To ascertain the nature of the, fatty acids, the melting-, or solidifying-point of the mixture should first be taken (see Oils, p. 1477). The apparatus suitable is shown in Fig. 1278. The fatty acid ia in the inner tube B, surrounded by water, which is gradually heated over a lamp ; tempera tures are observed by thermometers C D. To detect the presence of, and estimate, coco-nut-eil or palm-lverneboil, the method recommended in Oils, p. 1746, may be adapted. On the same page, will be found hints for the determination of the quantity of roain ; another method (proposed by Dalican) is here given, with the remark that Sutherland'a process, by which the rosin is oxidized by nitric acid, and Rampal's process, in which the rosin is precipitated in a finely-divided state by throwing an alcoholic solution of the fatty aeids into water, are both unreliable. Dissolve 10 grm. of the soap hi 100 grin. water, add enough concentrated soda leys to precipitate the soap (p. 1769); some resinates remain in the liquor, which is neutralized, evaporated to dryness, and the resin extracted with alcohol, which may be distilled off, and the rosin w( ighed (A). Then dissolve the precipitated soap in water, and add excess of barium chloride; collect and dry this baryta soap, and extract It with ether, which dissolves out only the resioates. Evaporate off the ether, and treat the resivates with boiling water and sulphuric acid, which sets the rosin free ; it may then, if necessary, be similarly dissolved out by alcohol, or may be merely collected on a weighed filter, and its weight (B) noted ; A +B is the weight of rosin in the soap. The portion insoluble in ether may then be suspended in water, decomposed with sulphuric acid, and the fatty acids collected, dried, and weighed.
For other methods of examining the nature of the fatty acids in soaps, c,onsult Oils—Detection and Analysis, pp. 1462-1477, to which may be added a reference to the amount of information that may be derived from examining, by polarized and ordinary light, under a microscope, the manner of crystallization of thin layers of fatty acid mixtures allowed t.o cool between two pieces of glass pressed together ; some very remark able results of this method were shown by Price's Candle Co. at the Paris Exhibition ef 1878.
Shorter, but less reliable, methods than the above have been frt quently proposed for determining the value of soap. To shorten the operation of weighing the fatty acids, many methods have been proposed for measuring them, by collecting them in a long-necked flask, graduated, or in a graduated tube attached thereto. Whenever this is clone, the weight can only be arrived at from the estimated sp. gr. of the fatty acids, and as this is very variable, the method is at best an approximate one, though useful in the factory when that sp. gr. is known. Buchner decomposes 16.66 grm., and measures the fatty acids to the *cc., multiplying by 0.93 to get the weight in grm. Ho also gives the following useful table, on tho basis that 100 lb • fat produce 155 lb. soap and about 6.25 lb. glyeeline ; the three last columns are of general use,
when the " fatty acids per eent." are determined by weight :— The Industrial Society of Mulhouse awarded a prize to Cailletet for a method of analysing soap without more weighings than that of the soap itself. Minute instructions are given in the Bulletin of the Society, No. 144, Tome xxix., p. 8. Suffice it to say here that, in the first place, much information may be gained for industrial purposes by attentively observing the behaviour of the soap with hot and cold water ; 10 grin. of the soap are then decomposed by excess of standard acid in presence of a measured volume of turpentine-eil, the increase in volume of which, multi plied by the sp. gr. of the fatty acids, gives their weight. The acid solution is titrated back with soda, and tho soda per cent. is calculated. It is stated that the turpentine does not dissolve the rosin, and that thus the presence of rosin may be detected, and even.estimated.
A short way of ascertaining whether there is much besides pure soap, and water, in a sample of soap, is to treat it with strong xvarm alcohol, which dissolves nothing but the soap, and excess ”f caustic soda, if any ; this last may be removed by a stream of carbonic acid gas. The insoluble residue may be collected on a tared filter, washed with alcohol, dried. and weighed.
The determination of the other constituents of commercial soaps has now to be considered.
Unsaponified fat.—This is very rarely present ; 10-20 grm. of the soap is cut into tine shavings, and dried at 100° (212° F.); then treated with warm benzol or pctrob um-spirit, which is decanted or filtered off into a tared flask ; the solvent is tlaen distilled off, or evaporated, when the unsaponified fat, if any, is left behind and weighed.
Glyeerine.—A weighed portion of the soap is dissolved in water, and decomposed with slight excess of sulphurie acid ; the fatty acids are removed, and the acidulated solution is evaporated to dryness and treated with alcohol ; the alcolaolic solution of glycerine is separated by filtration into a tared flask, from which the alcohol itself is distilled off.
Carbolic acid.—Mr. C. Lowe recommends the following process. Take 50 grain-measures of aqueous solution of caustic soda sp. gr. 1.345, dilute to 1000 gr.-m.; in this, dissolve by heat 100 gr. of the soap, then add 1000 gr.-m. saturated solution of common salt. Filter off, and wash with brine, the soap thus precipitated ; slightly acidify the filtrate and washings with bydrochloric acid, and add thereto enough bromine-water to make the liquid permanently yellow. Warm the liquid till the precipitate melts, then let it cool ; remove, carefully dry, and weigh the resulting mass, of whieh, 331 parts correspond to 94 parts of carbolic acid. If the inferior qualities of carbolic acid have been used, the precipitate, which is bibromoeresol, CJ:13131•30, forms a sticky mass, owing to the liquid nature of the cresylic acid it contains.