The polariscopic reading may now be taken, an observation on the 90 deg. control plate having been made immediately before as previously described. Then without altering the position of the instrument relative to the light, or changing the character of the latter in any way, the tube filled with the sugar solution is substituted for the control plate. The telescope is adjusted, if necessary, so as to give a sharply defined field, which must appear round and clear. (This condition must be fulfilled before the observation is performed, as it is essential to accuracy.) The milled head is turned until the neutral point is found, and the reading is taken exactly as previously described, the operation repeated five or six times, the average taken with the rejection of aberrant readings, the average figure corrected for the deviation shown by the control observation from the sugar value of the control plate at the temperature of observation as given in the table, and the result taken as the polarization of the sugar. When a series of successive polarizations is made under the same conditions as regards temperature, position of the instrument with relation to the high intensity, of the light, etc., the control observation need not be made before each polarization, one such observation being sufficient for the entire series. The control must be repeated at least once an hour, however, and oftener when the operator has reason to think that any of the factors indicated above have been altered, for any such alteration of conditions may change the zero point of the instrument. In the polarization of the quartz plates, as also in the polarization of very white sugars, difficulty may be experienced in obtaining a complete correspondence of both halves of the field. With a little practice this may be overcome and the neutral point found, but when it cannot, the ordinary telescope of the instrument may be replaced by another, which is furnished with the polariscope and which carries a yellow plate. This removes the difficulty and renders it possible, even for one not well accustomed to the instrument, to set it at the exact point of neutrality. SUMMARIZED SOURCES OF ERROR. The following principal sources of error must be especially guarded against: 1. Drying out of sample during weighing. 2. Excess of subacetate of lead solution in clarification. 3. Incomplete mixing of solution after making up to mark. 4. Imperfect clarification or filtration. 5. Concentration of solution by evaporation during filtration. 6. Undue compression of the cover glass. 7. Alteration of the temperature of room, position of instrument, or intensity of light while the observation or control observation is being performed. 8. Performances of polarization with a cloudy, dim, or not completely round or sharply defined field.