Finally, on the question of adequacy of machines, it must be pointed out that an insufficient number of machines must result in a poor product, since all rolls have to be used for all grades. Even with the greatest possible care it happens that pale crepe is sometimes spoiled because it is contaminated with foreign matter, resulting from the working of lower grades on the same machines. This is one of the great arguments in another direction for the installation of a scrap-washer. In conclusion, the writers can only give their opinion that one must not decide the question of adequacy by the number of existing machines, but by the time taken each day in working off the rubber, providing one can be satisfied that the best arrangement of the existing machines has been made. IDEAL ARRANGEMENT.--As to what this best arrangement may be, guidance can be obtained from the results of experience here given. It must be premised that the output of any factory preparing fine pale crepe is limited by the output of the smooth finishing rolls. Broadly, it will be recognised that if there is any excess of capacity in a battery it should be found in the smooth-roll machines. This sufficiency, or excess of capacity, may sometimes be attained by an alteration in the gearing of the drive of the rolls from the back-shaft, or by an addition to the number of machines. In the former case, there are practicable limits of speed, beyond which the second alternative measure must be adopted. SPEED.--The usual speed at which the back-shaft travels ranges from 60 to 70 revolutions per minute. Taking first the macerating machines, the intermediate gearing between the shaft and the rolls should give a driving speed of about 20 revolutions per minute on the faster-travelling roll. This is equivalent, with a 15-inch diameter roll, to a peripheral speed of about 60 to 65 feet per minute. The intermediate and smooth rolls can be arranged to travel more quickly, but the maximum comfortable speed for proper feeding and control appears to be about 25 revolutions per minute on even-speed rolls. In view of the fact that the rubber at each successive machine becomes longer and thinner, it will be seen that a smooth-roll machine could not cope with the output of a macerator in the same period of time. If, therefore, the macerator is fully occupied for the greater part of the time, an additional smooth-roll machine must be installed, even though the existing one has been "speeded up" to practicable limits.