The Philippine Sea Plate is the one major plate whose Tertiary motion
is poorly constrained and whose origin is problematical. Its southern
boundary is the Sorong Fault system which is part of a major left-late
ral fault system at the northern margin of the Australian plate. The s
outhern part of the plate in eastern Indonesia has been neglected in m
ost syntheses but includes some of the oldest rocks within the plate w
hich are separated from remnant arcs of the Daito Ridge province of th
e northern Philippine Sea by the West Philippine Central Basin. The ea
st Indonesian islands of the Halmahera-Waigeo region contain a good Me
sozoic and Tertiary stratigraphic record indicating a long are history
for the southern part of the plate. New palaeomagnetic data from thes
e islands define two sub-areas: an area forming part of the Philippine
Sea Plate north of the Sorong Fault, and an area within the Sorong Fa
ult system, The area north of the fault records a long-term clockwise
rotation history whereas that within the fault zone records local rota
tions interpreted as due to deformation at the plate edge. Rocks of Ph
ilippine Sea Plate origin within both areas record similar latitudinal
shifts. The rotation of the area north of the Sorong Fault is conside
red to represent the motion of the southern part of the Philippine Sea
Plate, The new data indicate large Tertiary clockwise rotations simil
ar to earlier suggestions for other parts of the plate but record a di
scontinuous and more complex motion history than previously suggested.
For the southern part of the plate there was 40 degrees rotation with
northward translation between 0 and 25 Ma, no significant rotation be
tween 25 and 40 Ma, and there was 50 degrees rotation with southward t
ranslation between 40 and 50 Ma. We show that the new palaeomagnetic d
ata form part of a single set with earlier palaeomagnetic data from el
sewhere in the plate. The translation history of the southern part of
the plate in eastern Indonesia can be reconciled with northward motion
s recorded elsewhere and can be used to determine rotation poles for t
he plate (15 degrees N, 160 degrees E for the interval 5-25 Ma, and 10
degrees N, 150 degrees E for the interval 40-50 Ma), Reconstructions
based on these poles predict that at similar to 45 Ma the Palau-Kyushu
Ridge had a WNW-ESE orientation which is very different from that pos
tulated by many models used to explain the widespread boninite volcani
sm in the Izu-Bonin-Marianas forearc at this time. The long are histor
y of the southern part of the plate and the reconstructions based on t
he rotation poles calculated from the palaeomagnetic data favour an or
igin for the West Philippine Basin by spreading in a backarc basin.