The structures of these two kinds of experiences are different, and the difference is not one of subject but of the relations _inter se_ of the several impulses active in the experience. A poem of the first group is built out of sets of impulses which run parallel, which have the same direction. In a poem of the second group the most obvious feature is the extraordinarily heterogeneity of the distinguishable impulses. But they are more than heterogeneous, they are opposed. They are such that in ordinary, non-poetic, non-imaginative experience, one or other set would be suppressed to give as it might appear freer development to the others. The difference comes out clearly if we consider how comparatively unstable poems of the first kind are. They will not bear an ironical contemplation. We have only to read _The War Song of Dinas Vawr_ in close conjunction with the _Coronach_, or to remember that unfortunate phrase ‘Those lips, O slippery blisses’! from _Endymion_, while reading _Love’s Philosophy_, to notice this. Irony in this sense consists in the bringing in of the opposite, the complementary impulses; that is why poetry which is exposed to it is not of the highest order, and why irony itself is so constantly a characteristic of poetry which is. These opposed impulses from the resolution of which such experiences spring cannot usually be analysed. When, as is most often the case, they are aroused through formal means, it is evidently impossible to do so. But sometimes, as in the above cited cases, they can, and through this accident literary criticism is able to go a step further than the criticism of the other arts. We can only conjecture dimly what difference holds between a balance and reconciliation of impulses and a mere rivalry or conflict. One difference is that a balance sustains one state of mind, but a conflict two alternating states. This, however, does not take us very far. The chief misconception which prevents progress here is the switchboard view of the mind. What conception should be put in its place is still doubtful, but we have already (Chapters XIV and XX) discussed the reasons which make a more adequate conception imperative. The rest of the difficulty is due merely to ignorance; we do not yet know enough about the central nervous system.