The second ground raised by the respondent for disputing liability was that the release was a release of the whole judgment and the sum of £800 could not be considered as a pro tanto payment of the debt. It was argued that the consent to the annulment was at least part of the consideration and, as the value of that could not be measured, no one could say whether any or how much of the £800 was part payment of the debt. The answer to that is that, on the construction both of the release and sec. 163, no part of the £800 can be taken to be the price of any consent to annulment. That sum was certainly the portion of the debt insisted on before the Bank would release the liability on the judgment. In effect the release was only for the balance - the £800 being a part payment. What, then, is the legal position apart from limitations? The Supreme Court differed, and, taking into account the learned primary Judge, was really equally divided. The matter is, therefore, by no means easy. The instrument of guarantee being "joint and several" contained in effect (leaving aside the third obligor) three possible obligations, namely, two several promises and one joint promise. The Bank chose to sue and get judgment against the appellant in respect of his several promise. The first problem is, what effect had that on the joint liability of both the appellant and the respondent? Certainly the Bank could never sue in respect of that joint liability. The principle can, I think, be traced through the cases, though it is not, so far as I can find, distinctly stated. Where the obligation is joint, there can be no doubt that judgment against one obligor puts an end to the original contractual liability of both - as against the judgment debtor, because that liability transivit in rem judicatam, so that he can never be sued again; as against the other obligor, because his right, if sued, to be sued in company with his co-contractor cannot be satisfied (King v. Hoare[4] and Kendall v. Hamilton[5]). But what is meant by transivit in rem judicatam? This is explained by Parke B. in King v. Hoare[6]. He says - referring to the maxim Transit in rem judicatam - "the cause of action is changed into matter of record, which is of a higher nature, and the inferior remedy is merged in the higher. This appears to be equally true where there is but one cause of action, whether it be against a single person or many." What is meant by "one cause of action"? This, to my mind, is the crux of the position. The "cause of action" is not, in my opinion, the liability to pay "jointly" or "jointly and severally" or "severally." Those terms signify merely certain conditions attached to the promise to pay. I freely confess that one may find observations difficult to square with this; but, on the other hand, when the leading case of Kendall v. Hamilton is carefully examined, what I have said will, I think, be found to be the underlying principle on which the decision is rested. I refer to the judgment of Lord Cairns L.C. He says[7]: - "I must say that the case of King v. Hoare2(1844) 13 M. & W. 494. appears to me to have been decided on satisfactory grounds. It is the right of persons jointly liable to pay a debt to insist on being sued together. If then there are three persons so liable, and the creditor sues two of them, and those two make no objection, the creditor may recover judgment against those two. But should he afterwards bring a farther action against the third, that third may justly contend that the three should be sued together. It is no answer to him to say that the other two were first sued and made no objection, for the objection is the objection of the third, and not of the other two. Nor is it any answer to him to say that whatever he pays on the judgment against himself he may have allowed in account with the others, because he may fairly require, with a view to his right of account or contribution, to have the identity and the amount of the debt constituted and declared in one and the same judgment with his co-contractors. If, therefore, when the third is sued, and required that the other two be joined as parties, the creditor has to admit that he cannot join the other two because he has already recovered a judgment against them in the same cause of action, this is equivalent to saying that he has disabled himself from suing the third in the way in which the third has a right to be sued." Lord Hatherley says[9]: "Each of the co-contractors has a right to be sued and to have the matter settled at once, instead of its being settled piecemeal." Lord Blackburn, in a most illuminating passages points out[10] that the defendant's objection, if his co-contractor is not sued, is not that there is a variance between the contract alleged and the proof. The objection was by plea in abatement. That view does, in my opinion, harmonize the cases, at all events the principal cases, including Isaacs & Sons v. Salbstein[11], Parr v. Snell[12] and Clarkson v. Davies[13]. If that be so, the judgment against Walker was one in which his obligation to pay £2,865 19s. 1d. (either jointly or severally) at the election of the Bank (see per Buller J. in Streatfield v. Halliday[14]) was entirely merged. That judgment, however, merely replaced by the higher instrument the liability originally existing by the original cause of action (Drake v. Mitchell[15]). All Walker's liability in respect of the secured debt was included in that judgment. True, Bowry had contracted to pay "jointly and severally" and not "jointly." He had thereby consented to the Bank suing and recovering judgment against Walker severally or jointly, and, even if severally, he undertook to be liable to pay the creditor severally himself. Even if several judgments are obtained, all would be liable. But he did not undertake that Walker should be released and still be liable himself to pay as if he had contracted "severally" only. So far as the Bank released Walker from his liability to pay the secured debt, the condition of "joint and several" liability, on the faith of which Bowry had entered into his obligation, would have been rendered impossible of observance. So far, the principle of Nicholson v. Revill[16] applies. But it applies only pro tanto. There was no release as to £800. The effect of the release was to relieve Bowry from the liability to pay the balance about £2,000. I say this apart from any right Bowry had to set up the Statute of Limitations, had he been sued. Apart from that and from the effect of that statute in this case, Bowry would, on ordinary principles of equitable contribution, be liable to recoup the appellant one-half of what he had paid, namely, a sum of £400. As to the Statute of Limitations I entirely agree with the view of the Supreme Court. No doubt, Walker had an equitable right, even before payment, to protect himself, but he had no right before payment to recover from Bowry what he is now claiming, namely, reimbursement for money actually paid.