November 21, 1955
KIDDE MANUFACTURING CO., INC., A CORPORATION OF NEW JERSEY, PLAINTIFF-APPELLANT,
TOWN OF BLOOMFIELD IN THE COUNTY OF ESSEX, A MUNICIPAL CORPORATION OF NEW JERSEY, AND BOROUGH OF GLEN RIDGE IN THE COUNTY OF ESSEX, A MUNICIPAL CORPORATION, DEFENDANTS-RESPONDENTS
On appeal from Superior Court, Law Division, certified to this court on its own motion.
For affirmance -- Chief Justice Vanderbilt, and Justices Oliphant, Wachenfeld, Burling, Jacobs and Brennan. For reversal -- Justice Heher. The opinion of the court was delivered by William J. Brennan, Jr., J.
[20 NJ Page 55]
The collapse of a stone wall forming the east side of Toney's Brook on plaintiff's property, and damage to one of plaintiff's buildings, resulted from the erosion of the gravel and silt bed of the stream which allowed the seepage of water under and behind the wall. Plaintiff sues the defendant municipalities, charging that when each built high smooth masonry walls along the sides of the brook above and below plaintiff's lands, and lowered and then paved the bed of the stream with a reinforced smooth concrete floor, but did not include in those improvements that segment of the stream which is on plaintiff's lands, the resultant increased velocity of the flow of waters passing plaintiff's property through the brook eroded the gravel and silt bottom there and caused the damage. Among other defenses, the municipalities contend that the erosion was due, not to increased velocity of flow attributable to their improvements, but to the faster flow caused by the action of the plaintiff in removing a stabilizing element in the stream within its lands by the levelling of a dam in the brook midway of its property, the removal of stones and boulders upstream of the dam, the cutting down of a corner of a curved concrete wall in the west bank at the dam site and the removal of a tree back of that wall. After trial without a jury, the trial judge entered judgment for the defendants. The judge found that, although the defendants were negligent in making their improvements, plaintiff's action contributed to the erosion of the bed and undermining of the wall. 28 N.J. Super. 355 (Law Div. 1954). We certified plaintiff's appeal to the Appellate Division on our own motion.
The trial judge's opinion considers the facts fully and they will not be re-stated here except where necessary to the determination of the appeal. We should note, however, that, by order entered October 19, 1954, the trial judge directed the following changes to be made in his reported opinion: the figure "550" in line 16 at 28 N.J. Super. 357 was changed to "135"; the word "condition" in the quotation
[20 NJ Page 56]
from Mr. Finley's report at line 14, page 359, was changed to "distance"; the year "1947" in line 30 at page 360 was deleted; and the year "1947" in line 33 of page 360 was changed to "1952."
Toney's Brook, about three miles long, rises in Montclair and flows south through Glen Ridge into Bloomfield where it joins the Second River which empties into the Passaic River. Plaintiff's lands, upon which it carries on a manufacturing business, are immediately south of the Bloomfield-Glen Ridge boundary line. With the exception of a small piece in Glen Ridge with a frontage of 27 feet along the west bank of the Brook, all of plaintiff's lands are in Bloomfield. The Clark Street Bridge in Glen Ridge crosses the stream about 300 feet above plaintiff's lands.
Flooding of areas bordering Toney's Brook has been a chronic problem for decades. Almost 90% of the area drained by the brook lies in Montclair, but topographically, according to Mr. Shanklin of the State Division of Water Policy and Supply, Glen Ridge and Bloomfield, which lie in the "low flood plane," are forced to bear the brunt of the waters descending from the "deep gorge upstream." The local problem is part of a greater problem involving some eight Essex County municipalities along the Second River and its tributaries. From its nature, the greater problem would best be solved by some master improvement plan. This has not been possible, however, within a framework of laws which, significantly for the purposes of this case, leaves local aspects of the problem to be dealt with locally. In 1938, following a widespread and particularly disastrous flood in Essex County, an engineering committee was constituted by a conference of officials of the State Water Policy Commission, the County of Essex and the eight affected municipalities to study the overall problem and recommend a master plan for its solution. Excerpts from the report of that committee point up the obstacle to such a plan:
"Since all construction work along or across streams must have the approval of the State Water Policy Commission, Engineer-in-Charge H. T. Critchlow [of the Commission] submitted a complete
[20 NJ Page 57]
report, accompanied by a progress map of all such work to date. However, Mr. Critchlow called attention to the facts that, while much effort was made to consider the problem as a whole when a design was prepared and improved, it must be borne in mind that each design [submitted, for example, by a municipality for some local project] was made for a specific problem requiring immediate solution, and that lack of time prevented adequate consideration to attain a comprehensive plan.
"* * * during intervening years, numerous floods have made it obligatory upon the municipalities and other agencies to make improvements in order to protect their respective properties. Since no comprehensive plan had been formulated and each authority was working within the limits of its own appropriations this protection, of necessity, had to be provided by means of channel improvement." (Emphasis supplied)
The municipal improvements to which plaintiff attributes its damage were channel improvements. The Glen Ridge improvement sped the stream through a 20 foot wide stone and concrete channel, and Bloomfield below matched Glen Ridge with an identical construction to get the stream past its limits and into the Second River beyond. But each stopped its improvement short of plaintiff's land. Bloomfield stopped about 110 feet south. The Glen Ridge paving, to the north, stopped about 160 feet above a constriction in the stream near plaintiff's footbridge. The segment intervening between the two improvements has neither a paved bottom nor a uniform width. After the stream leaves the Glen Ridge pavement, which ends some 50 feet past the Clark Street Bridge, the west bank of the Brook is in its natural state and the Brook substantially wider for a short distance; it then narrows in width to but 15 feet at a point in Bloomfield on plaintiff's lands between stone and concrete walls upon which rests plaintiff's footbridge at the northerly end of its property. Through plaintiff's lands the sides are concrete walls or timber grillage but the width is irregular and widest at the dam site previously mentioned. The dam was constructed across the brook in 1913 by plaintiff's predecessor in title, Consolidated Safety Pin Company. The dam originally was 16 feet wide, but before 1915 the banks on each side of it were recessed with curved concrete walls.
[20 NJ Page 58]
Between those walls and the original dam abutments gates eight feet wide were placed, so hinged at the bottom that they could be made to lie flat downstream in the bed to allow faster flow of flood waters. In 1942, following a serious flood of its property and the neighborhood surrounding Clark Street Bridge, plaintiff tore down the gates and the dam superstructure, leveled the abutments of the dam and a portion of the curved west wall to the then level of the stream bed, removed a tree from behind that west wall and closed the recess in the east curved wall with a new concrete wall across its base, using as fill behind the new wall some 30 to 40 tons of rocks and boulders removed from the stream bed upstream of the sill of the dam which was left in place.
None will gainsay the engineering opinion that the ideal situation would be to conform the bed and side walls of this segment of the brook to the side walls and pavements of the municipal improvements above and below it. Indeed, one of the principal items of damage sought by the plaintiff is a sum sufficient to defray the cost of such a conformation. But, so far as defendant Glen Ridge is concerned, the question immediately arises, what obligation, indeed what authority, did the borough have to spend its municipal revenues to extend its improvement across its boundary along lands lying entirely in Bloomfield? What it did do was of course done pursuant to statutory authority, N.J.S.A. 40:56-1(i), (m), (n) and (o), and with prior approval by the State Water Policy Commission as required by N.J.S.A. 58:1-26 (see also N.J.S.A. 13:1 B -50); but nothing in either statute obliged Glen Ridge to make an improvement beyond its territorial limits. The Glen Ridge improvement was started around 1934-1935 when the masonry walls were built; the paving came later and was completed in 1940. The trial judge rightly noted that a New Jersey State ERA survey, made in 1934 after consultation with the State Water Policy Commission and the County of Essex, contemplated the lowering of the bed of the brook downstream past plaintiff's plant; but we cannot agree that Glen Ridge's failure
[20 NJ Page 59]
to undertake the recommended extension has any bearing upon the question whether Glen Ridge was negligent in stopping its improvement within its territorial limits. Rather, we think the proper test is whether Glen Ridge used due care in stopping there when responsible officials of the borough knew or should have known that trouble might result for the plaintiff if steps, reasonable for the purpose, were not taken to avoid that trouble.
It is appropriate at this point to digress from the factual narrative and consider the argument of the municipalities that, as their improvements were made under a statutory authority, they are immune from liability for injuries incidentally arising from the doing of the work.
The briefs point out that plaintiff does not charge that proper engineering standards were not followed in the construction of the walls and pavements, but rather that the municipalities knew or should have known that a substantial increase in the velocity of the flow past plaintiff's property would result from the improvements and would erode the gravel and silt bed and allow water to seep under and behind its wall. This, the municipalities say, is merely a claim for injuries incidental to lawful and authorized improvements, and support their claim of immunity by reference to the decisions in Murray Rubber Co. v. City of Trenton, 103 N.J.L. 43 (Sup. Ct. 1926), affirmed on opinion below 105 N.J.L. 496 (E. & A. 1929); Beseman v. Pennsylvania R. Co., 50 N.J.L. 235 (Sup. Ct. 1888), affirmed per curiam 52 N.J.L. 221 (E. & A. 1889); Simmons v. City of Paterson, 60 N.J. Eq. 385 (E. & A. 1900); Wilson v. City of Plainfield, 4 N.J.L.J. 380 (Union Cir. Ct. 1881). But those decisions do not support the position the municipalities take here. In the Murray case, a count against the County of Mercer alleged the flooding of plaintiff's lands because of an impediment to the natural flow of a stream from the construction and maintenance of certain bridges which prevented the flow of water into the stream at its natural level. The count was dismissed for misjoinder of the City of Trenton as a party defendant. However, the former Court
[20 NJ Page 60]
of Errors and Appeals expressly stated that, "standing alone, we are not prepared to say that the negligent construction and maintenance of its bridges by the county, in such fashion as to impede the flow of water and cause its damming on the plaintiff's land, does not create a cause of action."
The Beseman case concerned the alleged liability of a railroad for annoyance and smells incident to its operation under its franchise. The Simmons case involved an attempt of owners and possessors of land along the banks of the Passaic River to restrain the City of Paterson from depositing or discharging its sewage through its drains or sewers into the Passaic River. Both decisions stand for no more than that the defendants could not be held responsible for the incidental damages that result from the "proper" exercise of legislative functions. In Beseman there was no evidence of improper exercise of the railroad's franchise. But in Simmons the right of some of the complainants to compensation for the diminution in value of their lands from the effects of sewage was expressly recognized.
Wilson v. City of Plainfield has no application to the problem before us. That case held merely that a municipal corporation having within its limits a natural watercourse may, by surface drains or sewers for carrying surface water, or by the alteration of the grades of its streets, carry the surface water which collects within its limits into that natural watercourse, and that anyone who is on the stream below is without remedy for any injury which may result to him from such passage of the mere surface water into the natural watercourse. See City of Passaic v. City of Clifton, 14 N.J. 136 (1953).
Nor do we perceive any merit in the alternative argument that here the remedy, if any, is by indictment. Plaintiff claims a private injury, and the case relied on by Glen Ridge, Waters v. City of Newark, 56 N.J.L. 361 (Sup. Ct. 1894), affirmed on opinion below, 57 N.J.L. 456 (E. & A. 1894), holds that, where public misfeasance has resulted, not in the creation of a public nuisance for which an indictment will lie, but solely in the infliction of a private injury to the
[20 NJ Page 61]
property of an individual, the remedy therefor is by a civil action by the party damnified.
We agree with the trial judge that, if the municipalities knew or should have known at the time that their improvements would so increase the velocity of flow through plaintiff's lands as to cause a material erosion of the gravel and silt bed to plaintiff's damage, and took no steps reasonably calculated to avoid that result, their actions constituted active wrongdoing for which an action lies under our cases. See Milstrey v. City of Hackensack, 6 N.J. 400 (1951), Kress v. City of Newark, 8 N.J. 562 (1952), Kelley v. Curtiss, 29 N.J. Super. 291 (App. Div. 1954), reversed on other grounds 16 N.J. 265 (1954). The approval of their plans by the State Water Policy Commission provided no protection, N.J.S.A. 58:1-26 expressly provides, "No such approval by the commission shall impair or affect any property rights, otherwise existing, which might be invaded by the construction or maintenance of any such structure." See City of Passaic v. City of Clifton, supra. And no reason of policy, certainly none in justice, suggests itself as justifying the claimed immunity. The courts of our sister states have found no difficulty in allowing suits against municipal corporations in such cases. Town of Jefferson v. Hicks, 23 Okl. 684, 102 P. 79, 24 L.R.A., N.S., 214 (1909); Mayor, etc., of Frostburg v. Hitchins, 70 Md. 56, 16 A. 380 (1889); Gross v. City of Lampasas, 74 Tex. 195, 11 S.W. 1086 (1889); Stein v. City of Lafayette, 6 Ind. App. 414, 33 N.E. 912 (1893); Ordway v. Canisteo, 66 Hun. 569, 21 N.Y.S. 835 (1893); Spelman v. City of Portage, 41 Wis. 144 (1876); Barden v. City of Portage, 79 Wis. 126, 48 N.W. 210 (1891); Annotations, 24 L.R.A., N.S., 214; 65 L.R.A. 250; 26 L.R.A., N.S., 199, and see Town of Union ads. Durkes, 38 N.J.L. 21 (Sup. Ct. 1875).
Our holding here that due care was to be exercised in making the improvements to guard against foreseen or foreseeable injury to plaintiff's lands in no wise conflicts with our decision in City of Passaic v. City of Clifton, supra. In that case the Division of Water Policy and Supply issued
[20 NJ Page 62]
permits to both municipalities authorizing each to make improvements in the section of MacDonald Brook within its confines. Passaic claimed that the nature of Clifton's improvement caused an increased velocity of flow through the brook in Passaic requiring Passaic to revise its improvement at an increased cost of $180,000. Passaic sought to have the Division impose upon the Clifton permit conditions adequate to relieve Passaic of the additional expense. Our decision was that, in the circumstances shown, the Division had not acted arbitrarily or unreasonably in refusing to impose such conditions. We noted that the Division's function is to be solicitous that the public is safeguarded against the danger from the waters impounded or affected by a structure, and that together the two improvements assured such protection to the public of both cities. We mentioned that any invasion of Passaic's property rights was to be redressed by remedies otherwise available to it but doubted that any such invasion was shown by a Clifton improvement which, like those in Wilson v. City of Plainfield, supra, and Town of Union ads. Durkes, supra, merely carried downstream surface waters collected within its confines and drained into the brook, a natural watercourse within its municipal limits. In the instant case, apart from the fact that Glen Ridge and Bloomfield sought to control much more than merely surface waters draining into Toney's Brook from within their confines, the improvements were made with the knowledge, indeed upon the premise, that comparable improvements would not be made in the brook through plaintiff's lands. Their duty to exercise due care in that state of facts is clear. The carrying of the municipal improvements through the brook on plaintiff's lands was not requisite to the proper performance of the duty. Reasonable care to avoid foreseen and foreseeable injury to plaintiff sufficed, and, as hereafter appears, neither municipality failed in its duty in that regard.
Returning now to the facts, -- for decades flooding at the Clark Street Bridge in time of heavy storms and rains ravaged adjacent lands and property, including plaintiff's lands. Plaintiff's expert witness, Arthur Barrett Miller, who
[20 NJ Page 63]
in 1913 constructed the dam on plaintiff's property, observed such floods as early as 1908. A report prepared by him in 1938 also mentions the aggravation of the problem on plaintiff's property attributed to the dam, and the effort of the previous owner to relieve the condition by recessing the banks and adding the eight-foot gates at the sides of the dam. In 1922 the County of Essex, owner of the Clark Street Bridge, raised the bridge 13 inches in an effort to provide sufficient clearance under the bridge for passage of flood waters. The effort failed and floods overflowing the bridge continued to wash out banks and flood cellars and streets in the neighborhood. The Consolidated Safety Pin Company then attempted in 1924 to restrain the Town of Montclair from directing artificial drainage into the brook. The town countered with a counterclaim for an order directing the company to remove its dam and thus eliminate that impediment to a free flow. Both the bill of complaint and the counterclaim were dismissed. Consolidated Safety Pin Co. v. Town of Montclair, 102 N.J. Eq. 128 (Ch. 1928), affirmed on opinion below 103 N.J. Eq. 378 (E. & A. 1928).
Glen Ridge constructed the masonry walls along the sides of the brook some 440 feet upstream of the bridge sometime after 1934-1935 when federal emergency relief funds were available to municipalities. The paving of the stretch did not come until later and after Mr. Finley, Essex County Engineer of Design, and Mr. Brooks, Acting Engineer-in-Charge for the State Water Policy Commission (the calculations in Mr. Brooks' letter were made by Mr. Shanklin) in 1937 expressed their written views of the source of the trouble at the Clark Street Bridge. Mr. Finley's letter said,
"The frequent floods at Clark Street are primarily due to the shallowness and rough condition of the stream. * * * it is recommended that a concrete floor be constructed through the bridge and extend above for a distance of 177 feet * * * fieldstone paving or riprap, to extend 50 feet below the bridge. For this plan the deepening will meet the present bed of the stream 650 feet below the bridge. Walls for this distance will probably need underpinning."
[20 NJ Page 64]
Mr. Brooks' letter stated:
"* * * it is our opinion that the trouble at Clark Street is due to the inability of the channel at that point to establish uniform flow. This condition results from what we believe to be a transition between low stage in the steeper reach of the channel some 400 feet upstream and the high stage flow in the flatter reach, which includes the Clark Street Bridge. This transition must be accomplished by a hydraulic jump * * * [which] * * * accounts for the very rough water surface which is found above the bridge * * * [and] reduces the approach velocity * * *. Our solution would be * * * paving the stream bed, provided that the paved section is made long enough to permit uniform flow through the bridge * * *. Downstream of Clark Street, flow will continue at low stage * * * to near the end of the paved section. At this point, flow must pass again into high stage. This transition must be another hydraulic jump, the location of which will depend upon the amount of flow, depth of flow, and required distance to overcome the higher velocity. Because of the variable sections of channel below this point, it is not practical to compute the location of this jump. We believe that if the paving is carried 100 feet below the outlet of Clark Street bridge, the jump will be formed sufficiently below the bridge so as not to interfere with the low stage flow through this opening * * *. Should the jump be located below the paved channel, it will probably damage seriously the building wall along the left bank. [This does not refer to plaintiff's building wall but that of a laundry building above plaintiff.] Paving of the channel through Clark Street will undoubtedly transfer the trouble at Clark Street to the downstream end of the paving."
Thus, in sum, Glen Ridge was told that a turbulence above the bridge, technically a "hydraulic jump," so slowed down the velocity of water coming downstream as to cause the waters to rise higher than the bridge and overflow the area, and that lowering the bed and paving it a distance upstream and through the bridge would transfer the "hydraulic jump" downstream and remove the cause of the floods. Glen Ridge was cautioned, however, that trouble must be guarded against at the point at or below the end of the pavement to which the turbulence or "hydraulic jump" would be transferred.
As finally completed in 1940, the paving extended from a point 440 feet upstream through the bridge to a point 75 feet below its center line. The work was done under the supervision of Mr. Eschenfelder, Glen Ridge Borough Engineer.
[20 NJ Page 65]
Any basis for the finding by the trial judge that Glen Ridge was negligent must depend on what Engineer Eschenfelder did in making the improvement. He was fully aware of the threat presented, where it came to rest, by a shifted hydraulic jump. But it must be remembered that at that time plaintiff's dam, a stabilizing factor in the stream, still existed and, also, that in the segment of the stream between the bridge and plaintiff's foot bridge the greater width of the brook was a factor slowing down the velocity of the waters as they passed enroute to plaintiff's property. Mr. Eschenfelder testified:
"* * * when we made our layout and submitted it to the State, we had in mind that we wanted to give adequate protection to any structure in the stream * * *. We therefore stopped our paving only 50 feet downstream from the Clark Street Bridge so we would have in between the end of that pavement a stretch * * * around 150 feet or so of natural stream bed * * * in addition there is only a wall on one side, on the east side [the laundry building wall] and on the west side you have a natural stream embankment overgrown with stubble, tree roots and very irregular which cuts in as it approaches Kidde & Company to the west * * * that offers considerable interruption to the velocity * * *. * * * when water is confined between two walls and suddenly * * * breaks off on one side into an open field, you set up a tremendous amount of turbulence which again would slow down the velocity * * * but perhaps the most important * * * thing that offers the greatest obstruction to * * * velocity is that [restriction] at the Kidde Company's footbridge * * *. These are the conditions which existed when we finished our work and as far as we knew those conditions were not to be changed."
And Mr. Eschenfelder read the letters of Messrs. Finley and Brooks as we think they are fairly to be read, that is, to caution that the hydraulic jump would be shifted below the bridge to the point where the pavement was ended, and that at that place the turbulence, while it would arrest the higher velocity of flow coming to it under the bridge, had erosive properties of its own which might do damage at its location. Mr. Eschenfelder said:
"That stopping of the pavement downstream of the Clark Street Bridge was done purposely, * * *. We knew that a hydraulic
[20 NJ Page 66]
jump would have to take place somewhere. The best opinion we got at the time, based on the County's design and calculations that the Water Policy Commission made, was that we would get a jump in the vicinity of where we stopped our paving * * *. * * * If we had gone down 100 feet we would have brought it that much nearer that troublesome section down below * * *." He therefore stopped the pavement where he did, rather than 100 feet downstream from the bridge because "* * * if there were a hydraulic jump * * * which might in any way do any damage, I wanted to give plenty of room for it * * *."
As an additional precaution he underpinned the laundry wall on the east bank, mindful of Mr. Brooks' caution that "it will probably damage the building along the left wall." He conceded that some increase of velocity past plaintiff's property would occur, but calculated that "it would be insignificant," not "material." He also testified that his decision not to use "riprap" in the 50 feet below the bridge, but to carry "the concrete paving right on down 50 feet," followed a conference with Mr. Finley's superior, County Engineer Stickel, when the two, and perhaps Mr. Finley, discussed the Finley recommendation before the work was started.
We have difficulty finding support in the record, in light of Mr. Eschenfelder's testimony, for the finding below that Glen Ridge failed to exercise reasonable care to avoid trouble for plaintiff from the making of the improvement. Our study of the record satisfies us that, if the improvement in fact accelerated the velocity of flow past plaintiff's land, which is doubtful at best, reasonable care was exercised, and we so find, invoking for the purpose our power under R.R. 1:5-4(b) to make new or amended findings of fact on review of any cause not determined by the verdict of a jury.
Moreover, we do not think that the evidence supports the finding that "the Glen Ridge improvements, completed in 1940, shifted the critical point from Clark Street downstream to a point near the footbridge at the north end of plaintiff's property." Mr. Eschenfelder testified that up until 1942 he saw no signs of the anticipated jump even at the end of the pavement. Mr. Shanklin, of the Water Policy Commission, who gave the opinion that "so long as the
[20 NJ Page 67]
hydraulic jump is located * * * upstream of the Kidde Company property, there could have been no increase in the velocity through the Kidde Company property except that due to the confinement of all flood waters within the channel at Clark Street," also testified that from personal observation the jump was "upstream of the Clark Street Bridge" even after the paving was done and did not shift downstream below the bridge, the paving "was not successful at all"; he first noticed that a shift had occurred when he inspected the scene after a 1948 flood "at which time Clark Street was not overflowed but the water went over the Kidde property. * * * indicating that that (at?) last [the] jump had moved downstream of the Clark Street Bridge." And the improvement did not prevent the 1942 floods at the Clark Street Bridge, which floods also affected plaintiff's land and led plaintiff to level its dam. Mr. Shanklin was of the opinion that the shift finally came about not because of the improvement made by Glen Ridge but because of the changes made by the plaintiff. The court's finding that there was no substantial erosion of the stream bed through plaintiff's lands until after 1942 (the erosion approximately doubled from 1942 to 1952 over the amount of erosion from 1934 to 1942), strikes us as support for Mr. Shanklin's conclusion.
The basis for the finding of negligence upon the part of the Town of Bloomfield has in our view even less basis in the record. After plaintiff leveled the dam and did the other things mentioned, great quantities of silt and debris were washed downstream. The brook below plaintiff's plant became so clogged that in 1944 Bloomfield cleaned out about 700 cubic yards of debris and silt and in 1948 another 754 cubic yards. Bloomfield removed a constriction at Glenwood Avenue and paved the channel in 1946. As a result, as the trial judge noted, "the velocity of the run-off increased, back waters were eliminated and consequently the flow upstream of Bloomfield's improvements and opposite the plaintiff's property was accelerated." But proof merely that Bloomfield must have been aware that the upstream flow would be accelerated does not prove a case of negligence against the municipality;
[20 NJ Page 68]
it must further appear that the town knew or should have known that material erosion of the bed of the stream through plaintiff's lands would result. We find no evidence to support that conclusion; the fact that 754 cubic yards of silt and debris came downstream after the paving was done does not permit that inference, since 700 cubic yards came downstream before the work was started. The more probable inference is that Bloomfield's improvement had little, if any, effect at plaintiff's site.
We do agree with the trial judge, however, that plaintiff's action in levelling the dam and doing the other things mentioned in 1942 was a source of the trouble. Our disagreement is with his conclusion that defendants' improvements also played a part. Upon the record as we read it, we are persuaded, and so find, that but for plaintiff's action there would have been no material erosion of the bed of the stream and that the damage which occurred was of its own creation. The work was done without the prior approval of the State Water Policy Commission because the company, then engaged in high priority war production work, felt it could not risk interruption of that work from floods. This may be a good reason for doing what was done, but plainly cannot justify shifting to the municipalities the responsibility for the consequences.