centre of each primitive octahedron in the first place, and next that which lies in the middle of the face of the complex octahedron.
The total change in volume of the crystal would thus depend primarily on the expansion of the atomic volumes, and next on the con- traction which follows from a change of configuration ; and it is from the supposed counteraction of these two opposing factors that we may find an explanation of the extraordinarily small coefficient of expansion which is characteristic of cuprite. Its value as measured by Fizeau is 0*00000093, that of sulphur (an homologue of oxygen) being 0-000064, and that of copper 0-0000169; copper in the free state thus possesses a coefficient eighteen times greater than that of cuprite, while that of sulphur is seventy times greater.
The variation in the coefficient of expansion was very carefully determined by Fizeau, and found to be 2*18. From this it follows that with increasing rise of temperature the effect of the expansion of atomic volumes gains upon that of the contraction due to change of configuration ; while with fall of temperature the contrary is the case, and a point may be looked for at which further cooling will cease to be accompanied by contraction ; at this point, determined as - 4 C. by Fizeau, cuprite will attain its maximum density.
If the atoms of copper and oxygen are to remain in contact, so that the six atoms of oxygen about the centre of a crystalline element shall touch each other as well as the atoms of copper with which they are associated, then when the configuration is that of maximum volume, the atoms of oxygen cannot have a less diameter than the atoms of copper. We have already seen, however, that at 15 C. the atoms of oxygen do possess a slightly smaller diameter than that of copper. It is therefore evident that further cooling if it tend towards pro- ducing the configuration of maximum volume will effect a separation of the six central oxygen atoms from each other, and this, though it will not destroy, will certainly diminish the stability of the system ; and consequently we may expect to find, if cooling be continued far below 15 C., that the existing configuration will be overturned, and the compound will acquire the symmetry of another crystalline system.
The cleavage of cuprite, as observed by Professor Miers, is most perfect parallel to the faces of the cube ; this may possibly be con- nected with the fact that these planes are situated at right angles to the tension which we infer to exist between oxygen and copper along the axes of the octahedron.
In concluding this account of cuprite attention may be directed to the interesting case of cupric oxide, which has been made the subject of frequent comment by students of atomic volumes, on account of the remarkable relation which seems to exist between the volume of oxygen as it exists in this compound, and in the lower oxide which we
