EARTHQUAKE PHENOMENA. 



419 



city, that (in the absence of better means) shall give the horizontal 

 velocity of shock within a norrow limit of error. 



Let there be constructed two similar sets of right cylinders — say, 

 each set six to twelve in number — all of equal height (a) and of the 

 same sort of material, but varying in diameter in each set, with a 

 uniform decrement from the greatest to the least. 



Convenient dimensions for earthquake observations of mean inten- 

 sity will be such that the cylinder of largest diamete^ shall have its 



altitude equal to three diameters, or b =z -' 7 and that the cylinder of 



Q 



least diameter shall have its diameter one -third of that of the greatest 

 one, or 6 = q. Any number of cylinders of intermediate diameters 



Fig. 6. 



X6 



may be interpolated between, and the greater the number thie-more- 

 accurate the instrument becomes. A series of six to ten in each set 

 will, however, be sufficient for any purpose. For observation of 

 shocks of extreme violence, larger diameters in proportion to- altitude 

 should be chosen for all the cylinders. 



The material of the cylinders is not important — cast-iron, stone, . 

 pottery, or other substances at hand — whose arrisses will not crumble- 

 away by being overthrown — may be used; but no material will be 

 found more convenient than some hard heavy seasoned wood, of uni- 

 form substance, straight grain, and equable specific gravity, from 

 which the cylinders can be formed in the lathe, and their bases 

 brought perfectly square to the axes with facility. 



Upon any horizontal and solid floor let two planks be placed, as in, 

 fig. G, with their directions in length respectively lying N. and S. and , 

 E. and W., each plank to be about three inches in* thickness, and in 

 width equal to the diameter of the largest cylinder, and its length. 



