TAMM, Circuit Judge:

Appellant was charged in a two-count indictment with taking indecent liberties with a minor child (22 D.C.Code § 3501 (1967)) and assault with intent to commit carnal knowledge (22 D.C.Code § 501 (1967)). He was tried by a jury in the district court and found guilty on both counts; however, the trial judge set aside the verdict on the indecent liberties count because the jury had been instructed not to consider this charge if it found him guilty on the assault count. Appellant subsequently was sentenced to imprisonment for three to nine years, and in due course he prosecuted this appeal. We affirm.

I. CORROBORATION OF THE CORPUS DELICTI

Appellant’s principal assertions of error in this appeal are based upon the sufficiency of the evidence which the Government introduced as corroboration for the complainant’s testimony. It has long been the rule in this jurisdiction that no person may be convicted of a sex offense when the only evidence against him is the testimony of the alleged victim. The reason underlying this doctrine is simple and of ancient lineage: 1 it has been felt, whether rightly or wrongly, that the risk of unjust conviction is particularly high in the prosecution of sex offenses, and thus the corroboration requirement has been added to the Government’s burden of proof as an additional safeguard for the defendant. In this jurisdiction the corroboration doctrine has evolved through the years into a twofold requirement that the prosecution introduce independent evidence showing both the commission of the corpus delicti and the identity of the defendant as the perpetrator of the offense. See generally Coltrane v. United States, 135 U.S.App.D.C. 295, 418 F.2d 1131 (1969). Appellant contends that the evidence adduced by the Government in this case failed to satisfy either half of the corroboration requirement.

A. The Evidence Adduced at Trial

The complainant, a fourteen-year-old girl, testified that she visited a record shop on her way home from school on the afternoon of October 3, 1967. As she left the shop, she was accosted by a man, later identified as the appellant, who “touched [her] on the behind” and attempted to engage her in conversation (Tr. 119). She further testified that the appellant followed her for several blocks, disregarding her repeated entreaties to leave her alone, and then hit her in the face with his fist and forced her into the basement of an apartment building. At this point the appellant told her that he wanted to kiss her, *706 pushed her to the floor, pulled off her underpants, and unzipped his trousers. She managed to resist his advances and escape, but sustained a kick in the eye before the encounter terminated.

An officer from the Metropolitan Police Department testified that the complainant reported the incident to the Sixth Precinct station between four and five o’clock that same afternoon, accompanied by her mother; at that time she was “very nervous, fast talking and nervous and generally upset” (Tr. 146). She had visible bruises and contusions on her face which were photographed by the police, and these photographs were introduced into evidence at the trial (Tr. 144-45). After the complainant calmed down, she gave the police a description of her assailant.

The defense chose not to contest the complainant’s version of the assault, 2 but rather relied primarily upon the reliability of the complainant’s identification of appellant as the one who had attacked her. The appellant testified on his own behalf, advancing an alibi defense. Although he professed no personal recollection of the afternoon in question, appellant stated that he customarily walked to his home, which was located near the scene of the crime, after leaving his job at two o’clock in the afternoon, and then slept until seven or eight o’clock in the evening (Tr. 176-78, 198-99). He further testified that he went to the record shop where the complainant said the attack began “[j]ust about every day” (Tr. 177) but that he had not seen the complainant until he was accused of this offense. Appellant called no other witnesses on his behalf.

B. Sufficiency of the Corroboration

Appellant contends that there was insufficient corroboration of the intent element to sustain his conviction for assault with intent to commit carnal knowledge, relying primarily upon Allison v. United States, 133 U.S.App.D.C. 159, 409 F.2d 445 (1969). The AUison opinion held that every material element of the corpus delicti must be corroborated, 3 and that there was insufficient corroboration for the intent to commit carnal knowledge even though the Government had produced evidence showing the complainant’s prompt reporting of the incident, distraught emotional condition, and lack of motive to fabricate, together with the testimony of an eyewitness who said that the appellant was “on top of” the screaming complainant. In the instant case, the appellant contends, the only evidence corroborative of the corpus delicti — complainant’s prompt reporting, distraught emotional condition, facial bruises, and lack of motive to fabricate — must be deemed insufficient corroboration of his intent *707 under Allison. This syllogism is appealing but we think that it rests upon a misconception of the manner in which the corroboration requirement must be applied.

As a basic principle, it seems clear that corroboration in a case involving an alleged sex offense is any evidence, outside of the complainant’s testimony, which has probative value — any evidence which could convince the trier of fact that the crime was committed. This fundamental proposition was forcefully stated in Borum v. United States, 133 U.S.App.D.C. 133, 147, 409 F.2d 433, 439, cert. denied, 395 U.S. 916, 89 S.Ct. 1765, 23 L.Ed.2d 230 (1969):

Even where the effort is to prove a fact, rather than less arduously to corroborate testimony, “[t]he competency of a collateral fact to be used as the basis of legitimate argument is not to be determined by the conclusiveness of the inferences it may afford in reference to the litigated fact. It is enough if these may tend, even in á slight degree, to elucidate the inquiry * * As Dean Wigmore has put it, “the general and broad requirement for Relevancy is that the claimed conclusion from the offered fact must be a possible or a probable or a more probable hypothesis, with reference to the possibility of other hypotheses.”

Although these tenets are well established, it must be admitted that some of our decisions have required a greater quantum of corroborative proof than others. At least part of this inconsistency can be traced to the rationale underlying the corroboration requirement: the desire to protect defendants from the risk of being convicted on fabricated or mistaken charges. Clearly, this risk will vary markedly depending upon the age and sex of the complainant, the existence of a previous relationship between complainant and defendant, and other circumstances of the case; thus it is well established in this jurisdiction that “[t]he need for corroboration depends upon the danger of falsification.” Thomas v. United States, 128 U.S.App.D.C. 233, 234, 387 F.2d 191, 192 (1967).

The factual element of the corroboration requirement looms even larger when the defendant’s intent is a material fact in the offense alleged. As the standard jury instruction in this jurisdiction recognizes, proof of intent is largely a matter of inference and deduction which is usually committed to the jury as trier of fact:

Intent ordinarily cannot be proved directly, because there is no way of fathoming and scrutinizing the operations of the human mind. But you may infer as to the defendant’s intent from the surrounding circumstances. You may consider any statement made and act done or omitted by the defendant, and all other facts and circumstances in evidence which indicate his state of mind. 4

Thus it would seem to follow that when the issue is corroboration of the intent element, the jury’s role is, if anything, greater than in the normal corroboration question; and, as our cases make clear, even under normal circumstances the jury must bear the major responsibility for applying the corroboration requirement. “While the matter of [the existence of any] corroboration is initially for the trial court, like any other question as to the legal sufficiency of the *708 evidence to warrant submission of the case to the jury, it is the latter’s function to decide whether the standard of corroborative proof has been met.” Borum v. United States, 133 U.S.App.D.C. 147, 152, 409 F.2d 433, 438, cert. denied, 395 U.S. 916, 89 S.Ct. 1765, 23 L.Ed.2d 230 (1969); see also United States v. Bryant, 137 U.S.App.D.C.-, 420 F.2d 1327, 1331, No. 22,511 (D.C.Cir. Dec. 11, 1969); 3 F.Wharton, Criminal Evidence 410 (12th ed. 1955).

In short, Allison cannot be read as a handy checklist of factors which are sufficient corroboration for the intent element of a sex offense; the decisions of this court have been consistently antithetical to such an approach, even in the brief interval. between issuance of the Allison opinion and this decision. 5 Our basic policy goal in applying the corroboration requirement was aptly stated in Coltrane v. United States, 135 U.S.App.D.C. 295, 299, 418 F.2d 1131, 1135 (1969): “we try to steer between the Scylla of a corroboration requirement incapable of attainment and the Charybdis of gutting the corroboration safeguard.” Acceptance of the appellant’s rigid interpretation of the Allison decision could well project us into the midst of the former peril, for the result would be that few factors short of the defendant’s confession could be used as corroboration of intent. 6 Cf. Kelly v. United States, 90 U.S.App.D.C. 125, 129, 194 F.2d 150, 154 (1952). We reiterate that “[tjhere can be no absolute test or concrete guidelines set down as to what constitutes corroboration * * *. Each case must be evaluated on its own merits.” Bailey v. United States, 132 U.S. *709 App.D.C. 82, 88, 405 F.2d 1352, 1358 (1968). In the instant case, we conclude that the Government’s independent evidence regarding the complainant’s bruises and contusions, prompt reporting, distraught emotional condition, and lack of motive to fabricate, viewed in light of the fact that the defense did not contest the complainant’s version of the assault, 7 was sufficient to warrant submission of the corroboration question to the jury under proper instructions.

II. THE IN-COURT IDENTIFICATION AND CORROBORATION OF THE APPELLANT’S IDENTITY

The question of whether the Government introduced sufficient evidence to corroborate the identity of the accused is inseparable from another argument urged by the appellant, the question of whether the complainant’s in-court identification of the accused was impermissibly tainted by two pretrial identifications which were excluded from evidence. At the pretrial suppression hearing, it was revealed that the complainant gave the police a detailed description of her assailant immediately after the incident, and that this description corresponded in salient details to the features and characteristics of the appellant (Tr. 15-17). She was then shown a series of photographs at the police station, but she did not select any of these as resembling her attacker (Tr. 62). Two weeks thereafter one of the police officers participating in the investigation became suspicious of appellant because of his involvement in a similar offense (Tr. 27-31). Police officers then took a set of seven photographs, including one of the appellant, to the home of the complainant, and she immediately and unhesitatingly identified him as the attacker (Tr. 14-15). However, further testimony raised the possibility that the complainant had seen some or all of the other photographs during her previous visit to the police station, and because of this risk that the circumstances surrounding the photographic identification were impermissibly suggestive, the court ruled that it must be excluded from evidence at the trial.

Another identification problem raised at the suppression hearing involved an accidental confrontation between complainant and the appellant at a preliminary hearing shortly after the appellant’s arrest. This confrontation took place while complainant, accompanied by a police detective, was waiting for the hearing to begin in a hallway of the Court of General Sessions. The appellant was being led into the courtroom by a marshal, and as they passed near the complainant she spontaneously identified Terry as her assailant to the detective who was waiting with her. Cross-examination of the detective indicated that it was possible he had exchanged remarks with appellant immediately prior to this identification, so it, too, was excluded from evidence. At the same time, the court ruled that these two impermissible identifications did not preclude the complainant from making an independent in-court identification based upon her observation of the attacker at the time of the assault. In support of this result, the court pointed out that the complainant was in close proximity to her attacker for a period of approximately fifteen minutes, both on a public street in broad daylight and in a well-lighted room; the court also noted that she gave police a consistent description of her attacker, and had not wavered in her identification at any stage of the proceedings (Tr. 102). At the trial she unhesitatingly identified Terry as the man who had at *710 tacked her (Tr. 102). We find that the trial court correctly assessed the relevant factors enumerated in United States v. Wade, 388 U.S. 218, 241, 87 S.Ct. 1926, 18 L.Ed.2d 1149 (1967) and United States v. O’Connor, 282 F.Supp. 963, 965-966 (D.D.C.1968), aff’d, 137 U.S.App.D.C. 76, 420 F.2d 644 (D.C.Cir. Dec. 19, 1969), and that the showing of independent source is at least as strong as that which was upheld by this court in Hawkins v. United States, 137 U.S.App.D.C. 103, 420 F.2d 1306 (D.C.Cir. July 9, 1969).

With the two pretrial identifications excluded from evidence, there was little that the Government could have introduced as corroboration for the appellant’s identity other than the testimony of a police officer showing that the complainant had given a description of her attacker when she first reported the offense (Tr. 146). 8 Nonetheless, this paucity of corroborative evidence is not necessarily a fatal defect under our precedents. It is clear that the issues of corroboration of identity and corroboration of corpus delicti “differ markedly in importance” 9 and that a lesser quantum of corroborative proof has always been deemed sufficient for the defendant’s identity because “the danger of an erroneous identification in a rape case is not of the same magnitude as the danger of a fabricated rape.” Franklin v. United States, 117 U.S.App.D.C. 331, 334, 330 F.2d 205, 208 (1964). Indeed, we stated in Franklin that “in the circumstances of a particular case, a convincing identification by the complaining witness based on adequate opportunity to observe need not be further corroborated * * * ” 117 U.S.App.D.C. at 335, 330 F.2d at 209. The Franklin rationale was explicitly extended to sex offenses less serious than rape in Coltrane v. United States, 135 U.S.App.D.C. 295, 418 F.2d 1131 (1969). This seems eminently sound, for it is likely that fewer kinds of corroborative evidence linking the defendant to the scene of the crime will be available in the less serious offenses ; for example, one very important type of proof tending to connect the defendant with the crime — scientific analysis of real evidence uncovered at the scene of the attack or on the person of the complainant — will only rarely be available to show that the defendant participated in one of the offenses less serious than rape.

The question remains whether the instant case is a proper occasion for invoking the exception to the corroboration rule developed in Franklin and its progeny. Guidelines for the resolution of this problem can be found in Thomas v. United States, 128 U.S.App.D.C. 233, 387 F.2d 191 (1967), as well as in Franklin itself; and practically all of the factors enumerated in these decisions point to the conclusion that this is an appropriate instance for relaxation of the normally rigid corroboration standards.

There was no evidence which cast substantial doubt upon the trustworthiness of the complaining witness, nor were there any significant shortcomings in her testimony on either direct or cross examination. The complainant had an ample opportunity to observe her attacker, and she gave police an initial description which was consistent with the appellant’s physical characteristics. It is true that there was no showing that appellant had an “opportunity” to commit the crime, in the sense that there was no independent evidence placing him at the situs of the attack during the relevant time period. However, according to his own testimony the appellant lived and worked in the neighborhood where the attack occurred, and frequented the *711 record shop near which the complainant was first accosted; he was also unable to offer any alibi witness or other independent evidence showing his whereabouts on the afternoon in question from 2:30, when he left work, until seven or eight o’clock in the evening. The identification issue was forcefully argued by defense counsel and submitted to the jury under proper instructions requiring the jury to find corroboration before conviction. 10 Under these circumstances, we do not believe that the appellant’s conviction should be overturned.

. See, e. g., the quotation from Hale’s Pleas of the Crown in 7 J. Wigmore, Evidence 342-345 (3d ed. 1940) :

“The party ravished may give evidence upon oath and is in law a competent witness; but the credibility of her testimony * * * must be left to the jury, and is more or less credible according to the circumstances of fact that concur in that testimony. * * * It is true, rape is a most detestable crime * * *; but it must be remembered that it is an accusation easily to be made and hard to be proved; and harder to be defended by the party accused, tho never so innocent.”

. In his summation, defense counsel stated :

Let us look for a moment at the facts. They don’t seem to be very much in dispute. It is pretty obvious that on October the 3rd [complainant] was assaulted. She was touched on the bottom, she was struck in the face and she suffered personal indignities. * * * ***** * * * She has been hurt and she’s been hurt in a way that scars may remain for the rest of her life, and she wants to get the person that hurt her. * * * Has the Government proved beyond a reasonable doubt * * * in your mind that this defendant, Tyrone Terry, is the person that assaulted [complainant] ?

Tr. 224-225, 227).

. Allison appears to be the first decision in this jurisdiction to hold explicitly that every material element of a sex offense must be corroborated; indeed, the cases cited for this proposition in Allison (409 F.2d at 449 n.13) do not even involve corroboration of sex offenses. The test formulated in the leading decision in this jurisdiction, Ewing v. United States, 77 U.S.App.D.C. 14, 17, 135 F.2d 633, 636 (1942) (emphasis added), cert. denied, 318 U.S. 776, 63 S.Ct. 829, 87 L.Ed. 1145 (1943) , spoke only of “circumstances in proof whihc tend to support the prosecutrix’ story,” and the other pre-Allison decision seem to use the phrase “corroboration of the corpus delicti” merely as a convenient method of distinguishing this requirement from the complementary rule regarding corroboration of the defendant’s identity.

. Instruction No. 43 (Proof of Intent), Jr. Bar Section of D. O. Bar Ass’n, Criminal Jury Intruetions for the District of Columbia (1966); cf. 5 Reid’s Branson, The Law of Instructions to Juries 135 (1962) :

It is not always possible to prove a purpose by direct evidence, for purpose and intent are subjective facts. That is, they are within the mind of man, and hence, in determining purpose, you may look to all the surrounding circumstances, including what was said and done in relation thereto; bearing in mind the presumption of law, that every one is presumed to intend the natural and probable consequences of his voluntary acts, unless the circumstances are such as to indicate the absence of such intent.

. See, e. g., United States v. Bryant, No. 22,511, 137 U.S.App.D.C. 124, 420 F.2d 1327 (D.C. Cir. Dec. 11, 1969), where sufficient corroboration for assault with intent to commit rape was found on the basis of a torn shoulder strap on the complainant’s dress, her prompt report to police, bruises, and trembling. It appears that the need for a strong corroboration requirement was, if anything, greater in the Bryant case than in the instant controversy : there, evidence had been introduced which tended to show that the defendant and the complainant had previously had an amorous relationship, thus increasing the possibility that the complainant had a motive to fabricate. See also Coltrane v. United States, 135 U.S.App.D.C. 295, 418 F.2d 1131 (1969).

. Illustration of this danger can be found in the Allison decision itself. At one point in that opinion there is a list of ten factors which our decisions have upheld as valid corroboration (409 F.2d at 448 n.8) ; if the Allison holding is applied to these factors in the strict manner urged by appellant, only two of them, condition of the complainant’s clothing and bruises or scratches on her person, would have any practical viability as corroboration of the defendant’s intent. It could be argued that Allison renders even these two elements insufficient a fortiori, since eyewitness testimony that the defendant was “on top of” the struggling complainant was deemed insufficient corroboration of intent in that opinion.

The Government urges us to take the same kind of approach to an opposite result in the instant case by limiting the Allison opinion to its facts. This argument is grounded on the Allison decision’s reference to the complainant as a “particularly impressionable” child (409 F.2d at 450 n.17). Although the age and prior experiences of the complainant obviously influenced the Allison court’s assessment of the need for corroboration (cf. Wilson v. United States, 106 U.S.App.D.C. 226, 271 F.2d 492 (1959)), the interpretation urged by the Government would require a strained reading of that opinion: the reference to “impressionability” in Allison was made only to show the weakness of one element urged by the prosecution as corroboration, the complainant's lack of motive to fabricate. The real difficulty with the “impressionable child” theory is more fundamental, however; in essence, it is an attempt to isolate one factor out of the complex mixture of factual considerations bearing on the corroboration requirement, and make it determinative. As we have indicated above, the question is not that simple, and it is not within the proper scope of our function as a reviewing court to impose an artificial simplicity upon this difficult problem. Rather, the primary responsibility for weighing and assessing all of the conflicting factual elements must of necessity rest with the trial court and the jury.

. By taking cognizance of the fact that the defense did not contest the complainant’s version of the assault, we do not mean to imply that the defense has an affirmative burden of proof on the corroboration question. This is merely a recognition of the general principle that the need for corroboration depends upon the risk of erroneous conviction, and that the court should look upon the corroborative proof with greater skepticism when there is some affirmative evidence casting doubt upon the credibility of the complainant’s testimony. See notes 5 and 6, supra-.

. No reason appears why a proper lineup was not conducted in this case. Evidence of such a lineup would have done much to allay the jury’s apparent concern as to pretrial identification of the appellant by the victim. See note 10, infra.

. United States v. Bryant, 137 U.S.App. D.C. 124, 420 F.2d 1327 (1969).

. The jury obviously had some difficulty resolving the identification issue, for after it had retired to deliberate the foreman sent a note to the judge asking whether the complainant had identified Terry prior to the in-court identification (Tr. 264). The court then properly instructed the jurors that they would have to make their decision solely on the basis of their recollection of the evidence introduced at trial. After approximately two hours of further deliberation, the jury returned a guilty verdict.