AUGUST 19, 1915] 

NATURE 
675 

more thoroughly carbonised, the central core 
became blocked with pitch and tar distilled for- 
ward into the coal mass, and the gas and tar 
vapour being in this latter stage forced to find 
their way through the outer shell of red-hot coke, 
became degraded to the utmost limit, and gave 
very large volumes of such permanent gases as 
hydrogen mingled with methane, the least easily 
broken up of the hydrocarbons, the result being 
that the modern tar is in reality a mixture of low 
temperature and very high temperature tar. 
Such a tar, obtained from a good gas coal, will 
contain approximately :— 
Benzol o'60 per cent. 
Toluene ‘ O15 os 
Solvent naphtha 1°00 n9 
Heavy naphtha 1°65 9 
Carbolic acid 06 . 
Cresylic acid o"9 6 
Creosote oil 30°6 Pe 
The remainder is naphthalene, anthracene, and | 
pitch. 
If the gas from the carbonisation be scrubbed, 
and the absorbed hydrocarbons be distilled out 
from the creosote oil and added to the light hydro- 
carbons obtained from the tar, the total yield is :— 
Benzol ... 
Toluene bc 
Solvent naphtha 
Heavy naphtha 
2-I per cent. 
271 
2-6 ” 
2°6 
It is found in practice that the carbonisation 
under the ordinary gasworks conditions of a ton 
of good bituminous coal yields about one-third 
of a gallon of toluene from tar and gas. 
In practice, the withdrawal from the gas of all 
the hydrocarbons that can be scrubbed out by 
creosote oil would reduce the heating value of the 
gas to a considerable extent, and various methods 
of getting over this trouble have been proposed, 
such as enriching the scrubbed gas by the benzene 
and xylene from the crude benzol after separation 
of the toluene by fractional distillation, or by so 
scrubbing the original gas that only about one- 
third of the crude benzol is withdrawn, as by 
such means the calorific value of the gas can be 
maintained at or above the statutory minimum of 
500 British thermal units. 
The tar and gas made in small country works 
as a rule contain more aromatic hydrocarbons 
than the products from large works, as the tem- 
peratures used in carbonisation are not so high; 
and should the Government need still larger 
quantities of toluene and carbolic acid for nitra- 
tion, the gasworks of the Empire could treble 
the output by reverting to the temperature and 
methods employed before the introduction of re- 
generation in the furnaces. This, however, would 
reduce the gas yield from the 13,000 cubic feet 
per ton now aimed at by the gas manager to 
10,000 cubic feet, but the higher calorific value 
of the gas would allow of the volume being made 
up by the addition of blue water gas. 
NO. 2390, VOL. 95] 


THE ELECTRONIC THEORIES OF THE 
PROPERTIES OF METALS. 
N a course of lectures delivered at the Caven- 
dish Laboratory during the Michaelmas term 
of 1886, Sir J. J. Thomson examined the dynamical 
result of assuming the passage of electricity 
through a metal to be of the same nature as 
through an electrolyte. The subject was dealt 
with more completely in ‘Applications of 
Dynamics to Physics and Chemistry” in 1888. 
The later discovery of the electron which might 
act as a carrier of electricity from molecule to 
molecule placed the idea on a much firmer footing, 
and in 1898 Riecke developed theories of the elec- 
trical and thermal conduction and thermo-electrical 
properties of metals, based on the existence bé- 
tween the molecules of carriers of positive and 
negative electricity. Two years later Drude 
worked out more systematically theories founded 
on the same basis, and it is his work which is 
usually quoted in accounts of electronic theories, 
generally with the simplification that only one 
type of carrier—the negative electron—is taken as 
moving freely between the molecules. These 
electrons are supposed to be produced by the dis- 
sociation of the electrically neutral metal atoms, 
what remains of the atom being left positively 
charged. 
The moving electrons are assumed to have the 
same average kinetic energy as a molecule of a 
gas enclosed in a cavity in the metal would have. 
The nature of the impact of electron on metal 
atoms is not discussed, but as the motion of eaeh 
metal atom is likely to be comparatively small, 
the gas laws which hold for the motions of the 
molecules of a light gas amongst those of a much 
heavier are applied and the electrons are said to 
have a mean free path of length A. The motions 
are distributed in all directions equally, and the 
electrons cannot escape through the surface of the 
metal unless their speed perpendicular to the sur- 
face exceeds a certain limit. The action of ultra- 
violet light on the surface facilitates the emission. 
When an electromotive force is applied to the 
metal there is superposed on this to and fro motion 
of the electrons a drift up the electric field, the 
relation of which to the field is such that the 
specific conductivity of the metal is proportional 
to nd/./T where n is the number of free electrons 
per cubic centimetre and T is the absolute tem- 
perature of the metal. Since the electrical con- 
ductivity of a pure metal is known to vary 
approximately inversely as the absolute tempera- 
ture, this implies that nA must vary inversely as 
VT. As there is nothing on the one hand to 
suggest so considerable a change of the free space 
between the metal atoms with temperature, while 
on the other the facts of thermoelectricity are 
against any considerable decrease in the number 
of electrons per c.c. as the temperature rises, it 
sems difficult to reconcile the law of variation of 
nd with experimental facts. 
If a slope of temperature exists in a metal, the 
