552 S. Tab'er — Growth of Crystals. 



the consequent reduction in the concentration of the solution 

 tending to diffuse in under the crystal. A concave surface is 

 less soluble than a plane surface, and, therefore, hollow faces 

 are soon filled up except in those cases where crystal growth is 

 so rapid as to reduce the concentration of the solution faster 

 than diffusion can supply material to the less exposed surfaces. 



The form of the needle-like columnar crystals of copper sul- 

 phate obtained by the writer is determined by the fact that 

 each crystal is in contact with the solution furnishing the 

 material for its growth only at its base. The material for 

 growth is supplied through capillary openings so small that 

 their surface tension is sufficient to prevent the solution from 

 creeping up the crystals. The enlargement of these columnar 

 crystals takes place only in the direction of greatest pressure, 

 but this phenomenon is due merely to the circumstance that in 

 other directions the crystals are not in contact with the solu- 

 tion. The long slender columnar form is never found in crys- 

 tals of copper sulphate that have grown normally with the 

 solution in contact with several surfaces, because this form is 

 not the most stable. The total surface energy of one of the 

 columnar crystals is great as compared with the mass, and, 

 therefore, they are dissolved by solutions that are supersatu- 

 rated with respect to more stable crystal forms. If the col- 

 umnar crystals are placed in saturated solutions with normal 

 crystal forms, the latter tend to grow at the expense of the 

 former in the same way that large crystals replace smaller 

 ones. 



The same argument applies to the formation of ice columns. 

 The ice columns are usually capped by sand grains, small 

 pebbles, etc., which, as pointed out by Prof. Abbe,* have prob- 

 ably acted as refrigerating surfaces and served as nuclei for the 

 crystallization of the ice. These nuclei are not essential, how- 

 ever, and are sometimes absent even in nature. JSTumerous 

 experiments carried on by the writer with different kinds of 

 material, varying quantities of water, and different methods of 

 insulation and refrigeration all lead to the same conclusion : The 

 essential condition for the formation of ice columns is that the 

 water for the growth of the ice column must be delivered 

 through a small capillary opening to the base of the growing 

 crystal which must not elsewhere come in contact with water. 



There are of course other ways in which elongated, columnar 

 and even fibrous crystals may be formed. In some crystals, 

 such as rutile needles, the form is probably due solely to the 

 molecular forces controlling the formation of crystal faces ; 

 while in other cases external forces have been the controlling 



*Abbe, Cleveland, Ice Columns in Gravelly Soil, Monthly Weather 

 Review, vol. xxxiii, pp. 157-158, 1905. 



