delivered the opinion of the Court.
Once again we are confronted with the question of racial discrimination in the selection of the grand and petit juries which have respectively indicted petitioners and found them guilty of the offense of murder. The claim is that Georgia's system of jury selection resulted in the systematic exclusion of Negroes from both the grand and petit juries in that its law required jury commissioners to select the names of prospective'jurors from the books of the county tax receiver which were maintained on a racially segregated basis. Ga. Code Ann. § 59-106. The grand jury question is raised in both these cases and we consolidated them for argument and do likewise on disposition.
No. 253 is an interlocutory appeal from a judgment denying petitioners’ claim as to the grand jury which in-
*547
dieted them. Georgia law authorizes such an appeal, Ga. Code Ann. § 6-701, and it was first perfected to the Supreme Court of Georgia which transferred it to the Georgia Court of Appeals. That court affirmed the denial of the claim of discrimination.
Following affirmance by the Georgia Court of Appeals of the interlocutory appeal, the trial court proceeded to try petitioners’ cases on the merits. After a challenge to the array of petit jurors was denied, petitioners were put to trial and were convicted. The Supreme Court of Georgia affirmed.
Whitus
v.
State,
I.
The petitioners have been, here twice before. They were originally convicted in 1960 and the Supreme Court of Georgia affirmed.
Davis
v.
State,
II.
After the Court of Appeals set aside the first convictions, Whitus v. Balkcom, ibid., the Superior Court of Mitchell County directed the jury commissioners for the county to revise the jury list. Georgia law requires that the six commissioners appointed by the Superior Court “select from the books of the tax receiver upright and intelligent citizens to serve as jurors, and shall write the names of the persons so selected on tickets.” Ga. Code Ann. § 59-106. They are also directed to select from this group a sufficient number, not exceeding two-fifths of the whole number, of the most experienced, intelligent, and upright citizens to serve as grand jurors, writing their names on other tickets. The entire group, excepting those selected as grand jurors, constitutes the body of traverse jurors. The tickets on which the names of the traverse jurors are placed are deposited in jury boxes and entered on the minutes of the Superior Court. Ga. Code Ann. §§ 59-108, 59-109. The veniremen are drawn from the jury boxes each term of court and it is from them that the juries are selected.
The State admits that prior to 1965, the tax return sheets furnished by the State Revenue Department, Ga. Code Ann. § 92-6302, were white for white taxpayers and yellow for Negro taxpayers. The 1964 tax digest, and all digests prior to 1964, were made up from these segregated *549 tax returns. Furthermore, the jury lists for each county are required by law to be made up from the tax digest. Ga. Code Ann. § 59-106. The State further admits that the “revised” jury list from which both the grand and petit juries serving in these cases were selected, had been made up by reference to the old jury list, which the Court of Appeals had condemned, and the 1964 tax digest, which had been prepared from the white and yellow tax return sheets of that year. However, the jury commissioners did not use the 1964 tax returns themselves, nor the 1965 tax digest which had not yet been made up. The tax digest appears to have been in one volume but was segregated into two sections — one for white and the other for Negro taxpayers. The Negroes whose names were included in the tax digest were designated by a “(c)” being placed opposite their names as required by Ga. Code Ann. § 92-6307.
The three jury commissioners who appeared as witnesses testified that they were not aware of the letter (c) appearing after the names of the Negroes on the 1964 tax digest; that they never included or excluded anyone on the “revised” jury'list because of race or color; that they placed on the “revised” jury list those persons whom they knew personally from their respective communities; that there were around 600 selected; and that the “revised” list, which the commissioners themselves prepared, had no designation of race upon it.
III.
For over fourscore years it has been federal statutory law, 18 Stat. 336 (1875), 18 U. S. C. §243, and the law of this Court as applied to the States through the Equal Protection Clause of the Fourteenth Amendment, that a conviction cannot stand if it is based on an indictment of a grand jury or the verdict of a petit jury from which Negroes were excluded by reason
*550
of their race.
Strauder
v.
West Virginia,
It is undisputed that the “revised” jury list was made up from the 1964 tax digest, the old jury list and the personal acquaintance of the commissioners with persons in their respective communities. It is admitted that the old jury list had been condemned as illegal by the Court of Appeals when it reversed petitioners’ first convictions. It is conceded that 27.1%' of the taxpayers in the county are Negroes; that the county had a population in 1960 of 10,206 people over the age of 21 years, of whom 4,706 were male, 1 with 2,004, or 42.6%, of this latter number being Negroes; that 33 prospective jurors were drawn for grand jury service for the term of court during which petitioners were indicted, three being Negroes, of whom one actually served on the grand jury of 19 persons; that a venire of 90 persons was used for the selection of the petit jury which tried petitioners, of which number at least seven were Negroes; and, that no Negro was accepted on the petit jury.
*551 Furthermore, it is obvious that the 1964 tax digest was required to be made under the same segregated system as were the previous digests, and suffered the same deficiency. Indeed, the State employed the same procedure which it concedes resulted in discrimination in the petitioners’ first trial.
We believe that this proof constituted a prima facie case of purposeful discrimination. While the commissioners testified that no one was included or rejected on the jury list because of race or color this has been held insufficient to overcome the prima facie case. Norris v. Alabama, supra, at 598. The State also insists that the revision of the jury list made evidence of the former practice of exclusion irrelevant. However, as we have seen, this revision was suspect. At the least it was based on the old jury roll which had been specifically condemned by the Court of Appeals and the 1964 tax digest which was suspect because of the system by which it was required to be prepared. The Court of Appeals condemned this same system in reversing the original convictions.
We believe that the circumstances here are akin to those condemned in
Avery
v.
Georgia,
Under such a system the opportunity for discrimination was present and we cannot say on this record that it was not resorted to by the commissioners. Indeed, the disparity between the percentage of Negroes on the tax digest (27.1%) and that of the grand jury venire (9.1%) and the petit jury venire (7.8%) strongly points to this conclusion. 2 Although the system of selection used here had been specifically condemned by the Court of Appeals, the State offered no testimony as to why it was continued on retrial. The State offered no explanation for the disparity between the percentage of Negroes on the tax digest and those on the venires, although the digest must have included the names of large numbers of “upright and intelligent” Negroes as the statutory qualification required. In any event the State failed to offer any testimony indicating that the 27.1%. of Negroes on the tax digest were not fully qualified. The State, therefore, failed to meet the burden of rebutting the petitioners’ prima facie case.
It is contended by petitioners that in the event of a reversal of the decision below they should be set free rather than retried. This contention arises from language
*553
used by the Court of Appeals in reversing the original convictions. The court expressed its “present opinion that a period of eight months . . . will be sufficient to afford the State an opportunity to take the necessary steps to reindict and retry the petitioners.”
Whitus
v.
Balkcom,
“A prisoner whose conviction is reversed by this Court need not go free if he is in fact guilty, for Texas may indict and try him again by the procedure which conforms to constitutional requirements.”
See also
Patton
v.
Mississippi,
The judgments are, therefore, reversed for further proceedings not inconsistent with this opinion.
It is so ordered.
Notes
Women, while qualified to serve, are not compelled to serve and may be excused upon request. Ga. Code Ann. § 59-124.
While unnecessary to our disposition of the instant case, it is interesting to note the “probability” involved in the situation before the Court.
The record does not indicate how many Negroes were actually on the “revised” jury list of approximately 600 names. One jury commissioner, however, Said his best estimate was 25% to 30%, which is in close proximity to the 27.1% who were admittedly on the tax digest for 1964. Assuming that 27% of the list was made up of the names of qualified Negroes, the mathematical probability of having seven Negroes on a venire of 90 is .000006. See Finkelstein, The Application of Statistical Decision Theory to the Jury Discrimination Cases, 80 Harv. L. Rev. 338 (1966).
