GORDON, J.

The Carter Oil Company began this action in the district court of Tulsa county, Okla., against the Independent Torpedo Company, to recover damages alleged to have been sustained by plaintiff in error by reason of the negligent shooting of an oil well. The parties will be referred to herein as they appeared in the trial court.

In the petition it is alleged that plaintiff had drilled its well to a depth of 2,124 feet, and that when the well had reached this depth it employed the defendant to shoot it by means of a shell or torpedo loaded with nitroglycerine. The defendant accepted this employment and sent its agent to do this work. It became necessary to lower two shells or torpedoes to the bottom of plaintiff's well and there explode the same. The defendant lowered one shell to the bottom of the well and put the same in place. It then began to lower the second shot, but when said second shot had reached a depth of approximately 400 feet, the same exploded causing the practical destruction of the well, together with part of the casing and doing certain damage to the rig of the plaintiff used in the drilling of the well.

In the petition it is alleged that nitroglycerine is a highly dangerous explosive and requires in its manufacture and in its handling a high degree of care and skill: that it will not explode spontaneously or prematurely when properly manufactured and handled in a skillful and careful manner. The sole allegations of negligence in the petition are as follows:

"That the said defendant in connection with the manufacturing of such explosive, and in the shooting of the said well, as aforesaid, failed to use the proper degree of care and skill required in the manufacture and handling of the said explosive: and that the said premature explosion, as aforesaid, was due solely to such carelessness and negligence of the said defendant and of its servants, as aforesaid, and would not have occurred except for such negligence and carelessness nor if the said explosive had been properly manufactured and handled in a careful and skillful manner."

It is alleged that by reason of the facts stated, the defendant has made a breach of its contract to shoot the plaintiff's well, at the proper point in the oil sand and in a skillful and proper manner, and that plaintiff has been damaged thereby in the sum stated.

Defendant demurred to this petition upon the ground that it failed to state facts sufficient to constitute a cause of action against the defendant. which demurrer was by the court overruled. The defendant then filed its answer, consisting, first, of a general denial; and second, of an allegation that prior to the time of shooting this well it had notified the plaintiff that all wells were shot by the defendant at the owner's risk, and that plaintiff with this notice employed the defendant to shoot said well and defendant pleads that under such state of facts the risk of explosion was assumed by the plaintiff. Plaintiff filed its reply to this answer, denying every allegation of new matter. Upon these pleadings, the cause went to trial before a jury. At the close of the plaintiff's evidence, defendant interposed its demurrer to the evidence, and this demurrer was by the court sustained. The petition of plaintiff was dismissed and judgment entered in favor of the defendant; due exceptions were saved by plaintiff, motion for new trial was filed and overruled, and the cause is now here for review upon the action of the trial court in sustaining the demurrer to the plaintiff's evidence.

It is evident that the plaintiff framed its petition upon the theory of the application of the doctrine of res ipsa loquitur to the facts here. It is well known that nitroglycerine, which is used in the shooting of oil wells, is a highly explosive substance and that a high degree of care is required in its handling. If there is a failure on the part of the one using this substance to use a degree of care commensurate with the danger and the failure to use such care proximately causes damage to another, the party so causing such damage is liable to the other so injured. In the case before us, and in other cases of like character, the difficulty is found in determining the cause of the explosion when it occurs prematurely as in this case. Plaintiff evidently labored under this difficulty in preparing his petition and appears to have relied largely upon the theory that the fact that the defendant was using a dangerous substance and that the appliances used to lower the substance into the well were wholly within the control of the defendant, would enable it to invoke the doctrine of res ipsa loquitur, thereby placing the burden upon the defendant to so explain the cause of the explosion as that it would not appear to be due to the negligence of the defendant. But this doctrine is applied only where the instrumentalities out of the use of which the accident occurred are wholly within the control of the party charged with negligence. In this case it is true that the shell or torpedo was the instrumentality of the defendant and that the defendant had absolute control over the handling of the same. The nitroglycerine was in the exclusive charge of the defendant and the apparatus for lowering the shell filled with nitroglycerine was also in defendant's charge and owned by it. The plaintiff had prepared and furnished the receptacle into and through which this shell or torpedo should be lowered. The well partly cased and partly not cased became an integral part of the entire apparatus necessary to accomplish the purpose of the parties and when the plaintiff furnishes a part of the apparatus and the defendant furnishes the remainder, the doctrine invoked here cannot be made to apply, but it devolves upon the plaintiff to allege and prove negligence of the defendant, either as to its apparatus or the use thereof, and that such negligence was the proximate cause of the injury. The authorities, we think, are not at variance upon this proposition.

It appears settled that in order to recover by reason of the rule of res ipsa loquitur, plaintiff must aver and prove such a state of facts as that the instrumentalities which have occasioned the injury were at the time under the control and management of the defendant; that the injury was caused by a failure of defendant's appliances to properly function, and that the cause of such failure is wholly in the power of defendant to prove, and beyond the power of plaintiff so to prove.

In Thompson on Negligence, vol. 1, sec. 15, page 16, it is said:

"The principle is generally expressed in the Latin formula `res ipsa loquitur', `the thing itself speaks.' The meaning was thus expressed by Erle J., in giving his judgment in a noted case. Where the thing is shown to be under the management of the defendant or his servants and the accident is such as in the ordinary course of things does not happen if those who have the management use proper care, it affords reasonable evidence, in the absence of explanation by the defendant, that the accident arose from want of care."

And in vol. 8, section 7635, pages 1022-4, the author says:

"Presumption — from the happening of the accident. Res ipsa loquitur. The rule of `res ipsa loquitur' is a rule of evidence only. It takes more than the mere happening of an accident to set the rule in operation. It must be shown that the act was of such a character, as, in the light of ordinary experience, it is without explanation, except on the theory of negligence. The thing causing the accident must have been under the control of the defendant or his servant at the time of the accident. The doctrine proceeds on the theory that it is easily within the power of the defendant to show that there was no negligence on his part. The doctrine is an expression of an exception to the general principle that the negligence charged will not be presumed, but must be affirmatively shown. The presumption is rebuttable, and is overcome by a satisfactory explanation by the defendant. There must be some proof of negligence, for the law does not presume negligence. The rule does not obtain in cases where upon proof of the occurrence, without more, the matter rests only in conjecture. Neither does the doctrine apply where the facts shown are equally consistent with the hypothesis that the injury was caused by the negligence of the injured person or by that of defendant, or by both combined."

See, also, Baliutis v. Hammond Co., 170 Ill. App. 598; Speer v. Boggs, 52 L.R.A. 953; Davidson v. Humes Torpedo Co. (Pa.) 41 A. 649; East End Oil Co. v. Penna. Torpedo Co. (Pa.) 42 A. 707; Cosulich v. Standard Oil Co. (N.Y.) 25 N.E. 259; Benedict v. Potts (Md.) 41 L.R.A. 478; Midland Val. Ry. Co. v. Fulghum, 181 Fed. 91, L.R.A. 1917E, p. 1; Musback v. Wisconsin Chair Co., 108 Wis. ___, 84 N.W. 36; Duhme v. Hamberg American Packet Co., 184 N.Y. 404, 112 Am. St. Rep. 615: Griffin v. Manice, 166 N.Y. 188, 59 N.E. 925, 52 L.R.A. 922; 29 Cyc. 623; Sweeney v. Erving, 288 U.S. 233, 240, 57 L.Ed. 815; Levi v. M., K. T. Ry. Co. (Mo.) 138 S.W. 699; Dail v. Taylor, 151 N.C. 284, 28 L.R.A. (N.S.) 949; McGrath v. St. Louis Transit Co. (Mo.) 94 S.W. 872; Koger v. Roden Coal Co. (Ala.) 73 So. 33; Standard Oil Co. v. Murray, 119 Fed. 572.

In our judgment, the cases cited by plaintiff in error do not disagree with the rule above expressed. In the case of St. Louis San Francisco Railway Co. v. Rushing, 31 Okla. 231, 120 P. 973, this court says:

"In Solts v. Southwestern Cotton Oil Co., 28 Okla. 706, 115 P. 776, the court quoted with approval from Patton v. Texas, etc., R. Co., 179 U.S. 658, 21 Sup. Ct. 275. 45 L.Ed. 361, to the following effect:

"`The fact of accident carries with it no presumption of negligence on the part of the employer; and it is an affirmative fact for the injured employe to establish that the employer has been guilty of negligence. * * * It is not sufficient for the employe to show that the employer may have been guilty of negligence. The evidence must point to the fact that he was.'

"That, unquestionably, is the correct rule in actions of this kind."

It is true that in the same opinion the general doctrine as to res ipsa loquitur is stated with approval, but the cause was not decided upon that principle; the petition in that case specifically alleged the negligence which caused the injury and proof was made in support of these allegations.

In the case of Muskogee Electric Traction Co. v. McIntire, 37 Okla. 684, 133 P. 213; the principle is applied. But this was a case where a passenger was injured by reason of the fact that the car of the defendant in which she was riding left the rail causing her injury. This principle has long and consistently been applied to the cases of injury to passengers who are being conveyed by a common carrier.

In Chiles v. Fort Smith Commission Co. (Ark.) 216 S.W. 11, 8 A.R.L. 493, it is held that:

"The maxim res ipsa loquitur is applicable to the wrecking of a building and the injury of an employe of a tenant by an explosion from unknown cause where all of the furniture and appliances containing the gas and ammonia in the building which alone must have caused the explosion are exclusively under the control of the owner of the property, against whom the action is brought."

In the case of Bradford Glycerine Co. v. Kizer, 113 Fed. 894, there was a spontaneous explosion of nitroglycerine. The plaintiff in that case was an oil well shooter. It was his duty to haul the nitroglycerine from the magazine of the defendant, where it was stored, to oil and gas wells, to place it in tubes or shells, lower it to the bottom of the well, and there explode it. At the time of the explosion the plaintiff had taken the nitroglycerine in a wagon to the well. When he began to uncork a can containing the nitroglycerine, he saw a blaze coming from one of the empty cans and ran out of the derrick just as the explosion took place. The negligence there charged was in the improper manufacture of the explosive material. There was proof that pure nitroglycerine would not explode spontaneously and that impure nitroglycerine would so explode. The court charged the jury that if they believed from the evidence that pure nitroglycerine would not explode spontaneously and that this nitroglycerine did so explode, they could take the fact of the explosion into consideration in determining the question as to purity of the nitroglycerine, but that in considering the question, if they found that pure nitroglycerine would explode, as well as impure nitroglycerine, the fact of the explosion could not be considered as bearing upon the question of purity. In that case the manufacturer or the furnisher of the material was being sued; the plaintiff had taken the material directly from the warehouse; there was no charge of any improper handling, but the explosion appears to have been purely spontaneous.

We must, therefore, consider whether or not there was any allegation of negligence, and if so, whether there was competent proof to sustain such allegation. We take it from the condition of the proof in this case that the allegation touching the manufacture of the explosive has been abandoned by plaintiff in error. No evidence was introduced as to the method of manufacture nor was this a case of spontaneous explosion. Plaintiff in error has predicated its case upon the assumption that the explosive was improperly handled, thereby causing the explosion, and improperly handled in this, that the shell containing the nitroglycerine was sent down into the well bearing nitroglycerine on the outside; which explosive coming in contact with the casing of the well by reason of the friction and the resultant heat caused the disaster. It is therefore, to this proposition only that we will give consideration. The petition containing this allegation was met by a demurrer which was overruled by the court.

In the case of Chiles v. Fort Smith Commission Company, herein before referred to, it was argued that the allegations of the petition were conclusions of law and not sufficient to support proof of negligence, and the court held that they were merely conclusions of law, but that the complaint was not demurrable because of the application of the doctrine of res ipsa loquitur. In fact, in the petition plaintiff sets out that he did not know the cause of the explosion.

The expressions used by the plaintiff in that case are, however, even more vague than those in the case before us and when coupled with the further statement that the explosion was caused by the negligence of the defendant in a manner unknown to the plaintiff, it appears that the court was justified in the statement that such an allegation amounted merely to a conclusion of law

The question upon the petition here is a serious one, but we are inclined to take the allegation of the improper handling of the explosive material in connection with that of the high degree of care necessary to prevent an explosion, and taking these two portions of the petition together, we are not willing to say that they amount to a mere conclusion of law. The demurrer interposed to the petition was overruled by the court; the defendant did not stand upon its demurrer, nor appeal from the decision thereon, but answered denying the negligence The office of a demurrer is not to reform vague or weak or indefinite statements in the petition. A motion to make more definite and certain in this case would have reached the evil and would doubtless have been sustained by the trial court.

We therefore come to the question of the sufficiency of plaintiff's proof to justify the submission of the cause to the jury, and whether the trial court erred in sustaining the demurrer to the evidence interposed by the defendant. Omitting immaterial portions of the evidence, the substance which concerns us here is: The well casing of the plaintiff through which the shot was run was 8 1/4 inches in diameter from the top of the well to a depth of 1.521 feet; the explosion occurred at a depth of about 400 feet, so that the uncased portion of the well does not concern us here. The bit used in drilling the well was pulled in and out of the hole about 600 times during the drilling operation. The first shot was lowered to the bottom of the well without accident. On the day of the explosion the shells were brought by the defendant from Pawhuska in a truck over a rough road. Before the shells were let down, a 5 1/2 inch bailer was run. The witness Clifton tells what the defendant did with the shells and with the nitroglycerine before letting the same down, and testifies that the shells should have been tested for leaks, either by immersion in a barrel of water, or by filling the shells with water and wiping off the outside in order to determine whether there was a leak. He testifies, further, that no such test was made; that friction will cause the explosion; that a drop of glycerine on the outside of the shell, together with friction will cause heat and explosion. The witness Hashberger had shot wells of this kind for about ten years. He testifies to practically the same tests as set out above, and, further, in response to a hypothetical question as to the cause of the explosion, he testified. "There might have been some glycerine there somewhere on the shell, or something — something happened." It is inevidence further from this witness that nitroglycerine will not explode unless someone is handling it or tampering with it; that it will explode from friction; that slivers or splinters are sometimes stuck on the casings, thereby puncturing the shell; that such an occurrence might explode the nitroglycerine even with the most prudent shooter.

The rule governing the trial court in weighing the evidence upon a demurrer to the evidence has been variously stated in the syllabi of decided cases in this court. In some of these syllabi there is an expression that upon a consideration of a demurrer to the evidence, it is the duty of the trial court to withdraw and ignore evidence unfavorable to the party against whom the demurrer is interposed. Such a statement may have been proper in the particular case under consideration but it is too broad to be correctly applied to every case, and this court has intended to say that it is only where the evidence is conflicting upon the particular point that the unfavorable portion upon the point is to be withdrawn from consideration. If evidence is introduced unfavorable to the plaintiff in the making of his own case, and there is no conflict as to the truth thereof, it would clearly be error in determining the duty of the trial court to require it to ignore such evidence brought out from the plaintiff himself or his witnesses. In the case before us it was proven by plaintiff's witnesses that the explosion here might have been caused by a splinter adhering to plaintiff's casing. This was uncontradicted and the trial court had a right to consider this in passing upon the demurrer to the evidence. However, it is our settled rule that a demurrer to the evidence admits every fact which the evidence tends to prove and all inferences or conclusions which may be reasonably and logically drawn from the evidence in favor of the party making the proof, and further, that if the inference to be drawn from the evidence is a reasonable one, supported by substantial testimony, although not a necessary one, the court will not invade the province of the jury by taking from it the right to pass on the facts to be deduced from such inferences.

In the instant case, plaintiff proved the highly explosive character of the material being used. It proved the methods most suitable in the opinion of its witnesses of inspecting the shell or tubes for leaks; that no such method of inspection was used by the defendant; that the first shell was safely lowered to the bottom; that the bit had been run up and down the hole 600 times, and that a bailer had been run before the shell containing the nitroglycerine was introduced, and further proved that a shell leaking and allowing nitroglycerine to be and remain on the outside thereof, together with heat caused by friction in passing down through the casing, might have caused the explosion which occurred. There was no evidence tending to show that the shell introduced by the defendant was leaky, nor that when the same was placed within the casing it had nitroglycerine on the outside thereof. There is nothing to indicate that an inspection of the shell would have revealed any defect therein, nor that such inspection would have revealed nitroglycerine on the outside thereof. Plaintiff reasons that when proof of no inspection is made, and proof given of the explosion, the explosion itself is evidence of the presence of nitroglycerine on the outside of the shell. But this is reasoning from effect to cause. The explosion itself was not evidence of negligence, nor was it evidence pointing necessarily to the presence of glycerine on the outside of the shell. The explosion occurred from some cause unknown. Whether it be from the presence of nitroglycerine on the shell or from a splinter, or from some improper condition of the material itself, is a matter of pure conjecture and this situation would prohibit a recovery by the plaintiff in an ordinary case of negligence. The presumption in such cases is that defendant exercised due care. Negligence is not presumed. Louisville Railway Co. v. Potter (Ky.) 194 S.W. 308; C., R. I P. Ry. Co. v. Foltz, 54 Okla. 556 154 P. 519: Ponca City Ice Co. v. Roberson, 67 Okla. 86. 169 P. 1111; Patton v. Texas Pac. Ry. Co., 179 U.S. 658. 45 L.Ed. 361.

The jury will not be permitted to reach a verdict by speculation or guess. Midland Valley Ry. Co. v. Fulghum, Adm'r, 181 Fed. 91; Cleveland T. V.R. Co. v. Marsh 63 Ohio St. 236, 52 L.R.A. 142: Duncan v. Ry. Co. (Kan.) 119 P. 356; Waters-Pierce Oil Co. v. Van Elderen, 137 Fed. 557; Levi v. M., K. T. Ry. Co. (Mo.) 138 S.W. 699: Houston, E. W.T. Ry. Co. v. Boone (Tex. Civ. App.) 131 S.W. 616; Hearell v. Ill. Cent. Ry. Co. (Ky.) 213 S.W. 561.

Considering plaintiff's evidence under these rules, can it be said that it is sufficient to justify a verdict of a jury against the defendant? Could a jury with any degree of certainty point to the cause of this explosion?

The witness Hashberger is the only witness introduced by plaintiff who claims by experience any expert knowledge as to the shooting of oil wells with this explosive material. He has had ten years' experience in the actual shooting of such wells. This witness was asked a hypothetical question which assumed the drilling of the well to a depth of 2,142 feet: that the bit had been pulled in and out of the well 500 times: that the bit at the lower end was almost as large as the casing; that a five-inch bailer was run; that the first shell lowered into the well went down properly and was lodged at the proper point; that the second shell, as it was being lowered, exploded at a depth of 400 feet from the top, and that the glycerine was properly made. Upon this assumption witness was asked, first, whether the explosion would be an unusual occurrence. Having answered that it would, the witness was next asked what in his judgment was the cause of the explosion. It will be noted that the question asked in substance is as follows: Given a well cased for a distance of 1,500 feet, in and out of which a bit has been pulled 500 times and in which a bailer has been run and one shell has been safely let to the bottom, what was the cause of an explosion of nitroglycerine which occurred in letting down a second shell filled with this explosive? Any answer to this question must be a mere guess. It was not shown that witness had ever seen such an explosion in his experience. The answer of the witness indicates the difficulty. His answer is, "There must have been some glycerine there somewhere on the shell, or something — something happened." The court then questioned, "You mean a leak in the shell?" Witness: "Yes sir." The Court: "And because of friction?" Answer: "Yes, sir." (Record 64). The witness further answered, "If your shell has glycerine on it and hits the casing, it is liable to go off." (R. 64-5). And further, on page 65 of the record:

"Q. Now is there any other reason that you know of why this explosion might have occurred? A. Well, I do not know of any reason without the shell was leaking or he was running his shell too fast to make it go off. Q. Now if the shell is run down rapidly, is it then likely to explode? A. Yes, sir; that is what I always figured."

This conclusion is by a witness who never had or saw an explosion in the shooting of a well. On page 69 of the record we find:

"Q. Now. you said, on direct examination that if you would let a little glycerine get out on the can it would put the thing off? A. Yes, sir. Q. Did you ever have that happen to you? A. No, sir. Q. Did you ever see it happen to anybody else? A. I never had an explosion. Q. Did you ever see it happen? A. No, sir." (R. 70.)

Further on page 70 of the record, we have from the same witness:

"Q. You know, as a matter of fact, do you not, that slivers sometimes are caught or casing and puncture wells? A. Yes, sir Q. That might put it off? A. Yes. Q. With the most prudent shooter in charge of it? A. Absolutely."

The result of this testimony is that the witness showed himself wholly unable to determine the cause of this explosion. He merely gives several things which in his judgment might cause it. For instance, he says that if the shells were lowered too rapidly, it might cause the explosion, but there is no measure of speed suggested for the lowering of the shell. There is, therefore, no measure fixed beyond which it might be said the defendant ought not to have gone as to the speed at which it was lowered, so that we may with an equal amount of reason say that this explosion was caused either by letting the shell down at too great a rate of speed, or by letting down a shell carrying nitroglycerine on the outside of it, or by the striking of this shell against a sliver in the casing. These are the only actual suggestions made by the witness, and when he has made these, not being sure of his conclusions, he says that the explosion was caused either by these or by something. We are in no better position to judge than was the witness. The jury had only his testimony. If this case be permitted to go to the jury the verdict must be founded upon the testimony of the witness Hashberger, as there is no other expert or witness testifying as to the cause of the explosion. The circumstances proven as surrounding the explosion afford no reasonable solution as to its origin. We have seen that a verdict must not be found by conjecture or guess. But the testimony of this witness as to the cause of the explosion is of itself conjecture and any verdict based thereon must be based upon such conjecture. The witness has never seen or known of an explosion occurring from any one of the causes set out in his testimony. We are of opinion that this evidence would not be sufficient to support a verdict in favor of plaintiff, and that the trial court did not err in sustaining defendant's demurrer thereto. The judgment of the trial court is affirmed.

NICHOLSON, BRANSON, LYDICK, and WARREN, JJ., concur. McNEILL, C.J., and JOHNSON, J., dissent.