viil 
Supplement to “ Nature,” June 9, 1923 


the same, in fact, as that of the boxes in Fig. 6. This 
is a very convenient form for packing, and it appears 
that a majority of known crystals pack together in 
this way. All of them show the symmetry that might 

Fic. 6.—Unit cells of naphthalene and anthracene drawn to the same scale 
(ro® to 1). It is apparently usual for cells of this kind to contain four 
molecules, but in these two cases they contain two only; that is 
because the molecules have some symmetry of theirown. The molecules 
shown in Fig. 4 fit into the naphthalene cell. If the unshaded mole- 
cules be supposed to be placed with one end at Q, its general lie is parallel 
to OC, The two shaded molecules of Fig. 4 would then be placed 
with their ends at Band F. ‘The anthracene molecule is like that of 
naphthalene except that there are three rings in a line instead of two; 
the axis OC is correspondingly larger. 
be expected. They are exactly alike on either side 
of a dividing plane: in other words, they are exactly 
like their reflection in a mirror, They have, too, an 
“axis of symmetry”; a half turn about the axis 
brings no apparent change. Fig. 7 is an illustration 

Fic. 7.—Phthalic acid, a monoclinic prismatic crystal possessing a plane of 
symmetry PQRS, and an axis of digonal symmetry AB, but there is 
no plane of symmetry through AB. Faces marked by the same letter 
are alike. The mutual orientation of the four molecules in its unit 
cell is that of the shoes in Fig. s. 
of a crystal of this kind. The X-ray methods show 
us that there are four molecules in the unit of pattern, 
and that they are arranged in the manner described. 
It is very interesting to observe the result of a 
different arrangement. Sometimes a set of four are 
arranged as in Fig. 8, like two pairs of shoes back 
Fic. 8.—The relative orientations but not the relative positions of the four 
molecules in the crystal unit cell in resorcinol. 
to back. The top and bottom are now unlike, but 
there is a greater symmetry in other directions. We 
have recently examined a crystal substance, resorcinol, 
which is built on this pattern; its external form is 
shown in Fig. 9. The fundamental molecule is a 
benzene ring in which two hydrogens have been re- 
placed by two hydroxyl (oxygen-hydrogen) groups. 


The crystal shows clearly different forms at its two 
ends, the difference of which is shown in another very 
interesting way. If the crystal is warmed, one end of 
it becomes positively electrified and the other negatively. _ 

We have been able to go some way to the actual 
determination of 
relative positions of the 
molecules: the results 
are shown in Fig. 10. 
There are two sets of 
planes perpendicular to~ 
the axis, which in the 
crystal occur alternately. 
In each plane there are 
two types of molecule, 
counting difference of 
orientation as difference 
of type. The plain lines 
in the diagram show the 
arrangement of the two 
types lying in the plane of 
the paper, and aremarked 
Aand D. One of them 
can be obtained from the 
other by a revolution 
; 
poe ene aa mane ween eee ns, 

Fic. 9.—Resorcinol, a rhombic pyra- 
midal crystal, having two planes 
of symmetry, CB and 
PQRCST, and an axis of digonal 
symmetry PC. = 
’ 
through 180° in the plane. They are joined by 
long and by short lines, the former containing the — 
oxygens of the hydroxyl groups. The arrangement 
of the next plane, above or below the first, is shown 
by the dotted lines: the molecules in this plane can — 
be obtained from those in the first by reflections, with 
proper translations, across planes that are perpendicu- — 

FiG. 10.—Probable arrangement of molecules in resorcinol. 
lar to the paper and cut it in lines parallel to either of 
the sides of the rectangle. The arrangement is clearly 
governed by the necessity of fitting the molecules to- 
gether so as to accommodate the hydroxyl attachment. 
Another case of great interest and importance is 
the two-molecule cell, the two being exactly alike. 
How would you pack a box with boots all right-footed 
or all left-footed ? You cannot find a way of packing 
the i 

* 
