396 JOURNAL OF THE WASHINGTON ACADEMY OF SCIENCES VOL. 13, No. 18 
structure of lithium iodide was not the anhydrous compound and 
that the resulting experimental data lead to an erroneous determina- 
tion of its crystal structure. From this standpoint, then, the fact 
that one or more compounds, none of which probably forms cubic 
erystals, could give as many as 20 lines which agree in position with 
a simple cubic arrangement is another and striking illustration of 
the dangers that arise from using unaided powder photographic data 
in determining the atomic arrangement in crystals. 
The structure of rubidium fluoride. At the time of their study of 
the structures of the other alkali halides, several unsuccessful at- 
tempts were made by the writers to obtain the structure of rubidium 
fluoride. The material produced by fusing preparations of rubidium 
fluoride was found to be essentially isotropic!® but so hygroscopic 
that it altered immediately upon exposure to air. For this reason 
it was impossible with the prevailing experimental facilities to ob- 
tain powder photographs showing lines belonging to the _ isotro- 
pic fluoride. The product used in the published determination? 
was a well-crystallized preparation made by a simple desiccation of 
a solution of rubidium fluoride. The writers’ experience with rubid- 
jum fluoride leads them to question the identity of this crystalline 
material with anhydrous rubidium fluoride. Two other factors also 
suggest that this structure may not be that of the anhydrous salt. 
Although the interatomic distances are additive for the caesium 
halides having the structure here ascribed to rubidium fluoride, there 
is no such agreement in this latter case (calculated distance = 2.80A; 
observed = 3.172A). Accepting the first assignment of structure! 
to lithium iodide, this disagreement would then be the only one 
among all of the alkali halides. 
The observed intensities of the reflections upon which this deter- 
mination of structure is based do not agree with those calculated 
for the stated atomic arrangement. This is shown by the data of 
Table III. The intensities of column 3 have been calculated with an 
expression analogous to the one used for lithium iodide.*® 
Summary. It is pointed out that the material used in a recent 
determination? of the structure of lithium iodide was not, in all prob- 
ability, the anhydrous salt, that the observed intensities of the dif- 
fraction lines obtained from this preparation conflict with those 
calculated for the assigned structure, and that its interatomic dis- 
tances do not agree with those to be expected from crystals of lithium 
10 The optical properties of these substances were obtained by H. E. Merwin. 
