As a measure of rate of cure we may take the actual measurements made on the record.[48] It is convenient to measure the elongation produced by a load of 130 kilos per sq. cm., as all fully vulcanised rings of soft rubber should give higher breaking load figures. For less cured or weaker samples a lower figure may be taken, such as 60 kilos. We have found that when fully vulcanised to give the maximal breaking strain, the elongation at a load of 130 kilos is in the neighbourhood of 850 per cent. (final length 950 per cent.). This applies to ordinary samples of estate rubber under the conditions of testing indicated above. If, however, the proportion of sulphur be considerably reduced, or mineral ingredients in a fine state of division be added to the mixing, or accelerators, whether organic or inorganic, be employed, the above relationship no longer holds. Nor does it hold with regard to plantation rubber prepared in an exceptional manner, as, for instance, matured coagulum or "slab." [48] Bulletin R.G.A., June, 1921, p. 246. There is a second method of determining the rate of cure--namely, by analysing a vulcanisate produced under standard conditions, and determining the amount of sulphur which has entered into chemical combination with the rubber. For this purpose the weighed sample is cut thin or creped thin, and exhaustively extracted with acetone to remove any "free" sulphur--that is, sulphur not in combination with the rubber. The sulphur remaining is then determined and calculated as a percentage of the raw rubber contained in the sample taken. This gives the so-called coefficient of vulcanisation. If we compare the coefficient with the time of cure at a constant temperature for an ordinary sample of plantation rubber, they are found to be approximately proportional, so long as the sulphur is in sufficient excess. The amount of combined sulphur is, therefore, an index of the time vulcanisation has been in progress (under standard conditions of temperature, etc.), and, therefore, the coefficient is a measure of the rate of cure. The change in position of the load-stretch curve is not directly proportional to the time of heating, and it therefore follows that it is also not directly proportional to the coefficient. For ordinary samples of crepe and sheet the relationship is, however, not very far removed from proportionality. This applies particularly to sheet rubber. The relationship is readily seen on plotting one against the other and tracing the curves. For sheet we get an almost straight line; for crepe there is some curvature.[49] For ordinary estate samples of sheet and crepe rubber the maximal breaking strain is obtained when the coefficient reaches approximately five units, so that this corresponds to the elongation of 850 per cent. at a load of 130 kilos. [49] Bulletin R.G.A., June, 1921, p. 246, October, 1921, p. 398.