(dissenting) :
I agree the failure to withdraw Jackson’s statement from the jury’s consideration was error. I do not view it under *174the circumstancse here as reversible error. The possibility or probability of prejudice is not the test to be applied when the evidence of guilt as in this case is overwhelming.
I would hold that the failure to withdraw the exhibit was harmless error in the sense that the evidence of guilt was so overwhelming that LaBarge could not have been prejudiced in the jury’s mind. Moreover appellant initially consented to the redaction of the statement and in any event its admission was discretionary.
“Whether a codefendant’s confession implicating a defendant can be effectively redacted so as to delete portions implicating the defendant and permit its introduction in a joint trial is for the trial court to determine.” U. S. v. Mitchell, 372 F. Supp. 1239 (D. C. N. Y.), appeal dismissed, Stans v. Gagliardi, 485 F. (2d) 1290 (2nd Cir. 1973).
Here, the testimony of the witness Gertrude Thomas was the State’s primary and principal evidence against appellant. She voluntarily confessed to her participation in her father’s murder as well as that of Jackson and LaBarge. She included appellant in the planning and placed him at the scene. She testified appellant and Jackson went into the house to rob her father while she remained outside in the car.
With the exception of Jackson’s confession placing appellant at the front door of the house during the commission of the crime, the confession was merely cumulative to Gertrude Thomas’ testimony and was harmless error. Chapman v. California, 386 U. S. 18, 87 S. Ct. 824, 17 L. Ed. (2d) 705 (1969).
As this Court stated in State v. Robinson, 238 S. C. 140, 151, 119 S. Ed. (2d) 671, 677 (1961) :
“[T]he only rational conclusion warranted by the evidence is that the accused is guilty, the judgement of conviction *175should not be set aside because of unsubstantial errors not affecting the result.”
In applying the harmless error test, we must consider the character and quality of the improper evidence as it relates to the other evidence bearing on the same issue, appraising the possible impact upon the jury of the inadmissible evidence. Schneble v. Florida, 405 U. S. 427, 92 S. Ct. 1056, 31 L. Ed. (2d) 340 (1972); State v. Atchison, 268 S. C. 588, 235 S. E. (2d) 294 (1977).
The U. S. Supreme Court in interpreting the Chapman rule in Harrington v. California, 395 U. S. 250, 89 S. Ct. 1726, 23 L. Ed. (2d) 284 (1969) affirmed the California Court of Appeals which held:
“[I]t is not reasonably probable that a result more favorable to defendant would have been reached had the [Bruton] rule been followed.”
The Court in Harrington held that the illegal evidence had been “cumulative” and that the State’s case against Harrington had been “so overwhelming.” This opinion suggests that Chapman now embraces an overwhelming evidence test for determining when an error is harmless.
I have no difficulty in concluding beyond a reasonable doubt that the alleged error did not influence the verdict.
I would affirm.