362 J. E. TODD — CONXEETIOXS AND THEIR GEOLOGICAL EFFECTS 



the stages in this process, a is the original intercretion, b an inter- 

 mediate stage when the forming shell is separated from the undissolved 

 remnant by a leached zone of the impurity left behind. A photograph 

 of such a stage is shown in figure 13, plate 50. It will be noticed that 

 the oxide forms in the joints of the cracked concretion as well as on 

 the surface. The finished product is represented by c, when the cavity 

 of the hollow iron pebble has been opened and the incoherent impurity 

 rattled out. Such may be found in weathered sand and sandstone. 



But the form most pertinent to our subject is the nodular form with 

 rhythmical structure, which may resemble certain forms of accretion. 

 This rhythmical structure Dana, with his customary insight, recognized 

 as analogous to the concentric rings formed by a drying emulsion like 

 milk upon a flat surface. We may suppose a similar rhythmical action 

 of forces in both cases. In the emulsion the drying accumulates a film 

 of the precipitated particles where the drying is most active, namel}^ on 

 the outermost edge. When the liquid loses volume, so that it can no 

 longer fill that limit, it is withdrawn by cohesive attraction considerably 

 within the first limit, where it again stands and deposits another ring 

 till there is a similar necessity of withdrawing again, so that the process 

 is repeated indefinitely. In the case of a concretion, however, we must 

 suppose a vibration between the different influences, for we can not have 

 a volume of the liquid playing against a cushion of air because of the 

 geometrical relations and the inelasticity of water; but we can suppose 

 either an alternating action between carbonic acid and oxygen in the 

 water as the water gradually withdraws by drying, or we may suppose 

 by varying volumes of water corresponding to different showers or other 

 fluctuations of climate. The writer has not seen good examples of this 

 structure, but concentric color lines in weathered rocks are so often 

 found that there seems no good reason for doubting concretionar}^ action 

 of this sort. 



There are, moreover, structures resembling those just described, except 

 that the practical surface is a cavit}^ communicating with the outer air. 

 The more common example may be found around root marks, which 

 maybe so much coarser than the pores of the surrounding deposits that 

 they ma}^ be emptied of water while the latter are still saturated. Hence 

 they form around these root marks concentric cylindrical films of iron 

 oxide having a vertical length equal to the difference in height between 

 the top of the ground water in the root marks and that in the surround- 

 ing deposit. Moreover, as these levels fluctuate, these elongated con- 

 cretions may have greater vertical length. Such concretions of minute 

 form are very common in the loess. A diagram of such is shown in 

 figure 15, plate 49. They are often of minute rhythmical structure, 

 but frequently are nearl}^ solid except the small hole in the center. 



