296 JAMES CRAWFORD WATT. 



body. At this stage some spherules are uniformly opaque (Fig. 

 24c?), others show a marked radial striation (Fig. 24*?). 



The final step in the change now occurs with a very gradual 

 disappearance of the spherule, apparently by dissolving into the 

 surrounding solution. It decreases in diameter and also in thick- 

 ness until the remnant, diminished very much in size, appears like 

 a nebulous ghost of the former spherule and eventually fades com- 

 pletely from view. During the solution the opaque center of many 

 spherules seems to dissolve early (Figs. 24/ and 32), converting 

 the body into a ring. 



In those cases where the spherule is uniformly opaque (Figs. 

 24J and /) solution is very uniform at the surface, the contour of 

 the spherule remaining perfect while the size contracts. Where 

 there is radial striation, however (Fig. 24^ and g), the periphery 

 dissolves unevenly, giving the spherule a rough, rather ragged 

 surface. 



Coincident with the beginning of the process of dissolving of the 

 spherules, crystals occur in these areas which formerly were en- 

 tirely composed of spherules, and which for several months' dura- 

 tion had not shown a single crystal. These crystals grow steadily 

 both in size and in number (Figs. 19, 20, 21), while the spherules 

 disappear. The passage of the material from the spherule into 

 solution is evidently very transitory, it coming out again imme- 

 diately to be deposited in the crystal. 



What is the interpretation of this phenomenon? It seems to me 

 that there is a very evident one, and that this is the same process 

 very much delayed in starting, and very much protracted in its 

 action, which occurs rapidly in ordinary aqueous solutions. It will 

 be remembered that in precipitating calcium carbonate in pure 

 water it is at first in a granular, amorphous form, which later re- 

 dissolves and comes out of solution a second time as spherules 

 and crystals, which grow rapidly, the spherules all being converted 

 into crystals during growth. 



Again in colloidal solutions the same process happens, but is 

 delayed somewhat, taking hours or days instead of minutes. In 

 such cases an area of spherules seen one evening after a precipita- 

 tion, when examined next morning is found to consist of crystals. 



