24 
D.J. Kitchener, L.H. Schmitt, P. Strano, A. Wheeler, A. Suyai%> 
Table 7 Nei's unbiased genetic distance between Rhinolophus simplex populations and R. borneensis from Java. 
NUSA PENIDA 
0.005 
LOMBOK 
0.034 
0.033 
SUMBAWA 
0.035 
0.041 
0.000 
MOYO 
0.038 
0.045 
0.004 
0.000 
SANGEANG 
0.034 
0.041 
0.000 
0.000 
0.004 
RINCA 
0.038 
0.034 
0.000 
0.004 
0.008 
0.004 
SUMBA 
0.038 
0.034 
0.000 
0.004 
0.007 
0.005 
0.000 
FLORES 
0.034 
0.033 
0.000 
0.000 
0.000 
0.000 
0.000 
LEMBATA 
0.039 
0.036 
0.000 
0.005 
0.009 
0.005 
0.000 
ALOR 
0.070 
0.062 
0.026 
0.030 
0.031 
0.035 
0.026 
TIMOR 
0.033 
0.038 
0.000 
0.000 
0.002 
0.000 
0.002 
SEMAU 
0.032 
0.031 
0.000 
0.000 
0.001 
0.002 
0.000 
ROTI 
0.043 
0.035 
0.002 
0.007 
0.010 
0.011 
0.002 
SAVU 
0.028 
0.020 
0.029 
0.038 
0.042 
0.038 
0.030 
JAVA 
0.171 
0.161 
0.162 
0.163 
0.174 
0.172 
0.159 
BALI 
NUSA P. 
LOMBOK 
SUMBAWA 
MOYO 
SANGEANG 
RINCA 
R. megaphyllus whereas the narrower cranium and 
nasal swellings and overall small size of R. nanus 
are more suggestive of R. simplex. However, J.E. 
Hill (pers. comm.) informed us that the holotype of 
R. nanus is very old and has a damaged skull. 
Consequently, he considers that the narrowing of 
the skull of R. nanus may well be an artefact 
because otherwise the skull is very like that of R. 
truncatus. 
Hill (1992:100) separated the R. megaphyllus 
species complex (in which he includes R. s. simplex 
and R. s. keyensis) from other IndoMalayan species 
in the ferrumequinum group principally on the 
supraorbital crests combining at a point behind the 
centre of the orbital cavity; such that the 
supraorbital length (from junction of crests to 
nares) is much greater than rather than slightly 
greater than or equal to the width across the 
anterior lateral rostral swellings (in our 
terminology NIL NIB) and the supraorbital 
depression being larger than it is wide. The 
association by Goodwin (1979) and Hill (1992) of 
the form parvus with R. borneensis and R. celebensis, 
respectively, depended largely on the supraorbital 
crests of parvus merging anterior to the mid point 
of the orbital cavity. However, in the sample of 
parvus available to us this junction point was 
variable; sometimes it was level with the point, or 
just behind or well behind it (an R. simplex 
character). Also the supraorbital length is 
frequently much greater than the width across the 
outer lateral rostral swellings (see also Figure 9). 
Clearly in the form parvus (and amiri) this latter 
character is too variable for it to be usefully 
diagnostic in terms of the association of parvus, 
although it appears to hold true for R. borneensis 
and generally so for R. megaphyllus and R. s. 
simplex. Goodwin (1979:104) further considered 
that R. simplex differed from parvus in being larger 
overall and in having "dentition (that) is somewhat 
more primitive. The vestigial premolars in both 
upper and lower jaws are generally not as 
crowded, but there is some individual variation in 
this condition" Further, "the sella of simplex is 
slightly constricted and the connecting process is 
not as prominent". In the specimens available to us 
there was considerable variation in the extent of 
crowding in both the upper and lower vestigial 
premolar, particularly the lower. The lower 
premolar in both parvus and R. s. simplex varied in 
its position from almost being in the toothrow to 
being completely extruded such that this first and 
second premolars were in contact. Further we can 
find no consistent difference between parvus and R. 
s. simplex in the shape of either the sella or the 
connecting process. 
We associate the form parvus with R. simplex 
rather than with R. borneensis, as suggested by 
Goodwin (1979) or with R. celebensis as considered 
by Hill (1992). This is because of its morphological 
closeness to R. simplex and because, as discussed 
above, the characters used by these authors to 
diagnose it from R. simplex cannot be substantiated 
by us. It is also relevant here that our 
electrophoretic study, using 30 loci, concluded that 
there was little or no detectable genetic difference 
between R. s. simplex, R. s. parvus and R. s. amiri. 
For example the population of R. s. parvus (Timor) 
is not genetically differentiated from several R. $■ 
simplex populations (Sangeang, Sumbawa), while 
two populations of R. s. amiri (Roti and Semau) are 
closer genetically to the majority of the R. s. simplex 
populations than they are to the third population 
of R. s. amiri (Savu). The significance of the 
apparent clusters within R. simplex based on the 
Nei genetic distance metric is tenuous because it i5 
the product of gene frequency variation at just one 
or two loci. Thus Alor differentiates due to 
