SCHAUER, J.

In this action for the wrongful death of plaintiff's 3-year-old son the jury returned a verdict in favor of plaintiff and against defendant Franklin General Benevolent Society, [fn. 1] a corporation (hereinafter termed the hospital), and against plaintiff and in favor of defendants Dr. Nellie B. Null and Dr. John Null. [fn. 2] Judgment was entered accordingly, and the hospital appeals.

It appears that on the morning of July 5, 1946, plaintiff's son entered defendant hospital for the purpose of having his tonsils and adenoids removed by the Drs. Null. During the operation the anesthetic, which was gas and ether, was administered by a nurse-anesthetist. The patient started to come out of the anesthetic on two or three occasions and the anesthetist administered more ether. After the third increase in ether Dr. John Null noticed that the blood in the patient's throat was dark, which indicated that something was wrong, and he also then found that breathing had ceased. He thereupon began to apply manual artificial respiration, while the anesthetist left the room to secure a mechanical resuscitator. She returned with a resuscitator three or four minutes later but neither it nor other means used was successful in preserving or restoring life to the child.

As grounds for reversal, defendant hospital contends that:

1. Plaintiff failed to prove certain material allegations of his complaint;

2. The judgment rests on the doctrine of respondeat superior, based on the erroneous theory that the anesthetist was an employe of defendant hospital;

3. The trial court erred to the prejudice of the hospital in the giving of certain instructions, particularly instructions proposed by the defendant doctors relative to the doctrine of res ipsa loquitur.

[1] The evidence, which must be viewed on appeal in a light most favorable to respondent (Estate of Bristol (1943), 23 Cal.2d 221, 223 [ 143 P.2d 689 ]), may be summarized as follows:

The child was taken to defendant hospital at the suggestion of defendant Dr. Nellie B. Null, who had previously examined him. The child's mother paid a fee to the hospital for use of the operating room and for the anesthetic.

The patient's tonsils were moderately enlarged and inflamed but were not in an acute condition; the operation was not an unusual case or an emergency nor did it involve major surgery. When the child was brought into the operating room he had already been placed in an "intermediate" or "preliminary anesthetic state"; after he was placed on the operating table further ether was administered, first through a mask by drops, and later "by tube, by gas, causing the ether to bubble through tubes."

Dr. John Null, who is the son of Dr. Nellie B. Null, testified that "bubbling ether through that tube while the child continues unconscious ... is a constant process, constantly watched, and varies in degree of how fast the bubbling occurs ... [A] great amount of discretion [is] necessary in the administration of that ether through the tube ... Unless the proper amount is given the child would not stay in the proper stage of anesthesia, and if it bubbles too slowly he will probably awake and be turning, and if it is given too fast, if he breathes it too fast it is apt to stop his heart ... [T]here is a grave danger which may well occur if the ether is sent too rapidly in the form of bubbles through that tube"; the witness further stated that ether "is a relatively safe anesthetic, and ... is one ... entrusted to the less experienced, in fact, that is what you train them on in giving anesthetics, is a tonsillectomy. It is considered safe, but any anesthetic is dangerous ... [T]he danger may reflect itself ... By the abolishing of the eye reflexes, which the anesthesia blocks; by the changes in the rate of respiration, and changes in the pulse rate and color ... [W]hile an anesthetic is being administered ... through the tube ... the anesthetist [must] ... Watch the color, the patient's pulse, respiration and reflexes, especially the eye reflexes ... it is her duty ... [T]he reflexes should [not] return at all during the course of surgery to a person who is supposedly in the proper state of anesthesia ... Several times during ... the operation [here involved] the child would partially regain consciousness so that the reflexes in the throat, the gag reflex would return, and she [the anesthetist] would have to increase the depth of the anesthesia before the operation could be continued"; the witness attributed "that return of the reflex action to ... too little ether being administered" and "two or three times" he or his mother asked the anesthetist to administer more ether, which she did "By opening the shutoff valve so as to let the flow of oxygen increase"; the witness further stated that it was "unusual for this child to awaken to the extent ... described ... occasionally it happens, but it is not supposed to ... This child had no unusual tolerance for ether" and the witness did not know of "any reason to expect those reflexes to return during the course of the operation."

While the child was "in the depths of anesthesia" Dr. Nellie Null started to remove the tonsils and Dr. John Null assisted by swabbing and using "the suction ... a mechanical device ... to pull the blood, mucous and phlegm out of the throat"; just after the second tonsil was removed and after the third increase in anesthesia he noticed the "blood was extremely dark ... [which] means there has not been a proper oxygen supply, and either the child was too sound asleep and not sleeping [sic] enough ... and I called to the anesthetist about it ... [The] anesthetist usually watches the blood, the flow and color ... [which] shows whether the child is breathing properly or not, and not asleep" and the witness assumed "she was doing this." At the same time he "also looked up at the child and I didn't see any respiration, which also told me there was something wrong ... [The anesthetist] apparently hadn't noticed it ... [It is my] opinion the anesthetist gave too much ether to the child ... More than indicated by the circumstances ... [and] that was the cause of the darkening of the blood ... [W]hen that kind of danger takes place ... the anesthetist usually immediately stops the flow of ether and artificial respiration is immediately instituted ... This time I turned the suction over to my mother and [began giving artificial] respiration ... with my hands, but the best way is by a mechanical resuscitator, which applies carbon dioxide and oxygen into the lungs ... the very best procedure is to apply the ... resuscitator ... at once ... [W]hile I was giving manual respiration I called for the resuscitator." The machine was not in the room and the anesthetist left the room and returned with it three or four minutes later; meanwhile artificial respiration was continued. Dr. John Null testified that he did not think that at that time "the child was actually alive, in the sense of awake. There was no way of telling whether it was beyond help of recalling the life or not, but it is possible that if we had had the mechanical resuscitator there we could have gotten the carbon dioxide and oxygen in there and it would have started breathing, but to all outward appearances it was dead. Lots of times that will happen and they are given artificial respiration and the child will come back to life." The resuscitator was used but failed to revive the child, and both doctors were of the opinion the child had died by the time the machine was brought into the room, and that the death resulted from an overdosage of ether by the anesthetist; Dr. John Null also testified that in his opinion a competent anesthetist would not "make a mistake of that kind," and Dr. Nellie Null stated that she had performed "hundreds of these tonsillectomies" since she had first been licensed to practice in 1906 and this was "the first case in which a death has ever occurred in one of" them. The doctors further testified that "it is up to the anesthetist to put the child to sleep, that is her job" and it was her duty if she sees the patient "moving around, to get him under so we can proceed."

The doctors stated also that the mechanical resuscitator was "standard equipment for hospital surgery rooms" which "we expect ... to be present in any kind of an operation," and it "should be immediately available, ... it is not always present right in the room, but should be there or close by." A doctor would normally "expect to find it in the room, although it is not always there." The hospital stipulated that "it is standard practice in the hospitals in San Francisco to have a resuscitator available for use in the operating rooms."

A coroner's autopsy report indicated that the child's death was caused by "inspiration of hemorrhagic material." The doctors testified that it was necessary to cease using the suction device in order to apply artificial respiration and that artificial respiration would have forced such material into the lungs.

Concerning the employment of the anesthetist the doctors testified that the hospital, rather than the doctor, provides the nurses, the anesthetist, the operating room, table, and instruments and "everything for the operation" and that the doctors did not select nor pay the fee of the anesthetist who worked in this case. Dr. John Null also stated that he did not employ the anesthetist and was not present "at any conversation in which" his mother employed her. Mrs. Stevenson, an employe of defendant hospital, testified that she is a "nurse-anesthetist" who was "in charge of all anesthetists" at defendant hospital at the time of the operation here involved; among the anesthetists "employed by the hospital" was the one who worked in this case; since October, 1946, a licensed physician who "specializes in the administration of anesthetics" had been in charge of the anesthetists; "throughout the past several years ... more and more hospitals have placed licensed physicians in charge of ... their [anesthetic] department"; the nurse-anesthetist who worked in this case "was in the employ of the hospital just as any other anesthetist was," her salary was paid "directly by" defendant hospital, she received no "fees or salaries from surgeons using the facilities of the hospital" but was "paid by the hospital alone," and she left the employ of the hospital in April, 1947.

Mrs. Stevenson testified further that she couldn't state "accurately" how many of the defendant hospital's seven operating rooms were in use at the time of the operation here involved, but it was her "best recollection" that all the rooms were then in use "because some of those rooms are running under local anesthetics where we are not involved"; that the hospital had only one mechanical resuscitator and she did not know where it was on the day here involved.

Dr. Cardwell, the physician in charge of anesthetists at defendant hospital since October, 1946, testified that he had been trained and had practiced his profession in Washington, D. C., prior to World War II; in 1945 while he was stationed in northern California during service with the United States Navy, he visited three hospitals in San Francisco and had observed therein only one resuscitator "in connection with one operating suite or one surgery suite"; the surgery suite at defendant hospital consists of eight operating rooms; in the other three San Francisco hospitals he had visited five rooms was the minimum he had observed to constitute a surgery suite; he did not know the number of hospitals in San Francisco; so far as he knew only two hospitals in San Francisco have resuscitating devices, but "they could have them in hospitals where" he had "not even visited"; "the idea of having a resuscitator present in or near a surgery room is so that if some emergency arises causing the respiration of a patient to disappear it can be used at once"; "no matter how far science may have advanced, the anesthetist must always be alert and vigilant during the operative procedure ... The greatest ... skill will serve the anesthetist naught if she relaxes her vigilance at any time ...; it is a difficult and dangerous field, in which vigilance must be added to scientific training."

1. Asserted Failure of Proof

In the complaint it is alleged (paragraph VIII) "That ... defendants ... so carelessly, negligently and recklessly performed the said tonsillectomy operation as to ... cause the ... child ... to suffer ... a severe hemorrhage therefrom, and did carelessly, negligently and recklessly allow hemorrhagic material to flow unchecked into the lungs of the ... child. [Paragraph IX.] That as a direct and proximate result of the carelessness, negligence and recklessness of the defendants ... said ... child was caused to ... suffocate and drown in its own blood ..."

Defendant hospital urges that although plaintiff alleged "that the child suffocated in blood ... Plaintiff didn't really prove anything but, if it can be said that anything was proved, it was that the child may have died from too much anesthetic," and that therefore, under the provisions of section 471 of the Code of Civil Procedure [fn. 3] plaintiff failed to prove his cause of action.

[2] It is apparent, however, from the evidence related hereinabove that the jury could properly, and must be presumed to, have concluded from the autopsy report that the immediate cause of death was the "inspiration of hemorrhagic material," that such inspiration resulted from the necessity to give artificial respiration and to cease using the suction device, and that that necessity was caused in turn by the erratic and excessive administration of anesthetic. Defendant's contention of failure of proof in this respect is, therefore, without merit.

2. The Doctrine of Respondeat Superior

[3] Defendant hospital contends that as a matter of law the nurse- anesthetist was the employe or agent of the defendant doctors, rather than of the hospital, and that the hospital cannot be held responsible for any negligence of such anesthetist. Again, it is apparent that the evidence amply supports the implied finding that the anesthetist was the hospital's employe in the operation here involved. The cases relied upon by the hospital (see Ybarra v. Spangard (1944), 25 Cal.2d 486, 491 [ 154 P.2d 687 , 162 A.L.R. 1258 ], in which this court stated "it appears" the anesthetist and special nurse were employes of the hospital owner; and Ware v. Culp (1937), 24 Cal.App.2d 22, 27 [ 74 P.2d 283 ], in which it was held the evidence failed to support a finding that a special nurse was the employe of defendant hospital) differ factually from the instant case and do not compel a holding here that the anesthetist was not a hospital employe (see Hallinan v. Prindle (1936), 17 Cal.App.2d 656, 662 [ 62 P.2d 1075 ]).

The hospital's contention that the operating doctors rather than the anesthetist were responsible for the proper administration and regulation of the anesthetic relates solely to a conflict in the evidence. The implied determination of the jury that the mentioned responsibility was at least primarily that of the anesthetist is supported by the testimony of the doctors which has been heretofore quoted.

3. The Doctrine of Res Ipsa Loquitur

[4] At the request of the defendant doctors the court gave a res ipsa loquitur instruction applicable against the doctors and the hospital. [fn. 4] As declared in Ybarra v. Spangard (1944), supra, 25 Cal.2d 486, 489 , quoting from Prosser, Torts, p. 295, the doctrine of res ipsa loquitur has three conditions: "(1) the accident must be of a kind which ordinarily does not occur in the absence of someone's negligence; (2) it must be caused by an agency or instrumentality within the exclusive control of the defendant; (3) it must not have been due to any voluntary action or contribution on the part of the plaintiff." In Engelking v. Carlson (1939), 13 Cal.2d 216, 221-222 [ 88 P.2d 695 ], where the peroneal nerve was severed in the course of an operation on plaintiff's knee, this court declared, "It is true that in a restricted class of cases the courts have applied the doctrine of res ipsa loquitur in malpractice cases. But it has only been invoked where a layman is able to say as a matter of common knowledge and observation that the consequences of professional treatment were not such as ordinarily would have followed if due care had been exercised. For example, it has been applied where a sponge was left in the body of the patient after closing an operative incision [citations]; where the patient was burned by the application of hot compresses or heating apparatus [citations]; where the patient was burned through the operation of an X-ray machine [citations]; and where the patient sustained an infection through the use of an unsterilized hypodermic needle [citations]. In each one of these situations the rule was applied because common knowledge and experience teaches that the result was one which would not have occurred if due care had been exercised."

"But the present case shows an entirely different situation. Here what was done lies outside the realm of the layman's experience. Medical evidence is required to show not only what occurred but how and why it occurred. That evidence establishes beyond question not only that the peroneal nerve may be injured even where due care is used but that this unfortunate result invariably occurs in a limited number of cases. The doctrine of res ipsa loquitur is, therefore, entirely inapplicable and no malpractice has been proved."

And in Moore v. Belt (1949), 34 Cal.2d 525, 530-531 [ 212 P.2d 509 ], after citing some fourteen cases on the proposition, it was stated that "In the cases cited where the doctrine was held applicable evidence that the defendant did not cause the injury was remote and it followed as a matter of common knowledge from the nature of the injury that the result would not happen without carelessness or negligence. In the present case the inference that the injury was not caused by the defendant, but from some source theretofore existing in the plaintiff's system, was not remote but could be drawn from substantial evidence in the record. On the evidence of the plaintiff's medical history the jury was not required to but could reasonably conclude that the prior infection, and not any negligent act on the part of the defendant, was the proximate cause of the trouble. The inference based on common knowledge is at the root of the res ipsa loquitur doctrine. Before it could be drawn under the facts of this case the jury would have to reject the hypothesis that the plaintiff's prior condition was the proximate cause. [Citations.]"

Neither party cites, nor has any case been discovered, in which res ipsa loquitur has been held applicable where a patient dies during a minor operation while under the influence of an anesthetic. Defendant hospital urges that Ybarra v. Spangard (1944), supra, 25 Cal.2d 486 , is distinguishable in that there the plaintiff, while unconscious during an appendectomy, received a traumatic injury to his shoulder--a healthy part of the body, not the subject of treatment nor within the area covered by the operation--which a layman could say would not have occurred in the absence of negligence, whereas here expert medical testimony is necessary to determine whether or not negligence caused the patient's death during the course of the operation.

Plaintiff argues, however, that expert testimony was relied upon by the plaintiff in the Ybarra case, to establish an essential link in the chain of causation, and quotes from this court's summary of the evidence, as follows (p. 488 of 25 Cal.2d): "Plaintiff also consulted Dr. Wilfred Sterling Clark, who had X-ray pictures taken which showed an area of diminished sensation below the shoulder and atrophy and wasting away of the muscles around the shoulder. In the opinion of Dr. Clark, plaintiff's condition was due to trauma or injury by pressure or strain, applied between his right shoulder and neck."

"Plaintiff was also examined by Dr. Fernando Garduno, who expressed the opinion that plaintiff's injury was a paralysis of traumatic origin, not arising from pathological causes, and not systemic, and that the injury resulted in atrophy, loss of use and restriction of motion of the right arm and shoulder."

In the instant case there was no suggestion at the trial that plaintiff's son died as the result of a preexisting condition, whether pathological or systemic in nature. The expert evidence is to the contrary and in this respect is wholly uncontradicted. It shows that, except for infected tonsils and adenoids and a slight temperature due to such infection, the child was normal and healthy, and that the tonsillectomy was not a major operation nor performed as an emergency. Dr. Nellie Null, as heretofore stated, testified that in her forty years of practice she had performed "hundreds of these tonsillectomies" and that this was "the first case in which a death has ever occurred in one of" them.

[5] Under the circumstances shown we hold tenable plaintiff's position that the evidence prima facie establishes, in the absence of explanation, that "the child's death was due to something which ordinarily does not occur in the absence of negligence, that it was caused by an agency or instrumentality within the control of defendants, and that it was not due to any [legally material] voluntary action or contribution on the part of either plaintiff or the child," and, consequently, that the res ipsa loquitur instruction was properly given. [6] Since it was properly given as to all defendants, it is not made erroneous as to defendant hospital by the fact that the jury accepted the explanatory evidence as being sufficiently exculpatory in respect to the doctors but not to the hospital.

A review of other instructions criticized by defendant hospital discloses no prejudicial error.

The judgment is affirmed.

Gibson, C.J., Shenk, J., Carter, J., and Spence, J., concurred.

TRAYNOR, J.

I dissent.

The doctrine of res ipsa loquitur is not applicable unless the accident is of a kind that ordinarily does not occur in the absence of someone's negligence. (See Prosser, Torts, p. 295.) Since the determination whether the accident is of that kind is usually made in the light of common experience and since medical knowledge is not within common experience, the doctrine of res ipsa loquitur is usually held inapplicable in cases of malpractice. It can be invoked, however, "where a layman is able to say as a matter of common knowledge and observation that the consequences of professional treatment were not such as ordinarily would have followed if due care had been exercised." (Engelking v. Carlson, 13 Cal.2d 216, 221 [ 88 P.2d 695 ].)

In Escola v. Coca Cola Bottling Co., 24 Cal.2d 453 [ 150 P.2d 436 ], the court recognized that the doctrine may apply even though expert testimony is necessary to establish that the accident was of a kind that ordinarily does not occur in the absence of someone's negligence. The court noted that in Honea v. City Dairy, Inc., 22 Cal.2d 614 [ 140 P.2d 369 ], it refused to take judicial notice of the technical practices of the bottling industry and therefore could not determine whether it could reasonably be concluded that a defect in a bottle was more probably than not the result of negligence. In the Escola case, however, there was expert testimony as to such practices and on the basis of that evidence the court held it could reasonably be concluded that it was more probable than not that the bottle exploded as the result of negligence. Thus, while ordinarily the occurrence of an accident is not in itself evidence of negligence, it may be evidence thereof when it can be said in the light of common experience that the accident would not ordinarily have occurred in the absence of negligence, or when experts in the field are able to conclude on the basis of their knowledge and experience that there is a balance of probabilities in favor of negligence as the cause.

There is no valid objection to permitting proof of negligence in malpractice cases by such circumstantial evidence. The law requires that physicians and surgeons shall " 'possess and exercise that reasonable degree of skill, knowledge, and care ordinarily possessed and exercised by members of their profession under similar circumstances.' " (Sinz v. Owens, 33 Cal.2d 749, 753 [ 205 P.2d 3 , 8 A.L.R. 757 ].) Experts may testify that the procedure followed by the defendant fell short of that commonly employed, and thus provide direct evidence of negligence. Apart from the fact that a certain procedure failed, however, what actually happened is often in doubt. In such situations experts may know as laymen would not, that it is more probable than not that the accident was the result of negligence. If they so testify the jury may properly be instructed that if they find on the basis of the expert testimony that ordinarily an accident of the kind in question does not occur in the absence of negligence, they may infer that the particular accident was the result of negligence.

In the present case a child died on the operating table during a tonsillectomy. An instruction on the doctrine of res ipsa loquitur would be proper only if it may be said in the light of common experience that such deaths do not ordinarily occur in the absence of negligence or if medical experts had testified to that fact. Neither basis for the application of the doctrine is present. Common experience teaches only that ordinarily persons do not die during the course of minor operations. In the rare cases where deaths occur the layman is without knowledge or experience to weigh the probabilities of and against negligent conduct as the cause of death. It has therefore generally been held that the doctrine does not apply when a patient dies under anesthesia during a minor operation such as a tonsillectomy or tooth extraction. (Mitchell v. Atkins, 36 Del. (6 W.W.Harr.) 451 [ 178 A. 593, 595 ]; Dolan v. O'Rourke, 56 N.D 416 [ 217 N.W. 666, 668 ]; Johnson v. Arndt, 186 Minn. 253, 257 [ 243 N.W. 67 ]; Loudon v. Scott, 58 Mont. 645, 656 [ 194 P. 488 , 12 A.L.R. 1487 ]; see, also, Nemer v. Green, 316 Mich. 307 [ 25 N.W.2d 207 ]; Eggert v. Dramburg, 197 Wis. 153 [ 221 N.W. 732 ]; anno's., 12 A.L.R. 1493 ; 162 A.L.R. 1265 , 1282-1284.)

There was nothing in the expert testimony relied upon in the majority opinion to support a conclusion that ordinarily deaths do not occur in the course of tonsillectomies in the absence of negligence. Dr. Null testified that she had performed "hundreds of these tonsillectomies" and that this was "the first case in which a death had ever occurred." Her testimony establishes only that such accidents are rare; it was silent on the question as to what are the probable causes when such deaths do occur. On the other hand, there was evidence that all anesthetics are dangerous; that the hazards of anesthesia are so well known to the medical profession that many of its members have specialized in that field; and that it is always a delicate procedure to produce anesthesia.

From the foregoing expert testimony the jury would be warranted in concluding, not that an anesthetic death was more probably than not the result of negligence, but that it resulted from unavoidable hazards attendant upon any anesthetization. (See Loudon v. Scott, 58 Mont. 645, 656 [ 194 P. 488 , 12 A.L.R. 1487 ].) Nevertheless the majority opinion holds that it was proper to instruct the jury that as a matter of law the occurrence of the accident gave rise to an inference of negligence against all the defendants, an inference that could be overcome only by affirmative evidence on their part explaining the cause of death or showing that it could not have occurred from any cause that due care on their part might have prevented.

By approving the instruction, the court in effect holds that solely because an accident is rare it was more probably than not caused by negligence. There is a fatal hiatus in such reasoning. The fact that an accident is rare establishes only that the possible causes seldom occur. It sheds no light on the question of which of the possible causes is the more probable when an accident does happen. Since common knowledge and experience shed no light on this question when a death occurs during the course of a tonsillectomy and since the record is devoid of expert testimony bearing on the subject, the doctrine of res ipsa loquitur is not applicable. I would therefore reverse the judgment.

Edmonds, J., concurred.

[fn. 1] 1. Named in plaintiff's complaint (second amended) as "Franklin Hospital."

[fn. 2] 2. Also named as a defendant, but not served, was a nurse-anesthetist.

[fn. 3] 3. That section provides: "Where ... the allegation of the claim ... to which the proof is directed, is unproved, not in some particular or particulars only, but in its general scope and meaning, it is not to be deemed a case of variance ... but a failure of proof."

[fn. 4] 4. The instruction reads as follows: "I instruct you that this is a case in which the doctrine of res ipsa loquitur, that is to say, the thing speaks for itself, is applicable to the Franklin Hospital, Dr. Nellie Null and Dr. John Null, and the defendants will not be held blameless except upon a showing either '(1) of a satisfactory explanation of the accident, that is, an affirmative showing of a definite cause for the accident, in which cause no element of negligence on the part of the defendant inheres, or (2) of such care in all possible respects as necessarily to lead to the conclusion that the accident could not have happened from want of care, but must have been due to some unpreventable cause, although the exact cause is unknown. In the latter case, inasmuch as the process of reasoning is one of exclusion, the care shown must be satisfactory in the sense that it covers all causes which due care on the part of the defendant might have prevented.' "