Permanence and Evolution. 25 



and investigated, and depending for the most 

 part on perfectly well known mechanical laws, 

 entirely expressible in terms of more or less ; 

 and he proved, or rather he pointed out, that 

 only suppose these operations to have gone 

 on long enough, they must have produced 

 appearances similar to those we see in the 

 earth's crust. Here, then, was nothing which 

 was not clear, closely reasoned, and scientific. 

 How different when the matters spoken of are, 

 as I showed before, not properly quantitative, 

 and still more when the modern changes ad- 

 duced are entirely conjectural. In science we 

 have nothing to do with what is probable as a 

 basis ; we may form hypotheses as guides for 

 investigation, or to group together instances, 

 but our bases must be proved facts. One man's 

 probable is another man's improbable. 



Lyell no doubt at times used incautious 

 language, as if the supposition of causes having 

 acted at one time of the world's life, and not 



