 
runs thus: Every assemblage of mutually-repellent particles 
will continually approximate to, or strive after, that relative 
arrangement of the particles composing it, in which it has 
come, at every part, to occupy a minimum of space under 
a given general pressure, or average repulsion, between the 
particles, This law acts on all assemblages, of the nature defined, 
however numerous the kinds of particles composing them ; but, 
for its effects to be traceable, a very limited number of kinds 
must be present. Passing from assemblages consisting of a single 
kind of particle, the author takes a very simple case of two kinds 
of particles confined to a plane, and shows what type of symmetry 
will be produced when equilibrium is realised. Very simple 
cases of particles in space are then taken, and it is shown that 
a large number of different kinds of symmetry are displayed by 
the equilibrium arrangements produced when there is variety in 
the relations between the repulsions. To illustrate ‘‘ close 
packing,” stacks of balls of various sizes are employed ; but it is 
pointed out that the conditions of statical equilibrium of the 
particles are not always adequately expressed in this way, 
although every case of the latter kind can be represented 
approximately by a case of the former kind, possessed of the 
same order of symmetry. Very slight variation in the relations 
between the repulsions, alters the form of the equilibrium, 
arrangement ; sometimes merely changing the angle without 
affecting the type; sometimes, when it passes some critical! 
point, bringing about an alteration in type. Changes of the 
first kind resemble the change in crystal form caused by varia- 
tion of temperature, whilst those of the latter kind, especially 
when associated with rearrangement of the particles, are 
analogous to polymorphism. In many cases, the arrangement 
of the particles is such that some may be removed without 
affecting the distribution of the remainder, and without disturb- 
ing the “close packing”; if, therefore, other particles, exer- 
cising a slightly less repulsion, be substituted for the removed, 
tnoperative, particles, the only resulting change consists in a 
diminution of the pressure on the particles surrounding them. 
A species of isomorphism is in this way realised. When the 
particles of an assemblage are partially connected by hypothetic 
linking in a symmetrical manner, similar groups are formed ; 
but, in order that the formation of such groups may not be 
arbitrary, the partitioning which is produced must have as. 
complete symmetry as that of the partitioned structure. In 
consequence of this, some kinds of groups are not directly 
obtainable by symmetrical partitioning of a homogeneous 
structure ; but it is always conceivable that they may be included 
in the larger groups of some more complex constellation, and 
that they may be subsequently separated to form an assemblage 
by themselves. Consequently, very intricate results may be 
reached by successive steps ; symmetrical intermixture, linking, 
and unlinking, succeeding one another until complicated groups 
are built up, for the production of which such an agency as 
‘close packing” appears at first sight inadequate. Having 
called attention to a large number of arrangements, some 
capable and some incapable of symmetrical partitioning into 
groups of a single kind, some linked and some unlinked, the 
author contends to have established the following two propo- 
sitions: (1) The nature of the symmetry displayed by a homo- 
geneous assemblage of mutually-repellent particles of different 
kinds, in equilibrium, depends on the relations subsisting be- 
tween the repulsions exercised by these particles. (2) The 
assemblages belonging to all of the thirty-two classes of crystal- 
line symmetry, result from the fundamental law of “ close pack- 
ing,” when the relations between the different repulsions take 
the widest possible range of variety. Links which restrain the 
action of the repulsions can be present between some of the 
particles in some cases. The author refers to crystal * twin- 
ning,” and points out that the action of dimorphism is com- 
petent to produce analogous ‘twinning of symmetrical 
