(concurring):
I concur fully in the result, both as to the reversal of the convictions and the dismissal of the indictments. In order to avoid a 4-to-4 deadlock and the unfortunate predicament of a Court being unable authoritatively to decide a case, Carter v. United States, 5 Cir., 1963, 325 F.2d 697 (en banc), cert. denied, 1964, 377 U.S. 946, 84 S.Ct. 1353, 12 L.Ed.2d 308, I cast my vote, weak as it is, for the thesis developed at such length that Congress intended the 1957 Amendments to be, not the minimum, not the maximum, but the sole standard of jury selection. But having so declared, I am not at all certain just what is decided.
I.
I characterize as “weak” my concurrence concerning congressional intention —what Congress thought it was doing, not what the law reads into what Congress was bound to think it was doing. I do this because too much is made out of too little. Almost as much as Judge Bell, I think the materials of legislative history are so very, very thin on which to raise such a mighty edifice.
But accepting this with all of its intrinsic infirmities, I think the Court’s opinion is too doctrinaire. For despite the assumed purpose of Congress to make the 1957 Act the sole standard, it is plain to me that this simply could not be.
There is, first, the command of the Constitution which overrides all. It is true, of course, as the opinion points out, that with respect to Federal juries Congress can prescribe legislative standards more exacting than what the Constitution requires. So, too, may Federal Appellate Courts give voice to' such standards in doubtful cases short of statutory violation through the exercise of supervisory power over the conduct of criminal justice. But neither Congress nor the Courts can disregard what the Constitution demands.
*73It is here that we see that the 1957 Act cannot be the sole standard. For it is now clear that there are two constitutional imperatives: (1) the jury must fairly represent a cross section of the community;1 and (2) to assure this jury selectors must acquaint themselves with the identity and availability of potentially qualified persons within significant elements of that community.2
This means that Federal jury selectors must be conscious of the “community” 3 and all of its significant components. These include not only racial groups but other significant elements such as urban, rural, wage earners, highly educated professionals, low income, poverty areas, and the like. Having become conscious of the community, they must then establish (or at least follow) a system or procedure which will in the jury “universe” represent a fair cross section. Even more vital, they must take steps to seek out, find and become acquainted with the identity and availability of, potentially qualified persons within such elements, especially where the background or experience of the selectors would not give them ready personal access to this information.
All of these problems become aggravated by two factors. The first is the relatively small number of names maintained in the jury box.4 The second, certainly for this Circuit, is the rapid expansion in the number and size of metropolitan areas. The latter presents an added complication since almost invariably the Metropolitan Division is made up of one, or perhaps two metropolitan counties plus up to ten or more non-urban or rural counties.
When the “universe” from which the jury box is to be filled numbers into the hundreds of thousands, indeed, at places, over a million, it is evident that there must be great selectivity. To begin with would be geographical area to assure some balance between the urban and non-urban sections of the Division. And within the metropolitan county (or counties), there must be further geographical distribution to the extent, as is frequently so, the areas of residence accurately reflect other significant distinctions such as race, economic status, education, type of employment, and the like.
Then, quite apart from geography, steps must be taken to assure that significant elements of the community inherent in a fair cross section are taken into account. And even random use of sources of large number of names such *74as city directories, telephone directories, voter lists, tax lists, public utility customer lists, etc., does not eliminate the preliminary, and constitutional, condition precedent to their use. Jury selectors must first be satisfied that any such source (or a combination of them) will give a true, fair picture of the community.
Of these elements, race is, of course, foremost in many of our cities.
Another significant factor essential to a fair-cross section is the educational background of the prospective jurors. The consciously sought blue ribbon jury is unacceptable for Federal Courts. Equally so is one of marginal literates. In today’s society of intense public education, a fair cross section is not afforded unless the system pursued fairly produces a balanced literate-intelligence spectrum.
And then on top of all of these selective steps there is the further and final one: selecting the limited number of names from the large source “universe” to provide and thereafter maintain the level in the jury box at the statutory, or larger court-prescribed, minimum.5 In
*75
*77the meantime, of course, the box is being partially emptied as grand or petit jury venires are drawn. Consequently, replenishments need to be supplied from time to time. These will normally be added in numbers even much smaller than the level prescribed for the jury box. Selectivity within this small lot is therefore needed unless, as an alternative procedure, some running inventory record is kept by which the names in the box retain a near cross-sectional quality through the successive addition of small lots from distinctive significant elements of the community.
At every step, therefore, there is a continuing demand for selective judgments. While in making them it may be assumed that as to individuals chosen they are required nominally to have only the qualifications prescribed in the 1957 Act, the fact is otherwise. To those must be added all of the other measuring tests to assure the fair community cross section. This means that whatever Congress might have thought it was doing, it did not, it could not, dispense with these additional judgmental standards by which out of hundreds of thousands, or occasionally a million or more names, the very, very tiny group comprising the box is selected. Judge Bell is most certainly this much right: the selection of a jury box in which all would barely pass the minimum literacy — read- and-write — test would inescapably be constitutionally defective.
There is a place therefore — indeed, the Constitution assures a place — for levels of intelligence above the marginal minimum literacy test. Thankfully, the Constitution rescues the judicial structure from the ironic assumption that the law cannot search for or try to get the sort of intellectual preparation for deciding today’s complex controversies.6 This is not to disparage the often profound, innate wisdom of illiterates or marginal literates. It is to say, however, that the community of today whose cross section must be found in the jury box is no longer — if it ever were — at the educational or intellectual level of the marginal literates. Nor is the society which law seeks to serve the simple, agrarian, frontier type. With billions of dollars of constitutionally generated public funds and added billions of private contributions being poured annually into our educational system, the aim of which is to produce better citizens for this nation, world and time, it simply defies reason to think that intellectual attainments and formal educational experiences must be totally disregarded as we select these very important public servants. It is also to say that a system which undertakes to require that the only applicable test is the lowest common denominator runs head-on into the Constitution.
II.
But there is no hesitation or weakness in my concurrence in the Court’s action. Pursuing one step further the approach discussed in Part I, it is certain that the Court’s decision is eminently correct. The statistical analysis of the number of Negroes in the box and those added in 1959, plus the testimony of the Clerk, his Deputy, and Jury Commissioner, *78make one thing clear. No matter how conscientious they were, the jury selectors did not add enough names of qualified Negroes because they did not know of them. Since the Federal Statute has been interpreted to reflect a legislative standard as favorable as that required by the Constitution,7 it is plain that from Smith v. State of Texas through Cassell v. State of Texas, and others,8 the resulting jury box is defective. A jury box produced by such a deficient procedure does not comply with the statutes. And the Court must, as it does, extinguish its actions because of prejudicial statutory error, not in the exercise of its supervisory power.
. Smith v. State of Texas, 1940, 311 U.S. 128, 130, 61 S.Ct. 164, 165, 85 L.Ed. 84, 86; Akins v. State of Texas, 1945, 325 U. S. 398, 409, 65 S.Ct. 1276, 1282, 89 L.Ed. 1692, 1699 (Murphy, J., dissenting); Fay v. People of State of New York, 1947, 332 U.S. 261, 296, 299, 67 S.Ct. 1613, 1631, 1633, 91 L.Ed. 2043, 2064, 2066 (Murphy, J., dissenting); Scott v. Walker, 5 Cir., 1966, 358 F.2d 561, 564 (en banc) [Mar. 31, 1966]; Billingsley v. Clayton, 5 Cir., 1966, 359 F.2d 13, 17 (en banc) [April 5, 1966, p. 7]. This same imperative was recognized in Glasser v. United States, 1942, 315 U.S. 60, 84-86, 62 S.Ct. 457, 471, 472, 86 L.Ed. 690, 707, Thiel v. Southern P. Co., 1946, 328 U.S. 217, 220, 66 S.Ct. 984, 985, 90 L.Ed. 1181, 1184-1185, and Ballard v. United States, 1946, 329 U.S. 187, 195, 67 S.Ct. 261, 265, 91 L.Ed. 183, 187, which in interpreting the statutory provisions governing the composition of Federal juries, adopted the constitutional cross-section principle of Smith v. State of Texas.
. Smith v. State of Texas, supra, 311 U.S. at 132, 61 S.Ct. at 166, 85 L.Ed. at 87; Hill v. State of Texas, 1942, 316 U.S. 400, 404, 62 S.Ct. 1159, 1161, 86 L.Ed. 1559, 1562; Cassell v. State of Texas, 1950, 339 U.S. 282, 289-290, 70 S.Ct. 629, 632, 94 L.Ed. 839, 847-49; United States ex rel. Seals v. Wiman, 5 Cir., 1962, 304 F.2d 53, 65-67; Scott v. Walker, supra, 358 F. 2d at 573; Davis v. Davis, 5 Cir., 1966, 361 F.2d 770, 771 (en banc) [May 31, 1966].
. The “community” may be the entire District if a single jury box, 28 U.S.C.A. § 1865, is maintained for the District, and at least the whole Division where separate Division boxes and Commissioners are ordered. (Cf. amended F.B.Crim.P. 18, effective July 1,1966, permitting trials anywhere within a District.)
. 28 U.S.C.A. § 1864 specifies only the minimum of 300 “qualified persons.”
. The following table reflecting data obtained from seven metropolitan area divisions reveals tangible figures demonstrating the need for continual selectivity by jury selectors. This is revealed by comparing the large populations of some of this Circuit’s metropolitan areas with the infinitely small number of names from such areas required for the jury box. For instance, the Houston Division of the Southern District of Texas (Line (1), Col. I) maintains a separate jury box. The Houston Division comprises Harris County, with 1,243,158 residents (Line (1), Cols. II, IV, V) and eleven smaller counties. Since the division currently uses approximately 2,000 jurors a year, this number of names must annually be placed in the box. Under the practice followed, this is accomplished by adding about 40 names about 50 times during the year. At each of these times, an effort is made to give representation to each of the 12 counties in the division based on its population in proportion to the population of the entire division. This means, that at each time 40 names are selected to be placed in the box, Harris County with 84% of the division’s population (Col. V) would have 34. In selecting these 34 names from 1,243,158 names (or about 1 from each 40,000), to secure a racial-economic-educational cross-section of the community the selectors must further consider the many smaller geographical areas, e.g., wards, precincts, subdivisions, residential districts, that make up the county. Whether done once (1700 names), the selectors must consciously (425 names), or fifty times a year (32 names), the selectors must consciously select these relatively small groups from known, identified significant elements of the community, such as race, wage-earners-laborers, executive-professionals, the poor, the wealthy, the marginal literates, the college educated, and the like. The same is true for Atlanta (Lines 7-12). Approximately 50% would come from that part of the city in Fulton County (556,326; Col. V, Line 7), and 25% from adjoining DeKalb (Col. V, Line 8). This small lot (913 plus 1560, Col. VI, Lines 7 & 8) must be chosen to assure a fair cross-section of this “community” of 800,-000. Likewise, to obtain 500 names for Miami (Lines 5, 6, Col. V) out of 1,268,-993, which fairly represents the Miami Metro, may be done only by consciously selecting portions from areas, classes, races, etc. or at random from broad sources (directories, etc.) which a conscious study demonstrates produces a like, smaller sample.
. The Judicial Conference of the United States, 28 U.S.C.A. § 331, does not think so. It approved the 1959 Report of the Judicial Conference Committee on the Operation of the Jury System, The Jury System in the Federal Courts, 26 F.R.D. 409, 421, which declares:
“In order that grand and petit jurors who serve in United States district courts may be truly representative of community, the sources from which they are selected should include all economic and social groups of the community. The jury list should represent as high a degree of intelligence, morality, integrity, and common sense as possible."
(Emphasis added.) In accordance with this suggestion, many districts now send a questionnaire (AO Form 178 (8-60) which corresponds to Exhibit 1 of the Report, 26 F.R.D. at 507-508) to prospective jurors which, in addition to inquiring whether the prospect can “read, write, speak and understand the English language,” (Question 8), asks the prospect to state what education he has had (Question 10).
. Glasser v. United States, supra; Thiel v. Southern P. Co., supra; Ballard v. United States, supra.
. See cases cited in notes 1 & 2, supra.