There are two areas where there is confusion over possible overlap between the two exclusive biogeographic markers proposed here. In the Omati to Chimbu Gorge region of Papua New Guinea (Figure 3, area 10 and "O") there is clearly some overlap, although this can be accomodated in the model proposed here, and the absence of Lacazinella from known low latitude outcrops remains an independent index of biogeographic origin. However some more work in the Eocene faunas of this region of PNG would be useful, if only to establish the scarcity of the APB fauna here. Bain and Binnekamp (1973) returned to the same area as Crespin's original samples and recorded common Lacazinella but noted that "no Pellatispira or Biplanispira ... occurs". Crespin's plates, however, appear clear on the matter, and the only confusion is over the extent of overlap.

In eastern Sulawesi there is confusion over sample locations. One interpretation wholly supports the model proposed here, but there is no solid reason to believe this is the correct one. New and accurately located Eocene samples from this area would clarify this issue and may better identify the tectonic suture in this geologically complex area.

Some notes on both these areas are given in Appendix 2.

Although not problematic to the model proposed here, a number of understudied Eocene records could be important applications of the APB - Lacazinella model in Tertiary plate reconstructions. Possibly of most importance, from a geologic perspective, would be the polymict conglomerate of the Wani beds (from the Tobelo Mts.) in North Buton, where there are records of general Eocene Nummulites and Asterocyclina/Discocyclina (van Bemmelen 1949 p. 148 and 420). Identification of APB or Lacazinella faunas in this area would be important evidence regarding the origin and time of drift of this plate and the adjacent Tukang Besi micro-plate. Van Bemmelen (1949, p. 148) notes that "Several large inliers of Eocene limestone are known from Buton." presumably basing this statement on his main quoted source - the work of Hetzel (1936). However a review of this older work and newer GRDC work suggests the Eocene is not as widespread as van Bemmelen suggests, and the task of finding Eocene pebbles in a polymict conglomerate would be difficult.

Data from the Eocene outcrops on islands along the current collision zone of east Indonesia would help geological reconstructions. The small islands of Raijua and Rotti, described by Verbeek 1908, both have carbonate faunas (cf. van Bemmelen) containing the Middle Eocene marker Alveolina but as yet neither the APB or Lacazinella fossils. In addition the Eocene larger foram facies on Buru lack any of the biogeographic markers, and more work here would substantiate the data from the adjacent island of Seram.

An Eocene fauna from Ocean Drilling Program (ODP) Leg 133 on the Queensland plateau has been studied by Betzler (personal commun. 1995). This Eocene carbonate contained Nummulites and Discocyclina but none of the biogeographic markers described here.

As mentioned earlier, some more detailed work on Late Eocene heterostegines may elaborate the geographic range and evolutionary relationships between Grzybowskia jasoni Lunt and Operculina eniwetokensis Cole. If more records of G. jasoni are to be found, the work here predicts a moderately low latitude palaeogeographic range.