MarcH 29, 1912] 
forward many good reasons for the view that 
atoms have a definite volume and are compressible. 
The number of negative electrons in atoms can 
be deduced from observations on the scattering of 
the 8 rays of radium or of Rontgen rays and in 
other ways, as Sir J. J. Thomson has shown. The 
results obtained indicate that the number is a 
small constant multiple of the atomic weight. 
The scattering of a rays led Rutherford to adopt 
the idea of a positive nucleus, since some a rays 
are turned through a larger angle than can be 
explained by the electric forces due to a charge 
equal to that on one electron. It may be, however, 
that other forces besides ordinary electric force 
act on a rays when moving through matter. The 
a rays are helium atoms which have a radius about 
10° em., so that they probably only get through 
by displacing the atoms of the matter. If we 
suppose the positive sphere of one atom can not 
penetrate into that of another then the scattering 
of a rays by matter can probably be explained on 
Sir J. J. Thomson’s theory. 
The most interesting application of Sir J. J. 
Thomson’s theory is the explanation which it af- 
fords of the relation between the atoms in series 
of similar elements like fluorine, chlorine, bromine, 
iodine. Sir J. J. Thomson supposes that the nega- 
tive electrons in the sphere are arranged in con- 
centric spherical layers and that each element in a 
series of similar elements is derived from the one 
before it by the addition of one more layer. The 
writer? has worked out this idea and shown that it 
is in approximate agreement with the atomic 
weights and that the number of electrons per atom 
ean be deduced approximately from the atomic 
weights. The result obtained was that the number 
of electrons is about eight times the atomic weight. 
Probably the most promising of the many ways 
of obtaining evidence as to the structure of atoms 
is by the study of the spectra of the light which 
they give out when set vibrating by different dis- 
turbing agencies. Most spectra, however, are so 
complicated that very little progress has yet been 
made. The fact that spectra contain lines of 
definite wave length suggests that the electrons in 
the atom vibrate about positions of equilibrium or 
else are moving in magnetic fields of constant 
strength. This seems to be a strong argument 
against theories of the planetary type, for on such 
theories the period of vibration is not fixed, but 
depends on the radius of the orbit. 
Any theory which explains spectra ought also 
to explain the Zeemann effect. 
? Phil. Mag., June, 1911. 
SCIENCE 
513 
An important question is whether all the lines 
in the spectrum of an element can be emitted by 
each atom or whether the different lines are 
emitted by different systems. The second view 
now seems the more probable. On this view the 
different lines do not correspond to the different 
possible modes of vibration of each atom, but each 
line is due to the vibration of a different system. 
Of course a particular system may give more than 
one line in some cases. Thus we might suppose 
a regular series of n lines to be due to the vibra- 
tions of molecules with one, two, three up to n 
atoms in the molecule. Other series for the same 
element might be due to molecules which had each 
lost say m electrons with up to m atoms per 
molecule. On this view the frequency of vibration 
would be a function of two integers n and m. The 
different series in the spectra of the alkalies can 
be represented approximately as functions of two 
integers, as is well known. 
The only theories of series spectra which have 
been developed to any extent are due to Ritz 
(‘‘Gesammelte Werke W. Ritz,’’ Paris, 1911). 
In his earlier papers he supposed the lines in each 
series to be due to different modes of vibration 
of an elastic membrane having special properties. 
Later he abandoned this view and supposed each 
line due to a different system. 
Ritz’s atomie vibrator consists of an electron 
vibrating in the magnetic field of a bar magnet 
at a point along its axis. The electron is sup- 
posed to stay close to a particular point on the 
axis and to vibrate in the plane perpendicular to 
the axis. The distance from the electron to the 
nearest pole is taken to vary by equal increments 
and the distance between the two poles also is 
supposed to vary by equal increments in going 
from one atom to another. This makes the fre- 
quency a function of two integers and the function 
found agrees approximately with the observed fre- 
quencies. However, to obtain exact agreement Ritz 
had to suppose the increments to be not always 
exactly equal. Ritz supposed the bar magnet to be 
made up of a row of nearly equal elementary mag- 
nets. In any atom some of the elementary mag- 
nets are in the row and the rest may be supposed 
arranged so as to neutralize each other. 
This idea of elementary magnets receives some 
support from recent work on the magnetic prop- 
erties of bodies by Weiss and others. The ele- 
mentary magnets of course may consist of elec- 
trons moving round orbits. 
