The weights are placed vertically and the duration of drying horizontally. It will be seen that the rubber was quite dry and fit for packing on the sixth day, as far as could be judged in the usual way by casual inspection. Peculiarly enough at this time it weighed slightly more than had been registered on the fourth and fifth days, but the difference did not amount to more than about 0·4 per cent. In examining the graph it should be borne in mind that the steeper the slope of the curve downwards the quicker the rate of drying, and that when the curve takes an upward direction there is an addition of moisture instead of abstraction. It will be noted that when drying takes place the slopes more nearly approximating the vertical represent the extent of day drying, and that often the night drying is represented either by a very flat curve or even by an upward curve which shows the addition of moisture. A striking feature of the experiment is shown by the rapidity with which drying takes place during the first few days and the comparative slowness with which the remaining moisture is got rid of. Thus, from the graph, it may be calculated that about 80 per cent. of the total moisture content was lost in the first two days, and over 93 per cent. in two and a half days. Yet three days had to elapse before the remaining 7 per cent. of total moisture was lost--_i.e._, before the rubber was judged to be ready for packing. It will be seen that after this stage had been reached the rubber alternately lost and gained in weight, with a tendency to increase. This increase was attributed to the presence of surface moisture after hanging overnight, when the rains had become frequent. Some light is thus shed upon a subject which has puzzled both shippers and receivers of crepe rubber. DIFFERENCES IN WEIGHT.--It will be obvious that if rubber is allowed to hang after becoming dry, and is taken down, packed, and weighed in the early morning, it will weigh more than when it reaches a drier climate. The loss in weight under such circumstances might amount to even 1 per cent. It may seem to some an unnecessary refinement to introduce, but it would appear from the graph that rubber should be packed for preference in the afternoon if the weights are to be more nearly correct.