1878.] I"-*- [Frazer. 



cite in the rock, a great number of specimens were tested for effervescence, 

 and all showed it in a prominent degree. As the above anatysis of Dr. 

 Genth gave no carbonic oxide, on inquiry., he writes : " The rock is full of 

 cracks and these are lined with a minute quantity of calcite. The portion 

 of which I sent you the analysis was as nearly as possible selected from 

 thai which did not show this coating. * * * Still some of the 'Igni- 

 tion ' may be Co 2 , " &c, &c. 



It is clear from the position of these calcite crystals that they cannot all 

 be due to infiltrated solutions of calcium carbonate through cracks in the 

 rock, because the individual crystals are isolated from each other. 



Their occurrence is peculiar and will be the subject of future study. 



Thi'ee separate, determinations of ignition were 3.65, 3.40, and 3.88 

 (average 3.64). 



Average determination of C0 2 = l,49 p. c. 



Annexed is the analysis, resolved into the ultimate constituents of the 

 rock (including 1 p. c. out of the ignition for C0 2 ). 



Analysis op Williamson's Point Trap. 

 Acid 



p. c. Oxygen p. c. 



Si 23.71] Oxygen 27.08 



C 0.27 I , 4 o " 0.73 . 0S1 



Ti 0.43 f 2iAb i " 0.27 f 



P 0.07 J " 08 J 



Al 7.55 \ 



Fe iy 2.69 i 



Basic 

 10.24 



Fe" 5.79 1 



Mn 0.37 | 



Mg 4.08 I 2Q00 



Ca 6.97 f ~ u,uu 



Na 1.40 | 



K 0.79 J 



54.72 



Oxvgen 6.64 \ 



1.15 / 



7.79 



Oxygen 1.65 ") 



0.11 | 



3.20 I 



2.78 ( 



" 0.49 | 



0.16 l 



S.M 



44.34 



Considering this collection of atoms as united into molecules in which 

 the oxygen performs partly a linking and partly a saturating function, we 

 may discover something as to the probable kinds of silicates contained.* 



The chemical units into which this analysis is resolved below represent 

 the amount of bond satisfying work which each atom performs, so that the 

 total amount might with propriety be considered the quantity of molecule 

 constructing work performed. It is calculated by considering the num- 

 ber of atoms of each element present, multiplied by the quauti valence of 



* On this subject see Report C, 1876, pp. 115 to 121. 



