The case finds that a burning match being applied, without fault of the plaintiff, to a cask of gunpowder in the attic of his house, the gunpowder took fire, exploded, set fire to a bed and clothing, charred and stained some of the wood-work, and blew off the roof of the house; and the only question in the case is-, whether the loss thus occasioned to the building is covered by the conditions of an ordinary policy against fire. The question may be generalized thus : By the ignition of gunpowder within a dwelling-house, damage is done to the house, that damage consisting in part of combustion, and in part of explosion. Is the whole damage covered by a policy insuring “ against loss or damage by fire ? ”
The very anomalous case of Austin v. Drew, has been adduced in argument and greatly relied upon, as having apparent analogy to this; but when that case is examined, the analogy disappears. The evidence there was, of a building of several stories, in each of which sugar, in a certain state of preparation, was deposited for the purpose of being refined; and a chimney running up through the building, formed almost one whole side of each of the stories ; and by means of this chimney, heat was communicated to the several rooms containing the sugar, and thus acted on it chemically. At the top of the chimney was a register used to shut in the heat during the night. The servant of the assured, in lighting the fires in the morning, neglected to open the register, in consequence of which, undue heat came out into the heating-room, and the sugars were thereby injured. And the action pending, was to recover damage for this under a policy of insurance against loss by fire. As reported in 6 Taunt. 436, the opinion *358of the court is as follows: Gibbs, Chief Justice, says: “ 1 think no loss was sustained by any of the risks in the policy. The loss was occasioned by the extreme mismanagement of their register by the plaintiffs.” And Dallas, J. says : “ The only cause of the damage appears to me to have been the unskilful management of the machinery by the plaintiffs’ own servants, and it is therefore not a loss within the meaning of the policy.” The case is also reported in Marshall; and there the language of the court is somewhat different. There Chief Justice Gibbs says : “ The damage was occasioned by the unskilful management of the machinery, and not by any of those accidents from which the defendants intended to indemnify the plaintiffs.” And Dallas, J. says : “ There was nothing on fire which ought not to have been on fire, and the loss was occasioned by the carelessness of the plaintiffs themselves.” 2 Marsh. 130. The conflicting and imperfect reports of this case have led to various and contradictory misapprehensions of its import. On the one hand it has been supposed that the decision in Austin v. Drew is put on the ground of carelessness of servants, (compare Hughes on Ins. 507-511,) and is thus in apparent contradiction with the decision of Dobson v. Sotheby, Mood. & Malk. 90, in which Lord Tenterden says, that “one of the great objects of insuring is security against the negligence of servants and workmen,”—which doctrine is now, in regard to fire policies, at least, the well settled law both in Great Britain and the United States. 1 Phillips on Ins. Ch. XIII. § II. 1049.
Another authority supposes the point decided to have been, that “ in order to recover upon a policy against loss or damage by fire, it is not sufficient to show that the property has been damaged by the heat of fires usually employed in manufacture, and incurred by the negligence of the insured, or his servants, beyond its usual intensity.” Ellis on Ins. 25. This construction of the case of Austin v. Drew is inexact; for it does not plainly indicate that the real question in controversy was of damage to the subject-matter of manufacture.
On the other hand, the decision in Austin v. Drew has been assumed to establish that “ to bring a loss within the risk in *359sured against, it must appear to have been occasioned by actual ignition, and no damage occasioned by mere heat, however intense, will be within the policy.” 2 Marsh. on Ins. (3d. ed.) 790. This proposition is not the point of the case; and it cannot be sound law; for it may well happen that serious damage, within the scope of a fire policy, shall be done to a building, or to its contents, by the action of fire in scorching paint, cracking pictures, glass, furniture, mantlepieces, and other objects, or heating and thus actually destroying many objects of commerce, and yet all this without actual ignition—that is, visible inflammation.
All these manifest errors, and the doubts they throw over the case of Austin v. Drew, are dispelled at once by the report of it in Holt and in Campbell, as it was tried at Nisi Prius. There it appears that the claim was for damage to the sugars by over-heating only. And Chief Justice Gibbs said: “ I am of opinion that this action is not maintainable. There was no more fire than always exists when the manufacture was going on. Nothing was consumed by fire. The plaintiffs’ loss arose from the negligent management of their machinery. The sugars were chiefly damaged by the heat. And what produced the heat 1 Not any fire against which the company insured, but the fire for heating the pans, which continued all the time to burn without any excess. The servant forgets to open the register by which the smoke ought to have escaped and the heat to have been tempered.” And when one of the jurymen suggested that fires arising from negligence of servants were covered by fire policies, Chief Justice Gibbs assented, and said it was not the case of a fire arising from negligence, for there was no fire except where it ought to have been; but it was the case of the damage of an article in the process of manufacture by the unskilful management of the fire used as an agent of the manufacture. Austin v. Drew, 4 Campb. 360; Holt N. P. 126.
If, in Austin v. Drew, the fire had been where it ought not to be, if, even with careless management, it had burned the building, and notwithstanding it was fire maintained only for the purpose of manufacture, then all the observations of *360the court go to show that, in this instance, as in that oí the whaleship mentioned in Emerigon (1 Tr. de Ass. 436) the insurers would have been held to be liable for the loss. This, therefore, and this only, as correctly stated by Beaumont (Ins. 37,) is decided by the case of Austin v. Drew, namely, that where a chemist, artisan, or manufacturer, employs fire as a chemical agent, or as an instrument of art or fabrication, and the article, which is thus purposely subjected to the action of fire, is damaged in the process by the unskilfulness of the operator, and his mismanagement of heat as an agent or instrument of manufacture, that is not a loss within a fire policy. This we apprehend is good sense and sound law. But it does not touch at all the present case.
It has been thought proper thus to analyze the case of Austin v. Drew, because having been variously reported by four different reporters, and presenting itself prominently in several of the text-books, but in nearly all of them with more or less of misconception, it has become the starting-point, in legal construction, of conflicting lines of argument leading to sundry false conclusions, and among others, that of a supposed application to the present question.
Some adjudications have also been cited of questions arising in the contingency of damage done by lightning. Thus, in Kenniston v. Merrimack Insurance Company, the supreme court of New Hampshire decided that damage by lightning, without any combustion to indicate the presence of fire, is not within the terms of a policy against “ fire by accident, lightning, or by any other means; ” the court, in a brief opinion, deducing the conclusion from the assumed premises that lightning per se is not fire. 14 N. Hamp. 341. The same conclusion, upon similar facts and upon the same words of insurance, “ fire by lightning,” is elaborately reasoned out in a recent case in New York, Babcock v. Montgomery County Insurance Company, 6 Barb. 637; where it is held, that to constitute a loss within the policy, there must be fire, or burning, and that damage by lightning in other forms is not the risk intended by the contract; because, though caloric may generate electricity, or electricity caloric, yet caloric and electricity are distinct things in nature.
*361The principle adjudged in the cases of this class will be readily seen by reversing the question. Suppose, not as fact but as mere supposition, a policy insuring against damage done through electricity generated by caloric. Obviously, this would not cover damage done by fire only, electricity not being evolved. So, in the actual case reported, of insurance against fire produced by lightning, if the effects be of lightning only, without exhibition of fire, it would not, according to the above decision, be within the policy. Or, suppose insurance on cattle against the risk of death by fire alone. In that assumption, if the cattle die, as they may, by a stroke of lightning, without a burn or any other action of fire on their bodies, it would not be the risk contemplated by the contract. Beaumont on Ins. 37.
The question of loss by lightning is' very summarily disposed of in the older authorities, by treating electricity as fire from heaven. See 1 Emerigon, c. 12, § 17, no. 1, and the authors there cited. But the progress of knowledge has led to juster notions of the nature of lightning, and of course to different conclusions touching its legal relations; which are correctly summed up by a late writer as follows, namely, that fire includes lightning if there be any mark of fire; but not otherwise. Beaumont on Ins. 37.
These cases of damage by lightning bear on the present question therefore, if at all, only by very distant analogy. Neither of them covers it, or has any direct relation to it. To the contrary of this, in New York, at least, the same courts, which decide that loss by lightning merely is not’covered by a fire policy, decide that loss by the explosion of gunpowder is. There is a series of cases precisely in point, which expressly decide, or by implication assume, that damage done by the explosion of gunpowder ignited within a building, as well as that done by its combustion, is within the risk of a fire policy. The case of Grim v. Phœnix Insurance Company was this : A vessel, insured against fire, was partly laden with gunpowder, which being ignited by carelessness, the vessel was blown up and totally lost. It was argued by eminent counsel and the opinion was given by Thompson, C. J.; and through-*362out the cause it seems to be assumed that the loss was, in respect to its cause, within the policy, and the decision was made to depend on other considerations. 13 Johns. 451. The same conclusion is also assumed in the case of Duncan v. Sun Fire Insurance Company, 6 Wend. 488. In the case of City Fire Insurance Company v. Corlies, the claim was on a fire policy for merchandise destroyed, not in burning, but through the blowing up of the building wherein it was stored, by means of gunpowder; and the court expressly adjudged this to be “ a loss by the peril insured against, within the meaning of the policy.” 21 Wend. 367. The same point has been ruled incidentally by the supreme court of the United States. Waters v. Merchants’ Louisville Insurance Company, 11 Peters, 225. Perhaps it may add a little to the weight of these authorities to say, that the same thing as to loss by gunpowder—sulphureo pulvere accenso—seems to have been holden by the older commercial jurists in Europe. Straccha de Assec. gl. 18, § 2.
This court, to be sure, is not bound by the decisions or opinions cited; but they are entitled to great consideration; and there is not, so far as we know, any contrary adjudication or opinion. Uniformity of decision is in itself a desirable thing. The question, we admit, is a nice one. Upon careful reflection, however, we have come to the conclusion that the received opinions on the subject, and the adjudications referred to, are in accordance with reason and principle. It seems not to be denied that actual combustion, produced by the ignition of gunpowder, is within the present policy. If, then, a combustible substance, in the process of combustion, produces explosion also, it is not easy to perceive why, of the two diverse but concurrent results of the combustion, the one should be ascribed to fire any less than the other. The plain fact here is, the application of fire to a substance susceptible of ignition, the consequent ignition of that substance, and immediate damage to the premises thereby It is no sufficient answer to say that some of the phenomena produced are in the form of explosion. All the effects, whatever hey may be in form, are the natural results of the eombus*363fcion of a combustible substance; and as the combustion is the action of fire, this must be held to be the proximate and legal cause of all the damage done to the premises of the plaintiff.
Our opinion excludes, of course, all damage by mere explosions, not involving ignition and combustion of the agent of explosion, such as the case of steam, or any other substance acting by expansion without combustion. See Perrin's Administrator v. Protection Insurance Co. 11 Ohio, 146. It likewise excludes all damage occasioned but remotely or consequentially through the agency of gunpowder, such as injury done to a house by falling fragments in the blasting of rocks, or the shattering of a house by the stroke of a cannon-ball, in which examples the shock of a projectile, and not ignition or combustion, is the proximate cause of the damage done. We recognize and accept, in the full force of its application, the maxim: In jure non remota causa sed próxima spectatur. Bacon’s Max. 1.
The legal relations of marine insurance have been copiously discussed in many express treaties of elaborate erudition, and are considered in a great number of judicial decisions, in which the whole subject has been explored with wonderful acuteness and comprehension of logic and of learning; while fire insurance, as a branch of legal knowledge, is, comparatively speaking, in its rudiments. The cases on marine insurance throw little if any light on the present question, except in so far as they attempt to prescribe a rule for distinguishing between what is remote and what is proximate cause. The conclusion reached in this discussion, as may be seen by the latest investigation of the point in Great Britain, Montoya v. London Assurance Co. 6 Welsb. Hurlst. & Gord. 451, is that, while for most cases it is practicable to draw the line, and to formalize a rule between the two classes of causes, yet in other cases, according to the general law of nature, the two classes approach and run into one another until the distinction vanishes; and within the limits of this debatable land of differences, it is necessary to apply judicial discretion to the particular questions as they arise, just as it is in the not infrequent inquiry *364whether a thing, or the use or measure of it, be reasonable or not. In Montoya v. London Assurance Co. it was determined that where the lower part of a cargo is damaged by sea-water, and,by the evolution of gases from the part thus damaged, or the propagation of heat arising from fermentation, the superior part of the cargo be damaged also, the loss on the latter is by the perils of the sea, the involvement of the secondary effect in the primary one being an example of ccmsa próxima.
In the present case there is no room for question concerning a series of causes, as whether primary or secondary, proximate or remote ; for the agent is one and the same throughout, namely, fire. The causa was burning powder; the ccmsa causans was a burning match; at each stage of causation it was the action of fire. Nay, to be exact, the burning of the gunpowder, like the burning of the match, was a succession of several complex acts of burning. Yet fire is the agent at each of these distinct stages of causation. Suppose there was a barrel of sulphur in the plaintiff’s attic, instead of gunpowder; and this being ignited with a match, afterwards the fire had passed from the burning sulphur to the substance of the house. This would be recognized at once as a case of fire. It does not change the legal relation of causes to substitute a barrel of burning gunpowder for a barrel of burning sulphur. The only difference in the elements of the question is, that the gunpowder, when ignited, consumes with more of rapidity than sulphur, and the combustion is accompanied or followed by explosion. Still, the agent is fire, though it acts in different ways upon the different successive subjects of its action, beginning with the match and terminating with the plaintiff’s house.
On the other hand, cases are conceivable, other than by the use of gunpowder, of explosion without any combustion, which, nevertheless, being the result of the action of fire, are still, it would seem, within the range of the general principle. Various mineral substances exist, of value in commerce and the arts, which explode by the action of the fire, without either ignition or combustion. In general, any close vessel, of whatever material composed, when filled with an expansive fluid, *365is liable to explode by the action of heat, though it may be that the vessel and its contents are alike incombustible. The same thing happens, under certain conditions, to some forms of wood; which, although combustible, may by the action of fire explode without ignition; or which, as in the present case, of a house, by having compressed within it some burning substance, which is explosive as well as combustible, like gunpowder, may suffer the double injury of combustion in part and in part of explosion.
If, however, the question of consequential damage needed to be explored for the determination of the present case, it would serve to confirm the conclusion to which we have on other premises arrived. Thus, in Great Britain, damage, which occurs consequentially in the case of a fire, by reason of confusion of mind, as in throwing fragile objects out of the window, or by sudden terror from alarm, as in leaving open the tap of a barrel, and thus wasting the contents, is held to be loss by fire, according to the usages of insurance offices or established legal principle. Beaumont on Ins. 41. So it is in the case of a beam, cornice or coving, removed to prevent the spread of conflagration. Ibid. We understand the same to be the rule, in the case, for instance, of a fire in the upper story of a building, and the destruction or damage of goods in a lower story, not by fire, but by the water thrown into or upon the building, for the purpose of extinguishing the fire. All these are fit illustrations of the question of merely consequential damage. Its legal relations may likewise be followed in the familiar case of the squib, falling on a party’s premises, and by him hastily thrown off, and so falling upon the premises of another, and thus giving rise to the inquiry, whether the first throwing or the second throwing should be taken as the responsible cause. Scott v. Shepherd, 2 W. Black. (2d ed.) 892 and notes; 3 Wils. 403.
In the hypothesis that fire is to be regarded as causa próxima -in the present case we can see but one supposable defect, namely, the suggestion that, though it be conceded that the explosion of burning gunpowder, and its effects, are the action *366of fire, yet this particular effect on the building is not exhibited in the form of igneous action. The cases above supposed, of the shrivelling of some masterpiece of pictorial art, the cracking or discoloration of a rich vase or gem, the bursting of a cask of wine through the expansion of its contents, these, it may be said, are distinctly cases of damage, without ignition it is true, but by the direct and specific action of heat as such; while it is denied that such is the fact in the present case of the blowing up of a dwelling-house by the ignition of gunpowder. We do not think the premises of this argument are sustained by the physical facts which occurred. If they were so, then the nearest analogy would be of damage by smoke, that is, the moisture thrown off by burning wood, and carrying with it ashes, empyreumatic oil, and other constituent parts of the wood, either in their natural condition, or transformed by the process of combustion. Now it is obvious that mere smoke, without any direct action of heat, may do great damage to many kinds of merchandise, such as delicate textile fabrics, esculent vegetables, articles of taste, and other numerous objects; and if a dwelling or a magazine taire fire, and some parts of it only be consumed, but the contents of apartments, to which the actual fire does not extend, are nevertheless damaged by the smoke penetrating into and filling them, can it be doubted that the damage thus done is a loss within the ordinary conditions of a fire policy ? Semble, per Gibbs, Chief Justice, arguendo, in Austin v. Drew, Holt N. P. 127. Yet, incontestably, damage by smoke is an effect, which is not in itself igneous action, though it be the result thereof; while, as we conceive, the explosion of gunpowder is igneous action.
In conclusion, we think the rule, which we propose for the present case, reconciles all the conditions involved in the question; is conformable to the nature of things; and constitutes a coherent and consistent doctrine, namely, that where the effects produced are the immediate results of the action of a burning substance in contact with a building, it is immaterial whether these results manifest themselves in the form of *367combustion, or of explosion, or of both combined. In either case, the damage occurring is by the action of fire, and covered by the ordinary terms of a policy against loss by fire.
Judgment for the plaintiff.