sea-floor spreading, at a half-rate of some 6.5 cm/yr, in the eastern 

 Bismarck Sea. The spreading axis has been active for about the past 

 3,5 m.y. This spreading is probably of the inner- arc basin type, 

 formed over the NW dipping New Britain subduction zone, and is probably 

 responsible for pushing the Manus Arc, an Oligocene feature (Coleman 

 and Packham, 1976) northward from New Britain over the Caroline Plate 

 along the Manus Trench. 



It is not clear that such spreading presently occurs in the western 

 Bismarck Sea. Figure 15 shows the western extension of the Bismarck 

 seismic zone as a leaky transform fault. Connelly (1976) felt that this 

 feature may have been the site of earlier spreading, but magnetic anoma- 

 lies have not yet been identified. I agree with Connelly that this 

 feature was at some earlier time a spreading center, and I speculate that 

 it pushed the western segment of the Manus Arc over the Caroline Plate in 

 a manner similar to that on the east. That this segment of Manus Trench 

 may now be inactive is indicated by the decrease in seismic activity from 

 east to west across the arc (Fig. 2; Johnson, 1979, Fig. 9). 



I believe this segment of Manus Trench is now inactive, as indi- 

 cated on Figure 15, and the present Caroline Plate margin is coincident 

 with the E-W leaky transform on the eastern side of the Bismarck Sea. 



Northern New Guinea is generally considered to be the result of a 

 collision between the Australia-South New Guinea landmass and a north- 

 ward subducting island arc, moving relatively southward, in late 

 Oligocene-early Miocene time (Moberly 1972; Hamilton, 1979; Johnson, 

 1979). 



63 



