1918] Davis: The Radiolarian Cherts of the Franciscan Group 399 



appear very similar to certain structures in concretionary limestone 

 and advances a theory to explain these limestones as a result of diffu- 

 sion phenomena. 



In all the diffusion reactions, above described, it will be noticed 

 that the banding is due to the diffusion of a solution of one reagent 

 into a colloidal medium impregnated with a solution of the other 

 reagent. Liesegang showed that such reactions might occur, even in 

 non-colloid media. 



Cole 107 has recently used a somewhat similar hypothesis to explain 

 the banding of flints in the chalk. Liesegang had suggested that the 

 banding of flints in limestones was due to some sort of diffusion reac- 

 tion. Cole suggests that waters containing dissolved silica (derived 

 perhaps from the solution of siliceous tests, and spicules, originally 

 in the chalk) might diffuse down through the chalk during the early 

 stages of its uplift. During its downward diffusion such silica might 

 be precipitated in bands and nodules in the midst of the chalk. Later, 

 still other silica might precipitate with these bands as nuclei. This 

 combination of factors would account for the regularity of the band- 

 ing and also explain the replacement phenomena which are some- 

 times observed in chalk flints. 



This hypothesis involves a slightly different idea from that in- 

 volved in the experiments before mentioned. In this hypothesis only 

 one solution is involved — the solution of silicic acid. No mention is 

 made of the precipitant. Possibly the solid lime carbonate precipi- 

 tates the silica from its solution in a rhythmic way. 



In all the above experiments the banding was produced at the time 

 of formation of the precipitate. There was no movement of already 

 precipitated material. The hypothesis which was above set out in 

 explanation of the peculiarities of bedding of the chert involves a seg- 

 regation of material already in the solid state, and the objection might 

 be raised that there was no similarity between the two sorts of phe- 

 nomena. 



In certain experiments performed by the writer this latter effect 

 was produced. It was found possible to produce a rhythmic segre- 

 gation. 



In these experiments finely divided clay, which would remain in 

 suspension in quiet water for an hour; was used. A suspension of 

 this clay was mixed with a moderately strong solution of sodium sili- 



i" Cole, G. A. J., The Khytlimic Deposition of Flint, Geol. Mag., Decade vi, 

 vol. 4, pp. 64-68, 1917. 



