Nodules of the Trimmingham Chalk. 455 



furnished by the action of carbonic acid on the sand of the 

 formation, and in the following way : — Water holding silicic 

 acid in solution, both in the crystalline and colloid condition, 

 filters downwards through the beds of the Blackdown sand, 

 and in its way meets a shell turned like a basin to receive it. 

 Now whether the silica shall pass through the shell or not 

 depends on two things, the permeability of the shell and the 

 state, colloidal or crystalline, of the silica. The shell is gene- 

 rally permeable, and from its lamella? of membrane acts pre- 

 cisely as a dialyzer : colloidal silica in solution on reaching the 

 shell is stopped ; the water passes through, leaving the silica be- 

 hind. This process goes on continuously till the silicic acid is 

 so far concentrated that it sets and prevents any further action, 

 or, as in some cases, actually overflows the shell. This con- 

 centration of silicic acid from weak solutions by the power of 

 dialysis furnishes us with a very simple explanation of the 

 condition of the Blackdown shells, without invoking the aid 

 of hypothetical sponges. The crystalline silica, which the 

 percolating water carries in solution, passes through the shell, 

 and in some cases, under favourable conditions, crystallizes 

 out in long fine prisms closely apposed to form a mamillary 

 layer of chalcedonic appearance. Now let us see what happens 

 to the calcic carbonate of the shell itself. The action here is 

 one of loose chemical affinities. The water carries in solution 

 as much silicic acid as it can hold. Calcic carbonate has a 

 greater affinity for carbonated water than has silica ; conse- 

 quently the calcic carbonate unites itself with the carbonated 

 water, displacing some quantity of silicic acid, which takes 

 its place in the shell and eventually entirely replaces it. We 

 have on these grounds every reason, a priori, to believe that 

 the silicic acid now found caught inside the shell was once 

 colloidal, and that the silicic acid which has passed through 

 the shell was crystalline. What do we find on examining 

 sections of the silica in these two positions ? Not one colloidal 

 and the other crystalline, as we should anticipate, but both 

 crystalline. But there is this very important difference 

 between them, viz. that in the one we should expect to be 

 colloidal, crystallization has commenced from various centres 

 in the mass, scattered mostly on both upper and lower sur- 

 faces, as if the whole had once been a jelly in which centres 

 of crystallization were set up, from each of which crystalliza- 

 tion radiated in all directions throughout the mass, till the 

 crystals of different centres interfered with one another, en- 

 croached on one another, and the process was completed. But 

 the one* we might expect to be crystalline is in a very different 

 condition : in this crystallization was evidently not an after- 



