The only pigments of use in this connection to the colour-maker are such as have very intensive colouring power and at the same time are low enough in price. Such substances are ultramarine, smalt and coal-tar dyes. Smalt is the best because its colour is unalterable. In point of chemical composition, this substance is a very hard glass coloured blue by cobaltous oxide. For improving the colour of chalk or any other white, the smalt must be in an extreme state of fine division, and levigated to an impalpable powder. Ultramarine can be used for the same purpose, but is not so permanent. To ascertain the correct proportion of blue pigment, it is advisable to make a systematic experiment, which is easily performed. Exactly 90 parts of the chalk in question are triturated with 10 parts of blue pigment in a mortar until the entire mass has become a perfectly uniform pale blue powder, which contains 10% of the blue ingredient. Several samples, each representing one hundred parts of the white pigment to be corrected are carefully weighed out, 1 part of the blue powder being added to the first sample, 2 parts to the second, 3 to the third, and so on, and the mixtures are compared with a standard white substance, such as best white lead or zinc white, to see which most nearly approaches the standard colour. It is then easy to calculate how much of the blue requires to be added to 100 or 1000 lb. of the material to be corrected. The correction can be effected in several ways; for instance, by grinding the blue pigment directly with the bulk, by adding it at the levigation stage, or mixing it with the dry, finished product. The first two methods are attended with certain drawbacks which render it difficult to obtain a perfectly uniform product, owing to the specific gravity of the blue pigments being higher than that of the whites. Consequently, when the two are mixed in presence of water--as is always the case in grinding and levigation--the heavier blue pigment settles down more quickly, and several strata can be clearly distinguished in the sediment. The upper layers will still have a decided yellow tinge--the proportion of blue being too small for proper correction--whilst the next in order will be pure white--accurately corrected--and those at the very bottom will be decidedly blue, because they contain the largest proportion of the blue substance. The most satisfactory results are obtained by dry mixing; and this can be successfully practised when the colour-maker has a cheap source of power (such as water power) available. Where, however, costly power plant has to be provided, only the finest grades of white pigments can be improved in this way, the expense of labour being too high for cheap materials.