BUTTLES, J.

The plaintiff village in this action of tort seeks to recover damages alleged to have been caused to its water mains by the defendants by negligent use of explosives while engaged in highway construction or improvement under contract with the State of Vermont on U.S. Route No. 2, so-called, leading easterly from the Village of St. Johnsbury. Trial was by court with findings and judgment for the plaintiff, on exceptions to which the defendants come to this Court.

Some of the facts not in dispute are as follows: Stiles Pond had, for many years, been the principal source of water supply for the said village. In order to improve the water system, a new main was laid in 1934 from the pond to the village, which consisted of a twenty-inch cast-iron pipe for part of the distance from the pond and a fourteen-inch cast-iron pipe for the remainder of the distance. Over that portion of the route with which this case is concerned, the new pipe was fourteen inches in diameter and was laid in the same ditch with the ten-inch main, which also extended from the pond to the village and had been in use for about forty years. Work on the new pipe line was completed and the water turned on about the middle of November, 1934, and the line functioned thereafter until the time of the explosion herein complained of. The preliminaries having been completed, work under said highway construction contract was commenced at about the time the water job was completed, but the pipe line had been installed with reference to the proposed new highway, plans for which had been furnished the village authorities. A permit had been issued by the State Highway Department at the time the highway improvements were under contemplation authorizing the village to lay 3,400 feet of fourteen-inch cast-iron pipe along the southerly side of the highway adjacent to but not underneath the proposed concrete highway.

At the place where the explosion occurred, there was a knoll or rocky ledge underneath and on the northerly side of the highway, a portion of which the defendants were required to remove in order to widen and lower the grade of the highway. A short distance southerly of the highway at the ledge were the tracks of the Maine Central Railroad and a short distance southerly of the railroad was Moose River. From the ledge, both pipe lines descended easterly along the line of the highway for a short distance and then turned and went southerly under the river — the pipes under the river being 1,900 feet to 2,000 feet from the ledge. The pipes at the ledge were about twenty feet higher than the break under the river hereinafter referred to. From the south bank of the river, the pipes again ascend to the pond a distance of about 9,000 feet and to a height some 220 feet above the grade of the pipes in the river. The ledge extended easterly and westerly about seventy feet.

On the morning of December 18, 1934, the defendants had prepared and heavily loaded with dynamite about thirty-five drill holes in the easterly section of the ledge, several holes in the center section and twenty-six holes in the westerly section. These holes varied in depth from four feet to eight feet and in distance from plaintiff's nearest line (ten-inch pipe line) from twenty-three feet to six or eight feet. At about nine o'clock a.m. that day, all of the holes in the easterly section were simultaneously fired and about five minutes later, those in the westerly section. The resulting explosions were so great that pieces of rock requiring three men to move were hurled onto the railroad right-of-way approximately forty feet south of the southerly edge of the plaintiff's fourteen-inch pipe. Almost immediately water commenced to gush from the ground and a considerable stream ran down the bank onto the railroad right-of-way. The superintendent of water works for the village soon arrived and at once attempted to shut off the water by closing the gates in both the ten-inch and the fourteen-inch pipes located on the north side of the river and easterly of the ledge. He was successful with respect to the ten-inch pipe but was unable to close completely the gate in the fourteen-inch pipe. He then shut off the water in both lines in the air valve chamber located near the Cary plant a considerable distance westerly from the ledge and toward the village. Some water continued to issue from the ditch and ran down the bank until 3.00 or 3.30 p.m. when the gate in the fourteen-inch pipe on the south side of the river was closed. This apparently stopped the flow of water.

Investigation disclosed that the ten-inch pipe was quite badly broken at the ledge and at about the same place there was a longitudinal crack along the north side of the fourteen-inch pipe extending for about twelve feet. Repairs to the fourteen-inch pipe were rushed and completed on the morning of December 19. The gates on both sides of the river in the fourteen-inch pipe were opened for the purpose of turning on the water; a man being stationed in the air chamber westerly of the break for the purpose of opening that gate as soon as he should hear the water coming. There was a pressure gauge in the air chamber but at no time while the man was there did it register any pressure either of water or of air. When the gates were opened, there was a little spurt of water and an escape of air at the ledge lasting for an instant, but no water came through the pipe.

The need for water in the village being urgent, the break in the ten-inch pipe at the ledge was repaired as quickly as possible and upon completion of such repairs, that line was restored to service.

In the meantime, extensive investigation relative to the fourteen-inch line was made and on December 20 another break was discovered in that pipe under the bed of Moose River. This also was a longitudinal crack extending practically the full length of one section of pipe. The repairs to this break were naturally much more difficult and expensive than the repairs at the ledge had been and required the building of cofferdams and about two weeks of intensive work, but on completion of the job, the whole water system was found to be again in good working order.

The defendants' brief does not challenge the finding that the defendants were negligent in their blasting operations at the time and place involved, nor the finding that the plaintiff was guilty of no negligence contributing in any degree to the cause of the damage. Neither do the defendants question the finding that they were liable for such damage as resulted from the breaks at the ledge in the ten-inch pipe and in the fourteen-inch pipe. They rely solely upon the contention that the court's finding of causal connection between the explosion and the break in the fourteen-inch pipe under the river and two other findings which are subsidiary to this one are not warranted by evidence. The subsidiary findings excepted to are:

That "he [Cox] was able to close the gate [on the north side of the river] in the ten-inch line completely but was not able to do so as to the gate in the fourteen-inch line, as a complete closing was apparently obstructed by some obstacle — either a piece of stone or gravel."

That "after he [Cox] had closed these gates, water still came out of the ditch at the knoll but in much smaller quantities."

Exception is also taken to the failure of the court to make certain requested findings in lieu of and inconsistent with the subsidiary findings above referred to, the substance of said requested findings being that it was pressure of the water from above instead of an obstacle under the gate which prevented its complete closing, and that water still came from the ditch at the knoll but in somewhat diminished instead of much smaller quantities.

The argument is that if it was the pressure of the water above that prevented the closing of the gate, that fact tended to show that the pipe under the river could not then have been broken and that if any considerable quantity of water still continued to come up the twenty-foot rise and issue from the ditch at the ledge after the partial closing of the gate, that would be evidence having the same tendency.

It is elementary that this court does not retry questions of fact determined by the trial court. The findings of a trial court, like the findings of a chancellor, must stand if there is legitimate evidence fairly and reasonably tending to support them. White River Chair Co. v. Conn. River Power Co., 105 Vt. 24, 35, 162 A. 859; Peck v. City Trust Co. et al., 104 Vt. 20, 26, 156 Atl 403; Trask et al., v. Walker's Estate, 100 Vt. 51, 55, 134 A. 853; Houghton et al. v. Grimes et al., 100 Vt. 99, 105, 135 A. 15; Town of Bennington v. Fillmore Slade et al., 98 Vt. 405, 417, 130 A. 137; Morgan v. Morgan 82 Vt. 243, 73 A. 24, 137 A.S.R. 1006; Hyde Park Lumber Co. v. Shepardson, 72 Vt. 188, 47 A. 826; McGaffey v. Mathie, 68 Vt. 403, 35 A. 334; Kelton, Admr. v. Leonard et al., 54 Vt. 230.

Conflicting evidence introduced by the defendants in their favor cannot avail them in challenging the findings of the court. All conflicts must be resolved against the challenger if there is legitimate evidence fairly and reasonably supporting the findings. Lee v. Donnelly, 95 Vt. 121, 128, 113 A. 542.

The gist of the findings first above referred to is that Cox was unable to close the gate in the fourteen-inch pipe on the northerly side of the river, to which is added "as a complete closing was apparently obstructed by some obstacle — either a piece of stone or gravel." This is in accordance with testimony of Mr. Cox, who attempted to make the closing. He had had some five and one-half years' experience as foreman of the water system and testified that on perhaps a half-dozen prior occasions, he had found something under the valve which had to be cleaned out. It is to be noted that the finding is merely as to what apparently prevented complete closing. The opinion of the witness as to what made it impossible for him to close the valve completely on this occasion is something more than mere conjecture, surmise or suspicion. The facts are clearly distinguishable from those in the cases cited by the defendants: Wellman, Admr. v. Wales, 97 Vt. 245, 122 A. 659; Wellman, Admr. v. Wales, 98 Vt. 437, 129 A. 317; Goodwin, Admr. v. Gaston et al., 103 Vt. 357, 154 A. 772; Goulette's Admr. v. Grand Tr. Ry. Co., 93 Vt. 266, 107 A. 118.

Furthermore, such finding as the defendants request would clearly be based in part upon surmise. They say that the pressure of the water from the south which made it difficult to close the gate on the south side of the river made it equally difficult to close the gate on the north side. But this would be true only if the pipe under the river was unbroken at that time, which is the very fact sought ultimately to be proved. This finding of the court is sustained.

The defendants excepted to the finding of the court that after he (Cox) had closed these gates (all except the gate in the fourteen-inch pipe south of the river), water came from the ditch at the knoll but in much smaller quantities, and to the refusal of the court to find that the flow of water continued undiminished from four to six hours; that it continued sufficiently to wash out the ground and to wash out the railroad embankment opposite the knoll; that this flow continued undiminished to any appreciable extent until 3.00 or 3.30 p.m.; or that the water continued to come out of the ditch but in somewhat diminished quantities. It is conceded that some water continued to flow out of the ditch and there is no doubt but that water flowing down the bank would continue to wash it out to some extent. However, the washing out of the bank or of the railroad embankment is not here in issue and failure to find with respect thereto is not error. The ultimate fact in issue having been found, error does not appear in the failure of the court to state the effect given subordinate facts or to report evidence. Trask v. Walker's Estate, 100 Vt. 51, 65, 134 A. 853; In re Bugbee's Will, 92 Vt. 175, 182, 102 A. 484; Town of Mount Holly v. Town of Cavendish, 92 Vt. 38, 102 A. 60; Hall v. Windsor Savings Bank, 97 Vt. 125, 140, 121 A. 585, 124 A. 593.

In their brief the defendants concede that "it is very likely that the flow was diminished more than 50 per cent." It might be considered that less than 50 per cent is "much smaller quantity," in the words of the finding, than 100 per cent. Mr. Cox, the only witness who had opportunity to compare the flow of water before any of the gates were closed with that at the time here in question, testified that a little water then continued to come out and run down over the bank. At another time he testified that there was some water still coming through there. We think that this testimony, together with the conceded effect of closing the gates in both pipes below the break and the gate in the ten-inch pipe completely and in the fourteen-inch pipe partially above the break, support and warrant the court's finding. These exceptions are not sustained.

John M. Perham, civil engineer of forty-five years' experience and a trustee of the village of St. Johnsbury; Walter E. Boothby, resident engineer who supervised the laying and testing of the fourteen-inch pipe for the village and had been engaged in similar work for about fifteen years; Charles S. Sumner, village manager, who had personally supervised much of the work of laying the new main, and for nine years previous to coming to St. Johnsbury had had considerable experience of a practical kind as city manager of St. Albans in laying, testing, repairing and maintaining water mains; Frederick B. Hall, engineer of thirty years' experience, specializing in water supply and sewer systems, who had immediate charge of designing plans and preparing specifications for the improvements in the St. Johnsbury water system and was there at various times in connection with the work — all testified that in their opinion, the concussion upon the fourteen-inch pipe caused by the blasting was so severe as to temporarily check the flow of water and cause a water hammer, so-called, which set back toward the pond and caused a break in the pipe under the Moose River at the lowest point between the ledge and the pond and the point where the pressure of the water in the pipe would be greatest. This opinion evidence strongly supports the finding that the negligent blasting carried out by the defendants was the proximate cause of the break in the fourteen-inch pipe line under the Moose River to which the defendants except. Certain it is that some cause which operated not earlier than the time of the blasting on December 18 and not later than the time that water was attempted to be turned on on December 19, resulted in the break in the pipe under the river. The evidence disclosed no other cause which appears to have been sufficient to produce the result, but the defendants say that the physical facts are such as to preclude the possibility of the cause having been the one found by the court. This claim merits careful examination, because the testimony of witnesses cannot prevail against established physical facts that are inconsistent therewith. Widham v. Town of Brattleboro, 105 Vt. 210, 215, 166 A. 22; Riggie v. Grand Tr. Ry. Co., 93 Vt. 282, 107 A. 126. The physical facts upon which the defendants so rely are:

(a) The large volume of water coming out of the ditch at the ledge before any of the gates were closed.

(b) The continued flow of water from the ditch after the gate in the fourteen-inch pipe north of the river was partly closed, and the gates in both pipes at the air chamber and one above the break in the ten-inch pipe completely closed.

(c) They say that if the break in the fourteen-inch pipe under the river had been caused as claimed by the plaintiff, the ten-inch main would also necessarily have been broken at the same time and place.

No attempt was made to measure the amount of water coming out of the ditch at the ledge after the blasts and before any of the gates were closed. Unquestionably it was a large stream but we cannot say from the evidence that it was so large that it must have represented the full discharge from both pipes from both directions. Our attention is not called to any testimony, expert or other, tending to show that such was the case. Undoubtedly, some of this water was then coming through the fourteen-inch pipe under the river because some water continued to discharge after the ten-inch pipe had been closed off entirely and the fourteen-inch pipe had been fully closed below and partly closed above the ledge.

Plaintiff's exhibit No. 6 indicates that after the pipe in the riverbed was uncovered, the break therein was merely a crack. It appears that later this crack was wedged open and a piece of the iron broken out for the purpose of examination. We cannot assume that immediately after this break occurred, the volume of water escaping into the river was as great as it was at the time that the attempt was made to put the fourteen-inch pipe back into service after the repairs at the ledge had been completed. If the volume of water so escaping increased after the time of the break, obviously the pressure tending to send the water through the pipe would decrease proportionately. Even the defendants' expert witness, Earl, admits that some water might continue to go through the pipe past the crack shown in plaintiff's exhibit No. 6 and that if there were three feet of backfill around the pipe in the river when it was first cracked, that backfill, until it was washed out, would have something to do with the amount of water that would continue to flow through the pipe past the crack.

In support of their contention that the blast and resulting water hammer could not have broken the fourteen-inch pipe under the river without also breaking the ten-inch pipe, the defendants assume that the ten-inch main, having been in service for forty years, was necessarily "much weaker, less able to withstand shock and more susceptible to destruction" than the new fourteen-inch pipe. The plaintiff asks us to take judicial notice that this assumption is not in accordance with the fact and refers us to certain authorities, which it is said, indicate that "the facts of history prove that while cast iron will become quickly coated with rust, it will remain practically unimpaired in weight and strength for hundreds of years under conditions that prove absolutely fatal to wrought iron or steel in ten or twenty years."

In the absence of proof, we cannot assume, neither can we take judicial notice, that because of age, the ten-inch pipe would necessarily be much weaker than the fourteen-inch main.

The same is true with respect to the defendants' contention that the lesser distance of the ten-inch pipe from the concussion than the fourteen-inch pipe — a distance not exceeding two feet from center to center of the pipes — precluded the possibility of the breaking of the larger pipe under the river without the smaller pipe being also broken there. This circumstance may have tended to support the defendants' contention, but the court was not obliged to accept it as conclusive in view of other evidence.

Evidence, the weight of which was for the trier's determination, indicated that the fourteen-inch pipe had only a longitudinal crack at the ledge while the ten-inch pipe was shattered and literally blown out of the ditch and entirely destroyed for perhaps eight feet. Pieces of rock were blown into the space between the pipes and some of these were tight against the fourteen-inch pipe. It is quite apparent that the shattering of the ten-inch pipe would allow the immediate escaping of the water contained therein while much of the water in the fourteen-inch pipe would be unable so to escape. It is a fair inference that in the course of many years the ten-inch pipe had become so imbedded in the ground under the river as to afford much more support and resistance to internal pressure than would be the case of the fourteen-inch pipe, which appears to have been in service for not much more than one month. Again, it appears without dispute that the new fourteen-inch pipe was equipped with air relief valves at the high points which allowed air in the pipe to escape automatically but the ten-inch pipe had no such valves and air which accumulated at the high points would have a cushioning effect against the force of the water hammer. It is to this circumstance that the engineers, Hall and Perham, improved as witnesses by the plaintiff, attribute the fact that the ten-inch pipe was not broken under the river by the supposed water hammer while the fourteen-inch pipe was so broken.

We cannot say that the court was in error as a matter of law in any of these findings to which exceptions were taken, nor conversely that there was error in refusing any of the requested findings which were inconsistent with findings made.

Judgment affirmed.