78 
Transactions of the Royal Society of South Africa. 
from the (100) planes. We find at once that it will not explain the (100) 
spectra. 
The observed (100) spectra are nearly normal, and there can be only a 
small difference of phase between the two sets of planes. James and 
Tunstall* have shown that this can be accomplished in the following way. 
Divide the unit face-centred rhomb into eight equal rhombohedral cells. 
Place an atom at each of the unoccupied corners of the small cells and then 
Fig. 2. 
push each atom along the diagonal by an amount equal to 0"074 of the 
length of a diagonal of one of the small cells. 
The spacings we have determined for the (111) planes would make this 
amount to be 0 059 instead of 0 074. 
The spacing for the (100) planes would be 0 059 : 0 941. The two sets 
of planes differ in phase by about 21°, and would give a nearly normal set 
of spectra in agreement with observation. 
(c) (110) planes : 
We must now examine whether such an arrangement would suit the 
spectra from the (110) planes. The planes would contain an equal number 
of atoms, and the spacings are 0*118 : 0 862. 
* James and Tunstall, loc. cit. 
