Although at first sight the estimation of the colouring matter in a dye-wood by means of the colorimeter appears somewhat complicated, yet it yields the most accurate results with the smallest expenditure of labour and time. The value of a colouring material may also be estimated by preparing the pure colouring matter from a weighed quantity. This process is lengthy, demands considerable practice, and only gives good results when it is carried out with the most painful accuracy. The colouring matters in question are precipitated by lead salts. If the dye-wood extract contained colouring matter alone its amount could be found by observing the volume of a lead solution of known strength required to precipitate the colouring matter completely. The decoctions, however, contain other substances which form lead compounds, and are precipitated together with the colouring matter, so that if the precipitate were regarded as the pure lead compound of the colouring matter a very inaccurate result would be obtained. In order to obtain results with some pretensions to accuracy the lead compound of the colouring matter must be purified. The impure precipitate is washed and suspended in water, through which sulphuretted hydrogen is passed until all the lead is precipitated as lead sulphide, which is filtered off, excess of sulphuretted hydrogen expelled by boiling, and the solution again precipitated by a lead salt. This precipitate may be regarded without considerable error as the lead compound of the colouring matter. The weight of colouring matter contained in the quantity of wood used can be calculated from the weight of the dry precipitate. This method is somewhat complicated and tedious; the results are inferior in accuracy to those obtained by means of the colorimeter, which instrument furnishes the most suitable method for testing dye-woods for practical purposes. A thorough knowledge of chemistry is indispensable to the colour manufacturer who wishes to carry on his business on any extensive scale. It enables him to match any sample of colour submitted to him and to test his raw materials with ease. We have indeed given in a section of this book simple methods by which the majority of commercial pigments can be tested with tolerable accuracy by means of a few reagents, and for ordinary purposes these methods are sufficient. But when an accurate examination of a pigment is required, it must be conducted by the ordinary processes of analytical chemistry. The colour manufacturer has not only to carry out these occasional examinations, but has frequently to test certain raw materials which he uses in large quantities. Soda may be taken as an example. An estimation of the percentage of sodium carbonate in this substance is an exceedingly simple matter to the chemist, but can hardly be carried out without a knowledge of chemistry.