92 ON THE CONVOLUTIONS OF STRATA, 



and wider to an indefinite extent, producing a solid unstrati- 

 fiecl mass. At the same time, they must have made room for 

 that yielding, partly by heaving up the superincumbent mass, 

 and partly by propagating the motion horizontally along their 

 own beds ; which last motion will be opposed by their friction 

 and inertia. We should then have three forces more or less 

 opposed to each other ; the force of elevation of the liquid, the 

 superincumbent weight, and the friction of the strata. The 

 consequence would be, that the strata, to a certain limited ex- 

 tent, would be thrust horizontally ; and so far as that action 

 reached, would be placed in the exact predicament of the 

 pieces of cloth in the old experiment, or of the clay in that be- 

 fore us, (figs. 4. and 6. Plate IV.) We have every reason to ex- 

 pect, then, that our experiments have been a faithful repre- 

 sentation of what would happen in a similar case in nature. 

 And the results we have obtained, bearing an exact resem- 

 blance to the peculiar forms of the strata of killas, seem to 

 justify that conclusion. If we suppose two such invasions of 

 matter in fusion, to rise parallel to each other, and both un- 

 der the circumstances above described, their influence exerted 

 in parallel, though opposite directions, would conspire, and 

 the space between them would be still more completely con- 

 voluted than where one action alone took place ; or would be 

 carried to a greater extent, than merely the double of one of 

 them. Were the neighbouring veins inclined at any angle to- 

 wards each other, the amount of the effect produced, and the 

 situation of the axes of convolution, would be difficult to cal- 

 culate ; but it is certain, that their effects would still conspire. 

 The complication would be still farther increased along with 

 the amount of the action, if a third vertical burst of liquid 

 matter occurred, so as to enclose a district of the stratified 

 mass within a triangle. We are not possessed of data by 



which 



