There is a substance, the double nitrate of silver and thallium, which, while solid at ordinary temperatures, liquefies at 75° C. and then has a specific gravity of 4·5. Admixture with water lowers the density to any desired point. If a glass cell is taken containing this liquid diluted to a density of about 3·6, and in it is thrown pieces of the above-named minerals, all those whose density is lower than 3·6 will rise to the surface, while the denser minerals will sink. If now a little water is carefully added with constantly stirring until the density of the liquid is reduced to that of the diamond, the heterogeneous collection sorts itself into three parts. The graphite, quartz, beryl, mica, and hornblende rise to the surface; the garnet, corundum, zircons, etc., sink to the bottom, while the diamonds float in the middle of the liquid. With a platinum landing-net I can skim off the swimmers and put them into one dish; with the same net I can fish out the diamonds and put them in a second dish, while by raising a sieve at the bottom I can remove the heavy minerals and put them into a third. The accurate separation of diamonds from the heterogeneous mixture can be effected in less time than is taken to describe the experiment. The table shows that diamonds vary somewhat in density among themselves, between narrow limits. Occasionally, however, diamonds overpass these figures. Here is an illustration. In a test-tube of the same dense liquid are three selected diamonds. One rises to the top, another floats uncertain where to settle, rising and falling as the temperature of the sorting liquid is raised or lowered, whilst the third sinks to the bottom. Allowing the liquid to cool a degree or two slightly increases the density and sends all three to the surface. PHOSPHORESCENCE OF DIAMOND