﻿64 Canadian Becord of Science. 



and serpentinous layers appear to alternate, and occasional 

 fragments of Eozoon occur in both, while the smaller 

 forms resembling fossils are, so far as can be observed, 

 limited to the serpentinous layers. 



At Arnprior on the Ottawa a portion of the Grenville 

 Limestone presents dark graphitic layers parallel to the 

 bedding, and giving it a banded grey and white appear- 

 ance which has led to its use as a marble. An analysis 

 by Dr. Harrington shows that the graphitic layers contain 

 8.32 per cent, of magnesia, the lighter layers only 2.57 

 per cent., in the state of grains or crystals of dolomite. 

 Associated with the marble there are also beds of brown- 

 weathering dolomite, affording 42.10 of magnesia. The 

 graphite in this marble, under the microscope appears as 

 fibrils and groups of minute clots, and sometimes coats 

 the surfaces of crystals or fragments of calcite, the appear- 

 ances being not unlike those seen in carbonaceous and 

 bituminous limestones of later date. 



In both the above cases the magnesium carbonate is 

 evidently .an original ingredient of the bed, and cannot 

 have been introduced by any metamorphic action. It 

 must be explicable by the causes which produce dolomite 

 in more recent limestones. 



Dana has thrown light on these by his observations on 

 the occurrence of dolomite in the elevated coral island of 

 Matea in Polynesia,^ under circumstances which show 

 that it was formed in the lagoon of an ancient coral atoll, 

 while he finds that coral and coral sands of the same 

 elevated reef contain very little magnesia. He concludes 

 that the introduction of magnesia into the consolidatmg 

 under -water coral sand or mud has apparently taken 

 place— "(1) In sea-water at the ordinary temperature; 

 and (2) without the agency of any other mineral water 

 except that of the ocean;" but the sand and mud were 

 those of a lagoon in which the saline matter was in pro- 



1 " Corals and Coral Islands," p. 356, etc. 



