E. S. JDana—Chondroditefrom the Tilly-Foster Iron 3fine. 87 



Following- the view of Kammelsberg, that the higher values of each 

 constituent are nearest to the truth, Mr. Hawes's analysis becomes as 

 follows. For comparison the results obtained by vom Rath for 2cl 

 type crystals from Vesuvius and from Sweden are added. 



Chondrodite. 

 Brewster, N. Y., Hawes. Swed 

 Silica 34-10 



. Magnesia 53-72 



Ferrous oxide 7 '2 8 



Alumina 0-48 



Fhiorine 4-14 



99-72 



Silicon 15-91 



Magnesium 32-23 



Iron 5-66 



Aluminum 0-26 



Fluorine 4-14 



Oxygen 39*78 



97-98 97-47 97-61 



Transforming the iron into an equivalent of magnesium, as also the 

 alumina (2Al=:3Mg), Mr. Hawes obtains further: 



Silicon 15-91, Magnesium 35-00, Fluorine 4-14, Oxygen 39-78, 



From these values a formula is deduced, which is essentially that 

 of the Swedish mineral according to v. Rath, 20(Mg Si20g)+ 

 MggSigFl^g. The close correspondence between the three analyses 

 in the above table is certainly very remarkable. It would have been 

 extremely interesting to have added an analysis also of crystals of 

 the 1st and 3d types; but, as will be appai-ent from what follows, the 

 material was not to be obtained. 



In completing the description of this variety of the mineral, in 

 general it may be repeated that it occurs usually in narrow veins, and 

 when free from alteration has iiniformly a deep garnet-red color. A 

 cleavage such as exists in humite (parallel to the basal plane) and has 

 been observed by Kokscharow on chondrodite from Pargas, could in 

 no case be discovered. The fracture is always conchoidal, and the 

 only thing that suggested cleavage was the laminated structure of 

 the massive specimens described, and the fractures parallel to e^ 

 which were quite conspicuous on two or three specimens. The ma- 

 terial in hand Avas not well adapted for optical determinations- 



