3. Some-vertebrates having the nasal canal communicating with the mouth are All-reptiles.

4. Some-cold blooded are Some-vertebrates that lack the outward sign of the organ

hearing.

Negatives.

of

1. All-mammals are not All-flying vertebrates. 2. All-vertebrates having cochlea to the labyrinth are not Some-fishes.

3. Some-Silurian and Devonian are not All

fishes.

4. Some-reptiles are not Some-vertebrates having the scapular arch articulated to the occiput.

In the process of generalisation, we begin with the real individual objects. Logically regarded, they are severally indivisible: we cannot affirm of a, that 'Some-a is not Some-a,' or of a', that 'Some-a' is not Some-a" &c. We select from among those individuals a", a""', a' &c., being all that have the mark of (infima) species A"": we accordingly sublimate that group into the place of species. We now select from among the marks of the A"", that mark which is next shallower (and therefore wider) than A”", viz. A”, and we accordingly sublimate the same group again, viz. to 6th subaltern genus: but there are other individuals, a", a", a", &c. which, though not possessing so deep a mark A"", have the shallower mark A"; and all such individuals we group under that name, and sublimate at once to 6th subaltern genus (vide 2nd column). We now select from among the various attributes of the two groups A"", and A", the next shallower (and therefore wider) attribute than A", viz. A", and these two groups we accordingly sublimate to 5th subaltern genus: but there are other individuals, a', a', a", &c. which, though lacking the deeper marks A" and A"", possess the shallower mark A”. These we group, and sublimate at once to the 5th subaltern genus (vide 3rd column); and so on, until all our groups are found in the summum genus; and, as a finishing stroke, all the individuals a, a, a, &c. which lack the marks A"", A", A”, A", A", A", and A', but possess the attribute A, (the shallowest and broadest of all) are per saltum sublimated from the region of reality to the summum genus.

In employing this table it must be borne in mind that all propositions can be read in either whole, without changing the sign of quantity. Thus All-A" is Some-A', may be read either, all A" contains in it (as attribute) some A', (depth); or, all A" contains under it (as constituent) some A', (breadth).

Terms have the same extension, or are coextensive, when they stand for the same objects. Thus man, rational, are co-extensive, because, (speaking of course with sole reference to beings of this world,) every man is rational, and every rational is a man, that is, the same objects are denoted by both terms.

Terms have the same intension, or are co-intensive when they stand for the same marks. Thus, rood, acre, are co-intensive, because every mark (except that of the extension) possessed by a rood is likewise possessed by an acre, and vice versá; that is, the same marks are connoted by both terms. 'All-man is All-rational,' is the co-extensive relation between terms that are not cointensive. All-rood is Some-acre' is the nonco-extensive relation between terms that are cointensive. Terms have equal extension, or are equi-extensive when the objects denoted by each are equally extensive.

Thus:

and

Justices of the court of Queen's Bench, 1750,

Justices of the court of Queen's Bench, 1850,

are terms which are co-extensive; though, in point of fact, some only of the former are some only of the latter. Terms have equivalent intension, or are equi-intensive when the marks connoted by each are of equivalent degrees of generalisation. . Thus, triangular, trilateral, are terms which agree in connoting triplicity, but the one connotes angularity, and the other laterality, and angles and sides are of equivalent degrees of generalisation, viz. geometrical-of-two-dimensions.

In the first of the last two examples the terms are co-intensive, as well as equi-extensive; and in the second, the terms are co-extensive, as well as equi-extensive. We might have chosen differently. Thus, for the first, we might have chosen the relation between hours on the dial, and a jury, (no hours on the dial are any of a jury). And for the second, we might have chosen equilateral and equiangular which are not co-extensive, since for instance, an equilateral parallelogram (not a square) is not equiangular.

Now what is the bearing of these remarks on the last table? Carry the eye down the first file (or column). The groups A, A', A", &c....A"", are composed of exactly the same individuals

a", a", &c. They are merely sublimated, from the species gradatim to the summum genus, according to the attributes they possess. Accordingly we say, that no two of the groups, A, A', A", &c... A"", are co-intensive, but they are all co-extensive. Again, carry the eye along the first