174 
PACIFIC SCIENCE, Vol. IX, April, 1955 
almost to the elbow in all specimens. On 
the foot darker colouring is visible along the 
outer edge. These characters are semi-diag- 
nostic and, when combined with other char- 
acters, can be used to separate an exulans type 
rat from a young R. rattus or R. norvegicus . 
The fur on the back and sides contains 
many spines but in the ventral fur there are 
fewer thinner spines. 
Two black specimens were caught on the 
edge of the rainforest. As far as is known 
these are the first records of a black R. exulans , 
at least from Samoa. However, the measure- 
ments are not included. 
Skin Measurements 
Table 1 gives a summary of the measure- 
ments taken. Forty-four adults were exam- 
ined in all, but through damage either in 
trapping or previously, certain individuals 
were excluded from some of the measure- 
ments. 
It may be seen in the table that the figures 
for f length” and "foot” differ between Ma- 
nono and Upolu with a probability in each 
case of less than 0.01 that this is a chance 
variation. The figures for length of tail have 
P<0.05 which is usually taken as the "prob- 
ably significant’ ’ level of a statistical difference. 
Thus there is a statistically significant differ- 
ence between the populations on Upolu and 
Manono on the external measurements alone. 
Skull Measurements 
A summary of the skull measurements is 
given in Table 2. Because of the severe 
breakages caused by trapping with breakback 
traps only 31 skulls were worth preserving. 
Of these several were broken in some way and 
so the number of individuals measured for 
each dimension varies considerably. The 
braincase was the most damaged but loss of 
the auditory bullae was very common partly 
due to their falling out as the skull dried. 
Five of the measurements: condylobasal 
length, zygomatic breadth, breadth of brain- 
case, length of mandible, and length of bulla 
show a similar difference to the external meas- 
urements between the populations of Upolu 
and Manono each having a chance probabil- 
ity P <0.01. Four measurements differ with 
P <0.05 thus supporting the belief in a true 
difference. In all nine cases the rats of Manono 
give the smaller figure. 
Parasites 
Twenty-eight of the rats examined were 
infested with mites and three had lice in the 
fur. It is interesting that no fleas were found 
living on the Polynesian rat. 
Gut parasites were almost universal; 75 per 
cent of the rats examined had nematodes 
living in the caecum and of these 15 or 38 
per cent are listed as heavy infestations. Nem- 
atodes were also found commonly in the 
rectum and a few in the intestine. In the 
intestine cestodes occurred in over 50 per 
cent of the rats and in one two acantho- 
cephalans were found. One cestode was found 
in the caecum. None of these parasites has 
as yet been identified further. 
DISCUSSION 
Though the series of measurements is not 
as long as might be desired, it has a value 
because of the limited areas from which the 
specimens were taken. The two areas are 
within one archipelago and were perhaps 
colonised by the rats at more or less the same 
time. Thus it is especially interesting that the 
results indicate differing populations in the 
two areas. The difference must be tentatively 
accepted for the combination of the statistical 
differences with the visual difference in pellage 
is more than suggestive, though whether it is 
phenotypic or genotypic is arguable. 
Taking the measurements as a whole, it is 
possible to make comparisons with other pub- 
lished data. Tate (1935) divides the R. exulans 
types into six groups on general size and 
especially on the bulla and toothrow measure- 
ments. He gives Samoa a separate group but 
he only had a few specimens from Eastern 
Samoa for the basis of his conclusion. The 
