I am constrained' to dissent from the . opinion of the court. In my judgment, the policy in this case never attached, and the plaintiff is consequently entitled to judgment for the premium paid by him*. My brethren suppose that the policy attached, retroactively, on the goods laden on board, before the first of December; the insurance being at as well as from Bristol. The vessel is u •warranted to have sailed from the port of Bristol, betxveen the 20th of October and the 1st of December, 180S. The contract between the parties was entered into after the day when the ship was warranted to have sailed, to wit, the 21st of December, 1808, and the ship sailed from Bristol after the 1st of December, but before the policy was subscribed. It becomes necessary to consider the nature and effect of warranties in policies, in order to show that this policy never attached.
Marshall on Insurance, 248. (p. 346. 2d edit, book I. c. 9. s. 1.,) gives an analysis of warranties which appears to me to be able and perspicuous. He says, “ a warranty is a stipulation or agreement on the part of the insured, in nature of a condition precedent. It may be either affirmative, as where the insured undertakes for the truth of some positive allegation; as that the thing insured is neutral property, that the ship is of such a force, that she sailed or was well on such a day, &c.; or it may be promissory, as where the insured undertakes to perform some executory stipulation, as that a ship shall sail on or before a given day, that she shall depart with convoy, that she shall be manned with such a complement of men,” &c.
Again, he observes, “ The breach of a warranty consists either in the falsehood of an affirmative, or ihe nonperformance of an executory stipulation;” and he adds that “ in either case, the contract is void, ah initio, the warranty being a condition precedent. Whether the thing was material or not, whether the breach of it proceeded from fraud, negligence, misinformation, or any other *14cause, the consequence is the same. The warranty makes the contract hypothetical, that is, it shall be binding, if complied with. With respect to the compliance with'the warranties, there is no latitude, no equity ; the only question is, has the thing warranted taken place or not. If it has not, the insurer is not answerable for any loss, even though it did not happen in consequence of the breach of warranty.” The same principles are laid down by Lord Mansfield, in Hibbert v. Pigou, Marsh. 272. (p. 369. 2d edit.) and Park, 339. (p. 443. 6th edit.) Again, in the case of De Hahn v. Hartley, (1 Term Rep. 343.) Lord Mansfield, after stating the effect of a representation, proceeds, “ But a warranty must be strictly complied with. A warranty in a policy of insurance is a condition or a contingency, and unless that is performed there is no contract. It is perfectly immaterial for what purpose a warranty is introduced, but being inserted, the contract does not exist, until it is literally complied with.”
In the case of Henckle v. The Royal Exchange Assurance Company, (1 Vez. 318.) the warranty was that the ship insured was an Ostend ship, and it was held by Lord Hardivicke, that the fact being untrue, the ship was never brought within the terms of the insurance.
It appears, then, from the cases cited, that where the warranty is of a thing-past or present, it is an affirmative stipulation; and the contract between the parties being hypothetical, it operates as a condition precedent; and unless the fact warranted be true, there is no contract between them. If the warranty be executory," as that the ship shall sail on a given day, and the policy be at and from, then, inasmuch as the policy immediately attaches, and the risk commences and endures until the warranty be broken by a failure of performance of the fact stipulated; and inasmuch as there can be no apportionment of the premium, on a single risk, the insured would not be entitled to a return of any part of the pre? tpium, because the executory event did not take place.
*15I must not, for a moment, be understood as countehancing the idea, that an event already past may not be the subject of insurance. Nothing is more common, and nothing can be fairer; because, although the contingency may have happened, in point of fact, yet the parties being ignorant of the event, it is, as to them, contingent, and the assurer, in such a case, takes upon him the risk, on the hypothesis, that the subject insured may be lost. In the present case, had any of the goods insured been lost in port, by any of the perils in the policy, before the 1st of December, and had the ship sailed pursuant to the warranty, the defendants would have been on the policy, and answerable for the loss; but the fact of the ship’s sailing before the 1st of December, being a fact warranted by the insured, and being untrue at the time the contract was entered into, and the whole contract depending on the verity of the fact of the ship’s having sailed before the first of December, the subject matter of the insurance was never brought within the terms of the insurance, and consequently there never was a contract between the parties. There is no difference between a warranty that a ship sailed on a day past, which was anterior to the policy, and a warranty that a ship has a particular quality, as that she is an Ostend ship. They are both affirmative warranties, operating as conditions precedent; and it appears to me to be violating every rule, in relation to conditions precedent, to maintain that though they are not true, and are never performed, yet that the party who warranted them to be true, and stipulated that they had been performed, as the vei-y basis of the contract on the other side, is dispensed from their performance.
The case of Bond v. Nutt (Cowp. 610.) has been relied on by the counsel on both sides, as favourable to their respective clients. The insurance was upon a ship, lost or not lost, at and from Jamaica to London, warranted to have sailed before the 1st of August, 1776. The policy was effected on the 20th of August. The *16point was, whether the voyage was to be considered as j3egUn from st. Ann’s bay, from which the ship sailed the 26th of July, or from Bluefields, from which she did not sail till after the 1st of August. The ship was iost) and the action was to recover for the loss of the ship. The question of a return premium never arose, as the defendants paid the whole premium into court, which Was taken out by the plaintiff before the trial. (Doug. 785. note 1.) Lord Mansfield, in the course of his opinion, asks, “ had she or had she not sailed on or before that day ? That is the question; no matter what cause prevented her, if the fact is that she had not sailed, though she staid behind for the best reasons, the policy was void, the contingency had not happened, and the party interested had a right to say there was no contract between them.”
It is impossible to find language more emphatic than that employed by his lordship; and the cases are precisely analogous in principle. It is true the insured in that case recovered on the ground that the ship had commenced her voyage before the 1st of August. Had it been deferred until after that day, then the observations made by Lord Mansfield would have been decisive.
Lord Mansfield may be quoted in contradiction to himself, from what fell from him in Tyrie v. Fletcher. (Cowp. 669, 670.) The point was, whether the risk was entire or divisible; if the latter, then the insured claimed a return of part of the premium. He observes, u a case of general practice was put by Mr. Dunning, where the words of the policy are at andfrom, provided the ship shall sail on or before the 1st of August, and Mr. Wallace supposes that the whole policy would depend upon the ship’s sailing before the stated day. I do not think so; on the contrary I think with Mr. Dunning, that cannot, be. A loss in port before the day appointed for the ship’s departure can never be coupled with a contingency after the dayand he proceeds to give the inclination of hie. *17mind, that there would be two parts or contracts of insu1 ranee, the first on the ship in port, if lost before the first of August, and if not lost in port, then on the voyage, What fell from his lordship on that occasion, is correctly considered by Mr. Park, 390. (6th edition, p. 528.) as an obwcr dictum, ove rruled by himself in Meyer v. Gregson, (Park, 389.) But he was not considering the case of a warranty to sail on or before a given day. Marshall, (2d edition, pp. 667. 658.) in the notes, observes of this case that the word “ provided, made the sailing a condition on which the contract was to take effect, and one part of the case put as repugnant to the Other, and that the word “ at” in the policy was nugatory? and he agrees with Mr. Park, that the opinion was extrajudicial, hastily delivered, or, perhaps, not accurately reported, and that it was overruled by the case of Meyer v. Gregson. If his lordship meant to speak of a warranty, this dictum of his is overthrown by the subsequent case of Hore v. Whitmore, Cowp. 784.) in the decision of which the court, including his lordship, was unanimous. In that case the ship was warranted to sail on or before the 26th of July, 1776, before which day she was restrained from sailing by the governor of Jamaica, and detained beyond the day? the ship was insured free from all restraints and detainments of kings, princes and people, &c. and she would have sailed but for the restraint ; and it was adjudged that the warranty being express, that she should depart before a certain day, must be complied with, though the cause of the delay was one of the risks insured against. This case furnishes a pretty full answer to the idea that there were two distinct risks in this case, the one in port, and the other on the voyage. My doctrine is, that there was but one risk in port and on the voyage, depending entirely on the fact that the ship had sailed according to the warranty. I am again met with another dictum of Lord Mansfield’s, in the case of Bremon v. Woodbridge, (Doug. 751.) in which he says, “ In Bond v. Nutt it was held there were two risks? *18amaica, was one; the other, viz. the risk from Jaa, depended upon the contingency of the ship having sajjeci on or before the first of August; that was a condition precedent to the insurance on the voyage from Jamaica to London.” I do not intend any thing disrespectful to that great man, when I say, that neither his lordship, nor any of the judges, held any such doctrine in Bond v. Nutt. On the contrary, he emphatically said, “ that if-the ship did not sail before the first of August, the policy was void, the contingency had not happened, and the-party interested had a right to say there was no contract between them.” In the case of De Hahn v. Hartley, already cited, he expressed himself to the same effect as in Cowper's report of Bond v. Nut.
The risk in port is either divisible or entire. If entire, and the policy attached so as to cover the goods laden on board before the first of December, and while the ship was in port, I cannot conceive why the risk did not endure as well in port as on the voyage. But my brethren seem to think these two risks, according to Lord Mansfield's dictum in Tyrie v. Fletcher, the one in port, absolutely, and the other on the voyage, contingently, if the ship sailed by the appointed day; and holding that opinion, how can it be, that the plaintiff is not to have a return of part of the premium; for it is perfectly well settled that where the voyage is divisible into distinct risks, the premium is to be apportioned according to the several risks. (Marsh. 655. 3 Burr. 1237. 1 Bos. & Pull. 172.) The cases of Meyer v. Gregson, and Long v. Allen, (Marsh. 658. 660.) which are supposed to apply to this case, are very different. In those cases the risks were single, and the warranties were executory, or warranties to sail by a given day, and not warranties that the ships had sailed. The underwriters were on the policies for the voyages, and a risk was incurred between the time of subscribing the policies and the time when the future events were stipulated *19to take place. In the present case, the warranty was that r r , - an event had already taken place; and on Jhe truth ot that warranty the whole policy depended.
It cannot, I think, be seriously contended, that where no risk has been run, and no fraud practised, the assurer has a right to pocket the premium. “ The premium, says Marshall, 548. (2d edit. 638.) paid by the insured, and the risk which the insurer takes upon himself, are considerations each for the other ; they are correlatives, whose mutual operation constitutes the essence of the contract of insurance. The insurer shall not be exposed to the risk, without receiving the premium; nor shall he retain the premium, which was the price of the risk, if, in fact, he runs no risk at all. For wherever a man receives the money of another upon a consideration which happens to fail, or is never performed, he is under an obligation, from the ties of moral honesty and natural justice, to refund it;” and he cites a variety of cases in support of this opinion. To these cases may be added the opinion of Lord Hardwicke, in Henckle v. Royal Exchange Assurance Company, that, if the ship was never brought within the terms of the insurance, so that the insurer never runs any risk, the premium must be returned in an action by the assurer.
I am not aware of any adjudged case controverting the right of the insured to a return of premium, where there has been no risk, and where there exists no fraud. It would require something very authoritative to induce me to deny a recovery in case of such strong equity. I shall only add, that although this appears to the court to be a plain case for the defendants, to my understanding, the law warrants the plaintiff’s recovery.
Yates, J. was of the same opinion.
Judgment for the defendants..