542 W. A. Richardson — The Origin of Cretaceous Flint. 



precipitation of silver chromate bands in vertical tubes. Treating 

 these measurements in the same way as those from the gels the 

 rhythm-curve of Fig. 2b is obtained. Here the amount of separation 

 of flint lines is plotted against their distance below the top flint line 

 (since the curves of Fig. 1 suggest a downward diffusion). This 

 curve is slightly generalized to show the characteristics of all the 

 sections of Fig. 3. The rhythm may be analysed into three main 

 stages, starting from the entry — 



(a) AB, on the curve. Separation of the bands decreasing 

 downwards. 



(b) BC, dense banding. The bands separated by intervals nearly 

 constant, but varying slightly between 2 and 3 feet. 



(c) CD, the separation of the bands increasing downwards. There 

 is also a marked tendency to the development of granular zones in 

 stages 1 and 2. 



This rhythm, therefore, differs from that which obtains in the 

 gels by the addition of an earlier stage with decreasing separation of 

 the bands. On the other hand, it is so definite that it can hardly be 

 accidental. Moreover, the same rhythm will be repeated more or 

 less perfectly, not only in the sections figured, but also in any other 

 which is of sufficient length and where details are sufficient to apply 

 the test. It is interesting to treat a coal-shaft section in the same. 

 way for comparison. In Fig. 2c the rhythm curve from a South 

 Wales mine is given 1 — the intervals separating seams being plotted 

 against their depth below the first seam. If the occurrence of long 

 intervals separated by a series of short ones be taken as a kind of 

 periodicity, it is obviously of a character vastly differing from that 

 shown either in the Chalk or in the gels. At the same time this 

 ' comparison brings out still more strikingly the close similarity of the 

 phenomena in the last two cases. 



It is not my purpose to speculate on the nature of the chemical 

 reactions which took place. Liesegang supposes that at the time of 

 uplift the Chalk was permeated by a solution of silica, which in the 

 light of the results of the previous section we may suppose to be 

 furnished by a partial solution of organic remains in the Chalk. The 

 uplift caused a downward draining, and either the solution reached 

 a supersaturated state or accumulated a precipitant. In either case 

 the solution was precipitated rhythmically. In the early stage AB 

 of the flint-rhythm (applying Stansfield's results) there was probably 

 progressive increase in the concentration of the reacting solutions, 

 accompanied by decreasing separation of the precipitation bands. 

 There followed the zone of most favourable concentration and 

 densest precipitation, and later progressive dilution of the solutions 

 brought on a gradual lengthening of the separation interval. This 

 sequence in the banding is curiously reflected in the character of the 

 flints in the various zones. "Where flints occur in the mucronata 

 zone and the Chalk above they are generally small, often spongiform 

 (as though "sowing" had caused the precipitation of a solution in 

 the metastable state) ; and formed of immature flint. Scattered 



1 South Wales Coalfield (Mem. Geol. Surv.), pt. viii, p. 113, 1907. 



