pletely the nature of the thing defined.

scription would of course enable us to

Such a de

identify the

subject, and to distinguish it from all other notions. And therefore it must be applicable only to the subject, otherwise it manifests, not the peculiar nature of the thing defined, but its common nature, the qualities which it shares with other things. As being applicable to the subject and to no other notion, it is co-extensive with it, and therefore may change places with it in the judgment. It is just as true to say that " every rational animal is man as that "every man is a rational animal." But if we said that " man is a warm-blooded animal," or that "man is a civilized animal," neither of them would be a definition, nor could the predicate in either become the subject, without some limitation. The former is a description that applies to more than man, the latter. to a part only of man; and of course neither of them would enable us to apprehend exactly what man's nature was.

Property is not easily distinguished from definition. Indeed, Aristotle confesses that property (idiov) i. e. something peculiar to the subject, and essentially its own, is a name which would naturally include definition, and would mean some attribute which belongs to all the subject and to it only; but he adds

the special limitation "without declaring the essence or nature of the subject." Every quality then which belongs to all the subject, and to no other, is a property, provided it be not used in the definition. It is co-extensive with the subject, and can therefore change places with it in the judgment without logical fault. Thus, "Man is capable of learning to write and speak correctly," might become "Every being capable of learning to write and speak correctly is a man."

But this subtle metaphysical distinction between the definition and the property is as difficult to maintain as it is unnecessary for the purposes of pure logic. How can we rely on being able to separate our notion of the nature or essence of a thing from the properties which accompany that nature? Let it be the definition of man that he is "a rational animal," and the property, that he is "capable of speaking correctly;" and how can we say that the latter is not in the essence, yet necessarily follows from the essence of man? It is a part of the essence, for "rational' implies it. In like manner, all the properties seem to be implicitly contained in every perfect definition. No criterion can be given for distinguishing betwee the essence and the inseparable accompaniment of the essence; and a larger acquaintance with the nature of things makes it evident that what one science regards as a property another must consider as essential,

and that there is no one paramount quality which is absolutely essential and can never be degraded to the rank of a property.

The predicable Genus is a class of which the subject is a contained part. It declares, though not completely, the nature of the subject. A subject may be included in many different genera by different sets of marks; a man may be good, brave, rational, mortal, fallible, sick, learned, and so on. But some of these qualities, as wholly separable from the nature of man, are to be considered not as genera, but as accidents. Genus, as being of the very nature of the subject, is inseparable from it. As including the subject in common with other species, it is not coextensive with it. Hence the transposition of the subject and predicate in a judgment which predicates the genus, cannot take place; "all roses are plants " cannot become "all plants are roses."

Accident is a quality which belongs indeed to a subject, but can be taken away from it without destroying its nature or essence. We predicate accident when we say that " a man is speaking." Accident cannot change places with its subject, because it does not apply to the whole of that subject and to it alone. But a criterion is wanting to distinguish between accident and genus or species. It is an accident to the people of this country that they were born in it; because we might conceive them to have

L

been born elsewhere; but then it has modified their nature or essence, and we understand by Englishman not merely one who was born within the four seas, but a man of particular feelings, views, and privileges, which are parts of his very nature. Here accident and genus or property seem to become confused. It is an accident too that this nail is rusty and that guinea bright, but then it shows that the gold has a property of resisting oxidation-which the iron wants, and might serve to place them in two distinct species of metals. Aristotle actually speaks of man as an accident of the genus animal, although it is commonly represented as one of its species*; no doubt because we might conceive that species annihilated without the destruction of the genus. It does not appear then that the predicable accident can at all times be distinguished from the others, which would be a valid objection against retaining the doctrine in which it holds a place.

We propose to abandon, as at least unnecessary for logical purposes, the distinction between property and definition, genus and accident; and to form, as Aristotle has also done, two classes of predicables; one

* Cat. vii. 14. In quoting the passage Crackanthorp says: "Omnia inferiora accidentia sunt respectu suorum superiorum." See too Cat. vii. 13.; Pacius, marginal note.

of predicables taken distributively, and capable of becoming subjects in their respective judgments without limitation, the other of such as have a different extension. In the former, the predicable has the same objects as its subject, but different marks or a different way of representing the marks. In the latter there is a difference both in the marks and the objects. The former may be called Definition, or Substitute; the latter, Attribute.*

§ 70. Definition explained.

Every predicate which denotes exactly the same class of things as the subject, may be called a definition. Whether it unfolds the genus and difference, or the property, or only substitutes one symbolical conception for another, it is useful to mark out for us more clearly the limits of the subject defined, and is therefore capable of being employed as a definition for some thinker or other. Logicians have always allowed that in our definitions we are bound to consider, not merely what is absolutely the explanation