Mr J. J). Dana on the Origin of Trap Minerals. 267 



investigations only enable me to state that chabazite is usu- 

 ally covered by the rest (when associated with them), yet it 

 is sometimes superimposed on stilbite ; and apophyllite is 

 almost uniformly above all with which it may be associated ; 

 calc-spar is at different times above and below. We thus 

 arrive at the following, as the usual order of superposition : — 



1. Quartz. 



2. Datholite. 



3. Prehnite. 



4. Analcime. 



5. Chabazite, harmotome. 



6. Stilbite, Heulandite, scolecite, natrolite, mesole, Lau- 

 monite, apophyllite. 



It is a reasonable inference that the species which covers 

 the bottom of a cavity was first deposited, and, as a general 

 rule, that the others above were formed, either simultane- 

 ously, or in succession upon the lowermost, as their order 

 may indicate. Each is usually perfect in its most delicate 

 crystallizations, so that we cannot suppose that the minerals 

 first deposited often underwent change after their deposition, 

 though instances of this may no doubt be detected. 



It is also evident, that if there were any species formed 

 previous to the complete cooling of the rock, or if any require 

 for their formation an elevated temperature, they are those 

 first deposited — the first in the above series. A few consi- 

 derations will place this, if possible, in a clearer light. 



Quartz, as we have stated in a preceding page, and fully 

 remarked upon elsewhere, enters largely into solution during 

 submarine eruptions. This solution has been shewn, by ac- 

 tual experiment, to be a necessary consequence of such 

 action. This fact corresponds most completely with the 

 above deductions. Quartz usually forms the first lining of 

 the geode or amygdaloidal cavity, when it is found at all, 

 and, moreover, it is the most abundant of all amygdaloidal 

 minerals. 



Quartz may also proceed from decompositions of the rock 

 in the cold, and incrustations of this kind are known to oc- 

 cur ; but such an explanation does not account for its gene- 

 rally preceding all other species in filling cavities and seams 



