§ The enamel colours used on porcelain, have the permanent character required. Different manufactories excel in different colours. The first step therefore would be to invite each manufacturer to send tablets of porcelain of a given size, on which are to be painted a number of small squares, containing all the pure colours he employs. Besides these squares, a certain number of other squares should contain two or more combinations of these colours, two by two, or in such proportions as are usually employed. The comparison of these tablets would indicate where the purest and most useful porcelain colours could be obtained. The next step would be that a small committee of manufacturers and men of science, should decide on the number of combinations and shades of colour it might be desirable to bring together as permanent and standard objects of reference. The different makers of porcelain should then each receive an order for a certain number of tablets containing those colours in which they respectively excel. Each small square should be numbered. A sufficient quantity of the proper materials constituting each colour, should then be mixed in the proper proportion, and applied at the same time, to the same number on each tablet; and these tablets should be exposed to the fire under as nearly as possible the same circumstances of heat, and for the same length of time. Thus an extensive system of unchangeable colours might be obtained, and if 500 sets were made, they might be distributed in all the great cities and universities of the world. It might perhaps be found that certain colours were deficient, and this would of course stimulate discovery by making known the want. Thus, in the course of upwards of twenty-five years, during which the author has been collecting on a small scale, such tablets of colours, he has been unable to meet with any specimen of an enamel colour at all approaching to the pure scarlet of the common geranium. The utility of such sets of standard colours would be very great, enabling all nations to speak a language regarding colour at once accurate and universal. It might serve as the starting point and the test of many analogous collections of materials tinted by colours of more transitory duration, whose relative degree of fixity might thus be measured: as silks, cottons, linens, woollens, leather, paper, and many other materials. There are two coloured substances which seem to promise a higher degree of permanence than those just mentioned—sealing-wax and glass. For these admit of the renewal of their surfaces by grinding, in case atmospheric or external causes should have altered or impaired the superficial colour.