This change is due to the gradual oxidation of the constituents of the paper, the ligno-cellulose of the mechanical wood pulp being most readily affected by the high temperature, and the pure cellulose of the rag paper being least altered. The process of oxidation, brought about rapidly under the conditions of the experiment described, takes place in papers of low quality exposed to air in the ordinary circumstances of daily use, but of course at an extremely slow rate. The deterioration of such paper is not, however, due to the simple oxidation of the cellulose compounds, because other factors have to be taken into account. The presence of impurities in the paper on the one hand, and of chemical vapours in the air on the other, hastens the decay of papers very considerably. _Percentage of Cellulose in Fibrous Plants._--The value of a vegetable plant for paper-making is first determined by a close examination of the physical structure of the cellulose isolated by the ordinary methods of treatment. If the fibres are weak and short, the raw material is of little value, and it is at once condemned without further investigation, but should the fibre prove suitable, then the question of the percentage of cellulose becomes important. There are several methods employed for estimating the amount of cellulose in plants. The process giving a maximum yield is known as the chlorination method, the details of which are as follows:--About ten grammes of the air-dried fibre is dried at 100° C. in a water oven for the determination of moisture. A second ten grammes of the air-dried fibre is boiled for thirty minutes with a weak solution of pure caustic soda (ten grammes of caustic soda in 1,000 cubic centimetres of water), small quantities of distilled water being added at frequent intervals to replace water lost by evaporation. The residue is then poured on to a piece of small wire gauze, washed thoroughly, and squeezed out. The moist mass of fibre is loosened and teased out, placed in a beaker, and submitted to the action of chlorine gas for an hour. The bright yellow mass is then washed with water and immersed in a solution of sodium sulphite (twenty grammes of sodium sulphite in 1,000 cc. of water). The mixture is slowly heated, and finally boiled for eight to ten minutes, with the addition of 10 cc. of caustic soda solution. The residue is washed, immersed in dilute sodium hypochlorite solution for ten minutes, again washed, first with water containing a little sulphurous acid and then with pure distilled water. It is finally dried and weighed.