(dissenting)

I dissent. The case was thoroughly tried for upwards of 5 weeks' by able and experienced attorneys under the supervision of a careful and capable judge, and in my opinion no reversible error occurred. *156The only question of substance presented by the appeal is whether there is evidence fairly warranting the jury in finding that fires carelessly set or allowed to escape by defendant helped to consume or destroy plaintiffs’ property, or, in other words, were a material concurrent cause of the loss. Anderson v. Minneapolis, St. P. & S. S. M. Ry. Co. 146 Minn. 430, 179 N. W. 45. In my judgment there is.

In the forenoon of October 12, 1918, a strong wind prevailed from the west or northwest. This developed into almost a hurricane during the afternoon, the wind obtaining a recorded velocity at Duluth of 76 miles an hour for a short space of time, and during most of the late afternoon and evening from 60 to 65 miles. To the fire defendant set in timber products piled at Mile Post 62 on the railroad about 3 miles west of Brookston, plaintiffs traced the origin of the great conflagration that swept the country to the south and east thereof on October 12, 1918, destroying Cloquet to the southeast and the country east of St. Louis river between said mile post and Clo-quet. There can be no fair doubt of there being abundant proof for finding that the fire so set devastated the territory mentioned up to and beyond the line dividing ranges 16 and 15, and as far north as the south tier of sections in township 51. The evidence justified a finding that between Mile Post 60 and 61 a railway fire jumped the St. Louis river and swept easterly over the Cloquet river toward Grand lake and Pike lake.

It must be remembered that an immense quantity of dry timber products were piled at Mile Post 62, some estimate the amount at about 400 carloads. When this mass caught fire, a hurricane-like wind sent firebrands great distances, the heat would augment the velocity of the wind, and as the fire spread and its path widened north and south its progress would be in streaks, and not uniform, depending on the material feeding the fire and the ground over which it swept, so that in some sections the fire front might be miles in advance of other portions. With a wind velocity of over 60 miles an hour smoke and fire would travel at an exceedingly rapid pace, and very much more rapidly in places where embers were plentiful, and found lodgment in very inflammable stuff. These matters must be considered, also that witnesses were recalling oc*157currences and impressions experienced when they were under the greatest mental and physical stress and excitement, when they could take little notice of time, and when accurate observation was out of question because of smoke and dust.

It should also be borne in mind the weather bureau’s observer at Duluth was using standard time, and generally the people had their clocks set an hour ahead.

There is also abundant proof that the fire from Mile Post 62 swept across St. Louis river south of the mouth of Cloquet river about the same time or a little before Brookston was wiped out. There is also evidence that some hours previous thereto, when the wind was almost a gale, employes of defendant were setting fire to piles of ties between Brookston and Brevator and this criminal negligence would justify an inference that therefrom fire jumped across the river. One of the victims who got a train of the Duluth, Missabe & Northern Railway Company to stop and take on his family, a short distance north of Munger, and passing thence north to Burnett and then back again to Proctor, testified that fire was on both sides that railroad all the way almost into Proctor. 'Lacts in regard to this conflagration, which the witnesses and jurors had to deal with, may best be illustrated by brief references to the testimony.

The witness Mittunen, who sought protection in the middle of St. Louis river south of the outlet of the Cloquet, testified: I see some chunks of fire flying right over us across the river up in the air, quite high up. The flames went over us, and I noticed real fire chunks flying over close to the river.

Witness Willeck, who had taken refuge on Bloom’s place, 6 miles directly east of where Mittunen was, in relating what he went through inside an hour after the flames passed over Mittunen testified: We put out one [firebrand] and another catch right by you. Then came the hard fire. It came so fast I don’t know when it get there. It was sparks flying around, and fine sand * * * I couldn’t tell how high it was, because I can’t see how high it was. It was roaring, I thought all was the end of the world; the roaring scares you more than the fire * * * It was about ten minutes to four when I saw the sparks coming. The wind was blowing from *158the west. It was just about as much that you can’t hold yourself on the feet.

Bloom said: I couldn’t tell exactly how high the fire was, very high up in the air, probably hundreds of feet. In about another hour the big fire was some 6 or 7 miles farther to the east in range 15.

The witness Kusch testified: I saw a great ball of fire flying from the west side to the east in Stutchies’ forty, that is right north of me.

Hugo Leslie, who was a mile west of the Jackson School and about 6 or 7 miles west and a little south of plaintiffs’ property, testified: I saw the fire and it was blowing and the flames was so strong it just broke off of the balsams. I could see the flame was above us, just like waves or rolls, and every once in a while around you could see a sort of whirlwind. This was about 5:80.

Near there was George Fairley who thus describes what he noticed : It is rather hard to describe, I presume, in proper language any more than to say that the immediate territory around was very densely covered with smoke; and it seemed like the air or some gas formation just all ignited and blew up; and there was red and blue-red streams of fire clear into the air just like an explosion.

Some 3 miles east of the Jackson School was the Norman Dairy which the fire reached about 6 o’clock. John Norman testified: Fire seemed to be direct west of us in the woods * * * It seemed to me as though the flames were 100 feet in the air, if not more; for 10 or 15 minutes before the fire came sparks were flying around us, they were mostly small. The fire seemed to come all at once. We ran out of the house and took a dish towel and dipped it in some water and jumped down in the well. It was 18 feet deep. The flame soon came through the air. Timbers from the barn were blown 50 to 100 feet and were burned lying on the ground. His brother said when the flame passed over as they were in the well there was a roar like thunder. From the Norman Dairy the fire went straight northeast over the Norton road to Fred Johnson’s, a distance of about 3/4 of a mile. He thinks the fire reached his place from Norman’s in about two or three minutes.

*159The witness Lignell, a short distance east of Johnson on the Norton road, told about a small brush fire north of his land he had been helping a neighbor to put out, about 2J miles west of plaintiffs’ property, where he was until the big fire came. He testified: The fire came from the west, from Norman’s place, straight over. It came with a big roar in the air, everything was aflame, the air and everything as far as I could see. He and his wife jumped into a well. Wheeler, their neighbor, told a similar experience.

Oscar Klang, who was about a mile northeast from Lignell and Wheeler and nearly the same distance northwest from plaintiffs’, said he got home about 6 o’clock; the fire was just west of him and he had to get out; that the fire came rolling over him from the west.

Victor TJllan, living a block or two west of plaintiffs, first saw the fire coming from the west over the hill, and he thinks this was about 6:15 or 6:30. Live embers 12 to 18 inches long were then falling around him. He testified: I was on the front porch, and I saw the fire coming. Some of it was burning low and there seemed to be rolls in the air, rolls of fire.

An expert witness, a supervisor in the United States Forest Service, testified that, under the conditions existing when the wood products at Mile Post 62 burned, fires might be set by embers therefrom at a distance of 2 miles.

The foregoing are only some of the items in the testimony which may be regarded as tracing and identifying the fire set at Mile Post 62 with the destruction of plaintiffs’ property. There are many others. When such a conflagration as this, fanned by a 60-mile gale, started from immense piles of dry timber products and in its course in a few hours consuming, besides the great quantities of dry vegetation, rubbish, slashings and down timber, large numbers of dwellings, barns, cord wood and other timber products in a territory of this character, we may expect a disturbance in the wind currents and an increase in velocity that may work many queer results as to the spread and progress of the fire. So would also the lakes, wet ground and the heavy timber. The approach to Lake Superior, no doubt, had some influence on the wind direction as well as the intense heat from the fire itself. There is no doubt much in the record *160■warranting a different verdict from that found. Other fires were burning in the territory involved, but the jury could find these to have been confined or under control, And as appellant’s counsel aptly remarks, when this big fire came “it swallowed up” the small ones. To my mind the evidence justified the jury in concluding that this same fire from Mile Post 62 passed over Thorwall’s, Gill’s lands and spread further north than the part traced through Bloom’s, Jackson’s School, and Norman’s Dairy, and finally reached the place where the fire department had been busy in the afternoon north of plaintiffs’ premises in fighting bog or brush fires.

From the examination I have been able to make of the record my judgment is that this court is not warranted in holding this verdict without adequate support, approved as it is by a judge known to be cautious in reviewing testimony in cases of this character.

I agree with Justice Holt that the ultimate question was one of fact for the jury and that the evidence sustains the verdict.

This is a fire case against the defendant railway companies and James C. Davis as agent of the President under the Transportation Act of 1920. This action is based on negligence. Action dismissed as to all defendants except the Great Northern Railway Company and James C. Davis, agent of the President under Transportation Act of 1920. Plaintiffs were given a verdict and from an order denying their alternative motion for judgment notwithstanding the verdict, or for a new trial, defendants have appealed.

Plaintiffs’ property is located in the city of Duluth, 28 miles east and about one mile south, of Mile Post 62 which is at the place where plaintiffs claim the fire started.

For the purposes of this case we assume that through defendants’ negligence a fire was started in some wood products at Mile Post 62 on Thursday, October 10,1918, and though suppressed, it was not subdued, and burst forth with renewed vigor about noon on October 12. There was sufficient evidence on this branch of the case to justify the verdict.

*148The files disclose maps which show .the townships, rivers, town lines, sections, lakes, railroads, highways, villages, cities and other pertinent matters; and the map inserted here will materially aid in the reading of this opinion.

In township 51 we have towns in ranges 19, 18, 17, 16, 15, 14 and 13 and, beginning with range 18, they are named Culver, Industrial, Grand Lake, Canosia, Eice Lake and Lakewood. Then immediately below this we have township 50 with ranges 19, 18, 17, 16, 15 and 14 and, beginning with range 16 they are named Solway, Herman and city of Duluth.

The St. Louis river flows from west to east through sections 30, 29, 28, 27, 26 and 25 in the town of 51-19 and through sections 30, 29, 28, 27, 35 and 36 of the town of Culver, and then through sections 2, 1 and 12 of the town of 50-18 and then southeasterly several miles to the city of Cloquet.

The Cloquet river comes from the northeast and enters the St. Louis river on section 36 of the town of Culver.

The village of Brookston was located principally on sections 34 and 35 of the town of Culver.

The country between Mile Post 62 and plaintiffs’ property was timbered country with settlers here and there developing their land.

The fire at Mile Post 62 started to spread at 1:40 p. m. The general direction of the wind in the afternoon on this date, which was so terrible to the settlers in that locality, was from a little north of west. This fire reached Brookston, 4 miles away, at 4 p. m.

A fire was on the north side of the river at about Mile Post 60, and while the evidence does not directly connect this with the railroad fire, the jury was possibly justified in drawing an inference that it was of railroad origin.

The railroad fire caught Brookston and passed over a territory immediately to the south thereof, and crossed the St. Louis river at 3:30 p. m.

The fire north of Mile Post 60 apparently went easterly and crossed the Cloquet river in section 30 of the town of Industrial at about 2 p. m. and then passed on to where it destroyed one Thorwall’s house in the western part of section 29 at about 3 p. m. and then the evi-

*0

*149dence seems to lose the identity of this fire, although it may have been the source of fire at Grand Lake station on the Duluth, Mis-sabe & Northern Railway Company at the southeast corner of section 27 at 5 p. m. At this point this part of the fire is completely lost and, from all that the record shows, there terminated.

The evidence shows that a fire at 4 p. m. struck Bloom’s place on the east half of the northeast quarter of section 2-50-17, which is 5 miles east of where the railroad fire crossed the St. Louis river at 3:30 p. m.

A fire was in the northwest quarter of section 6 of the town of Solway (6-50-16) at 4:35 ip. m.

A fire was in the tree tops at 4 or 4:10 p. m. in section 1 in the town of Solway (1-50-16). This is 12 miles east of the place where the railroad fire crossed the river at 3:30 p. m. and 6 miles east of Bloom’s place which the fire struck at 4 p. m.

In the forenoon of the eventful day, from sections 1 and 12 of the town of Grand Lake (51-16) fire was observed in the north and a trifle to the east. This fire had been in existence for several days and had come from due north and had “traveled possibly several miles east of where it originally had been seen.” It was in the vicinity of Fish lake (not shown on map).

At 7 p. m. flames were jumping in the tops of the trees from hilltop to hilltop like torches, going east and possibly south in the vicinity of Rice lake.

A fire was observed on section 15 in the town of Herman (15-50-15) at about 3:30 p. m. which is about the same time the railroad fire was crossing the St. Louis river 16 or 17 miles to the west.

Jackson School was 'at the northeast corner of section 10 of Herman town (10-50-15). Between 3:30 and 4 p. m. a fire existed half a mile north of the Jackson School and was there all day and, at a point two miles north of Jackson School, fire had crossed the roads at 4:30 p. m.

Between 4 and 5 p. m. a fire crossed large portions of the Work Farm on sections 1 and 2, town of Herman (1 and 2-50-15), covering all of section 1. The main buildings on this farm were burned by *150another fire between 7 and 8 p. m. and this was at a time when the wind was from the north.

Under a north wind the Kimball and Clark homes on section 12, just south of section 1, part of Work Farm, were burned. This was about the same time that plaintiffs’ home was destroyed. It was apparently located on the northwest 40 acres of section 2-50-14. But the Norman dairy, located on the southerly part of sections 4 and 5 and on the northerly part of sections 8 and 9 in 50-14, was burned at 6 p. m. Almost straight west from the dairy, the fire came from the west to the Leslie home on the west side of section 10, town of Herman (10-50-15) at 6 p. m.

Was the fire at Leslie’s at 6 p. m. the fire that crossed over the Work Farm between 4 and 5 p. m? Fire was seen from the dairy farm at 5 p. m. approaching from the west, 100 feet high, and yet the fire from the west apparently did not reach Leslie’s, over 4 miles to the west, until 7 p. m.

This brings us to the immediate vicinity of plaintiffs’ property. To the north and northwest are platted lots. About a quarter of a mile due north of plaintiffs’ home was a lettuce farm on which peat land was burning all the afternoon of the day of this misfortune, and the fire department of Duluth was called and arrived about 3:35 p. m. They connected a hose with a city water hydrant and fought the fire with water. After half hour of unsuccessful efforts, men were called from other fires up on the Arnold road, half a mile north, to help. This fire, in spite of the human opposition, continued to spread and advance, and at 7 p. m. the men withdrew and left it almost north of plaintiffs’ home, and from this direction and this locality came the fire brands that brought the destruction.

Calvary road, traversing from west to east in the center of sections 32 and 33 and then southeasterly through section 34 in Bice Lake township (34-51-14), had between 6:30 and 7 p. m. fire on both sides of it along part of section 34. Kingston road, which is parallel with the west part of Calvary road and one-eighth mile to the north in sections 32 and 33, was blocked by fire at 6:45 p. m. To the north of the Kingston road is Luzerne road and from the north of this road fires were sweeping in from the north. These fires were approaching from the northwest between 5 and 6 p. m.

*151At about one o’clock that afternoon a fire was burning on tbe west part of section 20 of tbe town of Rice Lake (20-51-14). This was about half a mile east of Rice lake and the territory to tbe north was covered by fires on this day.

At 12:30, when tbe railroad fire bad not yet made a fair start and bad not left tbe landing at Mile Post 62, such clouds of smoke were blowing over tbe city of Duluth, 29 miles to tbe east, that tbe sun was obscured and tbe condition alarming. This fact alone is very persuasive as to tbe general fire condition in a large territory west and northwest of tbe city of Duluth in which locality a number of stray fires were located. At 4:30 in tbe afternoon tbe witness Kayser saw fire on both sides of tbe Howard Gnesen road about two miles from plaintiffs’ home, which bad “jumped tbe road.”

There were a number of fires in this burned area on tbe afternoon in question, which could not possibly have originated from tbe fire at Mile Post 62. This is particularly true of fires in tbe towns of Grand Lake, Canosia, Rice Lake, Solway and Herman. In fact, aside from tbe fire at Mile Post 62, these various fires were not of railroad origin, and their origin is unknown.

Fires have been known to smolder for weeks and then flare up and proceed to consume property. This fire occurred on a day when forest fires were extant in northern Minnesota and burned areas extended beyond tbe limits of this immediate locality. Anderson v. Minneapolis, St. P. & S. S. M. Ry. Co. 146 Minn. 430, 179 N. W. 45; Hall v. Davis, 150 Minn. 35, 184 N. W. 25. This fire occurred in an unusual and protracted dry period and on a day when a high wind was prevailing.

Tbe record shows that at 7 a. m. on tbe day in question, in tbe locality of Leslie’s place, smoke was in tbe air, indicating tbe presence of fire which caused this man to delay a trip to Duluth, having in mind the welfare of bis children, who were to be left on tbe farm alone. Smoke was in tbe air at this place for several days previous, indicating the presence of fire which was not spreading rapidly until perhaps put in motion by tbe gale on tbe afternoon of October 12.

Tbe buildings on tbe dairy farm caught from fire coming from tbe west, apparently from tbe fires in tbe locality of tbe Work Farm, *152and from this same apparent source, fires may have blown south to Kimball’s place and Clark’s place from about 7 to 7:30 p. m., which was about the same time that plaintiffs’ home was consumed by flames brought in a wind from the northwest. The Carver house, which was a block, or so to the southwest of plaintiffs’ premises, did not burn.

Emil Lignell owned the northeast quarter of the northeast quarter of section 5-50-14 in the city of Duluth, which is immediately to the northeast of the Norman Dairy, and was called to his premises by phone on the afternoon of October 12, as he says, “to fight the fire back of my land, the first fire,” and he arrived there about 3 p. m., when the fire was north of his land, and he stayed there all afternoon watching and fighting this fire and he continued in this service until the fire came from the west destroying the dairy buildings. The Lignell fire, like others, indicates or shows various sources and opportunities for spreading fires.

At 2 p. m. a fire was reported as being on or near section 6-50-14 and the owner appeared there between 3 and 3:30 p. m. and observed heavy smoke then coming from the west. This is apparently 20 miles from the place where the railroad fire was crossing the St. Louis river at 3:30 p. m.

From the west side of section 28-51-14, around noon, signs of fire were noticed to the north, and from this point of observation, buildings were burning at about 3:30 p. m. North of the Luzerne road was a fire between 4 and 5 p. m. and at 5 p. m. the large gate on the Howard Gnesen road leading to Calvary cemetery was on fire.

Homecroft School on the southwestern part of section 34-51-14, and about a mile from plaintiffs’ premises, burned about 6:15 p. m.

The various fires located over this devastated region in their peculiar locations — and considering the direction of the wind — render it impossible to point out and identify which fire or fires caused plaintiffs’ loss.

It is the claim of the .plaintiffs that the fire was brought from Mile Post 62 by a west wind or at least by a wind a trifle north of west and, when this fire got into the vicinity of the towns of Solway and Herman, that there was a sufficient change in the wind to carry *153the fire in a northeasterly direction toward the location of ¡plaintiffs’ property, which property, together with Kimball’s home, Clark’s home and the Norman Dairy, was destroyed by a fire brought by a north wind.

If the fire which reached Leslie’s place on the west side of section 10 in the town of Herman at 6 p. m. was not the railroad fire, what fire was it? Surely it was not behind the railroad fire. If this was the railroad fire, how are we to conclude that it destroyed the Norman Dairy 4 miles to the east at 6 p. m? If it did not destroy the Norman Dairy, what reason is there to conclude that the railroad fire destroyed plaintiffs’ premises? It would seem more reasonable to conclude that plaintiffs’ premises were destroyed by fires from the north or northwest, though with possibly some connection with the fire that destroyed the Norman Dairy.

The evidence fails to disclose with any degree of probability that the fire, having its origin at Mile Post 62, constituted a material or substantial element in any combination of fires that consumed this property. The trial court, in a memorandum somewhat inconsistent with the order appealed from, says: “It may be that the fire reached plaintiffs’ property and aided in its destruction.” That is not sufficient. This very statement characterizes the ever present contrary probability born from this evidence.

On October 12, 1918, in this locality there were numerous other fires for which defendant was not responsible, and on this day a large area was swept over and a wind blew that would fan many slow and smoldering fires into flames which would either develop terrific fires, in themselves, or join the mighty march of other flames. It is plain that the railroad fire moved easterly and consumed Brookston and covered some adjacent territory, and crossed the St. Louis river, leaving its path of ashes and desolation. If this fire united with other fires, as it may have done, there is no evidence to sufficiently identify this fire, alone or in union with other fires, ever reaching plaintiffs’ premises. The evidence fails to show any probability that this fire was even a concurring element in the destruction of this property. The evidence leaves plaintiffs’ case in the realm of speculation and conjecture.

*154Suppose the wind did divert this fire to the northeast when or after it reached the locality of section 19-50r15. Where did it go? And what did it do? Did it carry on alone in its course of destruction or did it join other independent fires, which were present in that locality, heyond the peradventure of doubt, or did it get into the immediate vicinity of plaintiffs’ premises too late to join with the fire which consumed the holdings of the ¡plaintiffs? The evidence does not answer these inquiries. Herein is the weakness of plaintiffs’ case. The identification, which the law requires, is absent. Under our system of jurisprudence, remote possibility can never establish the existence of a fact which must be proved. The most that can be claimed for plaintiffs’ proofs is that it is barely possible that the railroad fire might have been, in whole or in part, the fire, which burned plaintiffs’ property. It may have been.

Mr. Justice Mitchell in the case of Baxter v. Great Northern Ry. Co. 73 Minn. 189, 192, 75 N. W. 1114, which is a forest fire case, said this:

“Under such circumstances it was necessary that the fire should be carefully traced and identified by the evidence from the right of way to plaintiff’s premises. The jury were not at liberty to arrive at the result by mere guess or conjecture, but must have had some substantial evidence on which to base their verdict. We have carefully read and studied the evidence, and are clearly of the opinion that there was no sufficient tracing of the fire, either by any one witness or by all the witnesses together, from the point where the sectionmen ignited the pile of ties and rubbish on the right of way, to plaintiff’s premises. Neither was it traced by proof of facts and circumstances from which the inference could be legitimately drawn that it was the same fire. The most that can be claimed for the evidence is that it is possible that the fire started on the right of way might have been, in whole or in part, the fire which consumed plaintiff’s property. Anything like an analysis of all the evidence would be impracticable and useless. A careful reading of it in the light of a sectional map of the country will, we think, satisfy any one that our conclusion as to its-probative force is correct.”

*155The Baxter case was cited with approval in Brennan Lumber Co. v. Great Northern Ry. Co. 77 Minn. 360, 79 N. W. 1032.

Defendant cannot be made liable upon a mere possibility, nor upon a remote probability, but there must be some positive proof which will permit a conclusion based upon some thing besides conjecture and speculation. Swenson v. Erlandson, 86 Minn. 263, 90 N. W. 534.

In the case last cited Mr. Justice Brown used this language [at page 267]:

“The origin of the fire must 'be established by evidence reasonably certain and convincing to an unprejudiced and unbiased mind, and until there is some such evidence there can be no recovery. Circumstantial evidence is, of course, competent and proper, and often the only obtainable evidence in such cases; but, as such evidence consists in reasoning from facts which are known or proved, to establish such as are conjectured to exist, the process is fatally vicious and defective if the circumstances from which we seek to deduce the conclusion depend also upon conjecture and speculation. 3 Rice, Ev. § 342.”

The circumstances furnish a basis only for a conjecture that the loss may have occurred as claimed, but the circumstances do not furnish a basis for the inference that the loss, in fact, did result from the railroad fire.

The evidence and circumstances upon which plaintiffs rely for a recovery are too improbable and conjectural to permit a verdict to stand. -'The authority and duty of the court in cases of this kind to direct final judgment notwithstanding the verdict of a jury, is well settled; and in the absence of lawfully required identification, which is necessary to support a verdict, the duty of the court to say so is imperative. The rights of the party entitled to such consideration cannot be ignored.

The order of the trial court is reversed and a new trial is granted.