12 



given rise to many ingenious speculations, but up to this time none 

 of them have been satisfactory. 



It has puzzled everyone who has given any attention to the subject 

 to explain how flints, consisting of silica, should be found here and 

 thsre imbedded in . the chalk which is almost pure carbonate of 

 lime ; it is sufficiently evident that silica must have been present, 

 but in small quantity, in the chalk seas, by what means then did it 

 become isolated in the manner we see it, instead of being deposited 

 in a state of mixture with the chalk ? The origin of flints is pro- 

 bably dialytical. Mr. Graham found that certain membranes, and 

 also parchment paper, when in contact on the one surface, with a 

 solution containing a mixture of crystalloid and colloidal substances, 

 and, on the other surface, with water, will permit the passage to 

 the water of the crystalloids, but not of the colloids. This mode 

 of separation is called dialysis. If we put into the dialyser a very 

 dilute solution of silica in combination with an alkali, in which 

 form silica in solution nearly always exists, and add a few drops of 

 some dilute acid, the crystalloid salt of which the alkali is the base 

 will be separated, and a solution of pure silica left ; this will 

 gi-adually become concentrated, and when sufficiently so will 

 aggregate round any organic nucleus with which it may come in 

 contact. In the chalk seas the soft mud would act readily as the 

 membrane does in our experiment, and by a dialytical process 

 the silica would be concentrated, and encircle a piece of sponge or 

 any other organism and form a flint. 



Mr. Stoddart, of Bristolj^who has in this way made some artificial 

 flints, tells me that they correspond in a wonderful manner with 

 the natural ones, even to the characteristic layers which have 

 hitherto been supposed to be due to successive depositions of silex. 



The necessity for being comparatively brief has precluded me 

 from entering into any arguments for, or objections to, the several 

 points I have raised, but this being so, I have been careful to assign 

 to each proposition its proper degree of probability, and in a theme 

 that must of course have much in it that is hypothetical, to take 

 nothing for granted but what may be fairly considered to be 

 proved. 



