dissenting. The apparent holding of the court today is that the state may prove that a substance is a drug solely on the basis of the testimony of an expert witness that the substance looked like a drug, even though the police have actual custody of the substance itself and no reason is given why a definite chemical test could not be performed. Such a sweeping dismissal of the need for positive proof is neither warranted by the facts of this case nor is it supported by the authorities cited by the majority. For that reason, and because I believe the defendant was unfairly prejudiced by the state’s rebuttal testimony, I would reverse the judgment and remand the cause for a new trial.
The majority opinion states in dicta that drug identification may be proved circumstantially and without expert testimony based upon scientific analysis. Standing alone, this statement is not objectionable, for in many cases circumstantial evidence will be both substantial as proof and be the only evidence available. Many of the cases cited by the majority, for example, involve prosecutions for sale of narcotics where a frequent drug user actually took the drug and was sufficiently expert to make an identification. E. g., People v. Robinson (1958), 14 Ill. 2d 325, 153 N. E. 2d 65; United States v. Gregorio (C. A. 4, 1974), 497 F. 2d 1253. Cf. State v. Hutton (1972), 7 Wash. App. 726, 502 P. 2d 1037.
In others, circumstantial evidence was held sufficient when the substance itself was excluded from evidence, People v. Galfund (1968), 267 Cal. App. 2d 317, 72 Cal. Rptr. 917, or had been thrown away by the defendant and lost. People v. Marinos (1968), 260 Cal. App. 2d 735, 67 Cal. Rptr. 452. In these cases, the testimony of a properly qualified expert has been held to be sufficient identification. But none of the cases cited by the majority support the argument that expert testimony can replace an actual scientific test when the substance in question is in the custody of the police, and when no adequate reason is given why it should not be positively identified.6 (See State v. *489Ciesieleski [1964], 18 Ohio App. 2d 85, 247 N. E. 2d 321.) Conceivably, the effect o£ the majority opinion eonld even be taken to mean that the police need not perform any tests on purported drugs and need not preserve a seized substance once they have decided to their own satisfaction that it is a narcotic drug.
If the court’s opinion applies only to identification of marijuana, I would require more proof than a quote from Corpus Juris Secundum that marijuana is so unmistakable, in all its forms, that expert testimony can be an adequate substitute for scientific tests, or that if tests are to be carried out, the burden lies on the defendant to do so for rebuttal, assuming he is even afforded such an opportunity.
If the majority opinion is based upon a procedural distinction between evidence introduced as part of a case in chief, and that introduced' for rebuttal, I must also disagree. Here, one sample of marijuana was properly identified by chemical tests and admitted in evidence as part of the state’s case in chief. The dispute in this ease is whether the rebuttal testimony of a police officer that he saw the contents of a pouch, seized by the police in defendant’s apartment on May 25th, was sufficient as identification evidence to prove the contents to be marijuana. That testimony was offered solely for the purpose of challenging the credibility of the defendant’s testimony, since the defendant is not here charged with any offense arising out of the May 25th inciden!. Defendant testified that he had no marijuana in his apartment during May. If the substance in the pouch was marijuana, it would rebut that *490testimony and would be admissible under Harris v. New York (1971), 401 U. S. 222. If the substance was not marijuana, the rebuttal failed to rebut.
It appears to me that in this case fundamental fairness requires that the identity of the substance be as conclusively proven as in a case in chief. The state’s failure to produce evidence in its custody inevitably reduces the possibility of effective cross-examination by the defense. Perhaps even more importantly, the admission of such testimony must strongly influence the jury. Any weak or doubtful case could be propped up by the charge that on a different occasion the defendant committed the same offense. The danger of prejudice from this sort of charge is so great and so apparent that simple fairness requires that the vital evidence of such charge at least be produced in court, and that the limited relevance of such evidence be carefully explained to the jury. The instant case is a clear example. Defendant may very well have been effectively convicted of possession of marijuana, not on the basis of the incident for which he was charged, but because of the testimony of an officer that a particular pouch contained marijuana, based solely upon the observation of that officer, while the pouch itself presumably lay somewhere in the police evidence locker. Simple fairness seems to me to require that the substance itself or a scientific test of it be presented in court, unless some good reason for not producing it is given by the state, regardless of whether as a procedural matter the evidence relates to the state’s case in chief or to rebuttal.
For the same reasons, the jury instruction requested by the defendant should have been allowed. That instruction asked both that the jury disregard the rebuttal testimony because of the state’s failure to prove the substance to fee marijuana, and that the jury not consider the incident on the 25th of May as supportive proof of the charge that the defendant possessed marijuana on the date charged, the 17th of May.
It requires little reflection to see that the admission *491in evidence of the testimony as to the May 25th incident could be highly prejudicial. Even assuming that the testimony was admissible at all absent a positive scientific test, the testimony was relevant solely for the purpose of rebuttal. It did not qualify as evidence of a similar act under R. C. 2945.59 and was not substantive evidence of the offense charged. Defendant was entitled to have the jury carefully instructed as to this difficult distinction, and that was substantially the import of the preferred instruction. Indeed, the defendant was entitled to a fuller instruction than was in fact requested, because the dangers of prejudice and confusion were so great.
Both for the admission of the unsupported identification testimony and for the failure to charge the jury as to the limited relevance of that testimony, the conviction in this ease should be reversed.
W. Brown, J., concurs in the foregoing dissenting opinion.The only case cited in which the best evidence rule is considered *489with respect to drug identification is Francis v. United States (C. A. 10, 1956), 239 F. 2d 560. In that case, the issue was whether the testimony of a chemist was sufficient evidence that certain marijuana seeds were capable of sprouting. The chemist testified that he had raised seedlings from the seeds, and the objection was that the seedlings themselves were the “best evidence.” The court properly ruled that the only real proof as to the results of the experiment was the testimony of the experimenter himself, since the probative value of any seedlings depended wholly on his testimony and that their admission would only be cumulative. In the instant case, the issue is the identity of the substance itself, not one of its properties, and no valid test of any kind was offered as evidence.