_Estimation of the Oxygen._——To determine this Angus Smith has recourse to the endiometer. Five or six of Bunsen’s endiometers were used at once and the mixed gases were exploded by means of a powerful battery and a Ruhumkorff’s coil. In his ‘Inorganic Chemistry,’ Miller thus explains the principle upon which the action of the endiometer is based: “By means of the endiometer various gaseous mixtures may be analysed with great exactness. Many different forms of this instrument are in use. One of the most convenient is Hoffmann’s. It consists of a stout syphon tube. (See next figure.) Into the sides of the tube, near the sealed end, two platinum wires (_a_, _b_) are fixed for the purpose of transmitting an electric spark through the cavity of the tube. The sealed limb is accurately graduated to tenths of a c. c. or other suitable divisions. Suppose it be desired to ascertain the proportion of oxygen in atmospheric air. The instrument is first filled with mercury, after which a small quantity of air is introduced; the bulk of the air is accurately measured, taking care that the liquid metal stands at the same level in both tubes, which is easily effected by adding mercury, or by drawing off the mercury if needed, through the caoutchouc tube, which is fixed upon the small inlet tube just above the bend, and which is closed by means of a screw tap (_c_). [Illustration] A quantity of pure hydrogen, about equal in bulk to the air, is next introduced, and the bulk of the mixture is then accurately measured. The open extremity of the tube is now closed with a cork, below which a column of atmospheric air is safely included. This portion of air acts as a spring, which gradually checks the explosive force, when the combination is effected by passing a spark across the tube by means of the platinum wires. The mixture is then exploded by the electric spark. The remaining gas now occupies a smaller volume, owing to the condensation of the steam which has been formed. Mercury is, therefore, again poured in the open limb until it stands at the same level in both tubes, and the volume of the gas is measured a third time. One third of the reduction of the bulk experienced by the gas will represent the entire volume of oxygen which the mixture contained. Liebig’s method is as follows. It is based upon the fact that an alkaline solution of pyrogallic acid absorbs oxygen: 1. A strong measuring tube holding 30 c. c., and divided into one fifth or one tenth c. c., is filled to two thirds with the air intended for analysis. The remaining part of the tube is filled with mercury, and the tube is inverted over that fluid in a tall cylinder widened at the top.