Opinion by
Mr. Justice Stewaet,We have here two appeals involving precisely the same question. The several appellants are the persons named in the following clause, indicated as the 29th, appearing in the last will of Admiral George W. Melville, deceased, bearing date 5th July, 1910, and duly probated:
“I do hereby appoint D'r. Max J. Stern, M.D., of Philadelphia, and Walter M. McFarland of New York, ex-officer of the Navy and at present writing in the employ *320of the Babcock & Wilson Boiler Company, of New York, cojointly to be the executors of this my last Will and testament to act without bond, and they to be equally my residuary legatees in all things relating to this my last will and testament.”
The will contained numerous devises and specific and pecuniary legacies, among others devises and legacies very considerable in value and amount to his two surviving daughters and their respective children. While it was testator’s expectation that his estate would be adequate to the payment of all these legacies, — he so expresses himself in the twenty-first clause of the will — it is evident that it was a matter about which he was more or less uncertain, since in the twenty-seventh clause he directs that,
“If it be found from any cause that the amount of my bonds, mortgages, and other belongings will not cover the monetary bequests I have made in this my last Will and testament, that monetary legacies or bequests be scaled down in amounts from $10,000, such amounts as will cover all other legacies; the amount in no case to be reduced below $5,000.”
The legacies so to be scaled down were the legacies to his sisters and their respective children. The fact referred. to is without consequence except as it affords some measure of the testator’s intended beneficence in making the appellants his residuary legatees. To this last will testator made four separate codicils, none of which call for special reference here excepting the last, which is dated 22d February, 1912, and gives rise to the present controversy. This codicil proceeds:
“Having had an expert Accountant assist me in making an inventory of my estate and finding that after all the bequests contained in my Will and several codicils there may remain about $150,000 of my undistributed estate. And being desirous of using the. residuary estate in some manner that will alleviate the sufferings of the *321indigent, deserving and aged poor, I desire the estáte to be used for such purpose.”
A devise and bequest of the entire residuary estate follows to the Pennsylvania Company for Insurance on Lives and Granting Annuities, the income to be expended in purchasing annually for aged and deserving poor entrance into desirable homes established for such purposes. The codicil concludes:
“In order to carry out the provisions of this codicil, clause No. 29 of my Will dated July 5, 1910, is hereby abrogated. The clause referred to relates to the appointment of Max J. Stern and Lieut. Walter M. McFarland as the residuary legatees of my original will. The fund thus established is to be chartered, if necesasry, in the State of Pennsylvania and is always to be designated ‘Estella Polis Melville Charity.’ ”
The testator died 17th March, 1912, within thirty days after the execution of the codicil, and the charity contemplated in the codicil therefore failed. The will and codicils were later proven, and letters testamentary on the estate were granted to these appellants. The question which now concerns us arises on the adjudication of the account of the executors, and, as stated by appellants, it is whether the revocation contained in the codicil of the twenty-ninth clause in the will appointing appellants residuary legatees was conditional or absolute. The Orphans’ Court in banc, hold the revocation absolute, and the charity having failed, awarded the residuary estate to. the daughters of the testator as his sole heirs and next of kin. The contention of appellants is that the revocation was conditional, depending upon the efficiency and sufficiency of the charitable bequest, and that the charitable bequest failing, the purpose of the revocation being thereby defeated, the residuary clause in the will remained unrevoked.
Whether we seek to resolve the question by ascertaining from the will and codicil the intention of the testator, or by applying settled rules of law, the result must *322be the same. The testator made the codicil because subsequent to the making of his will he had ascertained that his estate was $150,000 in excess of what he had supposed. He indicates in his will too clearly to admit of question, that in making appellants residuary legatees his thought was that the residuary estate would at most be inconsiderable. He was by no means certain that his estate would be adequate for the payment of his pecuniary bequests, and therefore provided for their abatement in case of a deficiency. What induced the making of the codicil was his discovery that, except as his will was changed, $150,000 of his estate would pass to these appellants under the residuary clause, whereas up to that time he had rested in the belief that whatever amount, if any, would pass under the residuary clause would be an inconsiderable part. So much we may safely conclude from what appears in the will and codicil. The substitution of another by the codicil to succeed to this enlarged residium was equivalent to a positive expression that those appointed by the will to take should be excluded because of his better understanding of conditions. The language of the codicil makes it clear to our mind that the new disposition made thereby was the result of a predetermination to revoke the old. We find nothing to indicate that testator in any event intended that appellants were to share to any considerable extent in his estate. It is not necessary, however, to decide the question on any such ground. We refer to it only to show that in this case the rule of law applicable does not defeat or disappoint in the slightest way any intention of the testator that can reasonably be derived from his will.
The codieil, executed with the same formalities as the will and duly probated, contains an express revocation of the twenty-ninth clause of the will, which appoints the executors residuary legatees,” the revocation leaves nothing to implication. True it is that the expressed purpose in the revocation is in order to carry out the *323provisions of the codicil, that is, the establishment of a charity. The dispositive part of the codicil failed, not, however, through any defect in the instrument, but because of something dehors, the death of the testator within a calendar month from the making of the codicil. Upon such state of facts can it be said that the revocation was conditioned on the efficiency of the charitable bequest, and that the charity failing the revocation fell with it? The authorities answer the question in the negative. They all recognize a clear distinction between failure of the dispositive part of the revoking instrument because of a defect in the instrument, and failure because of extrinsic circumstances; and there is entire concurrence of view that in the former case the revocation is inoperative, while in the latter it must prevail. Thus in Jones v. Murphy, 8 W. & S., 275, it is said by Rogers, J.:
“A will, though rendered inoperative by extrinsic circumstances, may revoke a former will. Thus if properly executed and tested to pass freehold lands according to the statute of Cha. 2, though it should be prevented from operating by the incapacity of the devisee or any other matter dehors the will, the former will is nevertheless revoked by it. So a will devising lands in fee to the heir at law, though void as to the purpose of the will, yet operates as a revocation if tested according to the statute.”
A leading case in oúr own reports is Price v. Maxwell, 28 Pa. 23, and peculiarly illustrative here. In that case the question was whether a revocation contained in a later will containing an express revocation was affected or impaired by failure of the devise contained in it, by reason of the testator dying within the time required by the act to give the devise effect. Lewis, C. J., in disposing of the case, says:
“If the will of 1856 contained no express clause of revocation of all former wills, a question might arise whether the will of 1841 was revoked by the devise afterr *324wards made. Where there are two wills in some respects inconsistent the latter revokes the former only so far as they are inconsistent with each other, unless there is an express clause of revocation: Jarman on Wills, 159. Where the second will is styled a codicil, or appears to have been intended for one, it is the duty of the court to construe them together as constituting one will; and in such case the revocation by implication will only take place where there is a clear inconsistency, and then only to the extent of that inconsistency, even in the case of a codicil, an express revocation of a former will must have its legitimate effect. But in the case before us, the property given to the school in the first will, is included in a general devise in the second will of all the testator’s estate to the same institution. There is, therefore, a manifest inconsistency showing that there was no intention that both wills should stand. The same thing is apparent from the appointment of new executors. But it is not necessary to regard these circumstances, because we have an express clause of revocation in the will of 1856, and that clause is not in any respect avoided or impaired by the Act of 1855. It stands in full force. The result is that the will of 1841 is revoked.
“But it is contended that this revocation was made upon condition that the devise to the school in the will of 1856 should take effect. How do we know this? Perhaps the intention to make the new disposition, induced the revocation of the old; and perhaps, the new disposition was only the result of a predetermination to revoke the old. There is nothing to lead the mind with anything like logical certainty to the deduction that either was the result of the other; and it is very clear that the heirs at law are not to be disinherited upon a mere peradventure. We have no right to add conditions not expressed by the testator, nor implied from his acts. He had it in his power to make conditions, but he made none, and we can make none for him.
“The rule in regard to revocations arising from incon*325sistent dispositions seems to be that where the second devise fails by reason of a defective execution of the second will, it is no revocation of the first: Jarman on Wills, 154; but where it fails from want of capacity in the devisee to take the prior devise is revoked: French’s Case, 1 Roll. Abr. 614; Roper v. Radcliffe, 10 Mod. 230; 8 Vin. Abr. 141, Tit. Devise R. Mr. Justice Rogers, in the case of Jones v. Murray, stated the rule correctly when he held, that if the second will was properly executed according to the statute, though it should be prevented from operating by the incapacity of the devisee, or any other matter dehors the will, the former will is, nevertheless revoked by it; 8 W. & S. 300.”
To the same effect are Union Lutheran Church Congregation’s App., 113 Pa. 32, Teacle’s Est., 153 Pa. 219, Wain’s Est., 156 Pa. 194.
Our cases are in entire harmony with the general rule on this subject, which is thus stated in 1 Jarman on Wills, 344:
“If, on the other hand, the new devise be ineffectual, on ac.count of the attestation being insufficient for a devising, though sufficient for a revoking will the revoking clause becomes inoperative on the principle before noticed that the revocation is conditional and dependent on the efficacy of the admitted new disposition, and that failing, the revocation also fails; the purpose to revoke being considered to be, not a distinct independent intention, but subservient to the purpose of making the new disposition of the property; the testator meaning to do the one so far only as he succeeds in effecting the same. But it seems, that, if the second devise fails, not from the infirmity of the instrument, but from the incapacity of the devisee, the prior devise is revoked.”
Page in his treatise on Wills, Section 271, thus states the rule:
“The later revoking instrument is much oftener a will or codicil. In the discussion of this branch of the subject a distinction must be noted between wills which *326contain an express clause of revocation and those which do not contain such a. clause. If the later will contains an express clause of revocation, the earlier will is thereby-rendered invalid, irrespective of the disposition of property made by the second will; and this is true even if the other provisions of the revoking will prove ineffectual.”
And again in Section 277 the author says:
“The testator may revoke his prior will by a later one which contains a clause of absolute revocation and is properly executed, and which by reason of something outside the will is ineffectual to pass the property sought to be devised. This differs from the case where the second instrument contains a. revoking clause and nothing more, for in the cases under consideration the instrument contains a revoking clause and a dispositive portion which fails of effect. The question is whether the revoking clause is conditioned upon the validity of the dispositive part or not. The general rule upon this point is that, a second will inconsistent with the first, perfect in form and execution, but incapable of operating as a will on account of some circumstance dehors the instrument, revoked the first instrument when the second contains a clause of expressed revocation.”
The authorities cited, to which others of like effect might be added, compel a conclusion adverse to the appellants’ contention. The several decrees are therefore affirmed, and the appeals are dismissed at cost of the respective appellants.