SLATES. 393 



confusedly mixed up in it ? It Avill be, be argues, and be 

 argues rightly, to place the plates with their flat surfaces 

 more or less perpendicular to the direction in which the 

 pressure is exerted. He takes scales of the oxide of iron, 

 mixes them with a fine powder, and on squeezing the mass 

 finds that the tendency of the scales is to set themselves at 

 right angles to the line of pressure. Along the planes of 

 weakness produced by the scales the mass cleaves. 



By tests of a different character from those applied by 

 Mr. Sorby, it might be shown how true his conclusion is, 

 that the effect of pressure on elongated particles or plates 

 will be such as he describes it. But while the scales must 

 be regarded as a true cause, I should not ascribe to them a 

 large share in the production of the cleavage. I believe 

 that, even if the plates of mica were wholly absent, the 

 cleavage of slate-rocks would be much the same as it is at 

 present. 



Here is a mass of pure white wax : it contains no mica 

 particles, no scales of iron, nor any thing analogous to them. 

 Here is the self-same substance submitted to pressure. I 

 would invite the attention of the eminent geologists now 

 before me to the structure of this wax. No slate ever ex- 

 hibited so clean a cleavage ; it splits into lamina? of sur- 

 passing tenuity, and proves at a single stroke that pressure 

 is sufficient to produce cleavage, and that this cleavage is 

 independent of intermixed plates or scales. I have pur- 

 posely mixed this wax with elongated particles, and am 

 unable to say at the present moment that the cleavage is 

 sensibly affected by their presence — if any thing, I should 

 say they rather impair its fineness and clearness than pro- 

 mote it. 



The finer the slate is the more perfect will be the re- 

 semblance of its cleavage to that of the wax. Compare 

 the surface of the wax with the surface of this slate from 

 Borrodale in Cumberland. You have precisely the same 



