dissent in a memorandum by Carro, J., as follows: Because the People failed to demonstrate that the vials admitted into evidence at trial were, in fact, the same vials received by the undercover officer from defendant *330in the course of a buy-and-bust operation, the vials and controlled substance contained therein, i.e., cocaine, were inadmissible. As a result, the conviction herein appealed from is predicated upon evidence insufficient to establish every element of the offense charged. Accordingly, because I would reverse the conviction and dismiss the charge, I dissent.
Before a defendant may be convicted of criminal sale of a controlled substance in the third degree, the People must adduce evidence sufficient to prove that the accused knowingly and unlawfully sold a substance containing one of the substances enumerated in Penal Law § 220.39. In this case, defendant was charged with selling four vials of a narcotic drug, cocaine, to undercover Officer Sheldon Dixon on February 9, 1988. Thus, in order for a valid conviction to be had, the People had to prove that what the defendant sold was, in fact, cocaine. (Penal Law § 220.39 [1]; 3 CJI[NY], at 1701.)
At trial, Dixon testified that after he purchased the four vials from defendant, he conducted what is known in police parlance as a "sign and seal”; that is to say, he marked each one of the vials with his initials and placed them in an envelope, which was then sealed.
At trial, Dixon was able to identify the envelope in which he had placed the vials he purchased and initialed; however, he acknowledged that his markings were not on a single one of the vials the prosecutor sought to introduce into evidence. Neither did Dixon’s testimony include any indication that even one of the vials bore any ink smudges. Regardless, in response to the prosecutor’s specific question "officer, can you tell the jury exactly what you did with People’s Exhibit Number Two for identification purposes on February 9th during the signing and sealing?”, Dixon unequivocally testified "they were—I initialed them, and then, they were vouchered.” Understandably, this conflicting testimony prompted vigorous objection from defense counsel, who correctly but unsuccessfully argued that there was no way to ascertain whether the vials presented in court were the same vials purchased by Dixon.
Monica Brooks, the chemist who analyzed the contents of the vials introduced at trial, testified that they tested positive for cocaine. Significantly, she recalled that the vials she tested were not initialed or marked in any way; had there been markings—of any sort—on the vials, she would have included that information in her report. Brooks claims she nevertheless remembered these vials, of the 1,000 similar vials per week *331she analyzed, because they were "wrapped” in tape. This was in stark contrast to Dixon’s testimony which was bereft of any statement that he wrapped the vials in tape. Defense counsel renewed his objections to the introduction of the vials; however again his arguments were unavailing.
I would note that while my colleague, writing for an affirmance, would urge that there is no discrepancy regarding the tape, the section of testimony wherein Dixon acknowledges that the vials "have tape on them” regards not the sign and seal procedure he followed, but rather the appearance of the vials at the time of trial, after the chemist conducted her analysis. Indeed, this section of testimony, in context, is reflected in the record as follows:
"Q. And is that [the vials] essentially in the same condition now as it was on February 9th?
"A. No, it’s not.
"Q. What’s different about them now?
"A. The top of it, I guess, during the lab analysis—
"mr. berman: Objection, and I move to strike.
"the court: Yes.
"Officer, try to answer the specific question that you were asked, sir, without guessing and just answer—
"Put the question again, Mr. Gordon to Officer Dixon.
"A. All right.
"Q. Officer, are the vials in the same envelope that they were on the day that you sign and sealed?
"A. Yes.
"Q. Is there a slit on the side for analysis purposes?
"A. Yes.
"Q. And are they—
"Are the vials now in an envelope in a sealed clear plastic bag?
"A. Yes.
"Q. Were they?
"A. Yes.
"Q. And the vials have tape on them; is that correct?
"A. Yes.
"Q. Are the vials themselves, Officer, in substantially the same condition now as they were on February 9, 1988?
"A. Yes.” (Emphasis added.)
Thus, it is clear the presence of the tape on the vials at trial *332was unrelated to Dixon’s testimony regarding what he himself did on the date in question, when he conducted the sign and seal.
On summation, both sides presented vigorous argument concerning the identity of the vials, with the prosecutor inexplicably urging that "whether or not Police Officer Dixon placed his initials on the vials that he purchased has nothing whatsoever to do with this sale.” The record reflects that during deliberations, the jury focused on the vials and, it may be extrapolated from their notes and requests, the question of whether or not the vials introduced were the same vials recovered from defendant. Court exhibit II was a request for the chemist’s testimony regarding the condition of the vials at the time she received them. Court exhibit III was a request for the vials themselves, along with a recharge of the elements of the crime charged and, most importantly, a readback of the chemist’s testimony regarding the tape which was on the vials when she received them. Although a subsequent note proclaimed the jury to be deadlocked, after further deliberations, which continued into the following day, the jury found defendant guilty of selling drugs to Dixon.
Defendant now appeals, correctly contending that the trial court erred in admitting the four vials into evidence. A reversal is therefore mandated, in view of the fact that the vials introduced at trial could not definitively be connected to defendant. Such a holding is necessary where, as here, the People fail to meet the requirements, as stated by the Court of Appeals, that "[w]hen real evidence is purported to be the actual object associated with a crime, the proof of accuracy has two elements. * * * [F]irst, that the evidence is identical to that involved in the crime; and second, that it has not been tampered with.” (People v Julian, 41 NY2d 340, 342-343 [1977].)
In the instant case, not only do the circumstances fail to provide reasonable assurances of the identity and unchanged condition of the evidence, but the police testimony was unwavering that while he had initialed each vial he purchased prior to placing them in the evidence envelope, not one of the vials introduced at trial bore his marking.* In People v Ruiz (162 AD2d 350), this court recently reversed a defendant’s convic*333tion of criminal sale of a controlled substance where the undercover officer did not actually identify vials introduced in evidence. In that case, the undercover officer testified that the vials he purchased had blue tops, whereas the vials introduced at trial, which tested positive for cocaine, had green tops. Although the officer was able to identify the "security envelope” into which he said he placed the vials he purchased from defendant when he vouchered them at precinct, the court declined to hold that the vials produced at trial were the vials received during the buy.
Similarly, in People v Gamble (94 AD2d 960 [4th Dept 1983]) the court reversed where there was no evidence to establish that the envelope given to the chemist was taken from the police locker or, in any event, that it was the same packet purchased by undercover officer. In another recent case, People v Steiner (148 AD2d 980, 981 [4th Dept 1989]), the court held that where condition of cocaine had changed in color and consistency, evidence presented by the People did not provide reasonable assurance of identity of the substance analyzed by two chemists, the second of whom testified that despite "proper procedure” for all chemists to initial each container when done, the first chemist’s initials did not appear on plastic bag containing drugs. (Accord, People v Heiss, 113 AD2d 953 [2d Dept 1985].)
The People cite several cases in an attempt to support the trial court’s ruling to admit the vials; these cases are, however, inapposite. For instance, in People v Newman (129 AD2d 742 [2d Dept 1987]), both the undercover officer, who obtained the narcotics from defendant, and the chemist who subsequently analyzed the drugs identified them from their respective notations on the packaging, packaging which also bore an intact seal upon arrival at the chemist’s lab. Similarly, substances containing methamphetamine and the results of associated laboratory tests were properly admitted in People v Hart (113 AD2d 966, 967 [3d Dept 1985]). That case was replete with evidence which reasonably assured the identity and unchanged condition of the substances, which "had been safely under police control at all times” (supra, at 967); this included unrefuted testimony that after concluding two separate buys from defendant, the undercover officer tagged the substance and brought it to the crime laboratory where the forensic scientist packaged, labeled, sealed and placed the drugs in a locked box, where they remained before and after testing. (Accord, People v McCutcheon, 122 AD2d 169 [2d Dept 1986] [testimony provided reasonable assurances that the *334packets analyzed were the same packets purchased from defendant].)
With the case law in mind, let us consider a hypothetical. Suppose that defendant was a charlatan rather than a drug dealer, and sold Dixon four vials containing talcum powder or baking soda. After lab tests were conducted on the contents of the vials recovered from him, he could perhaps have been charged with petit larceny, but not drug sale. This exercise presumes, as does the prosecutor’s theory of the case in the matter before us, that the vials recovered are the same vials tested and are the same vials introduced at trial. Unfortunately, neither my hypothetical nor the prosecutor’s hypothetical that defendant sold drugs could be put to the test under the circumstances presented in the case at bar.
Nor does the infirmity of the evidence merely go to the weight of the evidence, as the People argue in their brief. As now Chief Judge Wachtler observed in People v Connelly (35 NY2d 171, 175 [1974]), "inconsistent notations on the wrappers used to transmit evidence should be considered irregularities bearing only on the weight of the evidence * * *. On the other hand, the fact that the item was or might have been accessible to other persons not called as witnesses casts suspicion on the integrity of the evidence often rendering it inadmissible especially when it appears that the evidence was available to unknown persons”. (Emphasis supplied.) Thus, while "deficiencies in the chain of custody go to the weight of the evidence, not its admissibility, provided that the two basic requirements of proof and identity and unchanged condition are met” (People v Ramos, 147 AD2d 718 [2d Dept 1989]), since the identity of the evidence was not established in the instant case, the evidence is inadmissible as a matter of law.
Just as a reversal was required in Ruiz (supra, at 351) where the court unanimously held that "the jury * * * should have found that the vials so obtained [in a buy and bust] from defendant were not the ones introduced in evidence” in light of the fact that the tops of the vials become color uncoordinated, neither can the conviction in the case herein be sustained, where the initialed vials mysteriously became uninitialed. Indeed, I am of the view that facts in this case are more compelling than those in Ruiz, in view of the chemist’s testimony that she remembered the subject vials in the instant case because the vials she analyzed were wrapped in tape, an unusual identifying fact to recollect, given that Dixon never testified to wrapping the vials in tape in the course of his sign and seal. Thus, while neither the undercover officer *335nor the police chemist admitted to substituting the vials, neither was able to testify to their identical condition either, because, I would suggest, the vials purchased from defendant and subsequently marked were mixed up or otherwise interchanged with unmarked vials somewhere along the chain of custody. Consequently, there is no chain of custody linking defendant and the vials sold by him on February 9, 1988 to the vials later tested and introduced at trial. (See, People v Heiss, 113 AD2d 953, supra.)
Accordingly, the conviction should be reversed and the indictment dismissed.
Moreover, there is no evidence to support my colleague’s conclusion that "the initials, marked on the vials only in pen, were rubbed off in handling.” In any event, it strains the imagination to speculate that the initials on four out of four vials would be "rubbed off”, thus evaporating into thin air.