The usual method of determining moisture is to weigh 2 to 3 grams of finely shaved soap on a watch glass and heat in an oven at 105 degrees C. for 2 to 3 hours. The loss in weight is represented as water, although it is really impossible to drive off all the water in this way.
To overcome the difficulties just mentioned either the Smith or Fahrion method may be used. Allen recommends Smith's method which is said to be truthful to within 0.25 per cent. Fahrion's method, according to the author, gives reliable results to within 0.5 per cent. Both are more rapid than the above manipulation. To carry out the method of Smith, 5 to 10 grams of finely ground soap are heated over a sand bath with a small Bunsen flame beneath it, in a large porcelain crucible. The heating takes 20 to 30 minutes, or until no further evidence is present of water being driven off. This may be tested by the fogging of a cold piece of glass held over the crucible immediately upon removing the burner. When no fog appears the soap is considered dry. Any lumps of soap may be broken up by a small glass rod, weighed with the crucible, and with a roughened end to more easily separate the lumps. Should the soap burn, this can readily be detected by the odor, which, of course, renders the analysis useless. The loss in weight is moisture.
By Fahrion's method[13], 2 to 4 grams of soap are weighed in a platinum crucible and about three times its weight of oleic acid, which has been heated at 120 degrees C. until all the water is driven off and preserved from moisture, is added and reweighed. The dish is then cautiously heated with a small flame until all the water is driven off and all the soap is dissolved. Care must be exercised not to heat too highly or the oleic acid will decompose. The moment the water is all driven off a clear solution is formed, provided no fillers are present in the soap. The dish is then cooled in a dessicator and reweighed. The loss in weight of acid plus soap is moisture and is calculated on the weight of soap taken. This determination takes about fifteen minutes.
Test a freshly cut surface of the soap with a few drops of an alcoholic phenolphthalein solution. If it does not turn red it may be assumed free fat is present; should a red color appear, free alkali is present. In any case dissolve 2 to 5 grams of soap in 100 cubic centimeters of neutralized alcohol and heat to boiling until in solution. Filter off the undissolved portion containing carbonate, etc., and wash with alcohol. Add phenolphthalein to the filtrate and titrate with N/10 acid and calculate the per cent. of free alkali as sodium or potassium hydroxide. Should the filtrate be acid instead of alkaline, titrate with N/10 alkali and calculate the percentage of free fatty acid as oleic acid.