50 HENRY G. SMITH. 



For observing their perfect cleavage they were mounted in 

 balsam under a cover-glass, and both reflected and transmitted 

 light used. For analysis they were decomposed by fusion with 

 the carbonates of potash and soda. The protoxide of iron was 

 estimated from a portion decomposed by heating in a closed tube 

 with sulphuric acid, and determining with potassium bichromate. 

 The following are the results obtained : — 



20-642 



10-906) 



20-963 



10-05, 



Silica (Si0 9 ) = 38-704 



Oxygen. 

 = 20-642 



Alumina (Al 3 O s ) = 21*795 



= 10-256) 

 = -650 ) 



Ferric oxide (Fe 2 3 )= 2-168 



Ferrous oxide (FeO) = 27*750 



= 6-167 1 



Magnesia (MgO) = 9-725 



= 3-890 J 



Manganese trace 





100-142 





From which the formula may be written as, 



(l (Fe, Mg) 3 +1 Al) 2 Si 3 or 6f g$, 3 SiO a + 2 A1 2 3 , 3 SiO a 



The sesquioxides are a little high, although they balance with the 

 protoxides fairly well. If the whole of the sesquioxide of iron 

 found is calculated into the protoxide, the amount is 29-7012 per 

 cent., containing oxygen equal to 6*6, and this added to the oxygen 

 of the magnesia equals 10-490, an amount nearly that of the 

 alumina. The ratio is that of an unisilicate belonging to the 

 garnet group. The presence of such a large percentage of magnesia 

 is not usual in almandine garnets, and approaches the "pyrcpe^ 

 or Bohemian garnets, although the large quantity of protoxide of 

 iron retains them in the iron-alumina group. 



I am indebted to Mr. C. Wooller, of Harris Street, Ultimo, for 

 directing my attention to the existence of this garnet bearing 

 conglomerate, and its locality. 



