However, in the Northern Shan States, the early Ordovician lime- stones are not present (Mitchell et al., 1977), and the whole sequence is dominated by clastics. Given the otherwise striking continuity of the Sibumasu terrane, this difference is significant. Ordovician palaeontology in Burma still relies heavily on the descriptions of Reed (1906, 1915, 1936). These faunas have never been fully re- vised, and their palaeogeographical value is ac- cordingly to be treated with some caution, al- though we have revised brachiopod faunas (Cocks and Zhan, 1998) from the Shan States collected by Mitchell et al. (1977) in their survey. Our comments below on the trilobites are pre- liminary, but important to illustrate the problems involved. There is some evidence that Reed mis- associated some pygidia and cephalic parts.