On the other hand, there are one or two facts which are difficult although not impossible to fit in with theory. Thus, although the putrefaction bases are very easily soluble in water and acetone, they cannot be removed by washing on the creping rollers, or by acetone extraction. This may be due to the power of colloidal substances to retain other crystalloidal substances, such as the bases, which, in consequence, cannot be washed out. A parallel case is the retention of small quantities of water soluble substances in the soil. Also, the theory does not explain why rubber obtained by evaporation of latex at relatively high temperatures is fast vulcanising, although the possibility of putrefaction is excluded. As regards practical results, it follows that the rate of vulcanisation (or cure) of a sample of rubber will depend on the time allowed to elapse between the collection of the latex and treatment till the rubber is dry, as also on atmospheric conditions. Thus, slow drying will result in an increased rate of cure, for it gives an opportunity for putrefactive organisms to play a part. The results will, however, be influenced by the extent to which the rubber was washed previous to hanging, and so forth. Smoking is an antiseptic process and will, therefore, tend to inhibit the action of micro-organisms and produce a slower vulcanising rubber. On the other hand, sheet contains more serum than crepe, so that there is more food material for growth of micro-organisms. The net result is to give a rubber (sheet) which usually vulcanises a little faster than crepe. Among other factors controlling the rate of cure, special mention should be made of the nature and amount of coagulants. Weak "organic" acids, such as acetic, lactic, tartaric, etc., used in the minimal proportions (1 to 1,200 of standardised latex in the case of acetic acid), give the fastest vulcanising rubber; "strong" mineral acids, such as sulphuric acid, even when used in the minimal proportions (1 to 2,000), yield slower vulcanising rubber. Acid salts, such as alum, are intermediate in effect. Increased proportions of coagulant cause a reduction in rate of vulcanising with all coagulants, and the effect is least noticeable in crepe rubber, intermediate in sheet rubber, and most pronounced in "slab" rubber (discussed below).[55] [55] Bulletin R.G.A., July, 1919, p. 39; September, 1920, p. 343; November, 1920, p. 433; October, 1921, p. 393; March, 1922, p. 134.