THE STATE OF FLORIDA, APPELLANT,

v.

MARK CLARK AND BOBBY STOKES, APPELLEES

[*501] PER CURIAM.

The state challenges the trial court’s reduction of charges of trafficking in heroin to charges of possession of heroin.

The trial court’s order declared that the state violated defendants’ due process rights when the state’s chemist commingled the contents of tested capsules with the contents of untested capsules, thereby preventing defendants from analyzing the remainder of the untested capsules. The state’s chemist randomly tested capsules contained in separate packages in Stokes’ case and in a single bag in Clark’s case. Each tested capsule contained heroin, but the total weight of the tested substance was less than the required statutory amount. The chemist then commingled the contents of the capsules taken from each package prior to weighing, and concluded that the mixture contained enough heroin to violate section 893.135(1)(c), Florida Statutes (1985).

In Ross v. State, 528 So. 2d 1237 (Fla. 3d DCA), review denied, 537 So. 2d 569 (Fla. 1988), this court held that random testing of suspect controlled substances is permissible if samples from each packet are tested. The Ross court drew a distinction between cocaine powder wrapped in separate packets and pills, stating:

[Cjaselaw which allows random positive testing of one suspect illegal pill commingled in a single packet containing other similar looking pills ... as proof that the entire packet or bag contains illegal pills ... is totally distinguishable from the random testing of only one of many separately wrapped packets of suspect cocaine.

Ross, 528 So. 2d at 1240 (citations omitted).

As we stated in Bond v. State, 538 So. 2d 499 (Fla. 3d DCA 1989), “the gravamen of Ross was its recognition that a variety of powdery white substances resemble cocaine in powder form.” Ross, 528 So. 2d at 1239 (“[A] visual examination of untested packets of this weight is insufficient to convict because the white powder contained therein may be milk sugar or any one of the vast variety of other white powdery chemical compounds not containing cocaine.”).

Capsules containing a powdery white substance are not distinguishable from the plastic packets in Ross. Ross requires that the chemist test samples from each of the capsules and that the total weight of the randomly tested material equal or exceed four grams. See § 893.135(1)(c), Fla.Stat. (1985).

The testing method employed by the chemist did not follow the dictates of Ross, and the trial court’s reduction of the charge was proper. AFFIRMED.

BASKIN and JORGENSON, JJ., concur.

SCHWARTZ, Chief Judge

(specially concurring).

Although, for the reasons stated in my dissenting opinion, I disagreed — and continue to disagree — with Ross v. State, 528 So. 2d 1237 (Fla. 3d DCA 1988), pet. for review denied, 537 So. 2d 569 (Fla.1988),1 its holding represents the law of this district which, in the absence of an en banc review (which the court as a whole has declined to grant), every subsequent panel is bound to follow. See In re Rule 9.331, 416 So. 2d 1127, 1128 (Fla.1982). Since I believe that the rule adopted in Ross applies with equal force to the facts of this case and compels affirmance, I concur in that result.

Notwithstanding the attempt to do so by the prosecution here and by Ross itself, the packets of powdery substance involved in Ross cannot be meaningfully distinguished from the capsules in this case or the solid pills treated in Asmer v. State, 416 So. 2d 485 (Fla. 4th DCA 1982). I agree with the appellant that capsules containing small bits of drug product should be treated, in effect, as miniaturized containers of powdery substance which are more closely analogous to the packets of powder in Ross than to the pills in Asmer. Much more important, however, in terms of the basis of Ross itself, both the untested capsules and the untested pills in Asmer may just as easily contain harmless substances — the other capsules may be Contac; the other pills may be aspirin — as the untested powdery substance in Ross may be flour or sugar. (This case is in this decisive respect unlike Bond v. State, 538 So. 2d 499 (Fla. 3d DCA 1989), because one cocaine rock resembles nothing but another one.) Since the possibility that the untested materials are innocent is the very foundation of Ross,2 I agree, reluctantly but mandatorily following that decision, that we must affirm. In short, since we are stuck with Ross, we are stuck with ruling with the appellees in this case as well.

. The denial of review in Ross was by a four to three vote.

. It is, perhaps ironically, a disagreement with the validity of this very assumption that is the ground for my inability to concur with Ross. See Ross v. State, 528 So. 2d at 1241 (Schwartz, C.J., dissenting). Thus, my personal view is that reversal is the correct result here and, if my dissent in Ross had prevailed, I would cheerfully join in that result.