This is the second appeal in this matter, our first decision being reported in Costello v. United States, 9 Cir., 298 F.2d 99. We there remanded, directing that the trial court hold a hearing “confined to the reliability of the informant as it relates to the issue of probable cause for the arrest.” At that hearing, the name of the informant was disclosed, but she was not available to testify, being apparently outside of the jurisdiction. The government presented the testimony of four witnesses. The defendant presented no testimony. The court made findings as follows: The informant, during *261 a period from the fall of 1951 until 1953 or 1954, supplied representatives of the United States Marshal’s office with information on narcotics, prostitution and other matters, and the information thus supplied was found to be reliable. In 1952 the informant supplied the Marshal’s office, and particularly deputy marshal Thompson, with information which led to the arrest and conviction of one Cody on a narcotics violation and with information which was relayed by the Marshal’s office to a Treasury Agent at Anchorage, and resulted in the arrest and conviction of one Childress on a narcotics violation. The informant left Alaska in 1953 or 1954 and returned in November of 1957. At that time she was introduced to officer Calhoon of the Territorial Police by Thompson, who told Calhoon about his use of the informant and that he had found her information to be reliable. In addition to acting as an informant in the present case, she also furnished officers Calhoon and Sundberg with information pertaining to the names and addresses of prostitutes in Fairbanks, and this information proved to be reliable. The court concluded that the informant was reliable, that the officers, Calhoon and Sundberg, who participated in Costello’s arrest, knew or had reasonable grounds to believe that she was reliable, and that Costello’s arrest was based upon probable cause.

On this appeal Costello asserts that the findings are not supported by the evidence. In considering this question we apply the “clearly erroneous” rule by analogy to rule 52(a) F.R.Civ.P., 28 U.S.C.; United States v. Page, 9 Cir., 1962, 302 F.2d 81, 85. Appellant relies upon language used by us in our prior opinion in this matter, (Costello v. United States, supra, 298 F.2d at 101.) There we said that when a “tip” is relied upon to establish probable cause there must be a substantial basis for crediting the hearsay or a record of accurate and reliable information, in short, a previously reliable informant. Appellant also relies upon Jones v. United States, 1960, 362 U.S. 257, 269, 80 S. Ct. 725, 735, 4 L.Ed.2d 697 where the Supreme Court said that an officer “may rely upon information received through an informant, rather than upon his direct observations, so long as the informant’s statement is reasonably corroborated by other matters within the officer’s knowledge.” The-Court there cited Draper v. United States, 1959, 358 U.S. 307, 79 S.Ct. 329, 3 L.Ed. 2d 327. In Draper the corroboration consisted in the occurrence, under the observation of the officers, of certain events which the informant had predicted, and the officers’ observation of the events was held to be sufficient corroboration to warrant reliance upon the informant’s tip.

Appellant asserts that there is no such corroboration here, and we agree. The only real corroboration of this kind occurred after the arrest, and as we pointed out in our prior opinion, (298 F.2d at p. 101), the ultimate corroboration provided by the fruits of the search cannot be used retroactively to justify the arrest.

This, however, is not the end of the matter. Here, there was another kind of corroboration. The Court had before it the testimony of Thompson who had used the informant in the Cody case and found that the information that she gave was accurate. His testimony is that he told officer Calhoon about this experience and this testimony is confirmed by Calhoon. In addition, there is testimony by Thompson, Calhoon and Sundberg that each of them, before the arrest of Costello, had received information from the informant regarding prostitution in the Fairbanks area, which information they had cheeked out and found accurate. It was brought out on cross-examination that so far as the witnesses could recall, none of this information led to an arrest or conviction, and that the witnesses could not recall specific instances by name. The reasons given for this lack of recollection were two, first, that the witnesses were testifying in 1963 and the events involved occurred before the arrest of Costello in the spring of 1958, some five years before and, second, that prostitu *262 tion is a misdemeanor and, consequently, they did not keep the same kind of records as to the information gleaned from the informant that they would have kept in connection with more serious matters. It is urged that this destroys their testimony as to information given by the informant relating to prostitution matters. We think not. It goes to the weight of the testimony and, of course, the Court could have disbelieved the officers if it were convinced that their stories on this subject were fabrications. The Court’s findings, however, show that it believed them and we cannot say that these findings are clearly erroneous.

It is urged that the two specific cases to which the officers or some of them referred, namely, the Cody and Childress cases, cannot be taken into account as indicating reliability of the informant because they are too remote, the Cody case, at least, having occurred in 1952, and also because in the Childress case the officers did not know of their own knowledge that the informant’s tip had proved out, but had gotten this information from other officers in Anchorage. These contentions we must also reject. They, too, go to the weight of the testimony and the Court’s findings show that it believed the officers and accepted their testimony. We did not, in our prior opinion, use the word “record,” which appellant stresses, in any technical sense. The case of Worthington v. United States, 6 Cir., 1948, 166 F.2d 557, relied upon by appellant for the proposition that the Cody case is too remote, does not sustain that proposition. The question of the reliability of an informer was not involved in that case at all.

Our holding is that there are at least two means by which the credibility of an informant may be established. One is by corroborating external circumstances occurring in the course of the very case that is at issue, as happened in the Draper case, supra. The other is the fact that on prior occasions in other cases the informant has given information which turned out to be reliable. It is the latter type of corroboration that exists here, and that we hold to be sufficient. We are cited to no case to the contrary, although there is language to that effect, with which we do not agree, in the separate opinion of Judge Bazelon in Jones v. United States, 1959, 105 U.S.App.D.C. 326, 266 F.2d 924, 925. In Cervantes v. United States, 9 Cir., 1959, 263 F.2d 800, in fn. 8 at page 804, we stated:

“The rule formulated by the cases is that information received from an outside source * * * will not suffice to show probable cause unless such source is believed to be reliable, or unless it has substantial verification or supplementation in facts personally known to the officer or officers.”

See also Butler v. United States, 9 Cir., 1959, 273 F.2d 436, 441, 443.

Appellant also asserts, as he did on the prior appeal, that the presumption of importation in 21 U.S.C. § 176a 1 and the presumption stated in 26 U.S.C. § 4744 (a) 2 are invalid. Appellee asserts that these questions are no longer open because, in the prior appeal, our remand directed the district court to hold a hearing confined to the reliability of the informer. It is concluded from this that all other questions were impliedly decided adversely to appellant. This, we think, does not follow. In ordering a remand, we followed the procedure used by the Supreme Court in United States v. Shotwell Mfg. Co., 1957, 355 U.S. 233, 78 S.Ct. 245, 2 L.Ed.2d 234. We stated *263 that if the Court found the informant reliable it should enter a new final judgment based upon the record as supplemented by its further findings “thereby preserving to appellant the right to further appellate review.” The Shotwell opinion contained an additional clause: “ * * * including respondents’ right to have reviewed by the Court of Appeals alleged errors in the original trial which that court did not reach in the previous appeal.” While we did not use this language, we think that our prior opinion implies that the appellant should have the same right. In that opinion we did not discuss the question now presented because we ruled in favor of appellant, on the state of the record then before us, on the issue of reliability of the informant. The question of the validity of the two presumptions above mentioned was, however, briefed and argued on that appeal. It was not disposed of then and we are of the opinion that it is open now.

As to the presumption created by 21 U.S.C. § 176a, appellant’s argument is that the testimony in the case contradicts the presumption. He relies upon the testimony of two witnesses. One of them, Barkley, a state police investigator, was asked on cross-examination if he knew where marijuana comes from, and answered: “From a plant.” He was then asked if he knew where these plants grow, and answered: “Not exactly. I have heard that they grow all over.” The other, Ringstrom, a retired chemist of the Alcohol Tax Unit of the Treasury Department, identified the marijuana involved in this case and testified to certain chemical tests that he had used to verify the fact that it was marijuana. On cross-examination, he testified that he was unable to tell by chemical tests from what country any particular marijuana came and he then gave the following testimony :

“Q. Well, would you state what information you have relative to this general distribution and the growth areas ?

“A. It grows all over the United States, Mexico, Europe, the southern part of Asia, and northern part of Africa; as far as I know, it grows over the major part of the world in the warm climates and the temperate climates.

“Q. Is it a difficult plant to raise ?

“A. It is very easy to raise.

“Q. And does any part of the plant have commercial use?

“A. Where it is grown and harvested for definite purposes; yes, it does.

“Q. And is there any way known to you that it could be either chemically or botanieally determined whether a given stalk or portion of a stalk or leaves of marijuana was grown in the United States or in some foreign land ?

“A. Not that I know of.”

He further testified that the marijuana cigarettes in question contained “portions of the leafy portion of the plant rather than a portion of the seed and a portion of the fibrous structure of the plant itself.”

Costello makes two arguments, one, that the statute is unconstitutional on its face and, two, that in the light of the testimony that we have quoted, any presumption that the particular marijuana here involved was imported has no rational basis in fact. We reject the first argument; we have heretofore held that the statute is constitutional. (Caudillo v. United States, 9 Cir., 1958, 253 F.2d 513; Butler v. United States, 9 Cir., 1959, 273 F.2d 436; compare Agobian v. United States, 9 Cir., 1963, 323 F.2d 693.)

Appellant’s second argument is based upon the contention that the record shows (a) that marijuana “grows all over the United States * * * in the warm climates and the temperate climates,” and (b) that the particular marijuana here in question was “manicured,” i. e., made up entirely of leaves, and that in the Caudillo and Butler cases we took note of the fact that while “unmanicured” marijuana is seldom produced in the United States, “manicured” marijuana is seldom *264 imported. The record in this case contains no such evidence. In Caudillo we pointed out that we know of no medical ■or scientific use to be made of marijuana, ■save perhaps for occasional testing in order to make scientific comparisons with other nai'cotics, barbituates and amphetamines. We also noted that the growing of marijuana is illegal in several states including California (the state there involved), and that by far the larger part of all marijuana found within the United States is imported. In Alaska, as in California, the growing or possession of marijuana is illegal. (See Michie’s Alaska Statutes, §§ 17.10.010, 17.10.230(4), (12) and (13) ). The fact that the marijuana involved in Butler and Caudillo was “unmanicured” was simply an additional factor entering into the decision in those cases.

Those cases certainly establish the proposition that the mere fact that marijuana can be and is grown in the United States does not render the statute invalid. The only additional fact suggested here is that the particular marijuana appears to have been “manicured.” However, this fact alone is not enough to require a decision that the statute is ■invalid as applied to Costello. In DeRose v. United States, 9 Cir., 1963, 315 F.2d 482, we pointed out no attempt was made to make a record as to the conditions or restrictions under which the drug there involved, methadon, may lawfully be possessed or used, if at all, the quantity produced in the United States, the extent of importation as compared to United States production, or other facts on the basis of which a decision could properly be made. Such a record, in relation to ■cocaine-hydrochloride, was made in Erwing v. United States, 9 Cir., 1963, 323 F.2d 674, and we held a comparable statute invalid as applied to that drug. Whether a similar showing can be made in relation to “manicured” marijuana we do not decide, because, as in DeRose, no such showing was made here. We hold that on this record we cannot say that the statute is unconstitutional either on its face or as applied to this case.

Costello was convicted on three counts, the first two of which charged violation of section 4744 of Title 26 U.S. Code. The third charged violation of section 176a of Title 21. He was sentenced to five years on each of counts one and two, and to ten years on count three, all sentences concurrent. Under these circumstances, the sentence on count three being valid, we need not further consider counts one and two. Hence, we need not pass upon the validity of the presumption stated in 26 U.S.C. § 4744(a). (See Gonzalez v. United States, 9 Cir., 1963, 324 F.2d 223.)

Affirmed.