Townsend, J.
We have examined the voluminous record in this case with close attention, as well as the briefs and arguments of counsel, and at the outset are constrained to say that there is very great doubt in our opinion as to the origin of that fire. It appears that this case was tried before a jury, and resulted in a mistrial, by failure of the jury to agree. It was tried again, and withdrawn by reason of the sickness of a juror, and on this the third trial a verdict was for plaintiff (the appellee here.) It also appears that several other suits have been instituted, growing out of that fire, against the railroad company, the appellant here. One of them (Fuller vs Railroad Co.) was appealed by the railroad company to the circuit court of appeals (18 C. C. A. 641, 72 Fed. 467,) and thence to the United States supreme court (18 Sup. Ct. 944,) where the case was affirmed against the railroad company by a divided court, four of the judges being in favor of the affirmance of the case, and four against (the-court being equally divided); [89]*89but it also appears that the case at bar had much more evidence in behalf of the defendant than did the Fuller Case, and counsel for appellant confidently asserts that, had the facts in' the case at bar been in the record of the Fuller Case, reversal would have followed, without question. Counsel for appellant therefore insists that the trial court should have sustained his motion to direct the jury to return a verdict for the defendant.
The appellant introduced, as appears from the record, about 30 more witnesses on the trial of the case at bar than were introduced by it on the trial of the Fuller Case, and much of this testimony related to the direction the wind was blowing at the time the fire occurred. This was one of the most material questions presented to the jury. The appellee had 6 witnesses that testified that the direction of the wind was from the southeast, and the appellant had 16 witnesses that the wind was from the south, and on that point 12 more than in the Fuller Case. The wind had to be from the southeast to carry the smoke from the engine to the barn, as the barn was 300 feet west of the main track, and the measurement showed that from the point where the engine was located by appellee’s witnesses to the southeast corner of his barn was 359 feet, and appellee’s witnesses say the wind was strong from the southeast. If such were the case, it would necessarily carry the fire from the barn in a northwesterly direction; and here develops a physical fact of very great significance. The fire from the barn extended and destroyed the property north of the barn, and not northwest from the barn. The residence of John D. Moore was 661 feet from the main track of the Missouri, Kansas & Texas Railroad, and was located 4,260 feet from the barn. Hence in a short time the fire had extended on a line due north, and parallel to the main track of the Missouri, Kansas & Texas road, which ran north, 10 deg. east, and had only varied in that distance of 4,260 feet the difference [90]*90between 300 feet and 661 feet, while, if the witnesses of the appellee are correct, the wind had carried the smoke to the barn, 300 feet from the main track, in a distance from the engine to the barn of 359 feet. The physical facts sustain the testimony of appellant’s witnesses, and not the witnesses-of the appellee. Other evidence was introduced by appellant tending to show that the barn was frequented by people-who drank whisky and smoked pipes and cigars; that a small room was on the upper floor of the barn, which was-used for that purpose, and that immediately preceding the fire Dole Baugh took his horse out of the stable, and started east across the railroad track, met Homesley and Springston, and was talking to them; that he was smoking a cigar at the time, and was apparently under the influence of liquor; that, a very few moments after his leaving the barn, it was seen to be on fire, and Homesley and Springston called Baugh’s attention to the fire; that Baugh stated to a number of witnesses that he was sorry, and knew Wilder would blame him; that the fire started in the barn, just about where he got his horse; that he went in the barn, smoking, to get-his horse, and had not been out over 15 minutes when the fire was discovered. A number of witnesses testified to the statement that it was frequently mentioned in the town, while the fire was burning, that Baugh had set the barn on fire. Baugh, in rebuttal, denies these statements, but the number of unimpeached witnesses testifying to the statements makes his evidence of little value. Evidence was also given that during the fire Wilder, the appellee, accused Baugh of starting the fire. It is shown that Baugh was the stepson-in-law of Wilder, and it is also shown by evidence that Baugh was the one who developed the theory that the engine of the appellant started the fire, and he went to the telegraph agent and stated that he was accused of setting fire to the barn, and wanted the number of the engine that hauled the stock train that afternoon. Mose Alberty, wit[91]*91ness for appellee, testified that the fire burned two oat. stacks over 700 feet due north of the barn, and did not burn two haystacks less than 300 feet northwest of the barn, and that the fire burned all buildings located due north of the barn, but did not burn stores to the west of it, or Mrs. Grant’s hotel, to the northwest of it; and he, with William. Calvert and Francis Eller, all witnesses for appellee, testified they first discovered smoke and heat coming out from the east door of the barn, while appellee claims that the fire first appeared in the upper story, on the southeast corner, where hay was stored in the barn, and where his witnesses testify the smoke from the engine struck the barn. The appellee’s witnesses testify that smoke came out of the engine, and the wind drove the smoke to the southeast corner of the barn, but none of them testify that sparks or fire were emitted from the engine; and the jury, by their verdict, must have presumed that the smoke carried sparks of fire. There was no positive testimony to that effect. From the testimony presented by the record, it is only reasonable to conclude that the origin of the fire is in great obscurity; it being doubtful if the preponderance is not more in favor of other causes, than that it originated from appellant’s engine.
The appellee alleges in his complaint that the sidetrack was defective, on which the engine was moving a train of freighted cars, and in doing so ‘ ‘negligently let in and upon said locomotive engine such a high and heavy pressure of steam, * * * whereby a large volume of smoke, intermixed with sparks of fire, was driven out of the smokestack of said locomotive engine, which large volume of smoke, intermixed with said sparks of fire, was then and there blown by the strong wind aforesaid in the direction of, and against and through the cracks and openings in the south east corner of, this plaintiff’s aforesaid barn or livery stable, and thereby the hay in the loft of this plaintiff’s said barn of livery stable,- was set on fire.” The' engineer of [92]*92said engine testified: “It is almost impossible for an engine to emit sparks when great volumes of smoke are being emitted. Black smoke comes out when fresh coal is thrown in, and, the more coal that is thrown in, the more smoke comes out. The green coal, being of a lower temperature than the fire, prevents the sparks from coming out.” The engineer also testified that his engine slipped none on said side track. The appellee’s witnesses testified that the engine wheels were slipping, and large volumes of smoke were coming , out of the smokestack. Evidence was also introduced by appellant that the smokestack of said engine was in good condition both on the 20th of September and also on the 23d and 25th of September, being before and after the fire; that it was a Diamond smokestack; it had a screen in top, and was in perfect condition. William G-illis testified for appellant as follows: “Foreman boiler department M., K.,& T. Ry. at Parsons, Kansas. Said engine 16 had a Diamond stack. Explained same by a model. That screen had a four by four mesh. Four openings and four wires to the inch. That on some roads they use only a 3x3 mesh. That sparks go up and strike a cone first, and fall back and strike the side of stack, and then up and strike the netting. It is impossible for a spark to get out the first time it goes up through body of stack. I examined this engine 16 on September 25, 1891. I had top of stack taken off engine and put on the ground, and I examined it. I got up on top of stack and examined cone, and all was perfect. Put netting back, and engine went out on run with same netting. The ash pan was also in perfect condition. Diamond stacks have been in use ever since I was a boy, and they are in general use now, and are approved plan. Have been in this business for thirty years.”
[93]*93Negligence Burden of proof. [92]*92Such being the facts as presented by the record, the court very properly and correctly instructed the jury that: “The burden of proof is upon the plaintiff in this case to [93]*93show two things, before a recovery cán be had: First, that this fire had its origin from sparks of fire communicated from the locomotive ' engine of the defendant; second, that such fire escaped and ignited this property through the negligence of the defendant, its agents and servants, — that is, that they, the company, its agents and servants, were not in the exercise of that care and caution which a prudent man would exercise under like circumstances. ” The appellee, alleging negligence, must prove it, as it is said in the Nitro-Glycerine Case, 15 Wall. 587: “No one is responsible for injuries resulting from unavoidable accident while engaged in a lawful business. A party charging negligence as a ground of action must prove it. He must show that the defendant by his act or by his omission has violated some duty incumbent upon him, which has caused the injury complained of. The cases between passengers and carriers for injuries stand upon a different footing. The contract with the carrier being to carry safely, the proof of the injury usually establishes a prima facie case, which the carrier must overcome. His contract is shown prima facie at least, to have been violated by the injury. Outside of these cases, in which a positive obligation is cast upon the carrier to perform safely a special service, the presumption is that the party has exercised such care as men of ordinary prudence and caution would exercise under similar circumstances; and, if he has not, the plaintiff must prove it. Here no such proof was made, and the case stands as one of unavoidable accident, for the consequence of which the defendants are not responsible. The consequences of all such accidents must be borne by the sufferer, as his misfortune. This principle is recognized and affirmed in a great variety of cases: In cases where fire originating in one man’s building has extended to and destroys the property of others; in cases where injuries have been caused by fire ignited by sparks from steamboats or locomotives, or caused by horses running away, or by blast[94]*94ing rocks; and in numerous other cases which will readily occur to every one. The rule deducible from them is that the measure of care against accident which one must take to .■avoid responsibility is that wbieh a person of ordinary prudence and caution would use if his own interest were to be affected, and the whole risk were his own. ” Pierce (R. R. p. 437) says: ‘ ‘As negligence is the gist of the action against the company for injuries received from it while exercising its lawful right to conduct its trains, the burden of proof is •on the plaintiff to prove negligence. The fact of injury suffered by the plaintiff in consequence of the exercise of a right by the defendant does not raise the presumption of negligence, except in some peculiar cases, as in actions against innkeepers and common carriers, which are made ■exceptions lb the general rules on grounds of public policy, Hence the setting on fire of grass, fences, or buildings of landowners on a railroad by particles of fire which are proved to have issued from the company’s engines does not of itself justify the inference of negligence.” Wharton (Neg. § 869) says: “This is an essential industry; you are authorized to engage in it; and, as it is necessary that your engines should be driven by fire and steam, you are authorized to use fire and steam in your engines. Such being the case, the mere fact of the company emitting sparks from its engines is not negligence, unless it be proved that the sparks were negligently emitted. ’ ’ And in support of his text he cites Flynn vs Railroad Co., 40 Cal. 14; Rood vs Railroad Co., 18 Barb. 80; Railroad Co. vs Yeiser, 8 Pa. St. 366; Frankford & B. Turnpike Co. vs Philadelphia & T. R. Co., 54 Pa. St. 345; Jefferis vs Railroad Co., 3 Houst. 447; Railroad Co. vs Woodruff, 4 Md. 242; Burroughs vs Railroad Co., 15 Conn. 124; Sheldon vs Railroad Co., 14 N. Y. 218; Vaughan vs Railroad Co., 5 Hurl. & N. 679; Railroad Co. vs Brand, L. R. 4 H. L. 171, 201, 202. Rorer (R. R. p. 796) says: “In cases of loss by fire set out by an engine upon a railroad, and [95]*95involving the question of negligence, the mere fact of the injury does not amount to proof of negligence or wrong upon the part of the company. Such proof of setting out the fire, or escape thereof from the property of the company, is but one stage towards making out a case for recovery. The circumstances and facts relied upon as negligence are also to be proven, and are to be proven by the plaintiff. To recover, he must aver negligence on the part of the company, as also ordinary care on his part. To recover, the burden of proof is on him to prove what he thus necessarily avers.” In Jacksonville, T. & K. W. R. Co. vs Peninsular Land, Transp. & Mfg. Co. (Fla.) 9 South. 661, it is decided that “the mere emission of sparks from a locomotive, or the setting out of fires thereby, is not, per se, evidence of negligence on the part of a railroad company.” And after citing various authorities, including Whart. Neg ; Shear. & E. Neg.; Pierce, E. E.; Eedf. E. E.; and Eor. E. E., — in support of that holding, the court adds: “The fact that no instrument has yet been found which entirely prevents the escape of sparks from locomotives seems, when coupled with the fact that the use of these engines is both lawful and eminently useful, a sound reason for the view that the mere emission of sparks, or the simple setting out of fires thereby, is not, per se, evidence of negligence, and will not throw upon the defendant the burden of removing such presumption. ”
[96]*96Evidence. Origin of fire [95]*95In the case at bar there is no proof that sparks were emitted, and the jury necessarily had to presume that they were, in order to find the verdict. Isa jury authorized to presume that a fire originated from sparks, and, based on that presumption, further presume that it was through the negligence of appellant’s servants? In Sheldon vs Railroad Co. 29 Barb. 226, being an action to recover the value of a mill, located 67 feet from the track, with its windows open and highly inflamable material within, the court said: “It is not enough for the plaintiff to show a possibility that the [96]*96fire was communicated to the mill by sparks emitted by defendant’s locomotive. He cannot recover upon a possibility. Even if the evidence went further, and brought the fact sought to be proved to within a probabilty, still the plaintiff must fail, because, to justify a verdict, the law requires, not positive proof, it is true, but such proof as will leave no reasonable doubt of the existence of facts upon which it must rest. The rights of property and all claims to its possession and enjoyment are dependent on the existence of certain facts, and when they are disputed and become a subject of judicial investigation, if juries could assume their existence without sufficient evidence and render verdicts upon possibility, probability, and conjecture, the courts would be shorn of their legitimate authority, and the wise and just rules of common law, as they have been recognized and applied from time immemorial, would lose their principal value.” In Haley vs Railway Co., 69 Mo. 614, the court said: “It is a matter of inference only that the fire was kindled by sparks escaping from defendant’s engine. There was no direct testimony to that effect. It was not even shown that any sparks escaped from defendant’s engine as it passed. As there is no natural and necessary connection' between the running of the engine and the kindling of the fire, it ‘ would seem that some testimony should have been introduced, connecting the two events.” Philips (1 Ev. p. 598) says: “Where the facts proved are not the precise facts in issue, and the jury are to come to a conclusion upon the facts in issue by an act of reasoning from those other facts, the evidence in such cases is said to be presumptive. A presumption of fact is properly an inference of the fact from other facts that are known. It is an act of reason. Kay (Neg. Imp. Duties, p. 142) says: “Fires arise from many indefinable and undiscoverable causes, and are even spontaneous and without any known "cause, so far as any human agency is [97]*97concerned. In such cases they must he regarded as mere accidents. Where an • event takes place, the real cause of which cannot be traced, or is at least not'-apparent, it ordinarly belongs to that class of occurrences which are classed as purely accidental; and, in a case where a plaintiff asserts negligence, he must show enough to exclude tlie case from the class of accidental occurrences. ” Before an inference of negligence can arise, the accident must first be connected with the defendant by direct evidence. Higgs vs Maynard 12 Jur. (N. S.) 705; Smith, Neg. (Whittaker’s Ed.) 421; Welfare vs Railroad Co., L. R. 4 Q. B. 693; Wright vs Railroad Co., L. R. 8 Exch. 137, 42 Law J. Exch. 89; and Rice, Ev. § 51. Starkie, in his work on Evidence (page 754, note), in speaking of this class of cases, says: “Do they not afford temptations to juries, in hard cases, to trifle with the sacred obligation of an oath?” The inference which the1 jury is to draw is not a circumstance in proof, and is not' a, legitimate foundation for yet another inference or presumption. U. S. vs Ross, 92 U. S. 283;Dou glass vs Mitchell’s Ex’r, 35 Pa. St. 446; Rice, Ev § 34. In Lowney vs Railway Co. (Me.) 7 Atl. 381, the court say: “The burden upon the plaintiff, therefore, was to prove, not only that the fire was communicated by the engine, but also that the defendants were guilty of negligence and their negligencewas the cause of the communication of the fire. The communication of the fire alone does not import negligence. Nor does the proof of negligence alone import that it was the cause of the fire. The must be proved. Its relation as the efficient cause of the fire must also be proved. *' * * If it be suggested that it is difficult to prove negligence in such a case, and that the rules of law above stated are a- hardship upon the plaintiff and those situated as he is, it should be remembered that the defendants were pursuing in a lawful manner a lawful business, and one useful to the entire community. It is the general rule of law that one engaged in a lawful busi[98]*98ness and acting in a lawful manner is liable for sucb injuries only as are caused by his negligence. In some actions against individuals it may be difficult to prove that the defendant was negligent, and that his negligence caused the injury, but that difficulty would be no good reason for changing the rule.” In Railroad Co. vs De Graff (Colo. App.) 29 Pac. 664, the court say: “In regard to the proof necessary to establish the principal fact, viz. that the fire was communicated by the operating of the road, we think the doctrine of inference from circumstances has been carried as far or further than was legally allowable, without in any way criticising adjudicated cases. We think the rule should be limited, not extended. Country juries have invariably shown themselves inimical to lines of road operated adjacent to or through their property, and probably, in most cases, for satisfactory reasons; but, under an ironclad statute preventing a corporation from exonerating itself from blame in cases of fire, juries should not be allowed to infer or presume the fire to have been so originated, for want of possible proof of some other origin, and base a verdict upon inference or guess. The fact of the origin of the •fire, like any other material fact, should be established; and while the jury, within certain limits,, may-be left to infer the fact from the circumstances proved, such proof sho.uld be sufficient to rebut the probability of the fme having originated in any other manner. Corporations are certainly entitled to legal protection to this extent. ” In Peck vs Railway Co., 31 Mo. App. 123, the court quotes with approval from the case of Sheldon vs Railroad Co., as set out in this .opinion supra, and adds: “There were so many other means by which the fire in question may have originated, within the range of possibility and probability, as to render a judgment against the defendant too conjectural, on the evidence adduced by the plaintiff.”
Under the foregoing citations of authorities, which, in [99]*99in our judgment, fairly state the rule of liability for fires by railroad companies (and many other cases might be cited), we do not believe, under the proof in this case, that the appellee is entitled to recover; but the jury Lave found for the plaintiff, and the trial court has refused to set aside their verdict; hence we will consider further a few of the assignments of error.
Evidence. Negligence, The appellant has filed 54 assignments of error, and the fifth, sixth, and eighth assignments of error are substantially as follows: “The district court erred in permitting witnesses Waller and Farrell to testify that engines of the appellant, other than the engine identified by the complaint and appellee’s evidence in this case as the one that set out the fire in question, cast out fire and sparks, and erred in refusing to withdraw such evidence from the jury.” The engine that it was claimed emitted the smoke and caused the fire was fully identified by the evidence as engine No. 16, hauling train No. 102 on September 23, 1891. The burden being upon the appellee to establish the fact that the fire originated from the engine that it is claimed set out the fire, it is competent and proper to show that that particular engine on other occasions emitted fire and sparks, as showing the negligence of the appellant in operating and using a defective engine; but where the engine is fully identified, as in this case, it is improper to admit evidence that other engines of appellant emitted fire • and sparks upon other occasions, for the very good and sufficient reason that, if appellant is to be held liable for this fire, it must be for negligence in the use and operation of this particular engine, and not for negligence in the use and operation of defective engines generally. Such a rule would make the railroad company liable for negligence generally, when upon the occasion complained of the company might have had in use a complete and perfect engine, and one that was equipped with the best mechanical device known to engine builders [100]*100and operators for arresting fire and sparks. Where the engine or engines setting out fire are unknown and unidentified, the rule is enlarged, and the fact that engines of the railroad company have emitted fire and sparks on other occasions is admitted in evidence to establish generally the negligence of the railroad company in the use and operation of defective engines. The rule as above stated is clearly established by the weight of authority, and, for illustration of the two rules as above stated, see the following citations: ‘ ‘It is, of course, Indirect evidence, if it be evidence at all. In this case it was proved that engines run by the defendant had crossed the bridge not long before it took fire. The particular engines were not identified, but their crossing raised at least some probability, in the absence of proof of any other known cause, that they caused the fire; and it seems to us that, under the circumstances, this probability was strengthened by the fact that some engines of the same defendant at other times during the same season had scattered fire during their passage ’’ Railway Co vs Richardson, 91 U. S. 454. Also, Henderson, Hull & Co. vs Philadelphia & R. R. Co , (Pa. Sup.) 22 Atl. 851: “Testimony tending to show that other fires were set by the same engine about the same time, however, is the proper rule, and is undoubtedly competent; but when the loss or injury is shown to have been caused, or, according to the proof, may have been caused, by sparks from an engine unknown and unidentified, or by one of several engines, some of which are unknown and unidentified, then the rule of evidence is necessarily somewhat enlarged. The burden of proof in all such cases, in the first instance, is upon the plaintiff to show that the fire in question was communicated from the defendant’s engines. ‘It devolves upon the plaintiff to prove by a preponderance of the evidence that the fire was communicated by sparks or cinders from the railway engines. It need not be shown that any particular engine was [101]*101at fault, but it will be sufficient if the fire is proved to have been set by any engine passing over defendant’s railway, and the evidence may be wholly circumstantial, as — First, that it was possible, for fire to reach the plaintiff’s property from the defendant’s engines; and, Second, facts tending to show that it probably originated from that cause, and from no other.’” Also, Inman vs Railroad Co:, 90 Ga. 663, 16 S. E. 958: “Where the engine that caused the fire cannot be fully identified, evidence that the defendant’s engines frequently emitted sparks on former occasions near the time of the fire in question is generally held relevant and competent to show habitual negligence, and to make it probable that the plaintiff’s injury proceeded from the same quarter; but, when the engine is identified, the same reason does not operate, and évidence as to the condition of other engines, and of their causing fire, is clearly irrelevant. ” Also, Railway Co. vs Jones, (Ark.) 26 S. W. 595: “And it was inadmissible to show that other engines had set fire to materials on or near the right of way, as a circumstance to show that the engine which caused the fire on this occasion, or its appliances, were defective or in bad condition.” In the case of Railway Co. vs Johnson, 4 C. C. A. 447, 54 Fed. 474, the Court of Appeals for the Eighth circuit recognizes that in Railway Co. vs Richardson, supra, the Supreme Court was dealing with a case where the ■ particular engine which was alleged to have set out the fire in question was unidentified. Also, First Nat. Bank of Hoopeston vs Lake Erie & W. R. Co. (Ill. Sup.) 50 N. E. 1023: “The engine or locomotive which was alleged to have caused the fire was identified as the engine which drew the freight train passing the elevator near the noon hour of August 31, 1892. * * * In cases of this kind it often happens that the proof does not identify the particular engine which caused the loss, but is confined to negligence in the operation and- construction of the engines generally which run on the road. The rule seems to [102]*102be settled by the weight of authority, that when a fire has been caused by sparks from a particular locomotive which is identified, * * * ‘evidence of other fires kindled by different locomotives, before and after the fire complained of, is not admissible. ’ ” This case cites authorities from Wisconsin, Pennsylvania, Missouri and Maine, and also Shear-man & Redfield on Negligence, and the American & English Encyclopedia of Law. It also cites and relies upon the opinion of the Supreme Court of the United States in Railroad Co. vs Richardson. We are therefore of the opinion that the trial court, in admitting the testimony of Waller and Ferrell that engines of the appellant, other than the engine identified by appellees witnesses in this case, emitted fire and sparks, and in refusing to withdraw such evidence from the jury, was in error.
[103]*103Depositions. Exceptions. [102]*102The fourteenth assignment of error is as follows: “The district court erred in sustaining appellee’s motion to quash the deposition of Lem Springston. ” It appears from the record that this deposition was taken upon interrogatories prepared by the counsel of the respective parties, to be used by the appellant on the trial of five causes, all of them growing out of this fire; that the depositions were taken by W. S. Pollard, notary public, at Fayetteville, Ark., on the 1st day of October, 1897; that the five depositions of Lem Springston for use in each of the five causes were each duplicates of each other, and were by the notary mailed in one envelope addressed to the United States clerk at Muscogee, Ind. T., and indorsed on the envelope as follows: ‘ ‘Depositions in Cases of Ferrell et al vs. M. K. & T. Ry. Co. No. 2,363.” It further appeared that the clerk of the United States Court at Muscogee marked said depositions ‘ ‘Filed' ’ and placed them in the one envelope received from the notary, with the papers in the case of E. Ferrell vs Missouri, K. & T. Ry. Co., No. 2,363. It further appeared that the attorneys in the Ferrell suit and the suit at bar were the [103]*103same. It appears from the record that the trial of the cause at bar commenced on November 9, 1897, and the motion to suppress was filed' on November 12, 1897, which was during the trial of said cause. Mansf. Dig. § 2955 (Ind. T. Ann. St. 1899, § 2070), provides: “No exception, other than to the competency of the witness, or the relevancy or competency of the testimony, shall be regarded unless filed and noted on the record before the commencement of the trial.” It appears from the record that the deposition of Springston in this case had been filed by the clerk, but placed with the papers in the Ferrell case, and not with the papers in this case. This was the neglect or mistake of the clerk, of which appellant is blameless. We think the motion, under the statute, comes too late. The motion to suppress was not on the ground of the ‘ ‘competency of the witness or the relevancy or competency of the testimony, ” and therefore should not have been regarded, and in suppressing the deposition the trial court was in error.
The charge of the court in this case fairly and fully presents the law of the case; but in consideration of the doubtful preponderance of the evidence as set out in this opinion, and the errors of the court in admitting the testimony of Waller and Ferrell, and the suppression of the deposition of Lem Springston, which was exceedingly important evidence to assist the jury in arriving at the origin of the fire, and which might and probably would have had great weight with the jury in arriving at their verdict, we are of the opinion that this case should be reversed, and remanded for a new trial. Reversed.
Clayton and Thomas, JJ., concur.