1555 
gement of the molecules, these layers evidently remaining parallel 
to their original position. 
From a beautiful, colourless crystal of olivine, originating from 
Bohemia, a erystalplate parallel to (010), and 1.33 mm. was prepared, 
and then radiated through. The very nice RÖNrGeN-pattern of this 
silicate, whose composition is: ” Mg, SiO, + m Ire, SiO,, and whose 
parameters are: @:6:¢ = 0,4657:0,5865 is reproduced in fig. 8 
on Plate II, and in stereographical projection in fig. 8 on Plate V. 
Also in this case the normal symmetry can be discerned immediately, 
so that the results here obtained with these rhombic crystals can be 
considered to be a welcome completion of our previous experi- 
ments’), when we were able to prove the correctness of the theore- 
tical deduction also in the case of rhombic crystals. 
$ 3. Already previously we have drawn attention”) to the remark- 
able abnormality, which was observed by us in the case of the 
diffraction-image parallel to (1210), obtained with a erystal of quartz 
from St. Gothard; it consisted in the image not only possessing a 
single binary axis perpendicular to the photographie plate, but also 
two planes of symmetry, perpendicular to each other. On that 
occasion we expressed the surmise, that this abnormality could be 
explained by the fact, that the studied plate was prepared from a 
polysynthetie twin of quartz, according to the so-called brasilian 
law of twinning. Indeed, on this supposition the apparent increase 
of the symmetry of this image can be easily explained. But at that 
time we were not able to give any proof of the correctness of this 
view, because of the fact, that the rather imperfect Röntgen-image 
obtained with a plate cut parallel to (1010), seemed to be likewise 
symmetrical after two planes of symmetry. However since that time 
we have studied some other plates orientated with the utmost care 
parallel (1010), and have radiated them through during very different 
times of exposition. By these experiments we got absolute certainty, 
that the images are only symmetrical after a vertical plane, perpen- 
dicular to (1010). The photo in fig. 14 on Plate TV, and its stereo- 
graphical projection in fig. 10 on Plate V can prove this. By 
shortening more and more the time of exposition, we succeeded e.g. 
in making only the most intensive spots appear: it could be clearly 
seen then, that in the direction of the c-axis at the upper end there 
were two, at the lower end three of such intensive spots on the 
1) These Proceed., 18. 559 (Sept. 1915). 
2) These Proceed., 18. 550. (Sept. 1915) 
100 
Proceedings Royal Acad. Amsterdam. Vol. XVIII 
