360 J. E. TODD CONCRETIONS AND TPIEIR GEOLOGICAL EFFECTS 



4. The more positive evidence of our view is shown in figures 17, a, 6, e, 

 and 22 of plate 50. Compare also figure 5, i)late 49, where it is clearly 

 shown that the origin of the concretion is complex ; that it has been 

 built by successive additions in the order indicated, and each has shown 

 the same tendency to crack in the interior. In this case the theory of 

 shrinking by drying must be dismissed at once, for it would require not 

 only moisture succeeding drouth repeatedly, but in some way local con- 

 centration of moisture and of argillaceous material deposited again and 

 again without any obvious reason. If it be suggested that in some way 

 the shrunken portions were all moist or of peculiar shrinkable nature at 

 the same time, we must explain the more difficult point of seeming suc- 

 cession of crescentic members overlapping one another. We must explain 

 also the still more significant fact that in every case the cracks are 

 broadest toward the center of the whole mass, not toward the center of 

 each member, as we should expect if they were all formed contempo- 

 raneously. 



5. Another evidence-of our view are the signs which sometimes appear 

 around these concretions of their crowding the surrounding strata. In 

 the Benton and Pierre shales a rude, cone-in-cone structure is often de- 

 veloped around them. Again, in the Titanotherium beds, silicious con- 

 cretions are found to have split the fossil bones in which they form. 



6. Another argument may be derived from the analogous behavior 

 of films on evaporating solutions and also in the tendenc}^ of crusts 

 around the edge of a liquid to blister up and crack off from the vessel 

 containing them. 



The following is a more detailed statement of the process : 

 First, the collection of similar molecules as in an accretion ; but condi- 

 tions are such that the solidification takes place not entirely on the 

 surface but largely between the particles already deposited, wedging 

 them apart with force sufficient to separate the portion inside the ex- 

 panding zone, and also to resist and force outward the embedding 

 strata. This may result mainly from the fact that the outer portions 

 have the first chance to appropriate incoming material. It ma}'' be 

 thought incredible that molecular cohesion is powerful enough to do 

 this, but when we think of its force when resisting tension it may help 

 us somewhat to understand when it manifests itself apparently in an 

 opposite direction. We sa}^ apparently, for really in both cases it is the 

 attraction of molecule for molecule, and the apparent diverse effects are 

 due simply to the different mechanical relations. Indeed, when we 

 think how capillary action may force molecules of water into the pores 

 of wood with sufficient power to split rocks in the quarrj^, and freezing 

 water has still greater power, it may not be so difficult to believe that 



