HENRY, J.

This consolidated action concerns the liability of insurance companies under an electrical apparatus clause of fire insurance policies, which reads in part as follows:

"This Company shall not be liable for loss resulting from any electrical injury * * * to electrical equipment * * * unless fire ensues in electrical equipment * * * but if such fire does ensue * * * this Company shall be liable for * * * direct loss caused by electricity. * * *

"The words `fire ensues' and `ensuing fire' as used in the foregoing mean self-sustaining fire which continues after the electrical currents artificially generated have been interrupted. Electrical arcing and flashovers, caused by electrical currents artificially generated, shall not in themselves constitute an ensuing fire."

Plaintiff claims that on June 7, 1956, when two of its three generating plants were demolished and swept into the gorge of the Niagara River by a series of rock slides, electrical generators in its third plant (Station 3 A) were partially destroyed by electricity and ensuing fire. It claims that its loss therefrom in the amount of $1,196,104.90 is covered by the electrical apparatus clause.

On the trial it adduced evidence from which the jury might have found that electrical currents artificially generated were interrupted before 6:15 P.M. of the day the loss occurred and further evidence that one-half hour thereafter firemen observed fires in two generators which continued to burn for a further period of about half an hour while they attempted, unsuccessfully, to extinguish them. It claims that this evidence brought its loss within the coverage of the policies because it thereby established that generation of electrical currents had ceased and that a self-sustaining fire continued.

Defendants contend that electrical currents were not interrupted but continued to be generated during the time that fire occurred and that such fire was not an ensuing fire within the meaning of their policies.

The generators, and the turbines which turned them, were located in Station 3 A at the edge of the river. The turbines were turned by water which came from the forebay at the top of the bank. Water flowed therefrom through gates into penstocks and through them down into the turbines. It passed through the turbines causing them to turn the generators and was discharged through tailraces into the river. Water could also enter the tailraces through openings in the floor of the turbine room and generator room. As long as water ran through the penstocks and turbines the generators were turning and producing electricity. The speed at which the generators turned was controlled by governors. The operator in the control room could shut them down completely. When Stations 3 B and 3 C collapsed and fell into the river, control of the turbines and generators in Station 3 A was lost at 5:14 P.M. The generators then ran very fast. They were out of control, producing much arcing and sustaining electrical injuries. The only way that they could be stopped was by closing the gates in the forebay, thereby stopping the flow of water into the penstocks. The control room operator, Brennan, testified that such gates were all closed not later than 5:45 P.M.

Against this evidence that the gates to the penstocks were closed so as to prevent the flow of water to the turbines, defendants point to photographic exhibits which show water running out of the tailraces of Station 3 A, after the time that Brennan claims to have closed the gates and contend that the water was still flowing through the turbines when the firemen observed flames in the generators.

Plaintiff claims, however, that the water flowing through the Station 3 A tailraces came from the penstocks in Station 3 B which were broken when that station collapsed. Its hydraulic engineer testified that the rock slide sheared the penstocks in 3 B and that on the following day the water was running from the penstocks across the remainder of Station 3 B into the turbine room of Station 3 A, from whence it drained into the tailraces.

If the jury accepted plaintiff's evidence they might find that after electrical currents had ceased to be generated a fire continued to burn in two of the generators for approximately an hour.

Its expert witness testified that electrical arcs could occur only when electricity was being generated; that they would develop tremendous heat adjacent to the arc which would subside very quickly and in not more than one minute the heat would come down to a degree which would no longer sustain any volatilization or burning or anything of that kind. Thereafter the court received testimony of defendants' expert witness, over plaintiff's objections, that it would take several hours for the heat developed in a generator by an electrical arc to subside to the level of its temperature before inception of the arc. The provision of the policies that an ensuing fire is a self-sustaining fire which continues after electrical currents have been interrupted, rendered evidence respecting heat remaining in generators after termination of electrical arcs relevant insofar as such heat would tend to support combustion but when heat fell below that level it would have no relevancy on the question of self-sustaining fire. The policies do not require that a self-sustaining fire must occur after electrical injury is ended to be within their coverage, but their requirement is merely that such fire continue after electrical currents have been interrupted. Evidence of the time required for the heat to subside to the temperature that existed before the occurrence of the arc was not relevant. The error in receiving such evidence was compounded by the statement of defendants' attorney in his summation to the jury that it would take an hour or two before that electrical injury would be over. The error was emphasized further by the court's charge to the jury, which went far beyond the terms of the policy. Appellant took no exception to the charge although the court charged three conflicting rules, as follows: "[E]nsuing fire * * * would mean [1] one which had its inception following an interruption of the electrical current * * * [2] [A]lthough it might have in its inception been ignited by such disturbance * * * [3] a fresh combustion * * * after * * * a cessation or cooling of whatever heat force was originally generated by the electricity." We think that the testimony of defendants' expert witness: "For that copper to return to the * * * heat that it was before the arc, would take several hours" together with defendants' attorney's statement on summation: "[I]t would take an hour or two * * * before that electrical injury would be over" and the court's erroneous charge: "An ensuing, self-sustaining fire contemplates a fresh combustion * * * after * * * cessation or cooling of whatever heat force was originally generated by the electricity" resulted in the case being submitted to the jury on an erroneous theory, which precluded a fair consideration of it by the jury.

The charge is so inconsistent and contrary upon a material proposition that it is impossible to reconcile the different versions. (See Johnson v. Blaney, 198 N.Y. 312, 317.)

"While no exception was taken by the plaintiffs' counsel to the court's charge * * * the error, in our opinion, was a fundamental one and we take cognizance of it in the exercise of our power to reverse and grant a new trial in the interests of justice." ( De Joseph v. Gutekunst, 13 A.D.2d 223, 226.)

Because of the errors in the admission of evidence and in the interests of justice the judgment appealed from should be reversed on the law and facts and a new trial should be granted, with costs to appellant to abide the event.

WILLIAMS, P.J., BASTOW, HALPERN and McCLUSKY, JJ., concur.

Judgment as amended and order as resettled unanimously reversed on the law and facts and a new trial granted, with costs to appellant to abide the event.