Ten cubic centimeters of a normal saltpeter solution, containing two and a half grams of pure sodium nitrate in 100 cubic centimeters are placed in the funnel and, with continued boiling, allowed to pass, drop by drop, into the flask. When almost all has run out the funnel is washed three times with ten cubic centimeters of ten per cent hydrochloric acid and this is allowed to pass, drop by drop, into the flask. When no more nitric oxid is evolved the measuring-tube is transferred to a large jar filled with distilled water. The solution of the substance to be examined should be taken in such quantity as will give about the same quantity of gas as is furnished by the ten cubic centimeters test nitrate solution before described; _viz._, about seventy cubic centimeters. Eight or ten determinations can be made, one following the other, and at the end another determination with normal sodium nitrate solution should be made as a check. At the end of the operation all of the measuring-tubes are in the large jar filled with distilled water. The temperature of the surrounding water will soon be imparted to the contents of each tube and the volume of nitric oxid is read by bringing the level within and without the measuring-tube to the same point. The percentages are calculated for the given temperature and barometer pressure in the usual way; or to avoid computation the volume can be compared directly with the volume furnished by a normal nitrate solution, which is a much simpler method.