120 H. S. Washington — Basalts of Kula. 



much brown mineral was, if anything, rather against the 

 hypersthene theory. 



Causes of alteration. — We now come to the question of the 

 causes of the two kinds of alteration. The limits of the 

 present article will not allow us to go into detail on this sub- 

 ject but some space may be devoted to the phenomena to be 

 explained and a sketch of the possible explanation proposed. 

 The facts are as follows : The brown hornblende is more sub- 

 ject to both forms of alteration than the green, and it is often, 

 but not necessarily, changed first to u brown aggregate," and 

 this to the simpler augite-opacite aggregate. The changes 

 begin at the surface, or along cracks, of the hornblende, and 

 thence work inwards, and when along cracks no contact with 

 the molten magma exists. Both alterations occur with neces- 

 sarily no marked change of form in the original crystal. A 

 layer of fresh hornblende can be deposited on substance altered 

 to brown aggregate, but has not been observed on that other- 

 wise altered. Both alterations are more common in the basic 

 volcanic rocks, and are the more marked the less glassy the 

 groundmass. The "brown aggregate" alteration seems to 

 have taken place at a greater depth than the other. 



A number of explanations have been .given varying from 

 the vague general idea of " corrosive action of the molten 

 magma" to the statement that the hornblende crystal has been 

 melted or dissolved, and the new alteration product (?) depos- 

 ited in its place. Lack of space forbids an examination of 

 them here, but it may be remarked that the perfect preserva- 

 tion of the crystalline form precludes, in the writer's mind, 

 any possibility of fusion or solution of the hornblende in a 

 molten moving magma. To sum up it may be stated in brief 

 that all the theories proposed are too vague or too improbable 

 to be admitted as true. 



The writer's idea in regard to the formation of the brown 

 aggregate is that occluded hydrogen (produced by dissociation 

 of the water almost always present in lavas) acts as a powerful 

 reducing agent, changing the Fe a O, of the original brown - 

 hornblende to FeO, and under certain conditions of high 

 temperature and pressure (with probably others) giving rise to 

 a mixture of the ferrous magnesian silicate, hypersthene, 

 along with augite and magnetite. This process, it will be 

 observed, requires no fusion or solution of the hornblende. 



A chief factor in the formation of the augite-opacite aggre- 

 gate the writer looks for in a simple molecular change (similar 

 to that of monoclinic to rhombic sulphur, etc.) under certain 

 conditions, among which are probably the chemical constitu- 

 tion of the magma and a slow rate of cooling, with certainly 

 a high temperature. 



