The roasting takes place in a row of iron pans, each about 21 feet in diameter, which are built into a platform of masonry about 15 feet long and 4 feet in breadth, covered with flat tiles. The pans rest in an inclined position, partly against the back of the platform, which rises about a foot above its level, and partly on a small prop of brick work on the right side, an offshoot from the wall. Into the top of this prop a plate is sunk, in which a cloth saturated with wood oil or miniak-krtieng is kept. Behind each pan is an open furnace mouth, and a man constantly attends to the fires, keeping them supplied with a few billets of bakau wood, and regulating them with a long two pronged iron fork, so as to maintain a, moderate heat. The pan being gently rubbed with the cloth, a man who sits in front of it on a low stool placed on the platform, pours into it a quantity of granular sago. This he slowly stirs for a short time with a wooden implement called weah, having a sharp curved edge. More sago is poured in, until it amounts to about two chupa, when, as it hardens, he uses the weah more freely. After about three minutes' roasting, it is removed to a table and passed through a round sieve similar to that before described. The grains that adhere to each other are thrown aside, and those that pass through form a smoking heap, which is allowed to lie un disturbed for about twelve hours. The grains are about the same size as they were before roasting, and some retain wholly or partially their white and mealy appearance, but the greater part have become translucent and glutinous, and all have acquired a certain degree of toughness, although still soft. This change appears to be brought about in this way : The water contained in the granules being heated, first converts the mealy starch into a jelly, and then escapes by evapora tion, leaving the jelly tough. The second tume
faction drives out the remainder of the water. The final process is another roasting, which renders them hard and tough, and greatly reduces their size. The pearl sago thus prepared and fit for exportation, is put away in large open bins ready to be transferred to boxes or bags when sold.
This method of making pearl sago was intro duced into Singapore in 1819. It was taught to the first Chinese who tried it there, by a woman who came from Bukit Batu, a place,on the coast of Siak, facing the large island of Bencoolen, and famous for its great fishery of the trubu, the roe of which is so extensively used. In Malacca, however, manufactories had existed for many years before the establishment of Singapore. It is certain, however, that Malacca derived the art from Bukit Batu. In Singapore it appears almost from the first to have been conducted without any attempt at concealment.
Sixteen men would suffice for a manufactory such as described above, and they would produce about 450 pikuls per month. The original out lay is probably from 300 to 400 dollars. There were, about the year 1840, fifteen Chinese manu factories in Singapore.—Tour. Ind. Arch. ; Ben nett's Ceylon ; Low's Sarawak; Forrest's Voyage to New Guinea ; Crawfard, Hist. Ind. Arch. ; Hogen dorp, Coup cl'CEil Stir Java ; Dampier, Voyages, i. pp. 310, 311 ; Rumphius, Nat. Hist. Amboinensis, p. 80 ; St. John's Arch. ; TVallace, Arch. ii. p. 66 ; Roxb. Fl. Ind.; Voigt ; Poole's Stat. of Commerce; Seeman on Palms ; Madras Ex., 1855, Juries' Re ports ; Report Madras Central for Ex. of 1851 ; Dr. Pringle; Griffith's Palms ; Mason's Tenasserim ; Morrison, Compendious Description ; Crawfurd's Dict. Archip. ; Dr. TV: Jack, Malayan Miscellanies published at Bencoolen; Valentyn, Oud en N. 0.J. ; Ili. de Steur's Tijdschrift, Neer L., Ind. 8th year, part p. 367 ; Singapore Chronicle.