This inwardness is no less characteristic of Coleridge's treatment of the doctrine of atonement or redemption. It is intelligible so far as it comes within the range of spiritual experience. So far its nature and effects are amply described or figured in the New Testament, especially by St. Paul. And the apostle's language, as might be expected, “takes its predominant colors from his own experience, and the experience of those whom he addressed.” “His figures, images, analogies, and references,” are all more or less borrowed from this source. He describes the Atonement of Christ under four principal metaphors: 1. Sin-offering, sacrificial expiation. 2. Reconciliation, atonement, καταλλάγη. 3. Redemption, or ransom from slavery. 4. Satisfaction, payment of a debt. These phrases are not designed to convey to us all the Divine meaning of the atonement, for no phrases or figures can do this; but they set forth its general aspect and design. One and all they have an intelligible relation to our spiritual life, and so clothe the doctrine for us with a concrete living and practical meaning. But there are other relations and aspects of the doctrine of atonement that transcend experience, and consequently our powers of understanding. And all that can be said here is, “exit in mysteria.” The rationalism of Coleridge is at least a modest and self-limiting rationalism. It clears the ground within the range of spiritual experience, and floods this ground with the light of reason. There is no true doctrine can contradict this light, or shelter itself from its penetration. But there are aspects of Christian doctrine that outreach all grasp of reason, and before which reason must simply be silent. For example, the Divine act in redemption is “a causative act—a spiritual and transcendent mystery _that passeth all understanding. 'Who knoweth the mind of the Lord, or being his councillor who hath instructed him?' Factum est._” This is all that can be said of the mystery of redemption, or of the doctrine of atonement on its Divine side.