974. ANNUAL REPORT SMITHSONIAN INSTITUTION, 1909. 
have upon the faces of the crystal. The introduction of matter 
to one face only of a crystal causes it to develop unequally, and since 
all crystals of the same bath or magma are not influenced in the 
same way, they may present a number of different forms. The 
crystals formed at the bottom may be different from those which 
are deposited on the side walls or at the surface. 
Il. 
If these concentration currents can completely change the habit 
of a crystal by producing elongation in one direction, the nature 
of the faces is not modified; an octahedral crystal always appears in 
octahedrons. But there are two other influences which modify the 
faces. One of them is the rapidity of crystallization, the other the 
constant absorption of foreign matter by the crystal in process of 
growth. Still other factors may intervene, but they are only indi- 
“rectly concerned. 
It has long been known that crystals formed rapidly possess 
simple faces, while those which have grown slowly are more com- 
plicated. Thus in nature the crystals rich in inclusions, sometimes 
of large size, are poor in faces, while the small crystals of the same 
substance, but transparent, are generally limited by a large number of 
faces. These differences are due to the rate of crystallization, the 
influence of which has been made known by the experiments of 
Frankenheim, .Lecoq de Bosbaudran, O. Lehmann, and myself. In 
rapid crystallization the crystals have faces with simple symbols; 
in slow crystallization these same simple faces persist, but the angles 
and edges have been truncated and beveled, giving rise to new facets, 
and I have shown that in certain cases these facets make their ap- 
pearance always in the same order. With varying rates of crystalli- 
zation the dominant forms obtained in the case where the crystalliza- 
tion was rapid persist with more or less extensive development, but 
it may be otherwise in the case where the faces are modified by the 
regular absorption by the crystal during growth of foreign matter 
added to the mother liquor. This fact is easily made evident, as I 
have demonstrated, by adding a little coloring matter. 
Rome de l’Isle and Berniard have observed that the crystals of 
sodium chloride formed in urine are regular octahedrons insteac of 
oubes, such as crystallize from a pure mother liquor. Vanquelin anu 
Fourcroy showed later that this curious modification is due to the 
urea present. Boydant also established a few phenomena of the 
same class, and tried without success to ascertain why the mere pres- 
ence of a foreign substance can be thus effective. 
P. Curie developed a remarkable and attractive theory, which 
apparently furnished the key to this curious modification. He 
