Dissolve 3 g orcinol in 500 mL concentrated HCl, add 2.5 mL of a 10% solution of ferric chloride hexahydrate, and dilute to one liter with water; this is approximately 6 M HCl. The reagent is stable for months, but its yellow color gradually darkens and some precipitate forms; this doesn't seem to affect its reactivity. The "classical" Bial's reagent is made with a liter of concentrated HCl, undiluted with water. It gives a slightly stronger reaction, and considerably faster (30-60 seconds), but is much less stable than the recipe we've come up with, and the fumes are much more a problem with concentrated than with 6 M HCl. The reaction even seems to work, more slowly and with less intense color, if the final HCl concentration is only 4 M.
When 1 mL of reagent is heated with 5 drops of sample in a boiling water bath, a positive test for pentoses is formation of a green to blue color (not precipitate) in less than five minutes.