OF FOSSILS. 219 



Determinate figures bear a refemblance to geo- 

 metric bodies, and it is not without fomc degree 

 of probability that they are faid to be derived 

 from the 'nature of the matter : An opinion that 

 has long influenced many to believe, that cer- 

 tain figures were proper and eflemial to diuc- 

 rent fubflanccs. The folly of this doctrine I 

 have elfewherc demonstrated at large *. If 

 therefore regular figures, and thofe bell defined, 

 arc fallacious, we arc furely not to rely on any 

 fuperficial characters which arc very often com. 

 mon to fubilances of the moil oppoiite qualities, 

 and never uniformly conilant in the fame fpe- 

 cies. 



xxvn. Pbyfical Marks of Eartbs. 



NOR. arc we wholly to neglect the ' plyfical 

 rnarhi which, though they cannot be fully efti- 

 mated by the external fenfcs alone, yet may be 

 afccrtaincd by eafy experiments, without the 

 trouble of dccompofition. Such, in the firit 

 place, are hardncfs and fpeciiic gravity; to 

 \vhich, indeed, we may add the relation to the 

 magnet. 



xxvin. Haninefs. 



DEGREES of hardncfs may be determined in 

 various ways, by the nail, the knife, or by ilccl; 



and, 

 * a Eflays, vol. 2. 



