THE STATE OF FLORIDA, APPELLANT,

v.

AUGUSTINE LASARO ARROYO, MORGAN RANDOLPH TYNES, AND MARCOS ALFREDO ARCE, APPELLEES

DANIEL S. PEARSON, Judge.

The State appeals from an order granting the defendants’ motion to suppress. The issue before us is the admissibility vel non of an experiment conducted by the defense to impeach the testimony of police officers that two automobiles, each containing an approximate 300-pound load of marijuana in its trunk, appeared to be weighted down. We hold that the evidence of the experiment was inadmissible and reverse and remand for further proceedings.

Ordinarily, where a trial judge, sitting, as here, as the fact-finder, erroneously admits evidence, he is presumed to have disregarded the improperly admitted evidence, and the error of its admission is deemed harmless. Wythers v. State, 348 So. 2d 390 (Fla. 3d DCA 1977); Capitoli v. State, 175 So. 2d 210 (Fla. 2d DCA 1965); United States v. Dillon, 436 F. 2d 1093 (5th Cir.1971); United States v. Menk, 406 F. 2d 124 (7th Cir.1969); United States v. Krol, 374 F. 2d 776 (7th Cir.1967); United States v. Mitchell, 297 F. 2d 407 (2d Cir.1962); Teate v. United States, 297 F. 2d 120, 121 (5th Cir.1961).

Where, however, the record discloses that the trial judge relied upon the erroneous evidence, this presumption is overcome. United States v. Vaughan, 443 F. 2d 92 (2d Cir.1970); United States v. Turnipseed, 272 F. 2d 106 (7th Cir.1959).

See also Wong Sun v. United States, 371 U.S. 471, 83 S.Ct. 407, 9 L.Ed.2d 441 (1963) (possibility that trial judge erroneously considered co-defendant’s statement requires reversal).

In the present case it is abundantly clear that the defense experiment played a significant part in the trial judge’s ruling that the police were without any well-founded suspicion to stop the two vehicles driven by the defendants Arce and Tynes. During the argument on the defendants’ motion to suppress, the court commented:

“One aspect of this case that bothers me ... and that is, it seems like, as I recall, they stipulated to almost all of the facts, and the key would be driving into the warehouse, and then the car [sic] coming out.

“Now, I saw the demonstration, and the testimony was about three hundred and something pounds in each car, and that didn’t do much to the trunk. Then he had two fellows get in and they added another about 350 or 400 pounds, and the car [sic] didn’t drag. You know they make cars pretty good. They weren’t dragging the ground.”

In his written order granting the motion to suppress, the trial judge stated:

“In their ‘Organized Crime Bureau Report’ and sworn depositions, the officers who surveilled the warehouse testified that both vehicles were ‘heavily weighted in the rear.’ (Report at p. 4).

Lead Detective D’Azevedo, in fact, testified ‘as they got outside the warehouse is the first time that I really felt confident and assured that they had the narcotics because I could see the back of the vehicles weighed very, very heavily and almost dragging the ground. Prior to that time I didn’t feel I had probable cause.’ (D’Azevedo deposition p. 31).

This court has had the opportunity, pursuant to the Defendants’ demonstration, to observe the effects of weights equal to the weight of the marijuana in this case placed in the trunks of vehicles identical to those involved in this case. The demonstration showed, in fact, that the effects of 325 pounds in the trunks of these vehicles has a neglible [sic] effect on their appearance and that as much as 625 pounds does not cause the rear of either vehicle to even approach ‘dragging the ground.’ ”

While it is possible that the trial judge’s disbelief of the officers’ testimony that the vehicles seen by them “weighed very, very heavily” was not the sine qua non of his decision, his reliance on the experiment to impeach the officers’ testimony is inseparable from his various conclusions that there was no well-founded suspicion to justify the stops; that the stops were unjustified by any knowledge or information that the defendants were involved in any violation of Florida law; and that there was an absence of probable cause.

We turn now to the experiment. The automobiles which were stopped by the police were described as a rented 1979 Thunderbird and a rented 1980 Chevrolet Monte Carlo. The police officers observed these cars moving down a warehouse ramp. For all that appears in this record, the experiment admitted by the trial court was similar to the event to which it was being compared in only a single respect — a rented 1979 Thunderbird and a rented Chevrolet Monte Carlo of unknown year were used. As the State pointed out in objecting to the admission of the experiment and in urging error here, there was no showing, inter alia, that (1) the condition or type of the suspension system, shock absorbers or springs on the cars observed and the cars used in the demonstration was substantially similar; (2) the appearance of a vehicle with weight in its trunk is the same whether that vehicle is stationary or proceeding down a ramp; and (3) the distribution of weight in the trunks was substantially similar.1

Where the conditions attending the alleged occurrence and the experiment are not shown to be similar, evidence of the experiment to test the truth of testimony that a certain thing occurred is inadmissible. Alston v. Shiver, 105 So. 2d 785 (Fla. 1958); Hisler v. State, 52 Fla. 30, 42 So. 692 (1906); Spires v. State, 50 Fla. 121, 39 So. 181 (1905); Burgin v. Merritt, 311 So. 2d 688 (Fla. 3d DCA 1975).

The party offering the evidence of the experiment has the burden to lay a proper foundation for its admission by showing a similarity of the circumstances and conditions. Jackson v. Fletcher, 647 F. 2d 1020 (10th Cir.1981); Navajo Freight Lines v. Mahaffy, 174 F. 2d 305 (10th Cir. 1949).

As the Florida Supreme Court said in Hisler v. State: “The similarity of circumstances and conditions go to the admissibility of the evidence, and must be determined by the court. If, in the discretion of the trial court, such proffered evidence is rejected, the appellate court will not review the ruling, unless an abuse of discretion appears. But where such evidence is admitted over proper objections, and the rule as to similarity of circumstances and conditions attending the occurrence and the experiment does not appear to have been complied with in admitting the evidence, the appellate court will review the ruling, and if error be found therein and it does not appear from the whole record that no harm could have resulted to the defendant from the admission of such evidence, the error may cause a reversal of judgment.” 42 So. at 695.

Thus, a judge cannot simply “use his discretion to decide that despite a plain lack of substantial similarity in conditions he will, nevertheless, admit the evidence.” Love v. State, 457 P. 2d 622 (Alaska 1969).

In the present case, the defendants failed to show either that there was substantial similarity between the conditions of the experiment and the conditions surrounding the events observed by the police 2 so as to fulfill the requirement of basic relevancy or that this lack of similarity was inconsequential. This is not to say, however, that were the proper showing for the introduction of the experiment made that it could not be admitted. Since the trial judge’s precipitous admission of the experiment may have lulled the defendants into not presenting evidence which may have qualified the experiment for admission, we think it appropriate to reverse the order granting the motion to suppress and remand for further proceedings on the defendants’ motion to suppress.

Reversed and remanded.

. It may very well be that some dissimilarities can be shown through expert testimony to be without significance or to create, at most, a minor variance in the result. The point is that it is up to the proponent of the experiment to make this showing.

. Of course, the deficiency in establishing the condition of the observed cars could hardly be remedied by the trial judge’s act of erroneously taking judicial notice of “the fact that I have rented hundreds of cars ... but I have never had a defective car with shocks and things like that.” This fact is neither one “generally known within the territorial jurisdiction of the courts,” see § 90.202(11), Fla.Stat. (1979), nor one “capable of accurate and ready determination by resort to sources whose accuracy cannot be questioned,” see § 90.202(12), Fla.Stat. (1979), and therefore is not one which may be judicially noticed.