BarLtow—A Mechanical Cause of Homoyeneity of Crystals. 578 
result in its sudden rupture when a critical point is reached. If, 
however, it will bend till all curvature is done away we may look 
for ultimate production of a rectilinear perfectly homogeneous 
condition of the assemblage as the result of the addition of a large 
number of layers. If it affords great resistance to bending we 
may have a case in which the assemblage grows considerably in 
every direction, while still a curved assemblage. 
If skeleton growth, 7.¢., discontinuous accretion, takes place 
along a bent surface of the solidified portion of an assemblage 
of the modified nature now under consideration, it is evident that 
the curved branches formed cannot be congruent, ¢.e., cannot form 
part of a homogeneous whole capable of a congruent filling in of 
the interstitial spaces. In other words, an unhomogeneous branched 
assemblage and not a skeleton homogeneous one will result. 
The facts concerning curved and branched crystals exactly 
parallel the conclusions reached above. Bending and branching 
oceur generally when the crystals formed are extremely fine or 
thin,’ and consequently only when the conditions are appropriate 
for the production of such crystals. According to Lehmann these 
conditions are either rapidity of crystallization, viscosity of the 
solution, or a low degree of solubility of the crystallizing eee 
in the solution. 
Thus he found that when a hot saturated solution of ammonium 
chloride, which has been suitably thickened with gum, is cooled, 
the quicker the crystals are formed and the more the viscosity is 
increased, the more delicate are the forms of the crystals. And he 
further found that in order to obtain skeleton crystals of potassium 
chloride out of aqueous solution, either an extremely thin stratum 
must be employed or, which attains the object more readily, the 
viscosity must be increased by the addition of gelatine. 
As an instance of the effect of a low degree of solubility, the 
same writer says of silver chloride, “ Small skeleton crystals can 
be easily obtained by rapid evaporation of its solution in ammonia 
or sodium chloride solution. Very fine large, well-formed crystals 
resembling those of ammonium chloride are produced by the 
1 Thin, that is to say, as compared with their curvature. It must be remembered 
that the curvature, considered with regard to molecular magnitudes, is always extremely 
slight. 
