delivered the opinion of the court:
This is an action under the Administrative Review Act (Ill. Rev. Stat. 1973, ch. 110, par. 274) for review of a decision of the Illinois Racing Board finding veterinarian Walter D. Wilkey guilty of violating racing board rules and its order suspending and revoking his license to practice on Illinois race tracks.
On March 7, 1975, Dr. Wilkey was notified that he was charged with 119 violations of racing board medication rules during the 1974 racing season. Nine charges were subsequently added. The alleged violations concern four racing board rules:
Charges 1 — 52 allege possession and administration of drugs intended for human use without obtaining permission of the State veterinarian in violation of Rule 307;
*537Charges 53 — 57 allege failure to report administration of phenylbutazone in violation of Rule 309a;
Charge 60 alleges reporting of injections of phenylbutazone when none were given, a violation of Rule 310. (Related charges 58 and 59 were dropped during the administrative hearing.);
Charges 61 — 128 allege administration of apomorphine, an unpermitted drug, to certain horses shortly before races, a violation of Rule 301.
In an order dated November 12, 1976, after a full hearing on these charges, the racing board found Dr. Wilkey guilty of the violations charged and suspended his license for a total of seven months for the first three groups of charges, and revoked his license indefinitely for the last series of charges.
Dr. Wilkey filed a complaint for administrative review iii the circuit court of Cook County, and on January 10,1977, the circuit court entered an order reversing the racing board’s decision and its order of suspension and revocation. It is from this order that the racing board appeals.
For the reasons stated below, the judgment of the circuit court of Cook County is affirmed.
The most serious of the charges against Dr. Wilkey are those alleging that he injected the stimulant apomorphine into 68 horses shortly before their races. Injection of such a forbidden stimulant is a violation of Racing Board Rule 301.
The case against Dr. Wilkey may be divided into two parts: first, direct evidence that apomorphine was found in post-race samples of urine of horses under Dr. Wilkey’s care; and second, circumstantial evidence that Dr. Wilkey administered such injections. Because the facts relating to these two points are complex, it is useful to consider them sequentially.
We turn first to the direct evidence that apomorphine was present in post-race urine samples of horses under Dr. Wilkey’s care. The crucial portions of this evidence are as follows:
Post-race urine samples were sealed and kept in a laboratory assigned the task of performing tests to determine the presence of drugs. This laboratory performed these functions on behalf of the racing board.
For each horse there were two urine samples and one blood sample. The last of the samples with which we are concerned was received by the laboratory in November 1974.
During the period of February through April, 1975, certain tests performed on these urine samples indicated to laboratory experts the presence of apomorphine. In every case, however, these tests were performed on samples whose seals had been broken.
There was testimony that the reason these seals were broken was that these samples had been tested previously. The results of these earlier tests *538were not produced in evidence, but there was some testimony indicating that these earlier tests were of a kind that would not have yielded results as conclusive as those obtained from the later tests.
The evidence shows that between the first and second series of tests the samples were stored in a locked freezer to which a significant number of people had access. There was no evidence that the samples had been tampered with, but because their seals had been broken it was impossible for any of the witnesses to state positively that no tampering had occurred. There was testimony that, had someone added pure apomorphine to the samples, this would have been detected in the tests, but it was apparent that, had the contents of the samples been replaced with other urine specimens, this would not have been detected by the tests.
Meanwhile, for each horse in question there existed also a sealed urine sample and a sealed blood sample, neither of which was ever tested. A method for testing blood samples was operational prior to the racing board hearing on these charges, but the samples in question were not subjected to it because the laboratory was proceeding to test samples in chronological order from an earlier date and had not yet come to them. The sealed urine samples were not tested because they were “referee” samples, intended for use by owners, trainers or other persons disputing a finding by the racing board that a horse had been illegally drugged.
Dr. Wilkey urged the racing board to test these samples, but the board refused to do so. Instead, it offered them to Dr. Wilkey, so that he could have them tested.
Dr. Wilkey concedes that the racing board had sufficient grounds to find that apomorphine was detected- in the contents of the unsealed samples, but argues that under the evidence it was not proved that the contents of these samples was unadulterated urine taken from horses under his care immediately after their races.
In examining all of Dr. Wilkey’s challenges to the determinations of the racing board, we are mindful that a reviewing court will not set aside the findings of an administrative agency if they are supported by substantial evidence. (Drezner v. Civil Service Com. (1947), 398 Ill. 219, 75 N.E.2d 303; Community Unit School District No. 1 v. County Board of School Trustees (1956), 11 Ill. App. 2d 561, 137 N.E.2d 874.) As to the board’s finding that Dr. Wilkey administered apomorphine to horses prior to their races, we note that this charge alleged conduct in violation not only of a racing board rule, but also of section 36 of the Illinois Horse Racing Act of 1975 (Ill. Rev. Stat. 1977, ch. 8, par. 37 — 36), which makes such conduct a Class 4 felony. There is authority that, where the conduct charged also constitutes a crime, a determination of guilt by an administrative agency must be supported by clear and convincing evidence. Drezner v. Civil Service Com.
*539Viewed in isolation, the results of the second series of laboratory tests of the urine samples would certainly constitute substantial evidence to support a finding that horses under Dr. Wilkey’s care had been injected with apomorphine prior to their races. However, in determining whether evidence is substantial, it must be viewed in connection with all the circumstances of the case adding to or detracting from its credibility. Oakdale Community Consolidated School District No. 1 v. County Board of School Trustees (1957), 12 Ill. 2d 190, 145 N.E.2d 736.
In examining the evidence in this case, it is impossible to ignore that the results of the second series of tests are tainted by doubts because the seals on the samples were broken long before these tests were performed, creating a possibility of tampering. Further, the absence of the results of the first tests, which were the occasion for breaking the seals and which were also aimed at detection of drugs, creates confusion as to the significance of the second tests. And finally, the fact that samples still under seal were not tested in order to confirm the less than certain results of the second series of tests is in itself troubling.
We have no difficulty in saying that, under these circumstances, the evidence that the urine of horses under Dr. Wilkey’s care contained apomorphine was less than clear and convincing.
On the contrary, we find that it was not even substantial. Under Illinois law, the failure of a party to introduce evidence which would conclusively settle a doubtful issue gives rise to a presumption that that evidence, if produced, would be adverse to that party (Belding v. Belding (1934), 358 Ill. 216, 220-21, 192 N.E. 917). This rule is applicable in both civil and criminal cases (see, e.g., Central Waxed Paper Co. v. Industrial Com. (1965), 32 Ill. 2d 154, 203 N.E.2d 900; People v. Lewis (1964), 30 Ill. 2d 617, 198 N.E.2d 812), and is particularly appropriate here, where the decisive evidence — testing of the still-sealed blood and urine samples — is in the hands of the prosecuting party. Such a presumption arises even where the favorable or unfavorable character of the evidence is not known to the withholding party, as in personal injury cases where a party refuses to undergo a physical examination (Schlechte v. Chicago Electric Transit Co. (1910), 157 Ill. App. 181).
The logic of this rule is obvious: A party who seeks to convince a trier of fact of the truth of an allegation must put forth the best possible evidence for his cause. In the case of documents, this reasoning has led to adoption of the so-called best-evidence rule, under which inferior evidence may be altogether barred. See First Trust & Savings Bank v. Powers (1945), 325 Ill. App. 600, 60 N.E.2d 582, revd on other grounds (1946), 393 Ill. 97, 65 N.E.2d 377.
The general wisdom of this rule is widely appreciated (see 29 Am. Jur. 2d Evidence §§175-189 (1967)), and its application in a case where the party withholding pivotal evidence is the prosecutor is particularly apt, *540since the prosecutor’s position is one of public trust and he is not merely a partisan but rather an advocate of justice. Because the prosecutor is forbidden to conceal from a defendant in a criminal case any substantial evidence favoring his defense (Supreme Court Rule 412(c), 58 Ill. 2d R. 412(c), Ill. Rev. Stat. 1975, ch. 110A, par. 412(c)), it is especially appropriate to apply against the prosecution an adverse presumption where the prosecutor suppresses evidence potentially favorable to an accused. Likewise, since the prosecutor is forbidden to seek a criminal conviction of a person whose innocence is known to him (People v. Mortenson (1922), 224 Ill. App. 221) it is proper that, where the prosecutor could have determined decisively the guilt or innocence of a defendant, yet elects instead to prosecute on the basis of imperfect evidence, his efforts should be checked by an adverse presumption.
This case, to be sure, is not a criminal case, although it is a prosecution for conduct which constitutes a crime and for which the penalty of loss of the right to practice veterinary medicine in one’s chosen field — caring for race horses — can be imposed. However, the adverse presumption which arises against a party who withholds evidence is applicable even in civil cases and even against a defendant. (See Miller v. DeWitt (1965), 59 Ill. App. 2d 38, 208 N.E.2d 249, aff'd in part & rev’d in part on other grounds (1967), 37 Ill. 2d 273, 226 N.E.2d 630.) It is correspondingly less drastic in impact than the absolute bar applied to prosecutors in criminal cases because the presumption which arises may be rebutted by a reasonable explanation of the failure to present the evidence in question. Helbig v. Citizens’ Insurance Co. (1908), 234 Ill. 251, 84 N.E. 897; Dominick v. Behrends (1970), 130 Ill. App. 2d 726, 264 N.E.2d 297.
An examination of the record of the hearing before the racing board discloses, however, no reasonable explanation for the failure to test the still-sealed samples. The blood samples were not tested because they were not given priority treatment under a routine retesting program. The sealed urine samples were not tested because it was racing board practice to offer these samples to other persons connected with the horses in question for independent analysis. The racing board would not have these referee samples tested, but would permit others to do so.
Offering Dr. Wilkey the opportunity to have these samples tested does not prevent the presumption that, had they been tested, the results would have been adverse to the prosecution’s case because such a presumption arises whenever the evidence withheld was primarily under the control of the party failing to produce it. Belding v. Belding; Dollison v. Chicago, Rock Island & Pacific R.R. Co. (1976), 42 Ill. App. 3d 267,355 N.E.2d 588.
In the present case it should be noted that, had a test ordered by the *541racing board produced no trace of apomorphine, the absence of that drug would have been decisively established, whereas had the same results attended tests done at Dr. Wilkey’s behest, the impartiality of these tests would doubtless have remained an issue. Thus, had Dr. Wilkey accepted the samples and had them tested, the issue would not have been decisively settled and the purpose of penalizing the withholding party by erecting an adverse presumption would not have been accomplished.
Moreover, it must be remembered that, although this is not a criminal prosecution, the burden of proving Dr. Wilkey’s guilt was on the prosecution and the offering of the samples to Dr. Wilkey amounted to an attempt to persuade him to prove his innocence. A defendant is under no obligation to do so (Lucas v. Bowman Dairy Co. (1964), 50 Ill. App. 2d 413, 200 N.E.2d 374).
Accordingly, the explanation offered for failure to test the remaining sealed blood and urine samples is far from reasonable. The evidence shows that the racing board was aware of the desirability of further tests as a check on the accuracy of results achieved under its imperfect laboratory procedures and that samples were reserved for this purpose, yet the racing board refused to authorize confirmatory tests by its own laboratory, choosing instead to offer Dr. Wilkey an opportunity to generate further indecisive evidence.
The prosecution having failed to offer a reasonable explanation for failure to produce this evidence, the presumption that it would have been adverse to the prosecution’s case must stand.
This presumption renders insubstantial the prosecution’s evidence regarding the second series of tests of the unsealed urine samples because it must be presumed that, had the remaining sealed samples been tested, no apomorphine would have been detected. Sealed samples being more tamper-proof than unsealed samples, the presumptive results of the hypothesized tests would more than counterbalance the results of the tests on unsealed samples. It would be undeniably more probable that no apomorphine was present in the post-race urine samples than that it was.
The adverse presumption arising from failure to produce evidence does not, however, prevent a party from prevailing if there is sufficient evidence not tainted by this presumption. (People v. Scott (1967), 38 Ill. 2d 302, 231 N.E.2d 441.) It is therefore necessary to consider the prosecution’s second line of argument, based on circumstantial evidence.
The strongest evidence produced by the prosecution in this regard consists of curious notations on bills to many of the owners of horses whose unsealed samples yielded positive results and testimony that Dr. Wilkey administered injections of unidentified substances shortly before races to some of the horses whose urine tested positive.
The curious notation, “P.R. injections,” appeared on bills to owners of *542all but two of the owners whose horses tested positive. However, those notations also appeared on bills for treatment of 27 horses whose urine tested negative. Although the prosecution saw the possibility that “P.R. injection” could mean “pre-race injection” as cumulatively establishing these notations as a code for illegal prerace drug injections, we find the degree to which these notations coincide with affirmative test results unimpressive, so that the remainder of the prosecution argument on this point is mere speculation.
The testimony describing Dr. Wilkey’s giving prerace injections to horses came from two witnesses. Neither could identify the substance injected nor provide any details regarding these injections that would point to apomorphine as the substance injected. Accordingly, it is mere conjecture that these injections were of apomorphine.
We find that the circuit court was correct in holding that there was no substantial evidence that Dr. Wilkey administered illegal prerace injections of apomorphine to horses under his care.
The next set of charges against Dr. Wilkey alleges that on 52 occasions he possessed or used human drugs on race tracks without permission in violation of Racing Board Rule 307.
The case against Dr. Wilkey under these charges necessarily addresses two elements: possession or use of these drugs and absence of permission. We need consider only the latter issue because we find that the prosecution utterly failed to prove this essential element.
The sole evidence that Dr. Wilkey’s use of these drugs was without permission was testimony by Dr. Ronald Jensen, State veterinarian, who had the power to grant such permission. Although Dr. Jensen stated that he had never given permission to Dr. Wilkey to use human drugs, he also admitted making statements to Dr. Wilkey, which, in our opinion, are inconsistent with Dr. Jensen’s position that he never gave such permission.
First, Dr. Jensen conceded that he was confused regarding the validity of Racing Board Rule 307 in view of a letter he had received from the United States Food and Drug Administration stating in part:
“There is no restriction on a veterinarian if he chooses to use a drug that is intended for human use. If it satisfies his needs, his decision to use it falls within the freedom to practice veterinary medicine in accordance with his best judgment as a licensed veterinarian.”
Dr. Jensen also stated he made copies of this letter available to all veterinarians. He stated that he was not pleased with Rule 307, but said that he did not recall any conversation with Dr. Wilkey in which he expressed the view that veterinarians could, as Dr. Jensen saw it, administer any human drugs they chose.
Finally, when asked whether he may have conveyed his confusion *543regarding Rule 307’s effect to track veterinarians, Dr. Jensen replied that it was never his intention to do so, or to approve all human drugs.
This testimony indicates that Dr. Jensen was aware that his attitude toward Rule 307 may have been conveyed to track veterinarians in a way that would be interpreted as blanket permission to use human drugs. He was unable to say with any certainty that he had never granted such permission, but only said he did not recall doing so and did not intend to do so. Accordingly, Dr. Jensen’s testimony cannot be characterized as substantial evidence that Dr. Wilkey had no permission to use human drugs.
Because absence of permission was an essential element of the conduct charged and it was not established by substantial evidence, we find that the circuit court was correct in its determination that these charges were not supported by substantial evidence.
Five charges against Dr. Wilkey relate to a violation of Racing Board Rule 309a, which in part required reporting the administration of phenylbutazone within a race track enclosure. This requirement of Rule 309a was eliminated by a Board amendment on May 9,1975. The conduct alleged and the lodging of these charges occurred prior to that amendment. Hearings before the racing board began after it.
The pivotal issue regarding these charges is whether revocation of the reporting requirement of Rule 309a barred any action against Dr. Wilkey for violating this measure. Two apparently contradictory authorities are cited by the parties, one holding that repeal of legislation without an accompanying saving clause requires a reviewing court to dispose of cases under the law in effect at the time of its decision (Show of Shows, Inc. v. Illinois Liquor Control Com. (1967), 86 Ill. App. 2d 109, 230 N.E.2d 268), and the other upholding a suspension based on Racing Board Rule 309a where charges were filed after its amendment (Lamar v. Illinois Racing Board (1977), 55 Ill. App. 3d 640, 370 N.E.2d 1241).
In Lamar, the facts were identical in all important respects to those in the present case. The court acknowledged in its decision that the defendant had asserted the repeal of Rule 309a’s reporting requirement as a ground for reversal of the racing board’s suspension of his license for violating this rule. In upholding the decision of the racing board, the court rejected by implication this argument of the defendant.
In Show of Shows, the administrative agency involved was not the racing board, but rather the Illinois Liquor Control Commission. In other respects, however, the facts of that case were similar to those in Lamar and the present case.
In its decision in Show of Shows, the court not only acknowledged that the effect of revocation of legislation was in issue, but also discussed that *544issue at some length. The court held that a reviewing court must dispose of a case under the law in force when its decision is rendered, citing People ex rel. Eitel v. Lindheimer (1939), 371 Ill. 367, 373, 21 N.E.2d 318.
It is apparent that, of these two decisions by the same court, Show of Shows is the better reasoned in regard to the issue at hand. Moreover, Lamar cannot be reconciled with the Supreme Court decision in Lindheimer.
As a result, the rule enunciated in Show of Shows must be applied in the present case, so that, because Rule 309a’s reporting requirement has been repealed, the determination of the circuit court that the racing board’s action against Dr. Wilkey for violations of that rule cannot be upheld was correct.
The last of the charges against Dr. Wilkey is that he reported administering a phenylbutazone injection to the horse Mekin when in fact no such injection was administered, conduct in violation of Racing Board Rule 310. The evidence that the injection was not in fact administered consisted of the testimony of a single witness, Victor Kasemir, the trainer of the horse. His testimony was disputed by Dr. Wilkey.
Kasemir’s testimony was simply that he or his employees or his groom or his father were with Mekin at all times that Dr. Wilkey could have administered the injection on the day in question, and that he did not see Dr. Wilkey on that day. Kasemir also testified that he believed that Dr. Wilkey had double-crossed him by exposing a violation by Kasemir of a racing board rule on another occasion.
Neither the prosecution nor this witness suggested any motive for the conduct alleged. The prosecution merely characterizes this conduct as indicative of a disregard for racing board rules.
It must be noted, however, that only one incident of such conduct is alleged. This hardly demonstrates a disregard for reporting rules. No motive having been alleged, it must be assumed that the conduct charged would be due to carelessness. That Dr. Wilkey carelessly reported an injection that never was administered would be easy to believe, since many horses were under his care and keeping track of names, dates and medication would be sufficiently complex that an occasional error could occur.
But the guilt of an accused is never assumed, and the only reason the racing board was given to believe that Dr. Wilkey had made an erroneous report was the testimony of Kasemir. That testimony was effectively impeached by Kasemir’s admission that he held a grudge against Dr. Wilkey and the content of that testimony was somewhat vague.
We find that, because there was no circumstantial evidence whatever to bolster Kasemir’s vague and impeached testimony, that *545testimony did not constitute substantial evidence. Accordingly, we find no error in the circuit court’s disposition of this issue.
In view of our holding as to the charges against Dr. Wilkey, we need not reach his final argument, that the board committed reversible error in the conduct of the hearing on these charges.
Finally, we have before us a petition for leave to amend the record in this case. The racing board wishes to supply a page inadvertently omitted from the transcript of the basic hearing on the charges against Dr. Wilkey. Our disposition of this matter will not affect the outcome of this appeal, but this petition must be denied. Hartgraves v. Don Cartage Co. (1976), 63 Ill. 2d 425, 429, 348 N.E.2d 457.
For the reasons stated above, the circuit court decision reversing the decision of the racing board and its order of suspension and revocation is affirmed.
Affirmed.
GOLDBERG, P. J., and McGLOON, J., concur.