3 20 GEORGE F. BECKER AND ARTHUR L. DAY 



greater the crystal, the greater the weight supported upon its con- 

 taet surface (or rim), and the greater, a fortiori, the difficulty of 

 reaching the saturation concentration in any portion of the sup- 

 porting liquid layer and providing for farther growth from the 

 bottom. 1 



To make specific application of this analysis to our case, namely. 

 a single alum crystal resting upon a thin liquid layer in its own 

 saturated solution and supporting an outside load, the most favor- 

 able surfaces for growth will be the lateral faces, and there the 



1 That the description of these phenomena as observed by Bruhns and Mecklen- 

 burg differs in no essential particular from our own may be seen from the following 

 extracts from their paper (op. cit): 



P. 97: "The common view of crystal growth has hitherto been that a crystal 

 grows exclusively through additions of new matter from without. A crystal can there- 

 tore grow only where there is surface in contact with the solution, which offers room 

 for new material to be added, and where expansion can occur. A great deal of expe- 

 rience and many observations are in accord with this view." 



P. 100: "Unloaded crystals show continuous growth and the increase in weight, 

 which diminishes in the case of loaded crystals approximately in proportion to the 

 decrease in exposed area, proceeds normally; that is. where new material can find a 

 foothold there is growth, where it cannot there is none." 



P. 102: "A further circumstance which requires to be considered in connection 

 with the formation of the cup-shaped base [see Fig. 1] is this, that in a laboratory 

 experiment in a glass vessel the base of the crystal is not in contact with the vessel 

 hut rests upon a layer of water or of solution which adheres both to the glass and to the 

 crystal. The occurrence of such adhesion or adsorption layers is sufficiently familiar; 

 they are lacking only when the crystal grows fast to its support. By reason of this 

 liquid layer between the crystal and its support, the supersaturated solution is enabled 

 to diffuse under the crystal, even though the rate of diffusion in the capillary layer is 

 smaller than elsewhere. It will not advance far, however, for the molecules in excess 

 of the quantity needed to saturate the solution, when they attempt to pass close under 

 the crystal, will be quickly caught, that is. a rim [Wulst] will grow beneath the periphery 

 of the crystal, as observation in fact shows." 



P. 103: "Crystals which form on the. bottom of the vessel, without exception 

 show cup-shaped bases." 



P. 105: "If a large alum crystal is laid upon a smooth surface in a saturated 

 solution which is evaporating, its lower surface becomes cupped, but not in regular 

 steps like rock salt or bismuth. Instead of this the central portion remains practical!}- 

 tlat while a narrow supporting rim grows about it. Neither does long-continued growth 

 result in a stepped formation; the narrow rim moves outward while the inclosed area 

 continues nearly or quite flat." 



P. 105 : " The explanation of the phenomenon may perhaps be that the supporting 

 rim — which must of course bear the weight of the crystal — possesses a higher solubility 

 than the remainder of the crystal. Possibly also there is a difference in solubility in 

 different crystallographic directions." 



