The injury to the plaintiff in this case was caused by the slippery condition of the foot-way upon a bridge. This condition resulted from rain falling and freezing, thus making a thin coating of ice upon the surface of the walk. It is claimed at the outset that towns were not liable for such a condition under Public Statutes, chapter 76, section 1; and excerpts from the opinion in Hubbard v. Concord, 35 N. H 52, are relied upon to support the position taken. But the summing up of the reasoning in that case was that the jury should have been instructed that the town was not liable for such a condition “unless under the circumstances they ought to have repaired the defect before the accident happened, and had reasonable opportunity to do so.” Ib., 72.
It is a fact well known to those in practice before 1893 that suits against municipalities to recover for injuries caused by slipping upon icy sidewalks were then very numerous. After the decision in Hubbard v. Concord, supra, the question of liability for such a condition was not again raised for many years, but the assertion in the defendant’s brief that there is no other reported case upon the subject is evidently an inadvertence. The precise question was presented, and decided in conformity with the earlier case, in 1892. Chadbourne v. Exeter, 67 N. H. 190. The brevity of the opinion indicates that the question was not then considered to be a doubtful one.
*246Under the statute as it stood prior to 1893 the town might have been liable in this case, and the question presented is whether the changes then made in the statute so far limit the liability as to exclude this deficiency in the way. If Laws 1893, chapter 59, section 1, had merely limited the existing liability to injuries happening upon bridges etc., the plaintiff would prevail here, because of the construction which had theretofore been given to the language of the statute. But the legislation goes beyond that. It not only limits the parts of a highway for defects etc., as to the way over which towns are liable to travelers; it also limits liability to accidents caused by defects in the structure itself. An obstruction which is not a part of the bridge is not one for which the town is now liable. Under the old statute the question was whether there was a suitable way for the travel thereon. Under the new one the way must not only be shown to be unsuitable, but the lack of suitability must consist of an obstruction, defect, insufficiency or want of repair of the structure in question. If a bridge is in itself sufficient, but is obstructed by a log left across the way, the town might have been liable under the old law, but it would not be under the act of 1893. Wilder v. Concord, 72 N. H. 259.
While the precise question now presented was not involved in the case last cited, yet the reasoning upon which that decision is founded necessarily leads to the conclusion that there is no liability here. It is there said that the intent was to greatly limit, if not to wholly abolish, liability for faults in highways caused by outside agencies, — that the fault must be in the structure or there is no liability. Ib., 262. The point of what is there said is not that the identity of the agency which created the fault is the test, but that faults created by outside agencies and not altering the structure maintained by the town do not become parts of such structure within the meaning of the statute. A pile of refuse (Parsons v. Manchester, 67 N. H. 163) or of logs (Manchester v. Warren, 67 N. H. 482) left upon a bridge by a third person would not constitute an obstruction “of such bridge” under the law of 1893. But a hole through the bridge planking, although made by a third person, would clearly render the bridge defective within the purview of the present law. “ The fact that the depression arose from the wear of passing travel, instead of from faulty construction, is immaterial'. The statute makes no distinction of this kind.” Wilson v. Barnstead, 74 N. H. 78, 81.
The fact as to who places an obstruction may be of importance, if the question of intent to make it a part of the structure is involved; *247but aside from that, and in a case where an obstruction clearly is not a part of the bridge, there is no liability under Laws 1893, chapter 59, section 1, in any event. In such a situation it is wholly immaterial who caused the obstruction. And so if the structure be made defective it is immaterial (except upon the question of notice and opportunity to repair) who caused the defect. The test is whether a fault in the structure exists, and not who created the fault.
The derrick placed beside the culvert by the highway surveyor, in Iiardy v. Keene, 52 N. H. 370, was a fault in the way, created by a town official who intended to place it there. But it did not become a defect in the culvert, because it was not made a part of that structure. Wilder v. Concord, supra, 262.
The suggestion in the latter case that the intent of the legislature of 1893 was to entirely abolish the liability formerly imposed on account of faults not involving the structure, states the true construction of the statute. There is now no liability under the statute of highways (in the absence of the statutory notice of the fault, Laws 1893, c. 59, s. 2) unless the fault is in, rather than upon or near to, the highway construction maintained by the town.
The repeal in 1893 of Public Statutes, chapter 76, section 2, tends to strengthen this conclusion. That section provided that: “They are liable for damages happening from the snow encumbering the same, as from any other defect.” If there had been any intent in 1893 to retain a liability for encumbrances upon a bridge this section would have been included in the new act. Its repeal is convincing evidence that the intent was to exclude liability for that class of faults in the way.
The provision that the fault must be in the structure has been construed to limit liability to those things which are in fact of the class specified. The road surface upon a bridge is within the act, because it is manifestly a part of the structure. Wilson v. Barnstead, supra. But mere contiguity does not create unity. For this reason it was held in Wilder v. Concord, supra, that the road surface above a house drain was not to be treated as a part of the latter within the meaning of the statute. It will be noted that the connection between the two objects there referred to would be permanent, while in the present case it was only temporary. In that case both objects were parts of an entire structure maintained in its finished state as a way, while here the ice was something upon the way, not put there by the town and not designed to be a part of the structure.
Whether, in the case of an ordinary surface water culvert, defects *248in the roadway only come within the class of Wilson v. Barnstead, supra, or that of Wilder v. Concord, supra, was left undecided in Hickey v. Berlin, 78 N. H. 69.
For the reasons above stated, the suggestion at the argument that an obstruction caused by a third party would make the way insufficient, is unimportant. A town is liable only where “the matter complained of constitutes an obstruction, defect, insufficiency, or want of repair in a structure the defect in which is made a ground of liability.” Wilder v. Concord, supra, 262; Drew v. Bow, 74 N. H. 147.
The statement that an obstacle upon a bridge obstructs the bridge, and therefore is within the statute, is not an accurate description of the situation. It is the way that is obstructed. The obstruction is upon the bridge, but the thing obstructed is the passage of travel. The use of the term obstruction in Public Statutes, chapter 76, section 1, was strictly accurate, because under that statute any fault that rendered the way insufficient gave rise to the liability imposed. It did not need to be in the structure itself. It is manifest that the use of the term in the act of 1893 resulted from copying this portion of the earlier statute. Separating the provision as to this fault from its context, there would be much force in the argument that it imposes a liability for obstructions upon a bridge, although not a part thereof. But when all the specifications of grounds for liability are read together and in the light of the history of the legislation, it is evident that there was no intent to extend liability in this particular into a field which there was a definite purpose to exclude. The popular cry was that “towns should keep up their bridges.” The statute was manifestly designed to promote that end, and should be construed accordingly. For this reason the preposition “of” has usually been treated as having been here used in the sense of “in.” Wilder v. Concord, supra, 262, 263; Wentworth v. Pittsfield, 73 N. H. 358, 361; Wilson v. Barnstead, supra, 81; Drew v. Bow, supra; Hickey v. Berlin, supra, 71.
• It may be said that while it is a correct use of language to speak of a “defect in a bridge” it is incorrect to say “an obstruction in a bridge.” If this is true, it is also true that it is not strictly correct to speak of an obstruction as being “of the bridge.” .One is as faulty as the other. The provision in question cannot be construed by the application of strict rules of grammar or rhetoric. Its meaning is to be ascertained by determining as nearly as may be what the language meant to those who used it. Judged according to this test, it is reasonably plain that the use of the word obstruction does not extend *249liability beyond defects in the structure. It was apparently retained because of its use in the earlier statute and out of abundant caution, to the end that no insufficiency in the structure should be omitted. It was not the legislative purpose to retain any broader liability.
Can a temporary accumulation of ice upon a bridge be considered to be a part of the bridge, so as to be treated as a defect therein, as defined in Wilder v. Concord, supra ? The ice was merely something lying upon a sufficient bridge. A few hours of warm sunshine would remove the ice, leaving the still sufficient bridge in the same condition as it was before the ice formed. While the ice adhered to the bridge, it did not in any fair sense become a part thereof. The inertia of a heavy stone or beam left upon a bridge might present greater obstacles to removal than would the adhesion of the ice; but no one would think of speaking of either of the former as a defect in the bridge. In each instance the fault consists of the presence of something objectionable upon a sufficient structure. But the mere fact of physical contact with the structure is not the test prescribed. The deficiency must be in the structure. An encumbrance upon it is not enough. If the line of demarcation is not drawn here, it is difficult to see where it can be drawn.
If a bridge were to be treated as obstructed merely because the way over it was so, then the whole of the original liability as to bridges would exist, and the attempted limitation in the act of 1893 would come to nothing. Once it is conceded that the insufficiency may relate to the way and not to the structure of the bridge, the whole field of liability is open. Not only the pile of rubbish (Parsons v. Manchester, supra) or logs or lumber (Manchester v. Warren, supra; Manchester v. Quimby, 60 N. H. 10; Plummer v. Ossipee, 59 N. H. 55; Palmer v. Portsmouth, 43 N. H. 265), causing actual physical obstructions in the path of travel, but also sound-producing elements of the situation, like falling timber (Chamberlain v. Enfield, 43 N. H. 356) or squealing pigs (Bartlett v. Hooksett, 48 N. H. 18) and horse-frightening objects (Darling v. Westmoreland, 52 N. H. 401), could be treated as still being grounds for liability, since each would cause an unsuitableness of the way over the bridge. Nor would the somewhat crude test of actual physical contact with the structure save the situation. Such contact existed in nearly all the cases cited above. The conclusion announced in Wilder v. Concord, supra, that there was a legislative intent to abolish this class of liability in 1893 was correct, and the test there applied is the true one. The fault must be in, rather than upon or near, one of the structures named in the act.
*250The decision in Wilson v. Barnstead, supra, is not in conflict with this view. That case holds that the road surface is a part of a bridge. Non constat that something upon that surface is also a part of the structure.
The conclusion that ice is to be treated as something distinct from the structure of a bridge, makes the situation as to liability reasonably consistent. There is much force in the argument that the legislature evidently intended to abolish liability for slippery sidewalks, and that there is no good reason why towns should be made liable for such a walk upon a bridge and not liable for it if situated elsewhere. A different conclusion “would practically restore the repealed provisions of the Public Statutes. . . . Such a result is forbidden by the manifest purpose of the legislation of 1893, and cannot be reached except by annulling the terms of the section which creates the liability only when the unsuitability of the highway is caused by insufficiency in the particular constructions named therein.” Wilder v. Concord, supra, 263.
A liability as to bridges was undoubtedly retained because it was thought that there were peculiar dangers incident to the use of such structures. If the language of the act has failed to fully limit liability to such dangers (Wilson v. Barnstead, supra), there is no reason why the statutory limitation restricting liability to defects in the structure itself should be rendered still less effective by an interpretation which would treat as a part of the structure that which plainly is not so.
It follows from the conclusion that the ice was not a part of -the bridge that the plaintiff cannot recover in this case.
Judgment for the defendant.
All concurred.