UDALL, Justice.

This is a certiorari proceeding to review an award of The Industrial Commission of Arizona denying death benefits (A.R.S § 23-1046) to petitioner, Grace Sheldon Revles, widow of a deceased employee of W.L. Eyer and Alice Pauline Eyer, d.b.a. Coolidge Sand and Rock Company. The employer makes no appearance; it carried insurance with the State Fund.

The widow's claim for death benefits was denied by the Commission and she timely filed for rehearing. A rehearing was held, after which the Commission entered its decision affirming the previous findings and award, i.e., denying relief to the widow. We granted certiorari.

The undisputed facts are these: John Vernon Revles, the decedent, was an employee of the Coolidge Sand and Rock Company (hereafter referred to as the employer). On June 27, 1958, the date of the beginning of the events in question, he was 55 years old and had worked for the same employer for some 15 years. His work classification was "crushing plant operator", i.e., he was an operating engineer whose duties involved the operation and maintenance of the machinery and equipment used to crush and grade gravel for commercial use.

During a lifetime of hard work he had experienced several injuries — broken bones, the loss of an eye, and, as a result of a compensated industrial accident in California, he had undergone the amputation of his right arm above the elbow. In 1951, while living in Coolidge, he was hospitalized with what was believed to have been a stroke. In spite of all these misfortunes, he was a strong and apparently healthy man, and a good worker. His superintendent at Coolidge said he was "one of the best". Revles lived in a ranch house outside of Coolidge with his wife, the petitioner. The evidence indicates that he was a vigorous man who enjoyed working around his home in the evenings and on off days, particularly in his garden.

On June 27, 1958, Arizona was suffering through an unusually early hot spell. On that date the temperature in Coolidge reached 114 degrees. Decedent spent several working hours on that afternoon repairing and patching a "trammel screen", described as a long barrel or tumbler, approximately 4 to 5 feet in diameter and 27 feet long, made up of several sizes and layers of heavy steel wire screen. Revles worked inside the tumbler, cutting and placing the patches. He was cramped and worked in a crouching or recumbent position. Using a cutting torch, he would cut the steel wire patches to the proper size and then weld them in place with an electric welder. Both his working companion, Eleasar Marquez, and his superintendent, Ransom Stewart, who had performed the same job on occasion, testified that, because of the close quarters and the heat retained and reflected by the steel tumbler and the heat generated by the welding and cutting, it was considerably hotter where Revles was working than the 114 degree outside temperature.

Some thirty minutes to an hour before 4:30 p.m., the normal quitting time, Revles announced that he was getting sick, that it was "just too hot", and that he was going home. Marquez testified:

"He told me he was sick, can't make it.

* * * * * *

"He looked — he got too hot, you know."

It is not clear whether Revles left the premises before 4:30, since he did not follow the usual procedure of going by the office and filling out his time card. The employer's report of injury, which was not filed until September 17, 1958, states simply that at 3:30 p.m., June 27, 1958, "Employee working as crushing plant operator in 114° temperature, felt extremely ill, nauseated and dehydrated. Quit work at 4:30 p.m. and went home." It does appear, from the testimony of decedent's daughter, that, feeling sick, he arrived home early that day.

Neither decedent's family nor his associates at work had heard him complain of illness prior to the date and the events above described. From that day forward, however, Revles never regained his health and strength, and on July 26, 1958, — some thirty days thereafter — he died. As to this latter period, the following evidence was adduced: On the morning of June 28, 1958, the day following the initial illness, Revles telephoned Superintendent Stewart and said that he would not be to work, that he was ill. Stewart reported, "He said he didn't feel like working, that his — he felt like his stomach was all full, and he could hardly breathe." When asked whether Revles had said anything about being injured on the job, Stewart replied, "The only thing he said to me was that he thought it might be too hot * * * that he thought he had gotten too hot" on the job.

During this last month before his death Revles led the life of an ambulatory semi-invalid. He spent much of the time in bed, or sitting or lying around the house in his pajamas. Occasionally his wife would drive him into town with her when she went shopping. On such occasions she would leave him at the home of one of their two daughters and then pick him up when she was ready to go home. Both daughters testified that he seemed to have undergone a sudden and profound change, that he was always tired and listless. Revles did not drive the car, and, except for these excursions into Coolidge, rarely left the house. He stopped working in his garden. Through all this period he refused to see a doctor, insisting that he would soon get his strength back, that it was just a matter of time. He died at the end of the month.

The death certificate, prepared by Dr. David S. Burgoyne, who did not see decedent until after the death had occurred, attributed his death to two causes; the primary cause was stated to be "congestive heart disease"; the secondary or antecedent cause, "arteriosclerotic heart disease." As to each of these factors the examining physician gave his estimate of the time intervening between the onset of the condition and the death. For the former, he estimated two weeks; for the latter, several years.

Shortly after Revles's death, an autopsy was performed by Dr. George Scharf, of Phoenix, a specialist in pathology and a diplomate of the American Board of Pathology. Dr. Scharf submitted his report and also was examined at the Industrial Commission hearing. Based on this autopsy and on decedent's known physical history, the pathologist made certain findings, the gist of which is as follows: the arteries of decedent's heart were so clogged or "hardened" that very little room was left for passage of the vital blood (coronary arteriosclerosis). In some areas the coronary arteries had been entirely stopped up (occluded). In the left side of the heart one area had — by the above process — been entirely cut off from blood and oxygen and had died (a myocardial infarction). In this same area a thick clot had recently formed on the heart wall (mural thrombus). The heart was considerably enlarged (cardiac hyperthrophy, 425 grams).

Upon these facts, Dr. Scharf constructed his theory of how Revles had died, viz.: Revles had a long standing heart condition, a coronary artery disease. However, up to June 27, 1958, his heart had somehow compensated for its defects, i.e., it had managed to do the work required of it in spite of the obstacles. From that date on, the heart failed to compensate, or, to use Dr. Scharf's term, a "decompensation" set in. The result of this failure of the heart to overcome its deficiencies was a condition known as "coronary insufficiency". Symptoms of this deterioration were "dyspnea" and "edema", i.e., shortness of breath, and retention of fluids in the heart and elsewhere. Without medical treatment this condition ran on unchecked, ultimately causing death.

Dr. Scharf also gave his opinion as to what precipitated this "decompensation". The following appears in his autopsy report under the heading "Comment":

"From the condition of the heart and coronary blood vessels, and the clinical story as given to the prosecutor, we would have to conclude that the patient died of cardiac failure secondary to coronary insufficiency. The question arises as to whether there might be any relationship between the precipitation of this failure and the conditions existing immediately before the patient became ill. Again, referring to the condition of the heart and its blood vessels, we would have to conclude that the circulation of the heart was probably in a precarious state of balance, and that any conditions calling for additional blood flow, probably would result in acute coronary insufficiency. As noted in Bulletin No. 105 and the U.S. Dept. of Labor (1949), Dr. Arthur M. Master, reminds us that the first cause of acute coronary insufficiency would be anything that increases cardiac work. Among the variety of conditions listed are included extremes of temperature. We would therefore have to conclude that although the extremely high temperatures under which the patient was working certainly did not cause the basic pathologic changes of the heart, it is very likely that the same conditions were strongly influential in precipitating the terminal cardiac failure."

Of course Dr. Scharf did not insist that the theory set out above was the only possible explanation of the death. Under examination by the Commission's referee on rehearing, the following exchange took place:

"Dr. Scharf: With regard to the first part of your question as to what might have precipitated this, the autopsy itself gave us no direct information as to a precipitative factor, as far as the conditions under which this man lived.

"Q. You mean by that that you have no indication what precipitated it? A. Not from the autopsy. We have to put the autopsy together with the information given to us about the clinical state of this man.

"Q. By `clinical state,' now, what are you referring to? A. We mean the history in particular of his previous health, of his apparently becoming ill on a particular day under certain conditions, and of his condition after that illness. That is, that acute illness up to the time of his death.

"Q. And you say then that given that clinical history, it is possible that this precipitated the — A. Given that particular clinical history and knowing no other facts, it seems a good probability that the conditions under which he was working may very well have precipitated the cardiac decompensation that we have already referred to. We cannot prove it. It is a matter of opinion.

"Q. And given other facts or given additional facts, your opinion might change; is that correct? A. My opinion might very well change.

"Q. Your opinion is based on this clinical history without any other evidence as to other activities and so forth? A. That is correct.

"Q. It might have been something else; there might have been some strain at home, for instance, during the month after? If that was the fact, assuming that, anything could have caused it, any straining? A. We won't say `anything', but anything that would have caused additional strain on the heart, yes."

This, then, is the evidence which was before the Commission. None of it was controverted. No other medical evidence was introduced. No other incident was found which might have triggered decedent's decline. Questions posed by the referee failed to elicit any indication of some other strain — his family testified that he did not drink; Dr. Scharf testified that there was no sign of an infection, either in the heart or in the stump of decedent's amputated arm, which might have turned the balance toward decompensation.

Based upon the undisputed evidence summarized above, the Commission, in denying petitioner death benefits, entered the following findings:

1. That John Vernon Revles did not sustain an injury by accident arising out of or in the course of his employment on June 27, 1958.

2. That said John Vernon Revles died on July 26, 1958, as a result of Congestive Heart Disease due to Arteriosclerotic heart disease.

3. That the medical evidence indicates that the disability from which John Vernon Revles died was not the result of any injury by accident on June 27, 1958.

The crucial findings are those numbered 1 and 3, supra. We shall consider each in the light of the undisputed facts, the medical evidence, and the relevant decisions of this court.

1. Injury occurring on June 27, 1958. It is the settled law of Arizona that injury resulting from excessive heat can be compensable as "arising out of and in the course of employment" within the terms of A.R.S. § 23-1041, subd. A. Vukovich v. Industrial Commission, 76 Ariz. 187, 261 P.2d 1000; Watson v. Sam Knight Mining Lease, 78 Ariz. 114, 276 P.2d 536. In the latter decision we summarized the generally accepted law on this subject in declaring this to be the rule:

"* * * If it is shown that the conditions of employment result in subjecting an employee to exposure to heat or cold in excess of that of the general public in the same vicinity, and he is injured thereby, the injury is compensable." 78 Ariz. 120, 276 P.2d 539.

Numerous decisions from other jurisdictions to the same effect are collected in 1 Larson on Workmen's Compensation § 38.40.

The question, then, is simply one of fact, i.e., whether the excessive heat on June 27, 1958, "injured" the decedent, or, more particularly, whether on that date it was the operative cause of any illness or disability. The factual evidence in this regard has been set out above. None of it is disputed; in fact, several independent sources concurred in describing the symptoms of the malaise suffered by decedent and in attributing it to the severe heat in which he was working; and all sources agreed that the conditions under which he worked on that date were such as to subject him to heat far in excess of that suffered by the general public in the same vicinity. We find no evidentiary support for the Commission's finding to the contrary. On this same state of facts, we hold that the Commission erred in failing to find a compensable injury (the compensation to be determined on the basis of the employee's disability, if any, traceable to the injury).

2. Causal relation between injury and death. In considering the propriety of the Commission's finding number 3, several well established legal principles should be kept in mind. First, an injury or death does not become noncompensable merely because a pre-existing disease or condition of the employee has been a contributing factor thereto. As we said in Murray v. Industrial Commission, 87 Ariz. 190, 349 P.2d 627:

"* * * (I)n the field of Workmen's Compensation, the employer takes his employee as he is. In legal contemplation, if an injury, operating on an existing bodily condition or predisposition, produces a further injurious result, that result is caused by the injury." 349 P.2d 633.

Second, an industrial injury need not be the sole cause of death, in order to entitle decedent's dependents to death benefits, as long as it appears that the injury contributed to and accelerated the inevitable. The following statement from 1 Schneider on Workmen's Compensation § 138 was adopted by this court in Aluminum Company of America, on Behalf of Defense Plant Corp. v. Industrial Commission, 61 Ariz. 520, 152 P.2d 297, 300, and reaffirmed in Horn v. Industrial Commission, 68 Ariz. 323, 205 P.2d 1198:

"The courts, consistent with the theory of the Workmen's Compensation Acts, hold with practical uniformity that, where an employee afflicted with disease receives a personal injury under such circumstances as that he may have appealed to the act for relief on account of the injury had there been no disease involved, but the disease as it in fact exists is by the injury materially aggravated or accelerated, resulting in disability or death earlier than would have otherwise occurred, and the disability or death does not result from the disease alone, progressing naturally, as it would have done under ordinary conditions, but the injury aggravates and accelerates its progress, materially contributes to hasten its culmination in disability or death, there may be an award under the Compensation Acts."

Cf. Pierce v. Phelps Dodge Corp., 42 Ariz. 436, 26 P.2d 1017, overruled in Matter of Mitchell, 61 Ariz. 436, 150 P.2d 355.

Third, when medical opinions, based on matters peculiarly within the realm of scientific knowledge, are uncontroverted, such opinions cannot be arbitrarily rejected by the Commission. As we said in Jones v. Industrial Commission, 81 Ariz. 352, 306 P.2d 277:

"* * * We submit that the question of the causal relationship between the accident and death was necessarily within the singular knowledge of medical experts; in such case their findings are conclusive upon the commission. See, Ison v. Western Vegetable Distributors, 48 Ariz. 104, 110, 59 P.2d 649; Craig v. De Berge, 67 Ariz. 168, 172, 193 P.2d 442; Caekos v. Stanley Fruit Co., 55 Ariz. 72, 98 P.2d 471. The commission is not allowed to substitute its judgment on matters lying exclusively within the field of medical science. * * *" 81 Ariz. 358, 306 P.2d 281.

For a more recent dissertation on the matter of "proximate causation" see the excellent opinion, Murray v. Industrial Commission, 87 Ariz. 190, 349 P.2d 627.

Having these principles before us, we may now apply them to the facts in this case. As noted above, the Commission found:

"That the medical evidence indicates that the disability from which John Vernon Revles died was not the result of any injury by accident on June 27, 1958."

We have already determined, as a matter of law, that decedent did suffer an "injury by accident" on that date. The remaining question is whether "the medical evidence indicates that the disability from which [he] died was not the result" of that accident.

In several earlier cases we have had before us the question whether death by heart failure was "caused" by an industrial injury in the terms of the Workmen's Compensation Act. See, Jones v. Industrial Commission, supra; Gronowski v. Industrial Commission, 81 Ariz. 363, 306 P.2d 285; Hartford Accident Indemnity Co. v. Industrial Commission, 38 Ariz. 307, 299 P. 1026; Phoenix Bakery v. Industrial Commission, 78 Ariz. 188, 277 P.2d 745; Phelps Dodge Corp., Douglas Reduction Works v. Cabarga, 79 Ariz. 148, 285 P.2d 605; and Byers v. Industrial Commission, 80 Ariz. 406, 298 P.2d 1039. In the two most recent of the above cited decisions (the Jones and Gronowski cases, both decided unanimously in January, 1957) this Court considered two such deaths and found one compensable and one not. An analysis of the distinctions between those cases will serve as a yardstick in the resolution of the matter now before us. In the Jones case we held:

"* * * It is obvious * * * that no hard and fast rule can be laid down that governs all situations as each case must be determined upon its own peculiar facts. However, it is firmly established that for a case to be compensable, there must be a recognizable causal connection between an employee's employment and the accidental injury." 81 Ariz. 356, 306 P.2d 279.

In both the Jones and Gronowski cases the Industrial Commission held that the facts did not justify a finding that this requisite causal connection existed. In both opinions we recognized the well settled principle that the Commission, as trier of the facts, must be affirmed if its conclusion is reasonably supported by the evidence. Yet in the latter case, the Commission was upheld; while in the former, its award denying compensation was set aside. What, then, are the factors which distinguish the two decisions?

In each case the pivotal question involved the propriety of the interpretation which the Commission put upon the expert medical testimony. The Jones decision summarized this evidence as follows:

"It should be noted that the testimony of Dr. Marlowe is the only medical evidence in the case. As is shown by the above questions, the doctor did not waiver in his opinion or conclusion that the physical exertion of decedent immediately prior to his demise contributed to the cause of his heart attack and death, i.e., that it aggravated his disease and hastened or accelerated death. * * *" 81 Ariz. 358, 306 P.2d 281.

In Gronowski, this analysis of the evidence was given:

"We have attempted heretofore to fairly set forth the important medical evidence on the question presented. Dr. Barger's testimony is impregnated with substantial uncertainty. It is apparent that he was reluctant to even say decedent's efforts, assuming them to have been strenuous, were probably the precipitating cause. The value of this statement is weakened when taken in connection with the testimony of decedent's co-worker who did not remember that decedent was subjected to any unusual strain. The doctor also said he would hate to say it was probable and that it might have had little or nothing to do with causing death. We think taking his testimony as a whole that it is susceptible of an interpretation that the doctor is speaking more of possibilities than probabilities. Cf. O'Brien v. Industrial Commission, 90 Utah 266, 61 P.2d 418. Clearly, we cannot require the commission to find a fact on possibilities. Cross v. Industrial Commission, 81 Ariz. 222, 303 P.2d 710." 81 Ariz. 366, 306 P.2d 287.

It is our considered opinion that the facts of this case more nearly resemble those of the Jones case than those set out in Gronowski. Here the expert opinion, although somewhat equivocal, strongly supported the proposition that the excessive heat suffered by decedent on June 27, 1958, was a precipitative factor in his terminal cardiac failure. The opinion advanced by Dr. Scharf was not stated in terms of absolute certainty; however, he did take a position, and — on the basis of the known facts — he was willing to stand by it. His testimony indicates that his opinion was predicated upon certain given facts, i.e., the actual pathological condition of decedent's heart as revealed by his autopsy, the "sudden onset" of the decompensation cycle on the particular date in question, the heat and exertion which he underwent on that date, the symptoms of his condition subsequent thereto, and the lack of any evidence that some other external cause triggered the fatal decline. None of these underlying facts was questioned in any material respect. Under such circumstances the Commission was not justified in ignoring the only medical opinion before it. Jones v. Industrial Commission, supra. We hold that, on this record, the Commission's finding number 3 was not reasonably supported by the evidence.

Award set aside.

PHELPS, JOHNSON and BERNSTEIN, JJ., concur.