VANN, J.

The surrogate refused probate because the paper purporting to be a will was not "subscribed by the testator at the end" thereof as required by statute, and the decree was affirmed by the Appellate Division for the same reason. As the essential facts were stipulated and the original paper itself is produced as one of the facts agreed upon, the question whether it was signed at the end is a question of law. The entire printed form was treated below as the first page, the sheets numbered from one to six as succeeding pages, and thus the conclusion was reached that the decedent did not sign at the end of the paper. When read in this way the instrument does not read naturally or consecutively and does not make sense. The physical position of the six sheets, the place and method of attaching them and the closing paragraphs at the bottom of the printed form are substantially ignored. The natural order of reading the paper is subverted and an artificial order substituted, not to aid but to overturn the obvious intention. The natural order of reading it is to begin with the opening words on the printed blank and continuing with form and sense reasonably connected, to read the first of the numbered sheets and turning it over in the usual way with legal papers, to read the other sheets in their actual order and after the last has been read to turn that over also, when the closing paragraphs of the printed form follow, the end of the instrument is reached and no part thereof follows the signature. Who would read it in any other way unless he wished to destroy it as a will? Who would turn over the six sheets without reading them, read the testimonium clause and then turn back to read the rest? In reading an ordinary card calendar with the record of the month of December printed on the card itself and descriptive matter printed above, with the records of the other months on slips attached in the usual way, that of January being on top, who would not regard December as the end both of the calendar and the card? The slips become by the place and method of attachment virtually embodied in the card. So the six sheets of the paper in question are part of the body of the will, being physically incorporated therein and not, as in some of the cases, wholly without the body and merely referred to therein. The essence of the paper subscribed is not the printed form alone, but the printed form with the six sheets so inserted therein as to become blended therewith at the point of insertion. Thus the physical and the literary beginning, body and end of the instrument are the same, and the signature of the testator is found at the end.

The position of the courts below, however, finds some support in the adjudged cases, for we have gone far to protect testators from the danger of imposition and fraud through unauthorized additions and changes in their wills. That danger does not exist when, as in this case, the will is holographic, and while we do not regard that fact as controlling, it is worthy of remark that none of the six numbered sheets could have been changed without the co-operation of the testator himself. While the pins could be removed and other sheets substituted, this is true of many wills written on different sheets and fastened together by rivets or tape. There is no statute forbidding the use of separate sheets or directing how they shall be joined together. Cases where wills have been altered after execution are very rare, as the records of the courts show, while cases where the intention of the testator has been wholly defeated by a rigid construction of the statute requiring subscription at the end of the will are alarmingly frequent. Thus Judge CULLEN, writing for the Appellate Division in Matter of Andrews ( 43 App. Div. 394, 401), said: "With every disposition to uphold this will, we do not see how it can be done by this court. If limitations or qualifications are to be made on the cases on which this opinion is based, those limitations must proceed from the Court of Appeals. If the decision now made by us should be upheld by that court, it will follow that at least six wills (including one in the fourth department, 42 App. Div. 593), undoubtedly containing the true testamentary dispositions of testators and executed by those testators with the intent to conform with the statutes of this State, will have been held void because of the interpretation placed on the statutory requirements for safeguards against fraud. The rule which our decision enforces in this case operates only to prevent fraudulent additions to testamentary instruments, and not as a security against wills forged in their entirety. We think the decisions of the courts of this State will be examined in vain in the attempt to find six cases of alleged fraudulent additions to wills, or even half that number; and it must be conceded that as to this supposed danger the remedy has proved in practice far worse than the disease. In England a statute similar to our own, and construed as strictly by the courts of that country as our statute has been construed by our courts, was passed in 1837 (1 Vic. ch. 26). The evils resulting from it proved so great that in 1853 (15 16 Vic. ch. 24) it was modified. (1 Jarm. Wills [5th Am. ed.], 106.)"

The evil of fraudulent changes in wills is rare, while the evil of defeating wills altogether in the manner suggested is common. Hence, we think we have gone far enough in the direction of rigid construction and that the doctrine of certain authorities should not be extended, lest in the effort to prevent wrong we do more harm than good.

The Appellate Division, not without some reason, relied upon Matter of Whitney ( 153 N.Y. 259) and Matter of Andrews ( 162 N.Y. 1). In the earlier case the facts as stated by Judge EDWARD T. BARTLETT in his opinion were as follows: "The will is drawn upon a printed blank, covering only one page, and the testator and subscribing witnesses signed at the foot thereof. The subdivisions of the will, marked respectively `First' and `Second,' fill the entire blank space in the printed form, and at the end of the second subdivision are the words, `See annexed sheet.' On a separate slip of paper are written two additional subdivisions, marked respectively `Third' and `Fourth,' and this is attached to the face of the will, immediately over the first and second subdivisions, by metal staples, so that the slip annexed has to be raised up or turned back, in order to read the first two clauses." We held that probate of the instrument was properly denied, citing Matter of O'Neil ( 91 N.Y. 516) and Matter of Conway ( 124 N.Y. 455).

In the O'Neil case the instrument was drawn on a printed blank, the formal commencement being on the first page and the formal termination at the foot of the third page. The blank space was filled on the first, second and third pages and the last or thirteenth clause of the will was partly written on the third and the balance carried over to the blank fourth page. The names of the testator and the witnesses were subscribed near the bottom of the third page, below the formal printed termination of the will, and there only. The last clause was below and beyond the signature.

In the Conway case there was a similar state of facts, with the exception that at the end of the provisions in the body of the will were the words "carried to back of will," and upon the back of the sheet was the word "continued." Following this word were various bequests, and then below them were added the words "signature on face of the will" and there only was the will signed.

In the Whitney case, as is well said by the learned counsel for the appellant, the order of the sheets had to be disturbed if the will was to be read consecutively, while in the case at bar the order must be disturbed violently if one is to escape reading the will consecutively. The sheets attached in the Whitney will covered the preceding clauses thereof and prevented its being read consecutively; in the case at bar they covered only the subsequent clauses and so enable it to be read consecutively. In other words, we held in the Whitney case that when the instrument could not be read without turning the leaves backward and forward it did not comply with the statute. That is not this case, and while the facts are somewhat similar the principle involved cannot be extended without danger.

In Matter of Andrews a printed form was used consisting of one sheet, the two leaves of which were joined from top to bottom on the left side. After the printed formal part at the top of the first page was a blank space that was filled by the draftsman, who then skipped the second page, which had the attestation clause at the top, and filled the third, writing at the top thereof "2nd page." He then wrote on the second page at the top "3rd page," filled out the attestation clause directly beneath and it was subscribed and witnessed at that point only. Hence, the end of the will was found on the second page and the part written on the third page was neither subscribed by the testator nor incorporated by reference into the part that was subscribed by him. When read in its natural order the will did not run consecutively. It did not make sense without skipping backward and forward. As was said by Judge BARTLETT in his opinion: "We have here on one entire piece of paper, folded so as to make four pages, a complete will so far as form goes, on the first and second pages, and then follows on the third page of the blank and after the signatures of testatrix and witnesses on the second page of the blank a page marked `2nd page,' not connected with the will proper in any way, but complete by itself." (p. 10.)

We regard the decision in the Andrews case as extreme and as marking a boundary beyond which we should not go. The natural end of a will is where the draftsman stopped writing in the consecutive order of composition, which in the Andrews will preceded the third page and in the will before us succeeded the six numbered pages. The will before us when read consecutively as the mass of mankind would read it, has the signature at the physical and natural end thereof. We think that the will of the decedent should have been admitted to probate. We base our decision largely on the natural and consecutive method of reading, without turning backward, or skipping a part and then looking forward only to turn backward again, in order to have the sense connected and continuous. Form should not be raised above substance in order to destroy a will and the substantial thing in this case is a paper which reads straightforward and without interruption from the beginning to the end and when thus read the signature is found at the end.

The order of the Appellate Division and decree of the surrogate should be reversed, with costs in all courts payable out of the estate, and the proceeding remitted to the surrogate of Kings county, with directions to admit the will to probate.

CULLEN, Ch. J., HAIGHT, WERNER, WILLARD BARTLETT, CHASE and COLLIN, JJ., concur.

Order reversed, etc.