740 
for electrodes platinum wire encased in quartz, which, 
of course, becomes conducting in this flame. This 
arrangement gives an exceedingly perfect repro- 
duction of the sound—voice or music—far better 
than any carbon microphone. 
A small alcohol burner and flame can be employed. 
In this case I recommend as cathode a Nernst glower 
as supplying the necessary electrons. Or enrich the 
blue flame by potassium or sodium salts fed up 
through the wick, with the alcohol. 
Audion amplifiers must, of course, be employed 
if one wishes to use the “‘flame microphone” for 
broadcasting, radio-phone, or ‘‘ phonofilm ”’ purposes. 
The sensitiveness increases in general with the 
impressed P.D. across the electrodes—up to a limit. 
Care must be taken to guard against: (1) hissing 
due to too high voltage discharge ; (2) flame noises ; 
(3) air fluctuations ; (4) depositing of carbon upon 
the electrodes. 
I have not had time yet to make a careful scientific 
study of this phenomenon, but am persuaded it is 
chiefly a pressure effect, controlling ionisation and 
the ionised conductivity of the flame. 
LEE DE Forest. 


Molecular and Crystal Symmetry. 
Mr. T. V. BARKER has discussed in Nature of 
May 12 the theory advanced by Fedorov and Shearer 
with reference to the relations between molecular and 
crystal symmetry. According to this hypothesis the 
symmetry of the crystal includes the symmetry of 
the molecule with such additional symmetry as is 
afforded by the arrangement of the molecules, if 
there be more than one, in the unit parallelepipedon | 
or cell of the structure. 
At the reading of Mr. Shearer’s paper I mentioned 
some considerations which required to be taken into 
account in applying this principle (Proc. Phys. Soc., 
vol. 35, p. 99, 1923), and I propose to restate them 
here in somewhat more detail. 
In many cases there is reason to believe that 
WATURE & 


molecules have no existence in the structure of a | 
crystal. 
identity. 
In others they appear to maintain their 
It does not, however, follow that. they | 
retain the full symmetry they possess when in the | 
free state in a fluid; for the whole or a part of the 
symmetry may be destroyed by close packing in 
the crystal structure. Nor is it probable that the 
symmetry actually possessed by the unit cell formed 
of one or more molecules is always identical with 
that of the structure of which it forms part. 
In the first place, a number of primary cells with 
different but similar orientation (including in this 
expression a symmetrical relation between enantio- 
morphic forms) may be combined by what may be 
termed cell-twinning to form a greater cell with 
higher symmetry. These greater cells may of course 
be regarded for crystallographical purposes as unit 
or elementary cells; but it is improbable that they 
would always be recognised as such by means of the 
X-rays, which would in many cases not permit of 
discrimination between the differently orientated 
primary cells. The same crystallographic char- 
morphic forms with lower symmetry, formed usually 
acters would also result from ultra-microscopic | 
twinning on a larger scale, involving groups of 
differently orientated cells instead of individual cells. 
Repeated ultra-microscopic twinning of this char- 
_ by the writers in respect to the relation between the 
acter is believed to take place with the triclinic | 
mineral microcline, so as to give rise to the mono- 
clinic mineral orthoclase. 
Apart, however, from regular twinning, one would 
expect cells of low symmetry to build up in many 
instances structures of higher symmetry, but usually | 
NO. 2796, VOL. 111] 
| be too small to determine a trigonal symmetry. 
» 4 
[June 2, 1923 
belonging to the same system. Perfect identity in 
oe ep: 
~~ Tee @ i’ SF 
| cells is not necessary in the building up of a structure. 
A plagioclase crystal is formed of cells both of albite 
| and of anorthite, with quite distinct atomic composi- 
tion, and even to a limited extent of orthoclase which — 
differs both in system and in®*molecular volume. 
Where, therefore, the outward forms of cells of the 
same substance in different orientations (in the wide 
sense employed above) closely resemble each other 
but do not show absolute identity, it may be expected 
that the crystal structure will be built up indis- | 
criminately of cells with similar but not identical 
orientation. The result will be that the special . 
features characteristic of a lower symmetry will be 
| eliminated and only the highest symmetry of the 
system will remain. This is probably the:reason why 
crystals possessing the symmetry of one of the lower 
classes of a system are comparatively rare, and in 
some instances are not known to occur. 
These principles are well illustrated by the facts 
disclosed in a paper on the “ Relation between the 
Crystal Structure and the Constitution of Carbon 
Compounds, Part I., Compounds of the Type CX,” 
(Journ. Chem. Soc., vol. 123, pp. 71-79, 1923), by 
Miss Knaggs, of the Imperial and Bedford Colleges. 
She shows that in substances of the CX, type, where 
X is an element, the crystal usually belongs to the © 
cubic system. Those of the type CX;Y, where X and 
Y are elements, are as a rule trigonal or hexagonal, 
unless X is hydrogen, the atoms of which appear to 


































Those of the form C(CX;), are usually cubic, as the 
trigonal character of the CX, group enables all four 
trigonal axes of the cubic system to be preserved. 
Finally, substances of the form C(CX,Y), are in 
general tetragonal. In every case in which the 
symmetry of the crystal shows it to belong to the 
same system as that of the molecule, it must 
be referred to a higher class, usually that with 
the highest symmetry in the system. For sagen: 
the molecules €X, and C(CX;),, which are cubic, 
have no axial planes of symmetry, but wherever 
there is any definite crystalline form, the crystals 
possess such axial planes. In some cases, however, 
the cubic system is only recognised by the isotropic 
character of the crystals. Again it can be easily 
shown that the molecule C(CH,Y), belongs to a class 
of the tetragonal system with only a contra-direc- 
tional or inverse tetragonal axis, but the crystals have 
all a co-directional or simple tetragonal axis, such as 
is found in the higher classes of the tetragonal system. 
In many cases, on the other hand, there are iso- 
at lower temperatures. In these the atoms are 
apparently more tightly packed, and the molecules 
have either been distorted or have lost their identity 
altogether. Joun W. Evans. 
Imperial College of Science and Technology, 
South Kensington, S.W.7, 
May 15. 
In a recent letter to NATURE (May 12, p. 632) Mr. 
T. V. Barker takes exception to statements made 
symmetry of a crystal and that of its components 
(G. Shearer, Proc. Phys. Soc., 1923, vol. 35, p. 81, 
and W. T. Astbury, Proc. Royal Soc., 1923, vol. 102, 
p. 506). It appears to us that his criticisms are 
based on certain misapprehensions. 
Fedorov tried to prove (Zeits. Kryst., 1912, Vol. 
52, p. 22) that if m is the symmetry number of the © 
structural unit of the crystal, or, briefly, the crystal | 
