2. _Preparation from mercuric oxide._ This method is of interest, since it is the one which led to the discovery of oxygen. The oxide, which consists of 7.4% oxygen and 92.6% mercury, is placed in a small, glass test tube and heated. The compound is in this way decomposed into mercury which collects on the sides of the glass tube, forming a silvery mirror, and oxygen which, being a gas, escapes from the tube. The presence of the oxygen is shown by lighting the end of a splint, extinguishing the flame and bringing the glowing coal into the mouth of the tube. The oxygen causes the glowing coal to burst into a flame. In a similar way oxygen may be obtained from its compounds with some of the other elements. Thus manganese dioxide, a black compound of manganese and oxygen, when heated to about 700 deg., loses one third of its oxygen, while barium dioxide, when heated, loses one half of its oxygen. 3. _Preparation from potassium chlorate (usual laboratory method)._ Potassium chlorate is a white solid which consists of 31.9% potassium, 28.9% chlorine, and 39.2% oxygen. When heated it undergoes a series of changes in which all the oxygen is finally set free, leaving a compound of potassium and chlorine called potassium chloride. The change may be represented as follows: /potassium\ | | (potassium / potassium \ (potassium { chlorine } = { } + oxygen | | chlorate) \ chlorine / chloride) \oxygen / [Illustration: JOSEPH PRIESTLEY (English) (1733-1804) School-teacher, theologian, philosopher, scientist; friend of Benjamin Franklin; discoverer of oxygen; defender of the phlogiston theory; the first to use mercury in a pneumatic trough, by which means he first isolated in gaseous form hydrochloric acid, sulphur dioxide, and ammonia] The evolution of the oxygen begins at about 400 deg.. It has been found, however, that if the potassium chlorate is mixed with about one fourth its weight of manganese dioxide, the oxygen is given off at a much lower temperature. Just how the manganese dioxide brings about this result is not definitely known. The amount of oxygen obtained from a given weight of potassium chlorate is exactly the same whether the manganese dioxide is present or not. So far as can be detected the manganese dioxide undergoes no change. [Illustration: Fig. 4]