572 . EEPOBT 1880. 



fluid-cavities occur iu the constituent minerals, it appears to me very plausible to 

 suppose that those structures which are specially characteristic of volcanic rocks 

 are in great measure, if not entirely, due to the presence of associated or dissolved 

 vapour. The fluid-cavities prove that water was sometimes, if not always, present 

 as a liquid during the consolidation of granitic rocks ; and we can scarcely hesitate 

 to conclude that it must have had very considerable influence on the rock during 

 consolidation. Still, though these three extreme types appear to be thus charac- 

 terised by the absence of water, or by its presence in a state of vapour or liquid, I 

 think we are scarcely in a position to say that this difference in the conditions is more 

 than a plausible explanation of the differences in their structure. At the same time, 

 I do not know any facts that are opposed to this conclusion, and we should, 

 perhaps, not greatly err in thus correlating the structures, even though the water 

 was not the essential and active cause of the differences. 



Confining our attention to the more important crystalline constituents which are 

 common to the different types, we may say that the chief structural characters of 

 the crystals are as follows : — 



a. Skeleton crystals. 



b. Fan-shaped groups. 



c. Glass-cavities. 



d. Simple crystals. 



e. Fliud-cavities. 



Tiiese different structural characters are found combined in dilFerent ways in the 

 different natural and artificial products ; and for simplicity I will refer to them by 

 means of the affixed letters. 



The type of the artificial products of fusion may generally be expressed hj a + b 

 or b + c, that is to say, it is characterised by skeleton crystals and fan-shaped groups, 

 or by fan-shaped groups and glass-cavities. In like manner the volcanic type may 

 be expressed occasionally by b + c, but generally by c + d; and the granitic by 

 d+e. These relations will be more apparent if given in the form of a table, 

 as follows : — 



Slagtype . . { "^l^^ 



Volcanic type . . | ^^ , 



Granitic type . . . . d + e. 



Hence it will be seen that there is a gradual passage from one type to the other 

 by the disappearance of one character and the appearance of another, certain 

 characters the meanwhile remaining common, so that there is no sudden break, 

 but an overlapping of structural characteristics. It is, I think, satisfactory to find 

 that, when erupted rocks are examined from such a new and independent point of 

 view, the general conclusions to which I had been led are so completely in accord 

 with those arrived at by other methods of study. 



Conclusion, 



And now I feel that it is time to conclude. I have necessarily been compelled 

 to give only a general account of the subject, and perhaps, for want of adequate 

 description, many facts may appear more complex than they really are. Some are, 

 indeed, of anything but simple character, and their full explanation is, perhaps, 

 beyond our present power. The greater part are, however, much more simple 

 and easy to observe than to describe ; and, even if I have failed to make every- 

 thing as plain as I could wish, I hope I have succeeded in maldng the principal 

 point sufficiently clear to show that the structure of slags and of analogous artificial 

 products throws much light on the structure and origin of the various groups of 

 erupted rocks. I feel that very much still remains to be learned, and, as I think, 

 could be learned, by the further extension of this method of inquiry. What 

 strikes me most is the great necessity for the more complete application of experi- 

 mental methods of research ; but to carry out the experiments necessary to clear up 



