E. B. Tylor found it used by a travelling tinker at Pæstum. The second and improved variety of African bellows was described by myself during a visit to Yoruban Abeokuta. It deserves attention because it is a notable step in progress, leading to a further development; the troughs are a rudimentary cylinder, and the handles form an incipient piston.[424] ‘The two bags of goat-skin are made fast in a frame cut out of a single piece of wood; the upper part of each _follis_ has, by way of handle, a stick two feet long, so that it can be worked by one man either standing or sitting. The handles are raised alternately by the blower, so that when one receives the air, the other ejects it; the form is like that used on the Gold Coast; and there is a perpendicular screen of dried clay through which the nozzle of the bellows passes, supplying a regular blast.’ Evidently in this stage of the bellows, the lower halves of the leather bags are useless: the result would be the same if only the upper part of the wooden troughs were covered with skin, air-tight but loose enough to make play. This third step has been taken by the Djour (Júr) tribes of the Upper Nile, in north latitude 20°, and it is thus described by Mr. Petherick: ‘The blast-pipes are made as usual of burnt clay, and are attached to earthen vessels about eighteen inches in diameter and six inches in height, covered with a loose, dressed goat-skin, tied tightly round them and perforated with a few holes, in the centre of which is a loop to contain the fingers of the operator. A lad, sitting between two of these vessels, by a rapid alternate vertical motion drives a continuous current of air into the furnace.’ [Heading: _THE BELLOWS IN AFRICA._] [Illustration: FIG. 105.—PORTABLE AFRICAN BELLOWS.] This brings us to the fourth and last stage of African blast-improvement (fig. 105). Here the rudely-hewn wooden tube becomes a double-barrelled forcing-pump. The two air-vessels with their loose skin-coverings are attached to each base of the two central pipes that join into one. Such is the shape used in Madagascar, the cylinders being of bamboo, five feet long by two inches in diameter, and the piston a stick ending in a bunch of feathers. The bellows described by Dampier in Mindanao and elsewhere in the Malay Archipelago, is evidently borrowed from the Madagascar type; and into Borneo, Siam, and New Guinea a hollowed trunk takes the place of the bamboos. The sculptures in the Sukuh-temple of Java, attributed to the fifteenth century, represent smiths making Kríses (Creases), the bellows being worked by another man, who holds a piston upright in each hand. Colonel A. Lane Fox is of opinion that the sculptures ‘possibly point to a Hindu origin for this particular contrivance.’ I agree with him, but I would also trace the Asiatic article back to its old home in Africa—Egypt.