GEOLOGY OF TH£ CAMPSIH HILLS. 47 



lucentj and there shewed a muddy-brown colour. 

 In some parts of this obsidian, it is seen deviating 

 from its brilliancy, and from its compact vitreous 

 texture, and is evidently seen passing into a granu- 

 lar lava ; but the marks of this gradual transition 

 are soon lost, and the lava into which the obsidian 

 had passed, is then not to be distinguished in co- 

 lour, in fracture, nor in texture, from the general 

 lava of the other parts of the stream. Where 

 this obsidian appears in mass, and in a state of 

 perfect glass, it is very near to where it has been 

 first ejected from the side of the crater, and in a 

 situation where it must have undergone a rapid 

 cooling. In some parts of these congealed streams, 

 I could trace a transition of the obsidian into 

 pumice. In those places, the obsidian was seen 

 to contain scattered air-globules, which were al- 

 most always lengthened in the direction of the 

 stream. These globules gradually augmented in 

 number until the whole of the substance became 

 a light, fragile and frothy pumice. 



At the place where I discovered the great ac- 

 cumulation of obsidian and pumice, mentioned 

 above, I saw and obtained many specimens of ail 

 the transitions here described. In the various 

 streams of lava which had flowed from the prin- 

 cipal and central crater of this island, I found 

 many indications of pumice, but none of obsi- 

 dian. 



