OPINION
. Clеment Turner, appellee, was arrested and charged with criminal homicide in the death of Irwin Liggins, Jr., age 21 months. A non-jury trial was held before the Court of Common Pleas of Allegheny County. At the close of the prosecution’s case-in-chief, the trial judge sustained appellee’s demurrer to the evidence and discharged the appelleedefendant. The Commonwealth then appealed to this Court.
The test to be applied in ruling on a demurrer is whether, accepting as true all of the prosecution’s evidence and all reasonable inferences therefrom, it is sufficient to suppоrt a finding by the fact-finder that the defendant is guilty beyond a reasonable doubt.
Commonwealth v. Duncan,
Barbara Ann Miklos and the infant son, Irwin, were visiting appelleе, Ms. Miklos’ boyfriend, at his home in Pittsburgh. Early in the afternoon of June 10, 1979, the child began to misbehave by throwing some food. Appellee hit the child with his hand three or four times on the buttocks. When the child continued to misbehave, appellee escalated his discipline.
Later that afternoon, appellee beat the child about the legs and back with a belt which had a heavy buckle and a zipper pouch. Later still, after the boy spit some food on the floor, appellee again beat the child several times with the belt, finally stopping at Ms. Miklos’ insistence. After the child had placed his fingers near an electric fan, appellee beat him three or four times with his shoe, and apparently feeling this discipline insufficient to modify the young child’s behavior, appellee continued to beat him with a one-inch thick wooden stick against his buttocks, legs and *623 bare back. These multiple beatings raised numerous welts and bruises upon the child’s legs, buttocks, chest and back.
Ms. Miklos decided to return to her own hоme that night, and together with her son and appellee, walked to the bus stop. However, as she had missed the last bus, she returned with appellee to spend the night at his apartment. Ms. Miklos placed thе boy on the couch which was approximately 15 inches from the floor to the cushion, in appellee’s bedroom then went to bed at approximately 1:30 a. m. At that time, Ms. Miklos noticed no injuries оr marks on Irwin’s head or face.
At about 3:45 a. m., Barbara Miklos was awakened by appellee’s shouts that Irwin was not breathing. The child was lying on his back beside the couch. Ms. Miklos then noticed bruises on his head. Nеither paramedics nor hospital personnel could revive the child and he was pronounced dead at 4:12 a. m.
Death was attributed to cerebral edema, or a swelling of the brain. This was caused by neurogenic shock which resulted from blunt force injuries sustained by the victim. The autopsy revealed numerous injuries to the chest, buttocks, back, and legs, and at least fifteen separate areаs of contusions about the head of the child. Both the coroner, Dr. Steven Goldblatt and the chief forensic pathologist, Dr. Joshua Perper, testified that the child’s death was due to the combination оf the trunk and head injuries.
Pursuant to defense stipulation, evidence was received that blood stains found on the bedspread covering the couch cushions near where the child’s head lay as he slept, and a towel also found on the couch, were consistent with the blood of the victim. No blood stains were found elsewhere in the apartment.
Appellee demurred to the above evidеnce. The lower court granted the demurrer. The reasoning was two-fold: first, the court noted no direct evidence had been introduced to show that appellee struck the victim in the head, and found the circumstantial evidence was as consistent with *624 self-inflicted, accidental injury as it was with homicide. Second, the court found that the below-head injuries were not shown to have been, of themselves, the cause of death and therefore, in the absence of direct proof of above-head beatings by appellee, ruled that the requisite causation element had not been established beyond a reasonable doubt. We find the lower court’s reasoning faulty.
We are unable to comprehend, after examining the record how the lower court could conclude that the inferences arising from the evidence were as consistent with self-inflicted, accidental head injuries as they were with homicide. The reasonable inferences arising from the evidence overwhelmingly negate the possibility of accidental head injuries.
While Dr. Goldblatt acknowledged the remote possibility that a contusion to the child’s head could have been caused by a fall, he discounted that possibility. The following exchange took, place between the defense counsel (on cross-examination) and Dr. Goldblatt:
Q When you use the term blunt force injury, does that necessarily mean that one person striking another with an object or with their hand or something like that?
A No. One could get a blunt force injury from a fall.
Q You could fall and get a contusion?
A Striking something.
Q Are you able to state with some medical certainty the manner of blunt force injury which caused the injury to the head?
A Whether an object struck the head or whether the head struck an object?
Q Right.
A With a reasonable probability by the number and location of the blow, the only way it could happen, it’s mostly something striking the head. If it were to occur from the head striking the object, it would have to be on multiple occasions, repeated by at least from *625 the location of the injury. There would have to have —the head would have to strike something else probably at least ten different times, (emphasis added)
Notes of Testimony at 121-122.
In spite of this testimony, the trial court concluded “it is just as logical ... to beliеve that the child could have fallen off the couch and sustained the injuries to his head .... Likewise, it is just as logical to argue that this child could have gotten down off that couch, wandered around in this darkened, cluttered room and fallen over something, injuring his head.” These possibilities are fantasy — did the child fall off the 15 inch-high couch 10 times? Did he stumble and fall 10 times? Of course not. This Court has never required the Commonwealth tо disprove
every
possibility of accidental death or to prove that a hypothetical event did not take place.
Commonwealth v. Green, 477
Pa. 170, 175,
A very similar situation confronted this Court in
Commonwealth v. Paquette,
Admittedly, there were no eyewitnesses to the alleged beating; but the Commonwealth need not prove its casе directly. Circumstantial evidence can be as reliable and persuasive as eyewitness testimony. Commonwealth v. New,354 Pa. 188 ,47 A.2d 450 (1946); See I Wigmore § 26 (3d Ed. 1940).
[T]he medical evidence, offered by the Commonwealth, clearly establishes repeatеd and severe blows to the infant as being the origin of the factors causing death. Such a finding is completely incompatible with the defense’s theory that the injuries were either sustained by a fall from the cоuch to the floor or an epileptic fit while the child was in the bath. Considering the number of bruises, their *626 severity and their positioning about the head and the face, the factfinder was justified in rejecting the pоssibility of accidental or self-inflicted injury particularly when recognizing the mobility of a 6V2 month old baby.
Where, as here, an adult has sole custody of a child for a period of time, and, during that time the child suffers wounds which unquestionably are neither self-inflicted nor accidental, the evidence is sufficient to allow a jury to infer that the adult inflicted the wounds, (citations omitted) Id.,451 Pa. at 255 ,301 A.2d at 839 .
In the instant case, the prosecution’s еvidence demonstrated that Barbara Miklos went to sleep at 1:30 a. m., at which time Irwin’s head was unbruised, and awoke to appellee’s screams at 3:45 a. m., at which time the child’s head had sustained sevеre injury. Thus, the sole custody inference is applicable to appellee. Here, as in
Paquette,
the evidence is sufficient to allow an inference that appellee inflicted the head wounds.
1
See also, Commonwealth
v.
Blevins,
Notes
. The trial court distinguished Paquette on the basis that Irwin, a 21 month old child, was much more mobile than the 6V2 month old viсtim in Paquette and, thus, the likelihood of self-inflicted accidental injury was much greater in this case. We find this distinction insignificant in view of the number and severity of the blows to Irwin’s head, as well as the fact that, according to Dr. Goldblаtt, the injuries to the body alone would have initiated the shock, which shock would have decreased Irwin’s mobility.
. As we find the evidence sufficient to establish beyond a reasonable doubt that appellee caused the injuries to
both
the child’s head and his body, there is no need to address Judge Lewis’ causation analysis in detail. However, it appears this reasoning was also erroneous. Even though the lowеr court’s opinion points out that “both doctors testified that it was the combination of all the injuries that caused the child to go into shock and eventually die”, the proof of causation was insufficiеnt because the doctors could not say “with medical certainty that the blows to the body alone ... were a substantial factor in bringing about the victim’s death.” We think this is an inaccurate reading of the doctors’ testimony. While purporting to view the medical testimony “as a whole”, the lower court appeared, instead, to focus on one instance where Dr. Goldblatt testified he
*627
didn’t know how to qualify the wоrd “substantial”. (“The action of defendant must constitute a direct and substantial factor in causing the death of the victim ....
Commonwealth v. Johnson,