Mr J. D. Dana on the Origin of Trap Miner aU. 199 



writers, they result from the process of segregation ; — that 

 is, a separation of part of the material of the containing rock 

 during its cooling by the segregating powers of crystalliza- 

 tion ; and in illustration of the process we are pointed to the 

 many segregations of feldspar, quartz, and mica, in granite 

 and other rocks, the siliceous nodules in many sandstones, 

 the pearlstones in trachytes and obsidian. Others have 

 thought them foreign pebbles, enclosed at the time the rock 

 was formed. Again, they are described as proceeding from 

 the vapours which permeated the rock while still liquid, and 

 which condensed as the rock cooled, in cavities produced by 

 the vapours. By a few it is urged, admitting that the cavities 

 are inflations by vapours like those of common lava, that 

 tJiey may have been filled either at the time the rock cooled 

 or at some subsequent time, either by crystallization from 

 vapours, or from infiltrating fluids, but more generally the 

 latter. 



Of these views we believe the last to accord best with the 

 facts. MacCulloch, in his system of Geology, — a work which 

 anticipated many of the geological principles that have since 

 become popular, — dwells at length on this subject, and sup- 

 ports the opinion here adopted with various facts and argu- 

 ments. Lyell also admits the same principles. A review of 

 the facts will enable us to judge of its correctness. 



1. In the first place, the cavities occupied by the nodules 

 are in every respect similar to the common inflations or air- 

 bubbles in lava. These cavities are open and unoccupied in 

 common lava, and may be no less frequently so in the ejec- 

 tions under water ; and should they not be expected to fill in 

 some instances by infiltration \ They are the very places 

 where an infiltrating fluid would deposit its sediment, or col- 

 lect and crystallize, if capable of crystallization ; and such in- 

 filtrating fluids are known to permeate all rocks, even the 

 most solid, and especially if beneath a body of water. It is 

 evident, therefore, that we are supporting no strange or im- 

 probable hypothesis. On some volcanic shores one variety 

 of the process may be seen in action. The cavities of a lava 

 may be detected in the process of being filled with lime from 

 the sea-water washing over dead shells or coral sand, and at 



