CO129-556-18 Chinese tin- reports on analyses 1-7-1936 - 20-10-1936 — Page 9

3.

9

large slabs are produced, with a view to finding cut whether appreciable variations could occur. The tin is brought to Hong Kong partially refined and is there further refined by a liquation method to bring it over 3 pure. Up to the Summer af 1335 practically all the crude tin refined in Hong Kong was mined in Yunnan, and this tin contained only small amounts of impurities other than 1991. During the process of refining, the workers periciically withdraw small buttens of the molten metal and allow them to cool. They juige the fineness of the metal by the appearance of the oxide film that forms over the button as it solidifios, and, 98 far as the Yunnan tine sera concerned, they were extremely expert in juiging the extent to which the refining procees hat bear carried out. Now, however, a large amount of tin is coming in from Kwanggai, and this whilst being of a better quality than the Yunnan tin, running ovar 3^% unrefined, containg a far higher proportion of Antimony. (Saa footnote.) The tins from the 11fferent sources are frequently mixed during refining, and if a batch of Yunnan tin does not easily refine up to standari, as juige1 by the visual method ployed by the workers, then slabs of the purer antimonial Kanggai tin are glial to the malt, in order to bringthe mixture to the requirei degree of purity. It seems probable that the film formation on the cccling buttons is mainly dependent upon the amount of lead present as impurity, and further, the method of refining ices not appear as affective for antimony impurity as for leal. This process, whereby tins from different sources are mixed, ices result in a decrease in the proportion of lead present, and so the batch may be adjudged as sufficiently pure and cast inte ingots for export. But the decrease in the lead is nocesserilly accompanied by a smaller increase in the proportion of antimony present, and as this impurity does not appear to affect the film formation, ita presence is not allowed for by the workmen. During the latter half of 1935 an unusually large proportion of ting failed to pasa as No.1, and the presence of the slowly sclubla antimony was very noticeable ir nearly all 01880.

Whilst no market fluctuations in quality would be expected to occur over the pouring of a whole large batch of slaba under the old conditions, it was thought that this ixing of tins from different sources would cause a variation in the quality of the metal between different parts of the same glab, and also between different slabe in the same batch. The metal is taken cut in large 19dles and pouret into the rculis, and one ladle is 1ivided between five or six mculis, whilst the composition of the main melt 18 varying continually accoriing as more or less of the Kwangsal tin is alied. Thus each large slab is made up of layers formed by successive additions, and these layers may be of differing composition. Moreover, each pouring is allowed to solidify before the next layer is alted, the worker carrying on with the next lot of moulds, so that some time may alapse between successive additions; periciiomlly there may be a complete break whilst a number of fresh slabs of crude tin are allal to the malt and hastal un. Each moult is filled as soon as the solidifto- ation of the successiva layers will allow, and the same mouli may be used several times for different slabe during the production of a large batch. When, therefore, a lot of 500 slabs is refined, which is the usual practice, there is every probability that the slabs will vary markedly in compositier. Some will contain more antimony than others, and whilst all may mass the workmada visual check, in actual fact a large proportion will be failures.

It was determined ther, to ic a series of 189ays on large rumber of separate slabs from the game lot. Accordingly, parcal of 300 alebe wee sampled in the usual way by trilling 31 of the slabe, but the drillings from each slab were kept separate. A composite sample was made and 988ayed, and gava a figure of 22.05 # tin. The separate samples from aach slab were also assayed, and as expectel, varied considerably, the lcwest giving 38.76 tir, and the highest 32.321 tin. The averaga was 92.00%, which agreed