It is asserted that decedent suffered two industrial accidents in the month of October, 1946. He died at his home as the result of a coronary thrombosis on August 25, 1950, nearly four years later. This appeal is from an award of death benefits, and the question is whether there is substantial evidence in the record to sustain a finding that the death was causally related to the incidents to October, 1946.
Claimant was employed as a general baker. On October 3, 1946, a heavily loaded shelf started to fall and the decedent with others held it up. Following that he felt pain, but continued to work. On October 24, 1946, decedent was lifting a bag of flour and felt pain, causing him to drop the bag. He again continued to work. On October 26, 1946, decedent suffered a heart attack and was attended by his family physician, who testified in connection with a claim for disability compensation (but not on the death claim), that decedent suffered a coronary occlusion on October 24,1946. The family physician called into consultation a heart specialist, Dr. Crawford, who examined decedent and took an electrocardiograph on October 30, 1946, and continued to see him periodically thereafter. Dr. Crawford found that decedent had been suffering from coronary sclerosis prior to October 3,1946, and that he had been suffering therefrom *574for several years. Decedent returned to work on January 18, 1947, and worked until September 24, 1948, a period of nearly two years. On September 24, 1948, decedent arose in the morning feeling pains in his chest; he had felt ‘ ‘ sick ’ ’ the night before; he had done ‘ ‘ the usual work ’ ’ the day before, and he “ didn’t do anything ” that brought on the pains. Another electrocardiograph was taken by Dr. Crawford on February 15,1949, which showed no essential changes. On August 25,1950, decedent suddenly went into a state of collapse and died. Appellants have paid disability compensation to claimant during his lifetime, but this does not preclude them from contesting the cause of his death, and we must examine the medical testimony on that subject.
Dr. Crawford, a specialist who first attended decedent a few days after the above incidents in October, 1946, and continued to be familar with his case thereafter, testified that decedent’s death was not related to the original incidents, saying, in part: ‘ ‘ it would seem to me that was due to coronary sclerosis which he had, not due to his work, which is a degenerative disease, and that he went on with very marked coronary artery disease, and that his death was really due to the advancement of his coronary artery disease which precipitated the final occlusion but that it Avas not due to the original accident. His coronary arteries Avere getting narrower and narrower and finally he just had a coronary occlusion which killed him. ’ ’
Dr. Reiseh, another cardiologist, testified that decedent’s death was not related to the incidents of October, 1946; that decedent’s chest pains on September 24, 1948, when he stopped work, “ occurred on arising in the morning and most certainly had no possible connection with his labors ”. It is undisputed in the record that decedent had previously suffered from coronary sclerosis, and Dr. Reiseh testified that death was caused by the sclerosis, “ a natural, progressive degenerative disease ”.
It is true that the record contains a rather equivocal opinion by another doctor, not a specialist, who did not see decedent until August 23, 1949, that decedent’s death was related to the instances of October, 1946. Another doctor, who had never seen decedent, testified hypothetically to the same effect.
We are aware of the fact that there is frequently a wide variance in medical opinions, and that ordinarily a variance of opinion creates a question of fact. Here, however, two heart specialists, one of whom was familar with the case from the outset Avith opportunity for personal observations and who was called into the case by decedent’s physician, testify with positive*575ness and with sound explanations that decedent’s death was due to the pre-existing coronary sclerosis, which concededly was present. The surrounding facts and circumstances strongly corroborate them. Decedent worked for nearly two years and lived for nearly four years following the alleged incidents and suffered other attacks when not at work. There must- be a stopping point somewhere. We do not believe that bare medical opinion contrary to all the facts and circumstances and contrary to the opinion of those experts in the best position to know is enough to satisfy the substantial evidence rule. Evidence is not substantial if it consists merely of speculation. “ Viewed in the light of the record as a whole ” we do not think there is substantial evidence to sustain the finding that death was causally related. (Matter of Kopec v. Buffalo Brake Beam-Acme Steel & Malleable Iron Works, 304 N. Y. 65; Matter of McCormack v. National City Bank, 303 N. Y. 5; Matter of Peploe v. Burns Bros., 281 App. Div. 134.)
The award should be reversed and the claim dismissed, without costs.
Poster, P. J., Bergan and Imrie, JJ., concur with Coon, J.; Halpern, J., dissents, and votes for affirmance.
Award reversed and the claim dismissed, without costs.