dissenting:
I
The conviction in the case at bar cannot be sustained. The specifications dó not allege offenses. They do not allege offenses because the order of the Commandant of the Coast Guard applies only to “. narcotic substances, and other controlled substances” and these are tied by reference to schedules “as updated and republished under the provisions of that Act [Pub.L. 91-512 (91-513) ].” There are no such schedules.
II
The Coast Guard regulation involved here, by the use of the phrase “other controlled substances,” ties in the proscribed narcotic substances with the definition given for “controlled substances.” “Other” does not mean nothing, nor can it be manipulated out of the regulation. It means that what has gone just before is a specific of the more general which follows.
An attempt to explain it away by a theory that what was “clearly” intended was the establishment of three free-standing categories will not do. Just as clearly it must be seen that the regulator meant exactly what he said. In earlier days the Federal law itself was practically carved in stone and Coast Guard regulations tended to follow the general practice. When Congress moved toward recognition of dangerous substances other than narcotics, the Coast Guard regulations did the same. When Congress found that the marvels of modern chemistry could outstrip the legislative process, it chose to arrange all substances to be regulated across a spectrum, continuous but divisible, and divided into five categories or “schedules”. Some are narcotics, some are not, but all are to be regulated. It was logical for the Coast Guard regulation to be broadened to include the whole spectrum. It would be unthinkable, then, that there would be a narcotic known to the Commandant of the Coast Guard but not regulated under the Act. Further, since it is not a principal function of the Commandant to test ingestibles, classify them, and the like, it was entirely proper for him to leave this to the administrator appointed by Congress and furnished with the equipment to do just that.
A brave attempt is made to save three separate categories in the prohibition of the regulation by noting (1) that “narcotic substances” had always been banned by earlier Coast Guard regulations and hence, are banned “irrespective of their being included in the definition of ‘controlled substances’ ” or, alternatively, (2) that “since 21 U.S.C. 802(lc)[(16)(C)] statutorily defines cocaine as a narcotic drug it is proscribed by the regulation whether or not it is included in a Schedule of Controlled Substances.”
The first alternative here does not work because we are not dealing with Coast Guard regulations as they once were, or once were tailored to circumstances, but rather with 9-2-15 as it reads today. The theory does not explain the “other” in “other controlled substances.” The second alternative is ingenious but again it is not what the regulation says.
The Act defines “narcotic drugs.” The regulation does not say “narcotic drugs as defined in the Act.” It does not even use the term “narcotic drugs;” it speaks of “narcotic substances.” It does not define “narcotic substances” as it defines “controlled substances.” If we attribute to the drafter of the prohibition the subtlety that is imputed to him here, we might suspect that what he was actually doing was seeking to evade the definition of “narcotic drugs” in the statute as being perhaps too circumscriptive and to allow the condemning of other “narcotic substances” which *583might fall outside the scope of “narcotic drugs” (and also outside the “other controlled substances”). It is unacceptable that the regulation incorporates the definition of “narcotic drugs” in the Act while using a different term and while not referring to the Act at all in connection with “narcotic substances.”
Since the regulation itself does not define “narcotic substances,” and does not use the term in the Act which the theory attempts to import into the regulation, and does not associate its reference to “narcotics” with the Act other than through the phrase “other controlled substances,” and since it does inextricably unite its “narcotic substances” with “other controlled substances” as set forth in the Act, we-must read the regulation as saying just what it says, no more and no less. It merely, as is evident, groups “narcotic substances” in the class of “controlled substances.”
Acknowledgment of the ingenuity of the circumventions offered to construe the regulation serves only to point up that the majority opinion necessarily postulates ambiguity in the regulation, — the death knell of a criminal rule of conduct even if the justifiable explanation for the clear language offered here were not ready to hand.
The regulation as worded, then, fits beautifully and precisely into an efficient plan, reasonably to be adopted. Since the regulation can be read and understood under a plausible, indeed persuasive, explanation, there is no excuse to attempt to read some other meaning or intent into it by expelling a key word or pretending that words are there which are not.
Although it is not conclusive, it is of great interest that this meaning accepted for the regulation is precisely the theory under which the prosecution at trial proceeded. The trial counsel furnished, for the military judge’s notice, a copy of the regulation, a copy of the schedules in the statute, a copy of the current CFR publication (to show that the “executive branch” list as pertinent was still the same as the statutory list), and a page of an encyclopedia, to establish that cocaine is a narcotic (a point here conceded for the purposes of this case). This approach demonstrates clearly how the regulation is read by one appointed to prosecute its alleged violation.
There is nothing “wrong” about the regulation. All that could be “wrong” for the case at bar would be for cocaine not to be included in the statutorily authorized schedules, or for there to be no valid schedules. This latter happens to be the ease.
Ill
To approach the question of the status of the schedules to which the Coast Guard Regulation is tied, two provisions of the Comprehensive Drug Abuse Prevention and Control Act of 1970 furnish the route to discovery. They are sections 201 and 202 (21 U.S.C. §§ 811 and 812). Subsection 201(a) reads:
“The Attorney General shall apply the provisions of this [subchapter] to the controlled substances listed in the schedules established by section [812 of this title] and to any other drug or substance added to such schedules . . . The Attorney General may by rule:
(1) add to such a schedule or transfer between such schedules . . .; or
(2) remove any drug or other substance from the schedules . . .”
Subsection 202(a) provides:
“There are established five schedules of controlled substances . . . Such schedules shall initially consist of the substances listed in this section. The schedules established by this section shall be updated and republished on a semiannual basis during the two-year period beginning one year after the date of enactment of this [subchapter] and shall be updated and republished on an annual basis thereafter.”
Subsection 202(c) declares:
“Schedules I, II, III, IV, and V shall, unless and until amended pursuant to section [811 of this title], consist of the following drugs or other substances
Subsection 202(c) then goes on to set forth the initial Schedules I-V.
*584In addition to the appropriate language of the statute it is also necessary to set out the chronology of certain enforcement actions. The Act was enacted on October 29, 1970, with the first anniversary on that date in 1971. The administrator published schedules on April 24, 1971 (36 FR 7776). On May 12, 1972, he republished schedules “as of April 29, 1972” (37 FR 9545). On December 22, 1972, he signed documents promulgating schedules “pursuant to the mandate of section 202(a)” “as of October 29, 1972,” but there was no republication until January 8, 1973 (38 FR 953).
The dates April 29, 1973 and October' 29, 1973, both significant under the time scheme prescribed in the Act, were passed without action of any kind to “republish.” On June 20, 1974, the administrator republished an “annual” schedule “for 1974.” To date there has been no republication after that one.
Several items here provide food for thought, but comment is not appropriate at this time, since the first date of import to the case at bar is October 29, 1975. Even charity cannot, however, prevent attention to the irony that this date is the date set by the Act’s terms for the third “annual” republication of the schedules.
The significance of these dates is found in light of the Federal court decisions to be discussed.
In all, six Federal court decisions are of interest here. In order of decision, although not in the order in which they will be considered, they are:
(1) United States v. Nocar, CA7 (1974), 497 F.2d 719
(2) United States v. Infelice, CA7 (1974), 506 F.2d 1358
(3) United States v. Monroe, D.C.N.D. Cal. (1976), 408 F.Supp. 270
(4) United States v. Andrews, D.C.N.D. Cal. (1976), 408 F.Supp. 1007
(5) United States v. Eddy, CA9 (1976), 549 F.2d 108
(6) United States v. Monroe, CA9 (1977), 552 F.2d 860.
IV
On the question of the status of the schedules of “controlled substances” under the Act, this court says, “Accordingly we hold that despite the failure to republish schedules after June 20, 1974, cocaine remained a controlled substance within the meaning of the . . . Act . . .” It cites United States v. Andrews and refers us also to United States v. Eddy. What is not acknowledged is not only that these two decisions should not be mentioned in juxtaposition but that they are mutually contradictory.
Both the Andrews and the Monroe district court cases were up on appeal in the Ninth Circuit when the Eddy case was decided. The Eddy decision was cited as controlling when the convictions in the Monroe case were sustained.
The district court in the Andrews case, obviously less than enchanted with the earlier district court opinion in the Monroe case, brushed aside questions as to republication of the schedules and held that the original schedules in the Act itself were enough. This view is subscribed to by the majority here. Nevertheless, the court of appeals decisions in the Monroe and Eddy cases implicitly overrule the Andrews decision. In its Monroe statement, the court of appeals says, “. . . the republication requirement of section 812 is satisfied by the annual publication of the updated schedules in the Code of Federal Regulations. That requirement has been fulfilled.” It is obvious that if the statute requires a republication the Andrews opinion and this court are both wrong in saying that it does not.
This does not, unfortunately, make easier the expression of a dissent here, it does make the duty to dissent clearer but doubles the problem of expressing the reasons.
V
Since the majority here echoes the statements of the district court in the Andrews ease, the first proposition to be discussed is that no republication of the schedules is required under the Act.
*585The Nocar and the Infelice decisions must be set into the background. These cases involved acts committed on or before April 29, 1972. In the former case it was argued that the failure of the administrator to republish schedules between April 24, 1971, and the time of the indicted acts (the republication “as of” April 29, 1972, not having been made until May 12, 1972) resulted in there being no effective schedules. The court held that the time frame set by statute for the semiannual republications established April 29, 1972 (18 months after the effective date of the Act), as the date for the first semiannual republication. The publication of April 24, 1971, was held to have been not required by the Act and to have been not pursuant to the Act. The initial schedules provided in the Act were still effective at the time of the offenses and no action by the administrator had yet been called for.
The Infelice decision considered an attack based on a claim of retroactivity. The May 12, 1972, republication, purporting to reach back to April 29 in effectiveness, was urged as unconstitutional. Here the court found that the last of the indicted acts had occurred on April 29 itself and the initial statutory schedules were again held controlling.
While the Andrews decision relegated both of these decisions to footnote treatment and declared them, as to republieation, inconsiderable dictum, they are significant in showing how the Act is read by the reader who accepts it for what it says. This court of appeals, just as had the administrator himself in referring to the “mandate” of the Act, had seen the Act as requiring republication at stated intervals.
By the time of the Monroe and Andrews district court cases, the state of the schedules had become what it was for the case at bar, and is now. For October 29, 1972 (the second anniversary of the Act and the projected date for the second semiannual republication), the administrator signed on December 22, 1972, and published in the Federal Register on January 8, 1973, a set of schedules “as of” October 29, 1972, acknowledged to be pursuant to the “mandate” of the statute. Although it seems to have escaped the attention of a court, there was no to-have-been-expeeted third semiannual republication for April 29, 1973, nor was October 29, 1973 (when the fourth and last semiannual — and first “annual” — republication was to have come forth) respected in any fashion. The June 20, 1974, republication was declared to be the schedules “for 1974.” That was the last time schedules were republished by the administrator.
The Monroe district court decision actually stands for nothing since it rejects out of hand a series of defense protests dealing with the schedules but then, in each case, assumes the truth of the contention and goes on to predicate its action on a statement that nothing in Hotch v. United States, CA9 (1954), 208 F.2d 244, requires that the schedules be found ineffective. The Hotch case dealt with the problem of actual notice of an administrator’s order vis-á-vis failure of publication in the Federal Register. Since there was not a hint of actual notice of an unpublished rule in the Monroe matter, the opinion is of no moment. This was recognized in the Andrews opinion, given in the same Federal District, which would have recognized the Monroe decision as controlling, if it could have, but which, after paying lip service to it, went off on an entirely different theory.
For the case at bar, just as no note need be taken of the retroactivity from January 8, 1973, to October 29, 1972, or of the gap left by the omission of republication for the rest of 1973, the question whether the June 20, 1974, republication (“for 1974”) somehow survived to June 20, 1975, is not an issue. (The Monroe decision actually declares as a fact that a June 20, 1975, republication was called for and then, noting that there was none, drops the subject.) Since there has been in fact no republication since June 20, 1974, only two possible theories can support effective schedules after June 20, 1975. The one of immediate concern, the Andrews theory, brushes all need for republication aside. The Gordian Knot is cut. It is declared flatly that the *586schedules in the Act itself are still effective. Total reliance is placed on the “unless and until” provision of the Act. For the case at bar, the theory is that since cocaine is covered by the schedules in the Act and has not apparently been removed by the administrator it is still there. The theory cries out for scrutiny.
If no republication of any kind is ever required, there would be the initial schedules and two possibilities: (1) no amendments or (2) amendments. If there were no amendments (a “contrary to fact” condition), there would be no republication ever. It is clear, however, that when a statute admits of amendment by administrator’s regulation the public is entitled to fairly current notice of the fact of no change, and Congress recognized this. If there have been amendments — the fact — the opinion leaves open for speculation two more possibilities. The first is, again, no republication of any sort. This leaves the inquirer with recourse to the original schedules and an infinite possibility of Federal Register readings. The second is a kind of “smuggled” republication, but only of the amendments- — not of the unamended provisions of the original schedules. There would be no time set for republication of the amendments alone unless one went back to the statute, which is forbidden by hypothesis. Further, a deletion (which can occur under the statute either by direct act or by transfer from one schedule to another) cannot be “republished.” It is a void. There could be republication of a notice of deletion, of course, but that is not the same thing.
Further explorations are not in order. The statute commands republication of the schedules — not of amendments, nor of parts of the schedules, nor of notices about the schedules. The schedules themselves, as amended, whether by addition, deletion, or transfer, must be republished. In elevating the “unless and until” clause to solitary rule (and it would be captious here to insist that a literal reading of that provision alone would necessitate that Congress have built a “self-destruct” into the schedules — i e., that the first time the administrator amended them they would cease to exist), the court makes no provision for execution of the republication mandate. That the mandate meant something to Congress is shown by the fact that it not only called for republication, but republication at prescribed intervals, and it not only prescribed intervals but two different sets of intervals.
The court in the Andrews opinion cannot tolerate the prospect that in so momentous a matter as “drug” control there is no effective Federal regulation. The court is correct in asserting that Congress did not intend that there ever should be no valid schedules for enforcement of a painstakingly constructed comprehensive act. Unfortunately, Congress also never intended that its designated administrator should not carry out its mandate. Neither of these miscalculations can serve to validate that which just is not.
The opinions here and in the Andrews decision elevate some language in the statute above other language in the statute and then, from those heights, refuse to see that the other language is still there. It does no good service to Congress to attempt to vitalize one provision of the law by suppressing another. The provision as to updating and republishing cannot be read out of the Act. It is repeated that the “unless and until” clause cannot be construed as expunging the “update and republish” language. Even if Congress had given the administrator no authority to amend the schedules at all, if it said he must republish them annually he must republish them annually whether he considers it useful or not. When the concurring judges here say that “The time table was nothing more than an expedient legislative device by which the Congress authorized the administrative amendment of the schedules,” the point is missed as widely as can be. Amendments do not have to be made on any time table. The whole thrust of the amendment provisions is that amendment may be made at any time, subject to the notice and hearing requirements of administrative procedure. “Update and republish” has nothing to do with that at all. The schedules can be *587amended every day the Federal Register is published. They must be republished, now, only once a year.
It is also necessary to remember here that the Commandant, in incorporating the schedules into his regulation, did so “as updated and republished under the provisions of the Act,” a phrase given no attention by the majority.
VI
The alternative to the “original schedules still valid” theory, (also to be considered as the “republication not required” theory) is that adopted by the Ninth Circuit in its Eddy and Monroe decisions. The “Andrews ” case, source of the theory that the original schedules perdure, is now on appeal in the Ninth Circuit. There can be little chance that that court of appeals will withdraw from its holding in the “Eddf’ and “Monroe” cases. It may safely be predicted that its disposal of the “Andrews” appeal will be predicated on the same grounds. In effect, the Andrews decision can be considered overruled in its own circuit and cannot be acknowledged as the Federal law anywhere.
In United States v. Monroe, the court of appeals, quoted above, has acknowledged the duty under the statute to republish the schedules and has declared that publication of the Code of Federal Regulations is the republication required by the statute. Joining this opinion to those of the Nocar and Infelice decisions, we find a unanimity at least as to recognition of the force of the order to republish. As to how the republication is to be achieved, the circuits differ. The rationale implicit in the decisions of the Seventh can be demonstrated to be correct, as against what the Ninth accepts, which, in turn, can be shown to be unfounded in the law and a dangerous precedent leading to the undermining of long-standing principles of criminal justice.
There is great misunderstanding, both among the regulators and among the regulated, as to what the Code of Federal Regulations is. 44 U.S.C. § 1505 declares that, in general, documents which must be made valid as to classes of persons must be published in the Federal Register, and 44 U.S.C. § 1507 confers upon documents validly so published the quality of constituting constructive notice to all concerned. The Register, in its more than forty years of publication provides triple-columned pages of constructive notice running into the seven-figure totals, and more is added hundreds of times annually. In its wisdom, to give the regulated reasonable access to the mass of printed material (constantly affected in its application and content by amendments, rescindings, and the like), Congress provided for a key. The key is a supplement to the Federal Register. It is called the Code of Federal Regulations. By its careful use, the researcher can save hours, even weeks, of hunting, to find out what is the state of administrators’ regulations at a given instant. Careful use, itself, requires the utilization of “finding aids” provided with the grand “finding aid” of all, the Code.
There is a superficial analogy between the relationship of the CFR to the Federal Register and that of the United States Code (generally viewed) to the Statutes at Large. It can be misleading for, like most analogies, it limps. The regulators themselves, accustomed to publish documents in the Federal Register in the example format: “Title X CFR, Part y, is hereby amended by adding a new section ‘y + 2z’ as follows: .,” forget that the true entity is the Federal Register document, not “the Code,” and that the new pronouncement amends essentially an earlier Federal Register notice, not an independent quasi-statutory entity, the CFR.
44 U.S.C. § 1510, which creates this tool for entering the Federal Register system, speaks of it as a “supplemental edition.” The Administrative Committee (created by 44 U.S.C. § 1506) is directed to publish this supplement as the “Code of Federal Regulations” in a set of books. There is a further direction that: “Each book shall be either supplemented or collated and republished at least once each calendar year.”
*588It is necessary to note immediately that over a period of years there have been three basic formats used for issuance of the Code. Its first appearance was in a set of large black books with annual supplements to the set being provided. These annual supplements were twice replaced with a “Cumulative Supplement” for five year intervals, which was used in conjunction with the basic set. This form was replaced in 1949 with a multi-volumned set of smaller books, each volume of which was “kept up” by a familiar “pocket part” device. (At times, when a “pocket part” would become too big and reprinting the basic volume was not considered feasible, there would be a basic book, a separate supplemental volume, and a “pocket part” in the basic volume, supplemental to the supplementary volume.) The contemporary practice, opted for by the Administrative Committee from the two methods prescribed by Congress, is to reissue basic volumes annually. The total Code is divided into four parts and four (quarterly) compilation dates are assigned. (By the hazards, April 1 happens to be the compilation date for the Code title containing the schedules under the Act.)
Two important observations are pertinent here:
(1) the date (April 1, in this instance) is an “as of” date; it is not a date of publication, or of printing, nor of issuance, nor of availability, nor of delivery;
(2) The choice of reprinting the entire volume, which involves “collating and republishing” or of issuing an annual supplement to a basic volume, is the choice of the Administrative Committee and of no one else.
Galled upon to evaluate the legal significance of both the Federal Register and the compilation in the CFR, the Attorney General declared that the validity and constructive notice provisions of section 7 of the Federal Register Act (predecessor of 44 U.S.C. § 1507) do not apply to the compilation authorized and directed under section 11 of that Act (predecessor of 44 U.S.C. § 1510) 38 Op.Atty.Gen. 359 (1935).
Looking to the facts here, we see that the administrator has not republished schedules since June 20, 1940. The appearance of schedules in the 1975 and 1976 reprints of the schedules, at 21 CFR Parts 1300 to End, is accompanied by a “source” note cited to the text of each schedule reflecting “39 F.R. 22141, June 20, 1974” as the source publication in the Federal Register. Among the “Finding Aids” (“List of CFR Sections Affected”), there appears, as the last “affecting document” termed “republished,” for sections 1308.11-1308.15, the same reference to 39 F.R. 22141.
The CFR volume is “prima facie evidence of the text of the original documents.” 44 U.S.C. § 1510. This is acknowledged, in the language of the law itself, in the prefatory “Explanation” appearing in the CFR, title 21, book itself. The “Code” purports to show no more than a “republication” of a June 20, 1974, Federal Register document. This, of course, is not “republication” at all. It is only prima facie evidence of what the administrator published on June 20, 1974.
It is clear also that the republication required by the statute cannot be the product of a legal accident or purely fortuitous sequence of events. For example, if we assume a reprinting by the Administrative Committee of a book, “21 CFR 1300 to End” as of April 1, 1974, it would have contained “Parts 1300-1399” as “redesignated” from “Parts 300-399” by a notice at 38 F.R. 26609. From there, we would have found that “Part 308” (containing the schedules) had been “republished” at 38 F.R. 8255. (This is according to the “Finding Aids” provided in the CFR books). That, of course, was in 1973. Let it be accepted, for argument, that this is a valid republication in the Federal Register of unquestioned integrity. We know now that there was a republication in the Register on June 20, 1974. Again we assume a valid republication on this occasion, under the statute. That republished set of schedules will appear as the text of 1308.11-1308.15 when a new book is issued on April 1, 1975.
Now, let us suppose that by April 1,1976, the Administrative Committee elects not to *589reprint the entire volume containing “21 CFR 1300 to End” but to print instead a supplement comprising only the amendments made, if any, between April 1, 1975, and April 1, 1976. The holder, if one be so lucky, will have two objects in hand: the book (of 1975) and the supplement (be it “pocket part” or separate pamphlet) containing the April 1 — April 1 new material. He will not have an annually “republished” set of schedules. It does not matter that what he has might add up to what a real republication would have given him; nor does it matter that, in a particular case, one might be led down the garden path of “actual notice.” It is true also that by compiling his own daily addenda from the Federal Register himself he would have come up with an enumeration of substances. The simple, unalterable fact is that he would not have what Congress ordered, an annual “republication of schedules,” and he would never have it as long as the Administrative Committee chose to publish annual supplements, and he could never have it because the designated administrator just had never “republished.”
The truth is that a discretionary act by the Administrative Committee, even if similar to a proper act by the designated administrator, is not the act of the administrator who was charged by Congress to “apply the provisions of this Act to the controlled substances listed in the schedules . . .” If the Administrative Committee can be conceived as authorized under the Act to do the republishing duty of the designated administrator, it would follow that it could just as well amend the schedules as republish them.
While the court of appeals in the Monroe decision, declaring that the annual (accidentally so) reprinting of the CFR is the republication required by the Act, smoothly adds a footnote stating, “The schedules have been published annually in the Code of Federal Regulations since 1972 . . .,” and notes four such books, for 1972, 1973, 1974, and 1975, it gracefully omits one embarrassing fact. By direction of the Congress, from April 29, 1972, through October 29, 1973, annual republications were not enough. Semiannual republications were called for. There is no occasion to belabor a question whether once the initial schedules died without republication and the administrator could never again act to rehabilitate the system and get out an effective republication; the important thing is that the theory of the court is revealed as an unwarranted reliance upon a purely fortuitous coincidence of discretionary printing activity. ' The annual books of the CFR most certainly did not amount to the required semiannual republications when they were in order for a period of two full years.
VII
The dilemma of the dissent is now exposed. The theory of the perpetuity of initial schedules (although designed by Congress to be updated and republished), relied upon by the majority in accord with United States v. Andrews, has been rejected by the very court of appeals to which the “Andrews” case will be argued, while the theory espoused by that court of appeals in its Monroe decision is just as bad law as the Andrews decision itself.
VIII
One final note of frustration may be added, as indicative of the kind of bad law we can expect if the theory of the court in the Monroe decision ever gets beyond the Ninth Circuit. When the record of the instant case was reviewed, it became apparent that the exhibits marked for judicial notice by the military judge should be verified, especially those pages purporting to be from “21 CFR 1308.” It was then found, in January 1977, that in the set of Code of Federal Regulations in the Office of Chief Counsel, U.S. Coast Guard (the nation’s primary seagoing law enforcement agency, also performing Customs functions), the book presented as the current “21 CFR 1300 to End” was the book issued “as of” April 1, 1975. In the law library of the Department of Transportation, “parent” of the enforcement agency, the same was true. Not until . February 17, 1977, was the “as of” April 1, *5901976, book made available to the Chief Counsel of the Coast Guard. Not until March 2, 1977, was that “new book” available at the Department library.
“Constructive notice" is probably pushed to its limits when crimes are established in the Federal Register. At least, there, Congress has set the rule and we fail to observe it at our peril. At least, there the publication required follows a schedule as to which knowledge is broadcast and the legal status is defined. If a court, avoiding and twisting the language of Congress, can turn a mere “finding aid,” which for practical purposes can be tracked down only with relentless intent within the confines of a Federal department, into the equivalent of the duly established Federal Register, we verge on the unconstitutional abyss.
It cannot be overlooked here, either, that we are dealing with the case of a member of an armed force of the United States in a foreign country who is charged with violating a “constructive notice” regulation which incorporates by reference another “constructive notice” regulation. We cannot be satisfied with a shoddy substitute when the rules of the -game require that meticulous attention be given to the type and form of notice required from an administrator empowered to create crimes by regulation.
The correct course of action is so clearly laid out and the ease of execution so apparent that the strange sequence revealed in the chronology given above is inexplicable, and also intolerable.