The results given by the polariscope possess an accuracy rarely, if ever, attained by any other apparatus employed in the determination of practical commercial values.[57] The proportion of grape sugar intentionally added to cane sugar can also be determined by the use of the polariscope, certain modifications being observed in its application. As previously stated, cane sugar is converted into a mixture of dextrose and levulose, termed invert sugar, by the action of dilute acids. While the rotary effect of dextrose upon the plane of a ray of polarised light is constant at temperatures under 100°, that exerted by levulose varies, it being reduced as the temperature is increased; hence it follows that at a certain temperature the diminished levo-rotary power of the levulose will become neutralised by the dextro-rotary effect of the dextrose, _i.e._ the invert sugar will be optically inactive. This temperature has been found to approximate 90°. Since dextrose is not perceptibly affected by the action of weak acids, it is evident that by converting cane sugar into invert sugar and examining the product by the polariscope at a temperature of about 90°, the presence of any added dextrose (glucose) will be directly revealed by its dextro-rotary action. This is accomplished by a method suggested by Messrs. Chandler and Ricketts,[58] which consists in substituting for the ordinary observation tube of the polariscope a platinum tube, provided with a thermometer, and surrounded by a water-bath, which is heated to the desired temperature by a gas burner (Plate X. Fig. 4). The sugar solution to be examined is first treated with a little dilute sulphuric acid, then neutralised with sodium carbonate, clarified by means of basic plumbic acetate, filtered, and the polariscopic reading taken at a temperature of 86° to 90°. Since the results given by the foregoing method represent pure dextrose, it is necessary to first ascertain the dextro-rotary power of the particular variety of glucose probably employed for the adulteration of the sugar under examination, and then make the requisite correction. This process for the estimation of glucose is especially advantageous, in that the optical effect of the invert sugar normally present in raw cane sugars is rendered inactive. It is sometimes desirable to determine the relative proportions of the organic constituents which are present in commercial glucose. These usually consist of dextrose, maltose, and dextrine, all of which possess dextro-rotary power, but not in the same degree; that of dextrose being 52, that of maltose 139, and that of dextrine 193. An estimation of the amount of each can be made by first ascertaining the total rotary effect of the sample by means of the polariscope.[59] This is expressed by the equation P = 52 _d_ + 139 _m_ + 193 _d´_, (1)