290 MAMMALS. 



and rocky mountains, and the Dendrolagus for a residence in trees. An adaptation for a life 

 in trees could only be necessary where the soil on which thej^ grow, like the half-drowned coast 

 of New Guinea, afforded no rest to the sole of their foot. Another Kangaroo, not arboreal, 

 MACROPirs Bruni (Dorcopsis Bruni of Goiild) or the Filander, is found in the Island of Aru, 

 near New Guinea. Some say that it is found in New Guinea itself. This species is noted in 

 Johnston's " Physical Atlas," as found in Java, which, if true, woidd indicate that island as part 

 of the Australian group or system. It is a mistake, however; no marsupial animal is found in 

 Java or any of the islands west of the Straits of Macassar, and the error no doubt originated 

 in the first example that was met with, and ihat the first Kangaroo ever seen by Europeans, 

 having been observed by Le Brun at Batavia, and inferential^ supjjosed to be an inhabitant of the 

 land where he saw it. It appears from his narrative, however, that the animal which he saw was 

 not wild but in captivity. lie says, "Being at the country house of _ our general (at Batavia), I saw 

 a certain animal called Filander, which was somewhat remarkable. There were many individuals 

 )rifJi full freedom," (he woidd not have said that if they had been wild,) "running with some rabbits 

 which had their holes under a little hillock encircled by a balustrade."* This sjaecies, although 

 without the arboreal characters of the Dendrolagi, has yet some points of affinity with them, 

 as in the character of its fur, &c. ' 



Another of Gould's sub-genera (Onichogalea) has three species, one from the north-east, 

 and the other two from the interior of the east and the west. 



The Kangaroo Rats (sub-genera Lagorchestes, Bettongia, and Hypsiprymnus of Gould) number 

 about sixteen, of which only two (Lagorchestes) are found on the north or north-west coast, the 

 rest in East, Soiith, and West Australia, in nearly equal proportions, the south sharing two of 

 the species found in the west. 



Peramei.id.e. This is an insectivorous group, numerous in species, and universally dispersed 

 over Australia and Van Dieman's Land, which passes less easily from IIypsipryjinus than it does 

 by Tarsipes into Myrmecobius, and the Monotremes. It contains one species from New Guinea and 

 the neighbouring isles, one from North Australia, three from West Australia, three from New 

 South Wales, and one from Van Dieman's Land, and one of them is found in all the districts 

 except the north. 



Following these are three most remarkable genera, each consisting of only one species and 

 perfectly unique in tj^c, — the CniERGPUs castanotis, Tarsipes rostratus, and My'rmecobiits 

 FASCiATUs. These are all from the Swan River district of South-west Australia, apparently the 

 most ancient and peculiar part of the Australian continent. The last has spread a little into South 

 Australia, that is, one or two specimens have been found in that direction. 



Tarsipes and Myrmecobius both have long extensile tongues. Myrmecobius uses its tongue 

 for catching ants, as the ant-eater does. Tarsipes, not unlike a small harvest mouse, is said to use 

 its tongue in the same way as the humming-bird for extracting honey from flowei's, although 

 it also eats insects ; but it does not seem that because the same implement is applied to diflercnt 

 2)urposcs, our faith shoidd be shaken in it as an evidence of the affinity of the animals that possess 

 it. In the dentition and absence of the marsui^ial pouch of Myrmecobius, there is a marked approach 

 to the Monotremes, and more especially Echidna, which again on the other side holds out a hand 

 to the ant-eater section of the Edentata. 



* Le Brdn, "Voyage par la Moscovic en Perse et aux Tndes Orientalcs," vol. ii. p. 347, fig. 21.3. 



