Observations on the use of the double diffusion test in the diagnosis of vaginal candidiasis.