Not surprisingly, counsel for Tims did not take issue with the proposed course. Indeed he urged the judge not to give any warning at all in respect of Josevski's evidence. This was, no doubt, because Josevski had said that Tims was not, in any way, implicated in the escapade. It is, perhaps, surprising that Mr. Dann, who appeared for the applicant at trial, did not urge his Honour to give an accomplice warning - something which he conceded on this application. Perhaps both he and the judge were misled by the way in which the Crown was presenting its case - that is by charging only Wallis, Tims and the applicant with aggravated burglary of No.36 and by leading evidence from Josevski as a person not implicated in that crime. However, this was not a reason for failing to treat Josevski's evidence as the evidence of an accomplice if he was clearly particeps criminis in the offence charged, or it was open to the jury to find that he was. Where the evidence revealed, as I think it did, that Josevski was party to a common design with the others to return to the premises at Elizabeth Street, armed, and with a view to recovering money from the person who had "duped" them with such force as was necessary to achieve that object, I would have thought it clear that Josevski - whether he entered the house at No.36 or not - was acting in concert with those who did. Indeed I did not understand Mr. McArdle to submit otherwise. But, in any event, there was an issue for the jury to determine which, in itself, required the giving of a warning[12]. Although Mr. McArdle contended that the judge's direction to the jury that they should regard Josevski as a "potentially dangerous witness" and "should treat his evidence with caution" was sufficient to meet the circumstances, there seems to me to be a discrete difference between such a direction and a warning, given with the authority of the judge's office, that "it would be dangerous to convict upon the uncorroborated testimony of Josevski unless thoroughly convinced of its truth and accuracy". The former direction focusses attention only upon the potential unreliability of Josevski as a witness whereas an "accomplice warning" focusses on the unsafety of convicting upon such evidence - the core of the jury's function. The giving of such a warning would also have required attention to be given to whether there was any reliable corroboration of Josevski's evidence insofar as it implicated the applicant. Such corroboration could only have been found in the identification evidence which, as I have previously said, was anything but reliable so far as it concerned the applicant.