[12] "Preparation of Plantation Rubber," Morgan, 1913, chapters xii. and xiii. With a view to determining to what degree the drying of crepe rubber was hastened by the extent to which the rubber was rolled, experiments were made. It was hoped, also, that some idea would be gained of the particular stage in crepe rolling which had the greatest effect upon the rate of drying. In preparing crepe in the estate in the ordinary way the coagulum is passed through three sets of rollers, and the stages may be described as: (1) Rough rolling. (2) Medium rolling. (3) Smooth rolling. In the first the coagulum is broken down by passing through the machines until a thick rough crepe is formed. This passes to the intermediate rollers, where it is worked down to a medium crepe. The rubber finally goes to the smooth running at approximately even speeds. Passing through these a number of times it emerges as a thin uniform crepe, free from "lumpiness" and free from holes, which should dry in from ten to twelve days. In the experiment the rubber was passed through the machines with varying frequency, the number of times in each machine being progressively increased, while the working on the other machines remained constant. It was determined that the rate of drying was affected only by the extent to which the crepe was worked in the smooth rolls. The less often the rubber passed through these rolls, the slower the rate of drying. Beyond a limit in the other direction, increased rolling did not reduce the period of drying. It follows, therefore, that crepes which have a good thin finish should dry in a minimum period. [Illustration: DRYING GRAPH. PALE CREPE (THIN).]