Little, Justice.
The facts are set out in the official report.
The points of law relied on to reverse the judgment in •the present oas'e are controlled by the decision of this court in the case of Tittle v. Bennett, 94 Ga. 405. In the argument here, leave was asked 'and granted to review this latter decision. The exact question made is: whether a surety on a second bond given by the guardian under the provisions of section 2533 of the Civil Code, after making good a devastavit of 'the guardian which occurred prior to the release of a surety on the first bbnd executed, is entitled to contribution from such first surety on account of such devastavit. TO’ determine this question, it is only necessary to inquire and ascertain whether, with respect to a devastavit which occurs palor to the release of the first surety, the sureties on the first bond and those on the second bond are cosureties, inasmuch as the doctrine of contribution, as applied to sureties, is limited to cosureties. Civil Oodey §2992; 24 American & English Ency. of Law, page 811, par. 2 and authorities there cited; 1 Brandt on Suretyship and Guaranty, §254.
*120Broadly stated, the relation of cosnretyship may be said to exist where it appears that there are two or more sureties for the same principal debtor, and for the same debt or obligation, whether on the same or on different instruments. See Civil Code, §2992; 3 Pomeroy’s Eq. Jur. (2d ed.) §1418. ETor does it alter the relation or affect the right to Compel contribution, that the sureties are jointly or severally, or jointly and severally bonnd; nor that they are bound at different times and by different instruments, pee White & Tudor’s Leading Oases in Equity, part 1, vol. 1, *120 et seq., notes to case of Dering v. Earl of Winchelsea (reported in 1 Cox, 318).
In view of the fact that under our statute (Oivil O'ode, ■§2533) the surety on the first bond who applies for his release, is not and could not legally be released from any liability arising from a breach of the bond anterior to such release, so far as the rights of the obligee in the bond are concerned, and that the surety on the second bond by the terms 'thereof is made liable likewise for any devastavit occurring while 'the first bond was in force and prior to the release of the surety on the latter, it might be argued plausibly that as to such a devastavit 'there is a oomm'on liability, relatively to' the obligee of the respective bonds, as between the first and second, surety; that ’the first surety being still bound for all breaches occurring prior to his release, and the obligation of the second surety relating back so as to embrace liability also for such breaches, there is established a several liability, as between the first and second surety, springing from ‘the same principal and growing out of the same transactions; and therefore 'as to all such transactions the relation of cosurety would exist. As a general proposition, we are inclined to the opinion that, under such a state of facts, the doctrine would be applicable. The rule cannot, in reason, be limited to a case where the liability of each surety is equal as to amount and equally comprehensive as to the subject-matter upon which the suretyship oper*121.ates. The liability may be common as to a given transaction .¡and separate and distinct as to others. It makes no difference that they are bound in different sums, except that contribution could not be required beyond the sum for -which they had become bound. 14 Ves. 169; 2 Swanst. 192; 3 Bligh. 596. The fact that the obligations of the second ■surety embrace transactions which are distinct and foreign to the liability of the first surety, does not render the liability, relatively to the obligee in the respective bonds, for which the first surety is held liable, ¡any the less common ■to the first and second surety. The first and second surety .are severally liable for all defalcations prior to the release •of the first surety, -and the second surety separately liable for all defalcations, occurring after such release. But for •our statute, we should be inclined to hold that, as to a devastavit occurring prior to the release of the first surety, the first and second sureties were cosureties and hence entitled to contribution between themselves. However, under •our interpretation of the statute, we feel constrained to hold ■¡the converse -of the proposition. It is clearly the purpose of the. statute to release 'the outgoing surety as far as possible. It is obvious that, as between himself and the obligee of the bond, he cannot be released from liabilities springing from ■breaches of the bond occurring before his release; but whenever such circumstances ¡exist -as will allow him to he released from further liability, he is entitled to as full protection ¡as is obtainable at the time of his release.' The privilege to the gutodian to continue his trust by giving another bond and surety has the effect to postpone the right •of 'the first surety to have an ¡accounting from his principal, and thus possibly deprive him of means and measures of protection afterwards lost. The original bond being, presumably at least, ample to protect the obligee thereof for all devastavit occurring while it is of force, it could not he said that it was the purpose of the statute, in making the ■.second surety liable for past waste, to alone protect the in*122terests of the obligee of the respective bonds, but the principal object sought to' be accomplished by this retroactive feature of the second bond is to- afford to the released surety as much indemnity 'as possible, in view of the disadvantages to him attendant upon the postponing of his right to have his principal brought to an accounting. The release of the first surety could not operate to exonerate Mm from liability to the obligee of the bond for breaches made prior to-such release, for the rights and liabilities of the respective-parties were then fixed and vested, but there is no legal obstacle to imposing upon a surety who voluntarily assumes-liability under a second bond a primary liability, as between himself 'and the first surety, for all past as well as future-waste. Such a construction has heretofore- been placed on the statute now under review, by this court, in the case-of Sutton v. Williams, 77 Ga. 570, where it was held: “Where a guardian was appointed and gave bond, and subsequently one of the sureties, upon application, was released and discharged, and the guardian was required to give a new surety, which he- did, on a subsequent proceeding by bill in equity on behalf of the wards, to- recover from all the sureties for a devastavit of the guardian, the liability of the discharged surety and the second surety was not joint, but. several, both being primarily liable to the- wards, and, as-between themselves, the last surety being first bound.” In-commenting upon the statute, Justice Blandford says: “It is manifest that the last surety shall be liable first for any waste of the guardian, and while the- sureties on both bonds are liable to the wards severally primarily as between themselves, the last sureties on the bond are first bound.”' The case of Bobo v. Vaiden, 20 S. C. 271, is directly in. point, and is of like import with the Button cage supra. In a well considered opinion, Mr. Ohief Justice Simpson says: “As a general principle, it has been settled that the second bond becomes the primary security, and the first is at least suspended until the second is exhausted. True the first. *123■bond cannot be entirely discharged, so far as the parties interested in the estate are concerned; brut as between the sureties to- 'the two bonds, 'the second stands in the front rank and must protect 'the first.” Citing Glenn v. Wallace, 4 Strobh. Eq. 150.
It is apparent therefore, mot only that the first and last sureties are not cosureties, hut that the bond of the last surety stands as an indemnity to the first surety, and that if the latter is forced to pay off and discharge any devastavit of the guardian occurring before his release, he is entitled to recover the same of the second surety. Of course if it should appear that fraud or deceit in concealing a past devastavit had been practiced by the first surety upon the last surety, a different question would arise.
We think that the ruling made in the case of Tittle v. Bennett, 94 Ga. 405, supra,- is in harmony with our statute,- and therefore now adhere to and reaffirm the same.
Judgment affirmed.
All the Justices concurring.