FINCH, J.

The question whether the policy, as modified, by the insertion of the words "a sea," covered the shipment of cattle upon the steamer Greece, or took effect only upon later shipments, was one of fact, depending upon the inquiry when the modification was made, and whether before or after the risk by the shipment in controversy was assumed. That shipment was made on the 14th of February, 1878, and the steamer sailed with its cargo of cattle uninsured. An open policy with its blanks unfilled was at that date in the possession of the shippers, but could only become operative by a special indorsement describing the particular risk assumed. The day after the steamer sailed the broker of the shipper called upon the agent of the underwriters, with this open policy in his principal's possession, and in his own a memorandum denominated a "live cattle clause," with a view to procure an amendment of the policy. A live cattle clause had been in the policy during the business of a previous year, but had proved unsatisfactory to the shippers, and the broker's object was to procure its amendment before further insurance was taken. Apparently the parties came to an agreement as to the amendment to be made, for at the close of the conversation, one of the insurer's agents who had listened to it, modified the memorandum by inserting the words "a sea" among the causes of loss by which a liability was incurred, and dating the amendment February 15th. The day following this agreement to modify the terms of the open policy an application was made and accepted to insure the cargo of the Greece. How soon the words of modification were actually written in the open policy is somewhat uncertain, but upon the facts, the question was submitted to the jury whether the cargo of the ship was insured under the policy as it stood before the alteration, or as it was modified, and with their verdict in favor of the latter conclusion we must be content.

The proposition that the added words applied only to shipments on deck, and did not cover the risk in controversy, because the cattle were transported between decks, is answered by the finding of the jury that the cargo was insured under the modified policy, both parties well knowing that the cattle were carried under deck. And it encounters the further difficulty of attempting to contradict and modify the written contract by parol evidence. We have so recently considered this subject, and pointed out the difference between proof which alters the terms of the contract itself, and that which leaves it to stand unchanged, but establishes an additional and collateral agreement that we need not prolong the discussion. ( Chapin v. Dobson, 78 N.Y. 74; Eighmie v. Taylor, 98 id. 288.) The contract, as put in writing, covered by its terms all shipments accepted as risks, without regard to the place upon the vessel in which the cattle were carried. It purported to insure the safe transportation of one hundred and eighty-eight cattle upon the steamer, and that shipment was covered by the risk according to the terms of the contract. The parol proof seeks to establish the exact contrary, and to show that the shipment specified in the indorsement, was not covered by the risk because the cattle were placed between decks. The proof tended to modify the contract itself, to limit and restrain the legitimate scope of its meaning, and not to prove a new and collateral agreement consistent with its terms.

Nor can we dismiss this appeal upon the ground that the new trial awarded by the General Term may have been granted upon questions of fact, for, although a motion for a new trial was made at the Circuit and denied, no order was, in fact, entered from which an appeal could be taken, and the General Term held accordingly that the facts were not before them for review.

We are obliged, therefore, to meet the two final questions upon which the judgment against the insurer was reversed; and these are the true construction of the words, "by a sea," as used in the contract, and the inquiry whether the policy was or was not a valued policy. The words to be construed are susceptible of two meanings, the one general, and the other restricted and particular. "A sea" may mean a general disturbance of the surface of the water occasioned by a storm, and breaking it up into the roll and lift of waves following or menacing each other. When a captain reports that on a particular day he encountered a heavy sea, he uses a natural and appropriate expression, which we are not liable to misunderstand. If he says that a gale came from a particular direction, and raised a sea which delayed his progress, he properly describes the general disturbance of the water consequent upon a storm. On the other hand, if he reports that in a gale a sea carried away his boats, and another swept a seaman overboard, we understand him in each instance to refer to some particular wave or surge, separate from its fellows, which worked its own peculiar and special destruction. The latter is substantially the meaning of the phrase adopted by the General Term, while the former was the construction of the trial court. In his charge to the jury, the learned judge assumed such first construction as the true meaning of the phrase, and submitted the question whether the animals died from bad ventilation and foul air consequent upon the enforced closing of the hatches, or from the heavy sea which threw them down and inflicted mortal injuries. The verdict of the jury established the latter as the cause of the loss under the construction adopted by the court, but if that was wrong, the verdict may have been a consequence of the error.

Since the phrase inserted was thus ambiguous, we are at liberty to resort to the surrounding circumstances in aid of its interpretation. Before the words "a sea" were added the policy insured only against a loss occasioned "directly by stranding, sinking, burning or collision." These causes of loss evidently contemplate injuries to the ship which occasion as proximate consequences the death of the animals carried. If the ship strands, or sinks, is burned or collides with another vessel or some fixed obstruction, that has happened from which, if the animals die, the loss falls upon the insurer. When among these risks of the vessel another risk is placed, it must naturally mean also, a cause of loss from something which again has happened to the vessel. As it respects cattle carried between decks, where single waves could not reach them, it is clear that nobody contemplated, as a peril to be insured against, the direct impact of such a wave upon the animals, killing them by force of a blow which it could not directly strike, or by submersion in the water which would rapidly run from the decks. The risk contemplated was some effect of "a sea" upon the ship which among its proximate results should cause the death of the cattle. Both parties knew, when this particular risk was accepted, that the cattle were shipped between decks, and the modification effected by the new words inserted, must have contemplated the possible killing of the cattle by the action of "a sea" upon the ship. The jury found that such outward force applied to the vessel directly occasioned the death of the animals, and so the loss was covered by the risk, unless we are further to say that the death of the animals must be traced to the effect upon the vessel of one or more specific and particular waves as distinguished from the general commotion of the water. The ship rocked and pitched in the storm. That could be anticipated and the risk of resultant injury insured against. But the effect of any one wave upon the violent motion of the vessel could by no possibility be separated from the concurring agency of others, nor could any two or any fifty. Some one or more might be remarkable for size or force, but no man could measure their share in producing the motions of the vessel. It is scarcely reasonable, therefore, to believe that the words "by a sea" meant the blow of one or more particular waves upon the vessel, when loss from a cause so identified was almost impossible, where the cattle were carried between decks and killed by the tossing of the ship. The General Term do not limit the words "by a sea" to a single wave "specifically," but concede that it might mean more than one. If so, how many more, and how are they to be identified as causing that particular rolling of the ship which killed specific animals? A reference to the conversation which culminated in the modification of the contract, does not make the construction of specific and particular waves any easier to adopt, for that respected cattle shipped on deck. As to them, a surge which washed them overboard, or dashed them against each other or fixed parts of the vessel, might be identified, for it would be one which would come upon the ship and sweep its decks; but as to cattle under deck, no such injury could be reasonably contemplated, and that which would enter the minds of the contracting parties would be the tossing of the vessel in a heavy sea, by which the cattle might be thrown from their feet and dashed against the partitions or each other and so killed by "a sea" powerful enough to produce the result. In that conversation the plaintiffs' broker put several supposed cases to the underwriter relative to on-deck shipments, all of which the latter conceded should be covered by the risk. The first was a loss occasioned by the cattle being washed or pushed overboard by the sea, and the second by a sea boarding the vessel and killing the animals on deck. These cases contemplate an injury inflicted by one or more waves producing death differently. In the one the force of the surge sweeps the animal into the water and occasions death by drowning; in the other the same force throws the animal from its feet and dashes it against the deck or fixed obstacles, for the force of the wave alone, in and of itself, would not produce death. But a third case was put and received an equal assent. The broker suggested "if the animal should die by a collision by the sea." That did not mean a collision between a wave and the animals, because that case had just been put and answered. It could only mean a collision of the animal with some fixture or portion of the vessel occasioned by the action of "the sea" upon the ship, instead of a single wave coming on board. Unless it meant that, it was not a third case, but an idle and useless repetition of the second. So that, if we are at liberty to consider this conversation as a means of ascertaining what the parties really meant and intended by the modifying words afterward adopted, we must again believe that a loss occasioned by the motion of the ship struggling with a dangerous sea, and dashing the animals against partitions or each other, was within the intention and contemplation of the parties. The use of the word "directly" does not alter the interpretation. Between the shock of the waves upon the vessel and the death of the animals there was no intervening cause upon the theory involved in the finding of the jury. If a railroad train dashes against an approaching engine and a passenger is killed by being thrown against a seat or side of the car his death is the direct result of the collision, as completely as if the colliding engine had struck him. If in the present case, the storm had compelled the closing of the hatches, and the animals had died from foul air and insufficient ventilation, the loss would have been an indirect result of the heavy sea. But that theory is excluded by the verdict of the jury.

It was further claimed, and the General Term held, that this was a valued policy, establishing the value at the sum insured. In that also we think, they were in error. If this was a valued policy it became so by containing in its terms an agreement of insurer and insured, fixing absolutely for the purposes of the contract the value of the subject insured at the amount insured, If the policy expressed such agreement, and made the value of the cattle identical with the sum insured then since some of them were saved and landed in specie at the port of destination, and realized a value, it would follow, as the underwriters claim, that such salvage should be deducted from the insurance, and the balance only be recovered. The application of the shippers was for a policy valuing the cattle at the sum insured, but the question is what was granted rather than what was asked. The underwriters might still have given an open policy and the insured accept it, and we are to decide upon what was thus given and accepted, possibly taking into account the form of the application if there be ambiguity or doubt. We should naturally look for an agreed valuation in definite language and in the indorsement, since it is that which gives vitality to the policy and is the special repository of the intention of the parties. Usually in a valued policy after the statement of the subject insured the added phrase appears, "valued at ;" and the blank being filled, the agreed value is settled. (1 Pars. on Mar. Ins. 256; Hemmingway v. Eaton, 13 Mass. 108.) In this policy the technical phrase appears, but the blank is left unfilled, and where it appears it evidently contemplates a valuation in the indorsement if none is made in the body of the policy. The clause devoted to that subject provides "the said goods and merchandises hereby insured are valued at as indorsed." This means that the property is valued at the sum stated as its value in the indorsement, and not at the sum stated as amount of insurance which also appears in the indorsement. The policy remains open if its blank is unfilled and no valuation of the subject insured is specified in the indorsement. In the present case the blank valuation in the body of the policy was left a blank, and no valuation of the cattle was named in the indorsement, but merely the sum insured. That sum, however, it is claimed became the valuation by force of another clause of the policy which reads: "indorsements valued at the same, provided they do not vary from the cost more than per cent." Where this phrase occurs in the policy, it follows one which provides: "no shipment to be considered as insured until approved and indorsed on this policy by the assurer," and precedes the one already referred to, which agrees upon a valuation. If the phrase "valued at the same" means at the sum insured then this form of policy is never an open but always a valued policy; and then, too, the subsequent formal clause "the said goods" * * "are valued at" is superfluous and unmeaning, since already it had been said that the valuation was to be the sum insured. We think the clause relied on has an entirely different purpose and meaning, the force of which lies in the proviso. Its evident aim is to guard against an over-valuation, when it is made effective by filling up its blanks. It means that the property insured is valued "the same" as that value is stated in the indorsement, provided such value does not vary from the cost more than per cent. If a maximum of variation from cost, permissible in the valuation, is agreed upon, that furnishes a protection against an excess which might be in the nature of a wager, and largely exceed the real value. The clause, therefore, does not mean that the sum insured is to be the agreed value, and so in all cases the policy shall be a valued one; but contemplates an agreed valuation which may be stated in the indorsement, and which shall be the agreed value "of the goods, provided they do not vary more than per cent from cost."

We do not concur for the reasons stated with either of the conclusions which led the General Term to a reversal of the judgment, but the respondent further seeks to justify the reversal upon the ground of the refusal of the trial judge to charge "that the plaintiffs having shown only a partial loss, and not having given any notice of abandonment, cannot recover for a total loss." This request assumed that only a partial loss was shown, and as matter of law upon the facts. But there was evidence from which the jury might have found that the cattle surviving were in a weak and dying condition, so that a sale was necessary, and so that there was an absolute total loss rather than a constructive one. (2 Phil. on Mar. Ins., § 1487; McCall v. Sun Mut. Ins. Co., 66 N.Y. 505, 515.)

The order of the General Term should be reversed with costs, and judgment on verdict affirmed.

All concur.

Order reversed, and judgment affirmed.