The evidence of mere identifica ti on of the ac cuse d per son at the t rial for the first ti me is from its very natu re inherently of a weak char act er. T he purpose of a p r ior test identification, therefore, is to t est and s treng then t he tru stwor t hiness of that evidence. It is accordin gly considered a safe rule of prudence t o generally look for corroboratio n of the sworn testim ony of witne sses in cour t as to the identity of t he accused who are stra nge rs to them, in t he form of earli er identi fication pr oce edings. T h is rule of p rude nce, ho wever, is subjec t to exceptions, when, for example, the co u rt is imp r essed by a particular w itne ss on wh ose testim ony it c an safely r ely, without s uch or other corrob or ation. It is no doubt true that m uch evidentiar y value cann ot be attached to the identification of the accused in court where i d entifying w itne ss is a total stranger wh o had just a fleeting gl i mpse of the person identified or who had n o particu lar re a son to remember the pe rson concerned, if the identification is made for the first time in c ourt.