200 ON THE MEIONITE. 



fuffictemlr^ expert in this kind of proof. In the present case, the cha- 



decide the vacter borrowed from the form is sufficient. In fact, ac- 



point. cording to Mr. Mohs, only one primitive form can exist in a 



species : but the primitive forms of the feldspar and the 



meionite are distinguished from each other in all the forms 



with which we are acquainted : their dimensions have been 



ascertained by a rigorous theory, the accuracy of which is 



proved by the agreement of calculation with experience. 



These alone therefore suffice to distinguish the species: if 



they did not, they might agree with other species, the forms 



of which would be different, and then one species would 



have two different forms, which is contrary to the hypothe- 



Princlpleof sis, and implies a contradiction. Here we see clearly what 



5odr"^° distinguishes the method of Mr. Haiiy. It is founded on 



the smallest member of characters possible. That which is 



taken from geometry, which is precise, is always employed, 



and frequently alone. When the primitive form obtained 



by mechanical division is a limit, that is to say, a regular, 



or at least a symmetrical solid, some other character must 



be added, since it may agree with several species. How- 



The geometri- ever, it is not necessary to determine the molecule of a mi- 



cal character ,. j./-i , ..,, 



not necessary n^ral, m order to hnd to what species it belongs. This is a 



to the student, labour requisite only to the author of the method, who can- 

 not employ means too precise for the determination of spe- 

 cies. He whose object is merely to ascertain the species of 

 a mineral, will find in the method of Mr. Haiiy more ma- 

 nageable characters, that will guide him to his end. 



•Conclusion, From the details into which I have entered it will be evi- 



dent to all, who are acquainted with the theory of Mr. Haiiy, 

 that the forms of the meionite are incompatible with those 

 of feldspar, that the integrant molecules of the two differ 

 essentially from eacb other, and, in fine, that these two sub- 

 stances OBght to remain separate in the rainexalogicalsys^ 

 tern. 



IX. 



