delivered the opinion of the court, October 9th 1876.

In the fourth point of the defendants below, the court were asked to charge that “ if the jury believe from the evidence that the receipt given by Barr to Hendrickson was altered from May 3d to May 31st 1870, the plaintiff cannot recover.” The answer was: “ There is no evidence in this case of any alteration of the dates of the receipt. Barr swears the dates therein are the same now as when he received it from Hendrickson, the agent, and Hendrickson does not swear that they have been changed or altered. The jury need not consider the question raised by this point.” This question *34was the pivot on which this whole controversy turned. The application for the increased insurance was made by Barr on the 3d of May, unquestionably. If it was agreed between Hendrickson and himself that the policy was to protect the property from the 31st of May, and if, in accordance with that agreement, the receipt was post-dated, the plaintiffs below, as attaching creditors of Barr, had the right to ask that the policy should be reformed by substituting the date which was really fixed for the commencement of the contract, in lieu of that which was adopted by the company in consequence of the agent’s mistake. The act of the agent in effecting the insurance having been ratified by the execution of the policy, there is nothing in the general facts on the record to indicate that the discrepancy of twenty-eight days between the date agreed on and the date mistakenly adopted, could have had any essential bearing on the question of the good faith of the contracting parties, or on the obligation of the contract itself.

As the case stands, the material inquiry is whether there was any evidence of the alleged alteration of the receipt which should have been submitted to the jury. On the 31st of May 1867, an insurance on a dwelling-house, kitchen, smoke-house and certain household furniture, amounting to $4000, was procured by Barr from the defendants, through the instrumentality of John H. Zeller, at that time their agent at Mount Joy. In the autumn of 1869, George R. Hendrickson, who had been appointed agent in Zeller’s place, made a survey with a view to the insurance of other property belonging to Barr. On the 3d of May 1870, an application was made for a policy for $8800, embracing the $4000 covered by the policy of 1867, and $4800 on the other property. The insurance was specified to be “ for the term of three years from the 3d of May, 1870, at noon,” and the application was dated “May 3d 1870.” The policy, which was dated on the sixth day of May 1870, stipulated for the payment of damages.by fire to the property insured occurring between the third day of May 1870, and the third day of May 1873. The date of the expiration of the risk, “3d May 1873,” was legibly endorsed upon the policy. A fire occurred on the 19th of May 1873, resulting in the loss of property not covered by the insurance of 1867, but included in the risk taken in 1870.

On the trial, the plaintiffs produced the receipt of Hendrickson purporting to have been given “May 31st 1870.” It was alleged on behalf of the defendants that this receipt had originally been dated “ May 3d 1870,” and had been altered by the addition of the figure and letters “ 1st.” It was apparent that a vacant space had been left between the figure or figures expressing the day of the month and the figures which indicated the year, hnd that this space had been left without a connecting scroll and entirely blank. Barr testified that when he took the receipt it was “ with the understanding that the insurance was to take effect from May 31st, the expira*35tion of tbe old policy,” and that the dates originally inscribed remained unaltered. He also testified that he had not observed before the fire the discrepancy between the receipt and the policy. On the part of the defendants, Hendrickson, the agent, was called as a witness. He stated that the receipt of the 31st of May 1870, was in his handwriting, with the possible exception (he was not able to speak positively) of the figure “1” in both places (at the foot of the paper and in the cancellation of the stamp) where it occurred in the dates. He said that the letters “ st ” did not look to him like his writing. On cross-examination he added that “there was nothing on the stamp to indicate it was not in his hand except his reasoning.” He said also that when the new insurance was effected, he made an allowance to Barr of 55 cents for the premium on the cancelled policy of 1867, for the period between the 3d and the 31st of May 1870.

In view of these general facts, how can it be said that there was no evidence to affect the integrity of the date of the receipt in controversy ? According to Barr, the understanding was distinct that the risk was to begin on the 31st of May. According to Hendrickson, the understanding was that it was to begin on the 3d of May, and in pursuance of that understanding, he swore that he refunded the premium on the first policy for the twenty-eight days it had yet to run. Then the terms and date of the application, and the stipulations of the policy were explicit and unmistakable. While the evidence of Hendrickson was vague, it still raised a question whether the receipt produced in cpurt was in the precise form in which it was originally executed. The fire occurred more than three years after the insurance was effected, and no objection to the terms of the policy had been made in the interval. It is true Barr swore he did not discover the discrepancy, and it appears that the policy was assigned to S. H. Reynolds, a lien-creditor, on the 11th of September 1871, and was probably therefore not in the hands of Barr from that time till the accident happened. Still the fact remains that the policy was received and retained without objection.

All these were facts bearing directly on the point contested. But aside from these, there were surrounding circumstances which, as tending in some degree, to illustrate the issue, should have been submitted to the consideration of the jury. The property destroyed by the fire was not included in the policy of 1867. Evidence was given by the plaintiffs to prove that this property was protected by policies of other companies., which were dropped by Barr when he obtained the insurance from the defendants in May 1870. The policies so dropped were not shown, and no proof was made in regard to the dates of their expiration. Barr simply testified: “ I had my other buildings insured in other companies. I dropped them,- and took all in this company.” As the record stands, the theory of the defendants presents the peculiar case of an insurance to the *36amount of $4800 to take effect twenty-eight days after the date of the application, while nothing is offered to show that the property was not, during that interval, left wholly unprotected.

It is possible that a thorough investigation by a jury may enable them to account for the confusion in these instruments on the ground of accident, and not on the ground of an alteration of dates. In its general terms the receipt of 1870, is almost an exact duplicate of that given by Zeller on the 31st of May 1867. In the application signed by Barr, the date was stated as “May 3d 1870.” It was for the jury to say how and why the change was made from the numeral adj ective expressed in the word and figure “ May 3d ” in the application, to the ordinal number expressed by the word, figures and letters “ May 31st ” in the receipt. While the application and policy could be reformed by proof that the receipt showed the true terms of the contract, certainly the receipt could also be reformed by proof that the true terms were contained in the contemporaneous papers. One instrument would not necessarily control the other. The point was one which it was not the province of the court to decide, for it rested upon extrinsic evidence. It may be that in copying the receipt of 1867, the day of the month on which that had been given was copied also by Hendrickson through a natural inadvertence. This is mere speculation of course, for such a question could only have been passed upon by the jury, and to them, in connection with the general facts developed on the trial, it ought to have been referred. The other questions in the cause were accurately ruled, but there was error in the answer given to the fourth point of the defendants, and the ninth assignment of error therefore is sustained.

Judgments reversed, and venire facias de novo in each case awarded.