[95] (1) A definition, if it is to be called perfect, must explain the inmost essence of a thing, and must take care not to substitute for this any of its properties. (2) In order to illustrate my meaning, without taking an example which would seem to show a desire to expose other people's errors, I will choose the case of something abstract, the definition of which is of little moment. (95:3) Such is a circle. (4) If a circle be defined as a figure, such that all straight lines drawn from the center to the circumference are equal, every one can see that such a definition does not in the least explain the essence of a circle, but solely one of its properties. (5) Though, as I have said, this is of no importance in the case of figures and other abstractions, it is of great importance in the case of physical beings and realities: for the properties of things are not understood so long as their essences are unknown. (6) If the latter be passed over, there is necessarily a perversion of the succession of ideas which should reflect the succession of nature, and we go far astray from our object. [96] In order to be free from this fault, the following rules should be observed in definition:-- I. (1) If the thing in question be created, the definition must (as we have said) comprehend the proximate cause. (2) For instance, a circle should, according to this rule, be defined as follows: the figure described by any line whereof one end is fixed and the other free. (3) This definition clearly comprehends the proximate cause. II. (4) A conception or definition of a thing should be such that all the properties of that thing, in so far as it is considered by itself, and not in conjunction with other things, can be deduced from it, as may be seen in the definition given of a circle: for from that it clearly follows that all straight lines drawn from the center to the circumference are equal. (5) That this is a necessary characteristic of a definition is so clear to anyone, who reflects on the matter, that there is no need to spend time in proving it, or in showing that, owing to this second condition, every definition should be affirmative. (6) I speak of intellectual affirmation, giving little thought to verbal affirmations which, owing to the poverty of language, must sometimes, perhaps, be expressed negatively, though the idea contained is affirmative. [97] The rules for the definition of an uncreated thing are as follows:-- I. The exclusion of all idea of cause--that is, the thing must not need explanation by Anything outside itself. II. When the definition of the thing has been given, there must be no room for doubt as to whether the thing exists or not. III. It must contain, as far as the mind is concerned, no substantives which could be put into an adjectival form; in other words, the object defined must not be explained through abstractions.