dissenting:
The central issue in Payne’s case is whether extrinsic evidence substantiating the impeachment of a prosecution witness must be admitted when defendants seek to establish an affirmative defense solely by the tactic of impeachment. Ancillary to this issue is the question whether it is error to exclude evidence regarding the reliability of addict informers generally where defendants are seeking a jury instruction as to the lack of credibility of a paid informant who is also a drug addict.
I am not insensitive to the dangers to the judicial process inherent in the use of addict informants. But because so much of the crime plaguing us today is' attributable to narcotics, curbing the illegal drug trade is a primary problem facing our communities; in dealing with this problem, the use of informants is one of the most potent weapons in the public’s arsenal. Accordingly, although courts must be alert to protect the accused against abuses engendered by the practice of utilizing informants, it would be inadvisable to tamper unduly *581with the system where the result would be to sacrifice the security of society without a correspondingly significant advance in the protection of individual rights and liberties. The safeguards inherent in the traditional rules of evidence adequately protect defendants by enabling them to impeach the credibility of informants, yet they do not so weigh the balance that the use of such evidence by the prosecution becomes impracticable.
The facts in this case are not substantially in dispute. Robert Roscoe, who was awaiting trial on burglary and narcotics charges, agreed to assist the Bureau of Narcotics and Dangerous Drugs (BNDD) as an informant. Roscoe named both defendants as participants in the drug trade. Two months later, Roscoe went to see defendant Payne on several occasions to arrange a sale of narcotics to Agent Jackson of the BNDD, who was posing as a friend from North Carolina. Payne agreed to sell Jackson an ounce of heroin for $400.
The three men met at Payne’s residence and then left in Jackson’s automobile and proceeded to a bowling alley, where they met defendant Kinnard. Kinnard informed Payne that the heroin would be available at the Cadillac Carryout, and that they should meet there. Because of the police in the vicinity, the men then drove to the Shrimp Boat Carryout, where they again met Kinnard. Kinnard gave a manila envelope containing a white powder, later found to be heroin, to Roscoe who in turn gave it to Jackson. After a brief discussion regarding who would be paid, Jackson gave $400 to Payne, who handed it to Kinnard. Kinnard and Payne then agreed to sell Jackson $500 worth of cocaine the following week. Defendants proceeded to leave the scene together. The entire transaction was observed at a distance by another BNDD agent, Cooper.
At trial, although each defendant attempted to interpose the defense of entrapment, neither testified. Rather, they endeavored to establish their theory by attacking the credibility of Roscoe, who testified that Payne was a willing participant in the sale. Roscoe was asked about his use of narcotics. He admitted using narcotics “approximately eight times” during the last two years, and as recently as a month prior to trial. The court then refused requests that Roscoe bare his arms or that his arms be examined by a dermatologist to determine whether the use had been more frequent or more recent than Roscoe had admitted. Agents Cooper and Jackson were questioned about governmental favors Roscoe received, but the court would not allow Jackson to testify as to the testimonial reliability of addict informers generally. The court charged the jury with the standard cautionary instruction on informers and entrapment. Counsel did not object to the charge as given.
Defendants contend that the trial court abused its discretion by limiting the cross-examination of Roscoe as described above. However, this Court has noted:
“[I]f cross-examination of a witness has been extensive, repetitive and protracted, or if defense counsel has succeeded in bringing out several discreditable matters and further questions along this line would be merely cumulative, the trial judge might properly limit the scope of cross-examination without in any way harming defendants’ case.” United States v. Pugh, 141 U.S.App.D.C. 68, 70, 436 F.2d 222, 224 (1970).
Counsel for defendants, despite the limitation on cross-examination, were able to develop a considerable body of facts which would operate to impeach Roscoe’s credibility. Specifically, they showed that after Roscoe agreed to cooperate with the BNDD, the serious charges against him were dropped and Roscoe was allowed to plead guilty to misdemeanors; that Roscoe was paid for information; that he was still on *582probation; that he had lied in an identification hearing; that he had sold drugs; that he had used drugs as recently as a month prior to trial; and that he had a prior conviction for robbery.
The inquiry regarding Roscoe’s arms was clearly collateral. It did not go directly to defendants’ commission of the proscribed acts, their willingness to commit such acts, or to Roscoe’s credibility. At most, it would go indirectly to Roscoe’s credibility, by demonstrating that he had lied regarding the frequency of his use of narcotics. In such situations, the trial judge has rather broad discretion in limiting cross-examination. United States v. Pugh, 141 U.S.App.D.C. 68, 436 F.2d 222 (1970); Tinker v. United States, 135 U.S.App.D.C. 125, 417 F.2d 542, cert. denied, 396 U.S. 864, 90 S.Ct. 141, 24 L.Ed.2d 118 (1969); Howard v. United States, 128 U.S.App.D.C. 336, 389 F.2d 287 (1967). Generally, an inquiring party is precluded from offering extrinsic evidence to contradict a witness on collateral matters. Tinker v. United States, supra, 135 U.S.App.D.C. at 128 n. 16, 417 F.2d at 545 n. 16; Lee v. United States, 125 U.S.App.D.C. 126, 128, 368 F.2d 834, 837 (1966); 3 J. Wigmore, Evidence §§ 1000-1003 (3d ed. 1940).
The interpretation of needle marks on the arms of a narcotics user is not a matter within the common knowledge of a layman. Even if the trial judge had allowed Roscoe’s arms to be exhibited to the jury, the jury could not have drawn any inference therefrom without resorting to conjecture. Therefore, the request that Roscoe show his arms was properly refused because the demonstrative evidence standing alone would not have been probative even as to the collateral issue of the frequency of his use of narcotics. To make such evidence probative, experts, such as the dermatologist, or other physician with appropriate clinical experience, requested by counsel, would have to interpret the marks on the skin for the jury. But if a doctor so testified for defendants, the prosecution would undoubtedly call a physician or other expert to offer counter-testimony regarding the extent of Roscoe’s use of heroin, and a trial-within-a-trial, the very result the collateral impeachment rule is designed to prevent, would develop.
The purpose of impeaching Roscoe’s testimony was not to negate the fact of defendants’ participation in the sale, but rather was designed to create the inference that Roseoe had lied when he testified that Payne was predisposed to sell the drugs. Even if such inference could have been established by further impeachment, it would have allowed the jury only to disbelieve Roscoe’s statement and not to consider it because it was untruthful. The record is barren of evidence to the effect that Payne and Kinnard were other than willing sellers. Thus, discrediting Roseoe still would not enable the jury to reach the rational conclusion that defendants were entrapped into committing the criminal acts with which they were charged.1
*583Defendants also contend that the trial court committed error in not allowing cross-examination of Agent Jackson with regard to the unreliability of addict informers generally, and in refusing to give a cautionary instruction on the same subject to the jury. Counsel for defendants assert that it was not necessary to comply with the requirements of Rule 30, Federal Rules of Criminal Procedure, because to have done so would have been a futile act in view of the court’s ruling restricting the cross-examination of Agent Jackson. However, it is significant that counsel requested the standard instruction regarding informants and at one point stated: “I am not asking for an instruction on narcotics addicts or their reliability * *
Again, the purpose of the question asked of Jackson was to impeach Ros-eoe’s credibility. However, since the only evidence indicated that Roscoe, the sole informant, was not an addict, testimony pertaining to the lack of reliability of addict informants would be clearly irrelevant and not probative as to any issue in the case; therefore, it was not error for the trial judge to exclude such evidence. Similarly, assuming that the issue of the adequacy of the instructions was properly preserved for appeal, the addict-informer instruction was properly withheld, since the factual basis for it —that Roscoe was an addict, as distinguished from a user — had not been established.
Some may be troubled by the advisability of fashioning a rule that would preclude the impeachment of informants in all situations. For example, cases may arise where defendants offer affirmative evidence on the issue of entrapment. However, I express no opinion whether collateral impeachment would be proper in such situations.
. The theory advanced by Payne seems to be based on the following hypothesis: Although the jury evidently believed Roscoe with regard to Payne’s being a willing seller despite the terrible damage done to his credibility, disputed evidence that Roseoe was an addict rather than user might have been crucial to the total destruction of Roscoe’s credibility. After hearing such evidence and receiving the “paid-narcotics-addict-informer” instruction, the jury then might first have inferred that Roseoe lied about his use of drugs, and second, because Roseoe lied about the frequency of his own use of drugs, the jury might have believed he lied about Payne’s willingness, and if he so lied, Payne must not have willingly sold the drugs. Accordingly, even though the circumstantial evidence indicated that Payne was a willing seller and there was no evidence to the contrary, on the basis of the double inference drawn from Roscoe’s minimization of his drug use, the *583jury could rationally conclude that Payne’s desire not to sell drugs was overborne by Roscoe’s promises, threats, tricks or entreaties. To verbalize this hypothesis is to demonstrate its tenousness.