SPAETH, Judge:

Shortly after midnight on September 16, 1973, a young woman was walking from her home to an all-night restaurant in Erie, Pennsylvania. As she approached an overhead railroad bridge she noticed a black male pass her from the opposite direction, dressed in blue jeans and a dark T-shirt. Within a minute after passing under the bridge she heard behind her the sound of running. A man seized her by the mouth and throat from behind, pushed her under a truck, and from there up a small hill near the street, where he made her cover her head with her coat. Thus she never saw her assailant except for his arm when he first seized her, from which glimpse she knew only that he was black and wore a shirt with rolled-up sleeves. The man forced her to undress and lie on a large piece of cardboard, and then he raped her.

After her assailant left the scene the victim immediately sought help. Policemen took her to the hospital for treatment; then they went to the scene, where they retrieved two large pieces of cardboard and a pair of men's boxer shorts. Later at the police station the victim viewed an array of seven snapshots from which she identified appellant as one of two men who might have been the man who passed her on the street just before she was attacked. Around 3:00 a.m., responding to a radio call that resulted from this tentative identification, patrolmen arrested appellant in the office of a service station seven blocks from the scene of the rape. In his car they found two pairs of blue jeans and a dark blue T-shirt.

At the police station appellant was booked and strip searched. The search revealed that appellant was wearing no undershorts. He then was placed in an interrogation room. While he was there, two probative events occurred. First: As the police began to advise appellant of his Miranda rights, he told them that was unnecessary because it had already been done in the car on the way to the police station. At this moment the victim, who had been waiting in a room down the hall from the interrogation room, happened to be in the hall getting a drink from a water fountain. She heard appellant's voice and instantly stated to a policeman near her that she recognized it as her assailant's. (According to the victim's testimony at trial, the rapist had talked to her throughout the assault.) Second: Before the interrogation began, appellant asked the police, "What is this about, the rape of a girl named Linda [which was the victim's first name]?" According to all the police officers involved, no one had yet mentioned to appellant the crime under investigation or the name of the victim.

At trial, appellant defended himself in three ways. First, he offered an alibi defense. Second, he attempted to undercut the effect of his alleged question about the rape of a girl named Linda by denying that he had asked the question. (In this regard it may be noted that the investigating officer made no reference to the question in his report of the events at the police station.) Third, he presented extensive expert testimony on certain scientific tests, which will be discussed in more detail below; in brief, the results of the tests raised considerable doubt whether appellant could have been the rapist, for the tests indicated that only by remote possibility could he have been responsible for semen stains found on the victim's panties and on the piece of cardboard on which the rape took place. Notwithstanding this testimony, the jury found appellant guilty. However, because the testimony on the tests had not been entirely clear, and because certain of the expert witness's answers suggested that the tests positively excluded appellant as the rapist, the court en banc awarded a new trial.

Before the new trial began defense counsel requested production of certain physical evidence so that he could have further scientific tests made. This request led to the discovery that the police custodian had destroyed all the physical evidence; having read of appellant's conviction, he apparently assumed that the proceedings were at an end. Appellant thereupon moved for abandonment of prosecution, relying on Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963). The lower court denied the motion; a jury once again found appellant guilty; and post-verdict motions were denied. This appeal followed.

I. Scientific Evidence

Appellant argues that the lower court should have granted his motion for directed verdict because certain scientific evidence conclusively proved that he could not have been the rapist. Taken in its entirety, the record is against this argument.

A. Semen Typing Test

An FBI laboratory report showed that stains on the victim's panties and on the cardboard found at the scene were semen stains. A second test — an "absorption inhibition test" — of the stains showed that they contained a substance of the blood type "A"; since appellant's blood type is "B", he argues that the latter test excludes him.

The absorption inhibition test may be described as follows:

Under ideal conditions seminal fluid may be typed or grouped in much the same manner as blood is typed — as type A, AB, B, and O. However, unlike blood, it is not always possible to type or identify human semen as representative of one of the four basic blood groups. The person from whom the sample is taken must be a so-called "secretor". It is estimated that from 75 to 80 per cent of all human beings are considered to be "secretors"; that is, they possess, in fluids other than their blood, certain material properties that tend to be classified as would blood.

12 Am.Jur. Proof of Facts, Identification of Seminal Fluids § 12 (1962).

By testing appellant's saliva it was determined that he is a secretor. Therefore he argues that if he had done the rape the semen stains would have showed a "B" blood type. However, appellant's own expert witness (a pathologist) demonstrated that the test results are not so conclusive as appellant argues.

First, the presence of an "A" type in the stains does not necessarily mean that the rapist was an "A" type. As appellant concedes, since a saliva test of the victim showed her to be a secretor, the "A" in the stain could have come not from the rapist's seminal fluid but from the victim's vaginal fluid or blood, intermingled with the seminal fluid.

Second, on cross-examination appellant's witness testified that it was possible (although only remotely so) for a person to secrete his type in his saliva but not in other body fluids:

"Q Are you saying here that it is absolutely impossible for a person who is a secretor of the B blood type to have perpetrated a rape, had an ejaculation, and that that B blood type would not be found in that ejaculation? Is that absolutely impossible? Is that the indication?

A I would say it's an extremely strong possibility as to impossible, yes.

Q Totally impossible; one hundred percent?

A Well, that would be an extremely farfetched possibility, the possibility that a gene defect exists in that particular person that doesn't allow the normal response of the organism in producing blood substances which do appear on the red cells. However, a possibility of this kind would be an extreme; the odds would be extremely high. I couldn't tell you exactly what, but certainly in the neighborhood of one versus thousands and thousands.

Q So one in several thousands?

A Uh-hum.

Q But in which there is a possibility; an extreme one, but a possibility; an extreme possibility; one in several thousands?

A Yes.

Q So that would mean if we took a hypothetic group of very large numbers, say two or three hundred thousand, we might expect to find forty, fifty, sixty people within that group, a very small number, who might exhibit these qualities?

A Possibly.

Q And that would be that they might be, let's say, of a B blood group and secretor and yet their blood group would not appear in the seminal fluid?

A I suppose it could be possible"

N.T. 56-57.

The witness also testified that contamination of the stain by contact with air, water, or other materials could affect the accuracy of the testing. N.T. 61.

The possibility that a secretor's type will not always emerge from any stain of any of the secretor's body fluids was corroborated by the FBI laboratory technician who tested the stains:

Q So then a person who secretes into their saliva does not necessarily secret into their semen, is that right?

A Well, I found it's a rare occasion that they do not, but the saliva sample, of course, the reason we request a saliva sample is to give us an indication whether this person is a secretor or not, and we would call a person a secretor if we did find the blood group substance present in the saliva. However, we wouldn't draw the conclusion that they would have their blood group substance present in their other body fluids based on this examination. It would only be indicative of that situation.

Q So a secretor into saliva is exactly that; a secretor into saliva; and you cannot draw any further conclusions about their other fluids? A Well, we don't try to break it down as to secretor of certain body fluids. We consider; well, approximately eighty-five percent of the population of the country are known as secretors and we have no way of knowing what percentage secretes in all body fluids and what percentage secretes in just particular body fluids.

Q The concept that people may secrete in one fluid and not in others, is this something that you have developed also over your course of studies in the field, current material and literature?

A Well, again, I have found in my experiences in grouping stains that I was unable to detect blood group substances in stains where the individual was known to be a secretor. However, I can't say why I was unable to detect these blood group substances. I know that this possibility exists from my experience in the laboratory.

N.T. 83-84.

B. Gonorrhea Test

A test performed one week after appellant was arrested showed that he was suffering from acute gonorrhea. Vaginal smears taken from the victim at the hospital immediately after the rape showed no evidence of a gonorrhea infection. From these facts appellant argues that he could not have been the rapist.

On cross-examination appellant's treating physician testified that a vaginal smear from a woman who had recently had intercourse with a man infected with gonorrhea would not necessarily show signs of infection in the woman. A number of factors could account for the absence:

[T]he conditions of, first of all, the collection of the specimen at the time of the examination. How was it collected, the medium that it was placed upon, the method by which it was incubated, or examined, who examined it in terms of a technician, for instance, where it was drawn from at the time of the examination. I think there are a lot of qualifying things that go into it, I think, that reduce the incidence of finding it at that time.

Q. Would not another qualification be, for example, the number of bacteria that might have been in the specimen; if they were below a certain number they might not have been visible?

A. This is also possible based upon whether or not the individual who had gonorrhea at the time of the rape had not urinated before, for instance, and washed out his urethra where the large numbers of the bacteria might be present. This is a possibility.

Q. But if he had, for example, urinated shortly beforehand the urethra would be relatively clean and nothing would appear?

A. This is perfectly possible, particularly if the ejaculate was principally from the prostate and the seminal ducts and contained very little of the urethra washings at the particular time.

N.T. 381-82.

Thus far we have taken appellant's argument to be that the two tests discussed proved that he could not have been the rapist. As our summary demonstrates, testimony by appellant's own witnesses require rejection of this argument. Therefore appellant's motion for directed verdict was properly denied, since in passing on such a motion the test is whether viewing all the evidence in a light most favorable to the Commonwealth a jury could possibly find the defendant guilty as charged. Commonwealth v. Boone, 467 Pa. 168, 179, 354 A.2d 898, 903 (1975).

It may be, however, that appellant's argument really is that he is entitled to a new trial because the verdict was against the weight of the evidence. Whether a verdict is contrary to the evidence or the weight of the evidence so as to require a new trial is within the discretion of the trial judge. Commonwealth v. Zapata, 447 Pa. 322, 327, 290 A.2d 114, 117 (1972); Commonwealth v. Ashford, 227 Pa. Super. 351, 354, 322 A.2d 722, 723 (1974). Here, we cannot say that the lower court abused its discretion in denying appellant's motion for new trial. While the scientific evidence did raise considerable doubt whether appellant was the rapist, there was, nevertheless, evidence that the jury might have regarded — and evidently did regard — as strongly against appellant. The victim identified appellant as resembling the man who had passed her only seconds before she was attacked from behind; she recognized his voice; appellant knew the crime and the victim's name without being told; appellant was arrested near the scene not long after the rape; he was wearing no undershorts.

II. Destroyed Evidence

In Brady v. Maryland, 373 U.S. 83, 83 S.Ct. 1194, 10 L.Ed.2d 215 (1963), the United States Supreme Court held that "the suppression by the prosecution of evidence favorable to an accused upon request violates due process where the evidence is material either to guilt or to punishment, irrespective of the good faith or bad faith of the prosecution." Id. at 87, 83 S.Ct. at 1197. Later cases and commentators are agreed that this rule applies equally to a case involving destruction of evidence. United States v. Bryant, 142 U.S.App.D.C. 132, 439 F.2d 642 (1971), on remand, 331 F. Supp. 927 (D.D.C. 1971), aff'd, 145 U.S.App.D.C. 259, 448 F.2d 1182 (1971); State v. Wright, 87 Wn.2d 783, 557 P.2d 1 (1976); People v. Hitch, 11 Cal.3d 159, 113 Cal.Rptr. 158, 520 P.2d 974, vacated and rev'd upon rehearing, 12 Cal.3d 641, 117 Cal.Rptr. 9, 527 P.2d 361 (1974); Hale v. State, 248 Ind. 630, 230 N.E.2d 432 (1967); Note, "The Right to Independent Testing: A New Hitch in the Preservation of Evidence Doctrine," 75 Col.L.R. 1355 (1975). Considerable difficulty arises in such cases, however, when the defense is asked to show that destroyed evidence would have been "favorable to [the] accused." How can this be done, when the evidence no longer exists? In United States v. Bryant, supra, the United States Court of Appeals for the District of Columbia ordered an evidentiary hearing. We find a similar remedy appropriate here; to explain why, we must first discuss in some detail the nature of the evidence that was destroyed.

A. Nature of the Evidence

Appellant identifies the following items of physical evidence as destroyed by the police custodian:

1. Two pieces of cardboard: one bearing the semen stain (discussed in Part I, supra), the other showing a shoe print that did not match shoes taken from appellant.

2. A pair of shoes taken from appellant.

3. Two pairs of blue jeans taken from appellant.

4. A pair of men's undershorts found at the scene.

5. A pair of panties taken from the victim.

6. Soil samples from the scene.

7. A greeting card and envelope handled by the rapist.

8. Appellant's fingerprint card.

Appellant seems to complain that all these items would have been favorable to him, and that their destruction irreparably hindered his ability to use them to his advantage at the second trial. However, a closer look at each item reveals that appellant was in fact deprived of the usefulness of only one item, for with respect to all but that one, test reports or recorded observations were available to demonstrate what appellant presumably would have demonstrated had the evidence not been destroyed.

— Cardboard with stain. As discussed in Part I, supra, the FBI laboratory report was available, and a laboratory technician testified on the results of his tests.

— Cardboard with shoe print; shoes. An FBI laboratory technician testified that the shoe print on the cardboard found at the scene did not match the print of the shoes taken from appellant. We do not think that appellant's inability to ask the jury to judge the lack of similarity for itself — by looking at the shoes taken from appellant — could possibly amount to a denial of due process.

— Blue jeans; undershorts. These were the items that appellant requested in his petition that led to the discovery that the evidence had been destroyed. Apparently appellant intended to engage an expert to subject the jeans and shorts to a "fiber transfer test" — a procedure that demonstrates whether fibers from one garment have rubbed off onto another. By this test appellant hoped to demonstrate that the shorts found at the scene had not been worn under appellant's jeans. However, appellant was not harmed by his inability to have his expert make this test, for an FBI technician testified that such a test had been made and that it had revealed no transfer from the jeans to the shorts or vice versa. Furthermore, the test cannot prove conclusively that one garment was not worn with another; it may prove that one garment was worn with the other, or it may prove only (as here) that it cannot be demonstrated whether one was worn with the other. Thus the results stated by the technician were as good as appellant could have hoped for from a test made by his own expert.

— Panties. The results of an absorption inhibition test and a fiber transfer test were reported at the second trial.

— Soil samples. Tests indicated that the kind of soil at the scene of the rape could not be shown to be present on appellant's blue jeans or shoes. The test results were available at the second trial.

— Greeting card, envelope, and fingerprint card. Appellant's fingerprints were not observable on the greeting card or envelope. An expert's report, stating that there was no demonstrable connection between appellant and the card or envelope, was available at the second trial.

As to all the above items, we find applicable the rule of Commonwealth v. Mace, 234 Pa. Super. 463, 341 A.2d 505 (1975), cert. denied, 423 U.S. 996, 96 S.Ct. 423, 46 L.Ed.2d 370 (1975). There, in an involuntary manslaughter case, parts of the deceased's body were preserved after autopsy and sent to laboratories for further analysis to confirm the cause of death. Microscope slides were prepared from some of the body tissue that was imbedded in paraffin. After the laboratory reports were completed, the remaining body parts and some of the tissue imbedded in paraffin were destroyed by the laboratory. However, the slides and reports were available, as were the doctors who had performed the analysis. In response to the defendant's request for access to all the body parts, the Commonwealth explained that most had been destroyed; but it made available for the defendant's inspection all the hospital records, the slides, all the tissue in paraffin that remained, and it also made possible consultations with a doctor who had analyzed the tissues. We found no violation of Brady:

When all these factors are taken into consideration, it is apparent that the Commonwealth has not overborne the defendant by secretly amassing evidence or designing a case unfairly weighted against him. The good faith of the Commonwealth is attested to by the willingness with which it agreed at the hearing to make available its evidence and witnesses relevant to the cause of death. The validity of the findings can be tested by the defense by examination of the slides, remaining imbedded tissues, and hospital records as well as by cross-examination at trial of the prosecution's six expert witnesses.

(In a separate opinion, joined by Judge HOFFMAN, this writer concurred because "the failure to preserve the autopsy specimens appears to have done appellant no harm." Id., 234 Pa.Super. at 474-75, 341 A.2d at 511.)

The one piece of evidence that cannot be completely disposed of by the rule of Mace is the undershorts. Appellant wanted to do more than run a fiber transfer test on these; he wanted to be able to show them to the jury so that the jury could decide whether or not they seemed likely to be his. As defense counsel said in his summation:

I can't show you that pair of underwear. I can't show you. I can't wave that pair of underwear in front of you and say "This is a size twelve; this man is a size thirty-six" or I can't hold up a pair of fifty-six and say "These couldn't be his; he's thirty-six". I can't do that because they threw all the exhibits away.

N.T. 471.

B. Due Process

As stated at the opening of this discussion, the rule of Brady applies to cases of destruction of evidence as well as to cases of suppression of evidence. The good faith or bad faith of the prosecution is irrelevant, Brady v. Maryland, supra ("irrespective of the good faith or bad faith of the prosecution"). Here, the kind of evidence destroyed requires application of the rule of Brady. Under that rule, counsel's inability to use the undershorts in appellant's defense constituted a denial of due process if the shorts were "material to guilt" and would have been "favorable to [the] accused." We shall discuss these two factors separately.

In the absence of a clear standard of materiality enunciated by the United States Supreme Court, other courts have formulated various tests and reached varying results. Here, we see no need to decide on an explicit standard of materiality, for by any standard the shorts were material to appellant's guilt.

The guilty verdict must have been based at least in part on the evidence that appellant was wearing no undershorts when arrested and that men's undershorts were found at the scene. The other evidence of guilt — the photographic and voice identifications, and appellant's knowledge of the crime and of the name of the victim — was heavily counterbalanced by the scientific evidence, which went far towards exculpating appellant. It may therefore well be that the jury was tipped toward its finding of guilt by the evidence regarding the shorts. To deny appellant the chance to demonstrate that the shorts were not his was to deny him the only means he had to discredit important evidence against him.

We are especially persuaded of the importance of the undershorts by the fact that the Commonwealth itself was so persuaded. In his closing argument the District Attorney first reminded the jury of the various reasons why the scientific evidence did not conclusively exclude appellant as the rapist. Then, after reviewing the photographic and voice identifications, he argued:

Now, if that's all there were in the case we might have some doubts, but there's more, and it's the more that's in this case that ought to absolutely lock it up in your mind. John Chapman has no shorts on when he's brought in. There were a pair of shorts up at the scene with her hairs on it. Now, Mr. Schroeck [defense counsel] argues here an unusual argument;

"Lot's of people walk around without their shorts on; lots of people take their clothes off, I guess; young people like to go around without different pieces of clothing on", but, you know, I didn't hear Mr. Chapman tell us why he didn't have his shorts on. . . . There's no explanation on the record for that, and you can infer from that that the reason there is no explanation is because John Chapman's shorts on that night were up in that field where he left them after he raped Linda . . ., and there's other evidence to corroborate here when we're talking about corroboration, determining if people told the truth.

N.T. 508-09 (emphasis added).

Thus the District Attorney, recognizing the force of the scientific evidence, and the arguable uncertainty of the evidence of identification, focused the jury's attention on the shorts as the one piece of evidence that should lead it to be convinced of appellant's guilt beyond a reasonable doubt.

The question whether the undershorts would have been "favorable to [the] accused" is more difficult, for in answering it we must confront an aspect of this case that differentiates it from Brady and from other destruction of evidence cases: Here, appellant once had access to all the destroyed evidence. He therefore had an opportunity to observe the undershorts and to decide whether to use them demonstratively before the jury. At the first trial, however, appellant declined that opportunity. We must therefore assess how this action — or failure to act — should affect our decision.

As we undertake this assessment, at least this much is clear: defense counsel's failure to use the undershorts demonstratively at the first trial did not foreclose his ability to use them at the second trial. This is so because a new trial must be "new", that is, as if the first trial had never happened. "The granting of a new trial sets aside the verdict . . . and leave[s] the case as though no trial had been held." Commonwealth ex rel. Wallace v. Burke, 158 Pa. Super. 612, 613, 45 A.2d 871 (1946). Accordingly, at a new trial the defendant is not to be confined to the defense he used at the first trial. 5 Am.Jur. Appeal and Error § 955 (1962); 24 C.J.S. Criminal Law § 1426 (1961); United States v. Mischlich, 310 F. Supp. 669 (D.C.N.J. 1970), aff'd, 445 F.2d 1194 (3d Cir. 1971), cert. denied, 404 U.S. 984, 92 S.Ct. 449, 30 L.Ed.2d 368 (1971).

Furthermore, counsel's failure at the first trial to attempt to demonstrate by reference to the size and appearance of the undershorts that they were not appellant's is understandable. At the first trial the prosecution, although it offered the shorts in evidence, made no special reference to them. At the second trial, by contrast, the prosecution put a great deal of emphasis on them, as has been seen from the excerpt from the prosecution's closing argument quoted above.

To put all this another way: if the prosecution is free to change its strategy and emphasis at the second trial — as it is, and as it did here — so is the defense. In such a case, the United States Supreme Court has said that on a retrial the prosecution "is free to strengthen its case in any way it can by the introduction of new evidence . . . [but] of course [defendants] have a reciprocal opportunity." United States v. Shotwell Manufacturing Co., 355 U.S. 233, 243, 78 S.Ct. 245, 252, 2 L.Ed.2d 234 (1957) (emphasis added).

Appellant argues that because he was entitled to use the undershorts at the second trial, but could not because the Commonwealth had destroyed them, he should be discharged. This argument, however, skips a step, or begs a question: it assumes, or asks us to assume, that because at the second trial counsel wished to use the shorts, he would have used them, had he had them. If we knew, that is, if the record showed, that the shorts would have been evidence favorable to appellant, we might be willing to make this assumption. However, the record does not so show, and given that circumstance, to discharge appellant would be to impose upon the Commonwealth a sanction quite possibly disproportionate to its conduct in destroying the evidence. We agree with the Supreme Court of Washington that "[t]he range of sanctions available in suppression and destruction of evidence cases should be broad and, of necessity, will be developed over time." State v. Wright, supra, 87 Wn.2d at 792, 557 P.2d at 7.

Some attempt must be made to establish whether or not the undershorts would have been "evidence favorable to [the] accused." Brady v. Maryland, supra, 373 U.S. at 87, 83 S.Ct. 1194. In United States v. Bryant, supra, the evidence destroyed was a tape recording of a drug transaction, which the defendant requested for purposes of disputing the drug agent's version of the details of the sale. Rather than discharging the defendant, the Court of Appeals remanded for a further hearing. On this remand, the trial court was satisfied that the tape recording had been so inaudible or garbled that it could not have been of any use to the defendant. It therefore upheld his conviction, and on further appeal the Court of Appeals affirmed.

We choose a similar remand here. The lower court should hear testimony from all available witnesses to determine what can be remembered about the appearance of the undershorts. If the court is satisfied that appellant could not have plausibly demonstrated that they were not his — in other words, if the testimony shows that the shorts appeared to be appellant's — then appellant is entitled to no further relief, and the judgment of sentence may be reinstated. If, however, the testimony is such that the lower court cannot say with assurance that the shorts would not have been favorable evidence for appellant, or if the court is persuaded that their appearance actually would have shown that they were not appellant's then the appropriate remedy is to grant appellant a new trial under such conditions as the lower court may think proper. Without limiting the court's choice, we suggest consideration of two possible conditions: exclusion of all mention of the shorts, which might be appropriate if the court finds the evidence about the appearance of the shorts inadequate, see People v. Hitch, supra; People v. Harris, supra note 10, or a missing-evidence instruction, which might be appropriate if the court finds that appellant could have demonstrated that the shorts were not his, see Comment, "Judicial Response to Governmental Loss or Destruction of Evidence," 39 U.Chi.L.R. 542, 564 (1972).

We recognize that appellant may complain that the remedy of a new trial under either of these (or comparable) conditions would still deny him the opportunity to prove conclusively that the undershorts belonged to someone else, and perhaps thereby to persuade the jury that that someone else was the rapist. We think, nevertheless, that for his failure to investigate the shorts at the first trial, appellant must cede some ground. Due process is, after all, the legal concept that embodies our notions of what is fair. Wardius v. Oregon, 412 U.S. 470, 476, 93 S.Ct. 2208, 37 L.Ed.2d 82 (1973); Kinsella v. United States, 361 U.S. 234, 246, 80 S.Ct. 297, 4 L.Ed.2d 268 (1960); Rochin v. California, 342 U.S. 165, 169, 72 S.Ct. 205, 96 L.Ed. 183 (1951); Commonwealth v. Coder, 252 Pa. Super. 508, 514, 382 A.2d 131, 135 (1977) (Concurring Opinion by SPAETH, J.). Here we decide that it is fair, on the one hand, to refrain from penalizing the prosecution to the extent of discharging appellant, but on the other, to give appellant some relief in recognition of the possible hindrance to his defense.

Reversed and remanded for proceedings consistent with this opinion.

JACOBS, President Judge, and PRICE, J., dissent.

WATKINS, former President Judge, did not participate in the consideration or decision of this case.