developed when it is the minutest point discernible by the microscope as when it has attained its ultimate growth. I might add parenthetically that crystals are sometimes of immense size, one at Milan of quartz being 3 feet 3 inches long and 5 feet 6 inches in circumference, and is estimated to weigh over 800 pounds; and a gigantic beryl at Grafton, N. H., is over 4 feet in length and 32 inches in diameter, and weighs not less than 5,000 pounds; but the most perfect specimens are of small size, as some accident is sure to overtake the larger ones before they acquire their growth, to interfere with their symmetry or transparency. This you will see abundantly illustrated by the examples which I have prepared, as also the constancy of the angles of like faces. Chemically speaking, the crystal is always a perfect chemical body, and can never be a mechanical mixture. This fact has been of great value to the science of chemistry in developing the atomic theory, which has demonstrated that a body can only exist chemically combined when a definite number of atoms of each element is present, and that there is no certainty of such proportions existing except in the crystal. I hold before you a crystal of common alum. Its chemical symbol would be Al_{2}O_{3},3SO_{3}+KO,SO_{3}+24H_{2}O. If we knew its weight and wished to know its ultimate component parts, we could calculate them more readily than we could acquire that knowledge by any other means. But the elements of this quantity of uncrystallized alum could not be computed. Then we may define crystallization to be the operation of nature wherein the chemical atoms or molecules of a substance have sufficient polarized force to arrange themselves about a central attracting point in definite geometrical forms. Fresenius defines it thus: "_Every operation, or process, whereby bodies are made to pass from the fluid to the solid state, and to assume_ certain fixed, _mathematically definable, regular forms_." It would be folly for me to attempt to criticise Fresenius, but I give you both definitions, and you can take your choice. The definition of Fresenius, however, will not suit our present purpose, because the crystallization of wrought iron occurs, or seems to, _after_ the iron has acquired a _solid state_.