The next step is to dye some pieces evenly with green, violet, and orange, made by two of the primary colors, and then to try matching these with fresh, newly-mixed baths of the same dyes. It will be found here that success depends upon going slowly; and upon beginning with light shades and building the color up to the desired strength carefully, by means of successive dippings. Note that the color of cloth when wet is much darker than when dry. Some dyers hold the wet cloth to the bright sky and look through it, to get an idea of what the finished color will be like; but positively certain and satisfactory results are arrived at only by wetting the sample to be matched or drying the piece that is being dyed, so that both sample and piece are equally wet or dry, while their color is being compared. The real difficulty of color dyeing is not met with until the student tries to obtain shades embodying all three of the primary colors. A very few experiments will quickly show that with most modern dyestuffs it is hard to get soft, pleasant tones with the use of only two colors. Natural colors, as we find them in the sky, water, meadow, and woodlands, are never pure; they are invariably mixed. And our eyes are so accustomed to them that shades dyed with simple or pure colors look hard, cold, and inharmonious. Mixtures of two colors are better and softer than single colors, but still rather hard. But when the secondary shade resulting from the combination of two primary colors is mixed with even a small quantity of the third primary color, the result is invariably a soft and pleasing tone. The above statements presuppose that it is possible, in practice, to obtain good dyestuffs in each class, which are absolutely pure, clean shades of blue, yellow, and red without any admixture whatever. As a matter of fact, while the artificial dyestuffs are much more pure, and hence much more hard and brilliant than the best natural colors, they still in many, if not indeed, in most cases, when carefully studied, show shades that are mixed and not pure. It is very rare to find a blue that does not incline a little to the yellow (a Blue G as it would probably be labelled) or else contain a trace of violet or red (Blue R, or RR). The reds are almost invariably either scarlets, containing a trace of yellow, or crimsons containing blue. And the yellows, also, are very apt to tend towards orange or occasionally show a trace of green. This, of course, complicates the problem for the practical dyer greatly, and means that instead of being able to cover the whole range of shades with a red, blue, and yellow, it is frequently, if not always, necessary to have some mixed colors, giving sharp, clear shades of violet, green, and orange respectively, to obtain certain effects.