AND ON LOGIC IN GENERAL. 185 



name of no more than is contained under the three, A and B and C. That is, let X = A, B, C; 

 let Y = A-B-C, in which the h^iphens are frequently dropped. Then I say that X has more 

 extension than A, unless by casualty B and C be both contained in A : also that Y has more 

 intension than A, is more descriptive, more discriminative, more intense; unless, by casualty? 

 B and C should only be components of A. I call (A, B, C) an aggregate, and A one of its 

 aggregants ; A-B-C a compound, and A one of its components. And (A, B) is not A + B, but 

 A + B-AB; as pointed out by Mr Boole. The distinction may also be made as follows: — If 

 A and B be aggregants of X, it means that X contains both : if components, that X is con- 

 tained in both. 



A name may belong to an individual object of thought, objectively considered, or to the 

 class in which that object is placed by possession of a quality, objectively considered as inhering 

 in it. The name may also belong to the quality itself, and also to the class-mark, or attribute, 

 the subjective quality, the possession of the mind, by which it can think of the class subjec- 

 tively. These distinctions are fundamental parts of thought. Thus one name has no fewer 

 than four separate uses : and the battle of the four senses is a large part of the history of logic. 

 There are two reductions of plurality to unity, that of objects to class, that of qualities to 

 attribute. Both are often denied ; some maintain that we cannot think of class without the 

 individual object, nor of attribute without an object to qualify by it. But perhaps it is more 

 common in the general world implicitly to deny the subjective unify of class, and the objective 

 plurality of quality. However others may be constituted, I find myself conscious of full and 

 clear possession of both class and attribute, as second intentions. Illustration will be useful, 

 and none is better than the oldest. We are in a warehouse full of packages of all kinds : each 

 package has one or more seal-impressions on it, its qualities, by which alone we see the existence 

 of the packages. These packages represent, some the material, some the mental substrata 

 about the existence of which so much discussion has arisen : they are perfectly dark, and the 

 seal-impressions shine by a light of their own. We can put together all which have one 

 seal-impression, and so form a'class, the members of which might be scattered among other 

 classes, if we sorted by other seal-impressions. Perhaps all the impressions of one design come 

 from one original seal, universale ante rem : but this is the question of realism, and is purely 

 metaphysical : our affair is with the sorting &c. of the packages. Certain it is that for our 

 own use we take off on what we call memory a seal from each class of impressions, an inverse 

 process, that we may have the means of testing new impressions when we come to them, and 

 that we may always recall the design when we want it. Thus we make the quality, universale 

 in re, give us an attribute, universale post rem. 



The denial of unity in plurality is made by saying that man, for instance, thought of as a 

 class, means only the conception of an individuum vagum from among the class individuals, 

 who must be tall or short, fair or dark, &c. : also that attribute is only quality thought of 

 as appertaining to this individual. All but this, they say, is only grasping a name : the 

 conception of a class is only the imagination (calling up an image) of an individual. I 

 do not know how it is with others, except from their own report, which I cannot question. 

 But it is not so with me. Procrustes has hitherto been, in most cases, one of the names of 

 a psychologist ; a speculator who stretches or shortens the rest of his species to his own 



Vol. X. Part I. 24 



