The wild trees grow from 50 to 100 feet in height, and often have a girth of three feet at four feet from the base; the trunk being commonly without branches for a considerable height. The native ascends the trunk by means of a sling passed round it, to one end of which is a loop which fits over his right thigh and to the other a stirrup in which he keeps his left foot. By moving the rope up the tree the ascent is quickly accomplished, and when stationary, both hands are left free. The climber carries a gouge, or semi-cylindrically-faced chisel, with which he cuts vertical as well as transverse oblique grooves, forming a rough “herring-bone” pattern, which may be continued for the whole length of the trunk, and even extend on to the main branches. To make the cuts the tool is either used in one hand or is held in the left and driven by blows on the end of the handle, given with the palm of the right hand. The flow of latex commences at once, and is directed down the vertical channel into a calabash placed at the base of the tree, into which the latex is guided by a lip of clay or a chip of wood. The illustration shows a native in Ashanti tapping a tree (Fig. 18). The quantity of latex procured at the one tapping varies; an exceptionally large amount being about two quarts from a large tree. The tree is not tapped again for several months and until the wounds have healed. On the second tapping the same form of cuts is made upon the opposite side of the tree, and the transverse channels often intersect those made previously; besides this, owing to the want of regulation in the depth of the cuts, the inner or cambium layer of the bark may be so damaged that the intersecting cuts may ring the tree and cause death. Few trees probably survive the third or fourth tapping. [Illustration: “ODUM” TREES _CHLOROPHORA EXCELSA_. Fig. 19, p. 72.] [Illustration: PARA RUBBER TREE (_HEVEA BRASILIENSIS_), TAPPED AT ABURI. Fig. 20, p. 66.] [Illustration: RUBBER TREE TAPPED, HERRING-BONE SYSTEM IMPERFECT, ABURI. Fig. 21, p. 61.] If the herring-bone cuts were made carefully upon the lower part of the trunk, and attention given to their depth and direction, the tree might be preserved for a longer period. An illustration given shows the incorrect mode of tapping, where the transverse cuts are opposite instead of alternate (Fig. 21). Experiments seem to show that there is not the same “wound response” in the case of Funtumia trees as is found in that of the “Para” tree (_Hevea brasiliensis_); the paring of the wounds some days after tapping yielding no flow of latex similar to that procured in the first instance. On cutting the bark area between the transverse channels, the latex cells appeared to be empty and to indicate that a large area is drained on the first tapping. Further investigation is necessary in order to compare the results obtained from this tree with those of _Hevea_.