IRON CONCRETIONS OF REDBANK SANDS 251 



lect any number of suites of specimens having from one to a dozen 

 chambers in which the order of the formation of the different arches 

 may be made out with reasonable certainty. It will be observed that 

 for each new chamber added to the concretion, a new curve is added 

 to its outer boundary, so that in the more complex individuals this 

 boundary becomes exceedingly tortuous. 



The uni-chambered tubes of very large and very irregular cross- 

 section, such as those shown in Fig. 6, are to be accounted for under 

 two suppositions. The general appearance of their walls suggests 

 that they were originally very complex individuals which had been 

 built up in the regular way — i. e., by successive additions of secondary 

 chambers — but which have lost 

 their interior partitions through 

 resolution. While this explana- 

 tion of their origin has a certain 

 plausibihty, it is not regarded as 

 the correct one. It is more likely 

 that we have to recognize here 

 the results of a process analogous 

 to, if not identical with, the 

 famiHar ** twinning" by which 



^ ^.rc , i- i FiG. 8. — Diagrammatic cross-scction of 



the different parts of a crystal ^ three-chambered concretion. 



may be oriented in contrary ways. 



If it be granted that the molecular tendencies residing in the cement- 

 ing material be competent to cause it to assume the cylindrical form, 

 it may be taken for granted that the cross-section of the young con- 

 cretion will be an arc of a circle before it becomes a full circle. It 

 is well-known that in the process of crystallization growing crystals 

 are not always guaranteed against twinning; the molecular forces 

 which govern crystallization are capricious. In the case of the young 

 concretion there is no reason to suppose that the arc would always be- 

 come a complete circle. From what is to be seen in numerous cross- 

 sections, it appears that the centers and degrees of curvature were 

 subject to change at frequent but irregular intervals. It must be 

 regarded as in some degree accidental that the composite Hne formed 

 by the numerous arcs should eventually form a closed area. In fact, 

 in the majority of cases, the line did not so close upon itself. In 



