1903. ] KOENIG—ARTIFICIAL PRODUCTION OF CRYSTALS. 2235 
thought it to be until the analysis showed it to contain 72.9 per cent. 
of copper. The crystals even penetrated into the asbestos, and 
from this very extremity the material for the analysis was taken. 
This experiment was carried on from 8 a.M., January I1, 1900, for 
forty hours. Here was a phenomenon of molecular or ionic 
activity without parallel; at least to me extraordinary, for I had 
not seen any record of a similar observation. It is not difficult to 
understand the building up of crystals from a medium which con- 
tains the molecules in the liquid or gaseous state. But what I 
observed here implied a very different condition of things. Not - 
even the skyward growth of a tree, which somewhat resembles this 
stretching out of the domeykite toward the supply of arsenic, is 
comparable. For the cells draw their nourishment from the liquid 
sap and the gaseous air. The growing may happen from the root 
by pushing or by growing at the front. In the latter case the cop- 
per ions must be supposed to be going like the ions under the direc- 
tion of a current, but going in the solid condition, and this is the 
point at which the imagination recoils. Either alternative rests 
upon a push or a draw impulse. The present experiment would 
seem to point toward a push from the root as the cause, that is to 
say to a mobility of the copper arsenide molecule Cu,As. In Fig. 
3 is represented one of the results of a later experiment, which 
gives support to the notion that the copper ions are moving and not 
the molecule CusAs. Here C is a piece of copper turning. On a 
slender stylus S sits the large domeykite crystal D (the tabular type, 
three millim. in diameter). The crystal is incomplete on one side. 
From it leads a second stylus S', and upon this another somewhat 
incomplete crystal of domeykite D has been growing. All the 
material for the crystals must have come through the stylus S. 
Instead of all the material, 1 should say more correctly all the 
copper. The stylus habit for the crystallization is very common ; 
Fig. 3 merely represents an unusually fine specimen of this habit. 
Looking at the phenomenon of molecular mobility in the solid 
state merely as a physico-chemical process, aside from crystalliza- 
tion, I can see an analogous occurrence in the so-called cementa- 
tion process of steel or case-hardening process. In this process a 
bar of soft iron is exposed to red heat in a packing of solid char- 
coal, and becomes gradually converted into carbid Fe,C to the very 
innermost parts. It would seem that the solid carbon ions become 
mobilized, passing from one group of iron ions to the next until 
