ACCRETIONS AND INTERCRETIOXS 359 



clay may act as a nucleus, and after the formation of a more rigid shell 

 that subsequent shrinking by drying may i)roduce a cracked interior, 

 as Professor Dana and other writers have stated. The writer, after con- 

 siderable study and observation, is not aware of any clear proof of such 

 a case. In all concretions observed the cracked portion has not been 

 of such a character, but of similar composition to that of the rest of the 

 concretion. 



IN TERCRE TIONS 



On the contrary, from a study of such growths we have been led to 

 quite a different explanation for those showing a cracked interior. 

 Instead of the interior shrinking after the outside has become rigid, on 

 the contrary the outside or outer portion has expanded and become too 

 large for the interior, and it has done so with sufficient force to wrench 

 apart the interior. Moreover, this has been a gradual process, the grow- 

 ing zone passing gradually outward by contemporaneous accretion. The 

 cracks starting first in the central portion gradually follow the growing 

 zone outward. This process is theoretically shown in figures 3«-e, plate 

 49. To simplify the case, a globular form is assumed, but a lenticular 

 shape, similar to figure 4, is more common in the Cretaceous shale of 

 Dakota, and vertically elongated ones or rudely globular are more com- 

 mon in loess, as in figures 5 and 22a. In the Benton and Pierre cla3^s 

 they grow to a size of several feet, sometimes a rod or more in extent. 

 They are cracked into polygonal blocks 3 or 4 inches across. The evi- 

 dence in favor of the conclusion stated above may be given as follows: 



1. The cracking can not be due to drying, because (a) the material is 

 not clayey, but usually contains much sand, and is not distinguishable 

 from the outer portion of the concretion ; (6) because the concretions are 

 sometimes found by well diggers to be filled with water, and from this 

 fact, as well as from their shape, they are called "jugs." Moreover, the 

 crevices are often lined with crystals or filled with crystalline matter. 

 In the latter case the structure is known as a septarium. Such facts 

 show that the interior was saturated with water probably during all the 

 time of its formation. If the shrinkage was due to desiccation, why 

 should not the material when wet again recover its former bulk and the 

 cracks disappear? 



2. Nor can it be confidently referred to contraction from cementation. 

 It is difficult to believe that this can explain the amount of shrinkinsj 

 indicated in the figures, though we may frankl}'' admit our ignorance of 

 any quantitative evidence concerning the amount of shrinkage due to 

 molecular cohesion. 



3. Nor can we conceive of any chemical reaction producing such loss 

 of volume. 



