On Definitions. 271 



On the other hand, the definition of a compound body should 

 state merely the simple substances, and the proportions of them 

 which enter into the combination. These form the characteristic 

 differences that distinguish it from all other bodies, and serve this 

 purpose altogether independently of the new properties which the 

 compound body may possess. If we define copperas to be the sul- 

 phate of iron, or a combination of a certain portion of iron, with a 

 definite portion or quantity of sulphuric acid, this definition is com- 

 plete ; for it distinguishes the substance in question from every other 

 substance that can possibly exist. The properties of the compound- 

 ed mass may be few or many, simple or most extraordinary, but the 

 substance itself can never be any thing else, than the product of the 

 two constituent principles of which it is the combined result. That 

 compound bodies should possess properties so extremely different 

 from those of their simple elements, and even from the same ele- 

 ments in different proportions, is a very striking demonstration, added 

 to some others, that we are yet extremely ignorant of the constitu- 

 ent forms and properties of matter, and indeed that we are far be- 

 hind in our knowledge of the bodies in the universe. 



The definitions of instruments and operations, in this as in other 

 branches of knowledge, are best taken from the purposes which they 

 serve ; and, if there are different instruments for the same purpose, 

 it may be taken from the principle of their construction. These, 

 and similar terms in other departments of knowledge, do not admit 

 of logical definitions. They belong more properly to the class of 

 proper names, to be considered hereafter. 



In the three great branches into which Natural History is com- 

 monly divided, the want of accurate definitions is, if I mistake not, 

 a great and serious inconvenience, and one of the greatest difficulties 

 to be overcome in the prosecution of these studies. Where the ob- 

 jects of investigation are so extremely numerous, or rather innumer- 

 able, exceedingly little progress can be made without distinct classi- 

 fication, and classification is of Httle avail, without clearly defined 

 terms. To arrange all existing plants and animals into certain classes 

 and orders, and to distinguish these divisions and subdivisions by as 

 clear definitions as possible, so as to avoid all ambiguity in the descrip- 

 tion of species and individuals, has, if 1 am not mistaken, been the 

 object of all the systems of botany and zoology that have from time 

 to time been proposed. It is especially the object of the system, 

 which, for more than half a century, gained general ascendancy, that 



