TRANSACTIONS OF SECTION A. 387 
in the entire absence of a second order spectrum. That is to say, there is no reflection 
at an angle @ for which sin 6 = * though it does occur when sin 0 = 3 or a or - 
or <9 Fluor-spar shows the same effect, and zinc blende approximates to it. The 
2d 
interpretation is that the (111) planes are not equally spaced, but there is an alterna- 
tion of spaces, one of which is three times the other. The spectra also tell us the actual 
values of all the spacings. When all the information is put together we find that the 
element of volume of the diamond is a face-centred cube; a cube haying, that is 
to say, a carbon atom at each corner and one in the middle of each face. In the 
same cube are also four carbon atoms at the centres of four of the eight small cubes 
into which the large cube may be divided. ‘The structure may also be described as 
consisting of two interpenetrating face-centred cube lattices, one of which can be 
derived from the other by translating it in the direction of the diagonal through a 
distance equal to one quarter of that diagonal. 
In the same way zinc blende is shown to have exactly the same structure as the 
diamond, except that one of the two lattices consists entirely of zinc atoms, and the 
other of sulphur. In the case of fluor-spar there is a face-centred lattice of calcium 
atoms, fluorine atoms being placed at the centres of all the eight small cubes. Iron 
pyzites, sulphur, calcium, and other crystals have also been examined by these 
methods. Some of them show the most remarkable variations in the strengths of the 
spectra of the various orders. 
The method of the Laue photograph has also been applied to the examination of 
the diamond structure ; the results entirely agree with those of the reflection method. 
2. On Lightning and Protection from it. By Sir J. Larmor, F.R.S. 
The rationale of electric discharge in a gas is now understood. When a 
small region becomes conducting through ionisation by collisions in the electric 
field it should spread in the direction in which the field is most intense, which 
is along the lines of force. Thus the electric rupture is not a tear along a 
surface but a perforation along a line. This is roughly the line of force of the 
field: the electro kinetic force induced by the discharge, being parallel to the 
current, does not modify this conclusion. A zigzag discharge would thus 
consist of independent flashes, the first one upsetting adjacent equilibria by 
transference of charge. Successive discharges between the same masses would 
tend to follow the same ionised path, which may meantime be displaced by air 
currents. 
If the line of discharge is thus determined by the previous electric field, the 
influence of a lightning-conductor in drawing the discharge must be determined 
by the modification of this electric field which its presence produces. For a 
field of vertical force, such as an overhead cloud would produce, it may be 
shown that the disturbance caused by a thin vertical rod is confined to its own 
immediate neighbourhood. Thus while it provides a strong silent discharge 
from earth into the air, it does not assist in drawing a disruptive discharge 
from above—except in so far as the stream of electrified air rising from it may 
provide a path. It is the broader building, to which the rod is attached, 
that draws the lightning: the rod affords the means of safely carrying it away, 
and thus should be well connected with all metallic channels on the building 
as well as with earth. It is the branching top of an isolated tree that attracts 
the discharge; a wire pole could not do so to a sensible degree. Separate rods 
projecting upwards from the corner of a building do not much affect the field 
above it, but if they are connected at their summits by horizontal wires, the 
latter, being thus earthed, lift up the electric field from the top of the building 
itself to the region above them, and thus take the discharge which they help 
in attracting, instead of the building below them. Similarly, when the lines of 
force are oblique to a vertical rod, its presence does somewhat modify the field 
and protect the lee side; but generally the presence of a rod should not ever be 
a source of danger, unless the ionised air rising from it provides an actual path 
for discharge. 
(Kole? 
