native Carbonate of Alumina and Lime. 463 



which it lost no carbonic acid, but apparently a little of its 

 combined water. It gave the analysis in the second column. 

 Other specimens yielded the numbers in columns III. and IV. 





I. 



II. 



III. 



IV. 



Silicic acid . . 



622 



5-87 



5-41 



5-30 



Alumina . . . 



41-04 



39-58 



36-32 



40-51 



Carbonic acid . . 



10-91 



14-77 



18-15 



14-14 



Lime . . . 



7-37 



11-22 



11-62 



9*18 



Water . . . 

 Traces and loss 



. 33-16\ 

 . 1-30 J 



28-56 



29-16 



30-87 



100-00 100-00 100-66 100-00 



Now in each of these cases the carbonic acid is far more than 

 sufficient to neutralize the lime. Thus 



7'37 parts lime neutralize 5*79 parts carb. acid, leaving 5*12 



11-22 „ „ „ 8-81 „ „ „ „ ' 5-96 



11'62 „ „ „ 9-13 „ „ „ „ 9-02 



9-18 „ „ „ 7-20 „ „ „ „ 6-94 



In what way is this excess of carbonic acid combined ? It never 

 exceeds in amount that which would be required to form bicar- 

 bonate, but in three instances it nearly approaches that quantity, 

 hence the lime might be conceived as existing as such; or it 

 might be carbonate of alumina ; or a double carbonate of lime 

 and alumina; or collyrite in which part of the silicic acid is 

 replaced by carbonic acid. But each of these suppositions has 

 its difficulties. Bicarbonate of lime in a solid form is unknown ; 

 yet it is conceivable that alumina, by its remarkable power of 

 withdrawing other substances from solution, might have enabled 

 such a bicarbonate to exist in combination with itself. Of the 

 existence of any carbonate of alumina we have as yet no proof, 

 whether as a mineral or a production of the laboratory. A 

 double carbonate of lime and alumina was purely hypothetical. 

 The partial substitution of carbonic for silicic acid has not 

 hitherto been recognized, that we are aware of, in mineralogical 

 chemistry; and though the results of analysis of the first 

 and second specimens given above would accord very well 

 with that view, yet the third and fourth specimens show too 

 much silicic acid, unless indeed we suppose that they were 

 derived from a collyrite much richer in silica than those hitherto 

 examined. 



If in the above analyses we view the lime and carbonic acid 

 as wholly in combination, and reject them as adventitious, the 

 remaining mineral will have very nearly the same composition 



