6 M O U M O U India, has no less than eighteen mammte, nearly the largest number found among the Muridie. III. Golunda, Gray, like Mus, but with a distinct groove down the front of the upper incisors. There are only two species, one from western India, and the other from eastern Africa. IV. Uromys, Peters., differs from Mus in having the scales of the tail not overlapping, but set edge to edge, so as to form a sort of mosaic work. There are about six species of Uromys, spread over the northern part of the Australian region from the Aru Islands to Queensland. V. Hapalotis, Licht. Hind-limbs elongated. Incisive foramina veiy large. No coronoid process to the lower jaw. This genus is confined to Australia, where there are about fifteen species known. They are pretty little animals, with long ears and tail, and in many respects resemble the Jerboas, whose place they seem to take on the sandy Australian deserts. VI. Mastacomys, Thomas, like Mus, but with the molars remark ably broadened, and with only four mammae. The single species in the genus is as yet only known from Tasmania, though it has been found fossil in New South Wales ; it is somewhat similar in size and general appearance to the English Water-vole, but has much longer and softer fur. VII. Acanthomys, Less. Fur almost entirely composed of flattened spines. Coronoid process very small. There are six species of Spiny-mice known, all of about the size of the Common Mouse. They are found in Syria, Palestine, and eastern Africa as far south as Mozambique. VIII. Echinothrix, Gray, a very remarkable rat with an ex tremely elongated muzzle, all the bones of the face being much produced. The incisors are faintly grooved. The only species is E. leucura, an animal of about the size of the Common Rat, with its fur thickly mixed with spines. It is found in Celebes. The remaining genera belong to the Sigmodontes ; they are rather more numerous than those of the Mures, but, on the whole, present somewhat less strongly marked generic differences. IX. Hypogeomys, Grand., a very peculiar form of large size, with long ears, feet, and tail. There is only one species, H. antimena, a fawn-coloured rat about 9 inches long. X. Ncsomys, Peters., contains two species of long-haired rats, more or less rufous in colour, about the size of the House-rat. XI. Brachytarsomys, Giinther, contains only B. albicauda, a pretty velvety-haired fawn-coloured rat, with short feet and a long tail. XII. Hallomys, Jent. The only species, H. audeberti, is very like a Nesomys, but has much longer hind-feet. This and the last three genera are confined to Madagascar. XIII. Hesperomys, Waterh. Molar structure as shown in fig. 2, B. The Mus of the New World, containing the great mass of the rats and mice of America, and having no very special generic charac ters common to all its members. This large genus is composed of at least seventy distinct species spread over all America from Canada to Cape Horn, of which none are quite as large as Mus decumanus, while several are considerably smaller than Mus musculus. They have been split up into ten sub-genera, of which perhaps the best marked is Rhipidomys, a small group containing about ten species, remarkably like Dormice in their habits and general appearance, having soft woolly fur and long hairy tails, and living entirely in trees, bushes, or in the roofs of houses. The other Hcsperomys are all terrestrial in their habits, much as the Old- World rats and mice are. One only, H. spinosus, a native of Peru, has as yet been found with spines in its fur, a rather remarkable circumstance when we remember how many of the tropical species of the allied genus Mus have more or less spiny fur. XIV. Holocheilus, Brandt, like Hesperomys, but with the third upper molars proportionately larger and the skull more stoutly built. This genus, confined to Brazil, contains about six species, some of which are the largest indigenous rats of America. Two species are aquatic in their habits, and have therefore developed short webs between the toes of their hind-feet. XV. Sigmodon, Say and Ord, differs from Hcsperomys in the pattern of the molar teeth. It contains one species only, the Rice- rat, S. hispidus, which ranges from the United States to Ecuador. XVI. and XVII. Reithrodon, Waterh., and Ochetodon, Coues., more or less like Hcsperomys, but with grooved upper incisors. The first of these is a South-American genus, and contains four rat-like species, one from Venezuela and the other three from Patagonia. The second consists of three North American mice, of about the size and proportions of the English Wood-mouse, Mus syh-aticus. XVIII. Neotoma, Say and Ord, a peculiar North - American group, in which the teeth have the prismatic appearance of those of the Arvicolie (see VOLE). There are four species known as "Wood -rats," all of about the size of Mus decumanus, one of them, N. cincrea, having a tail almost as bushy as a Squirrel s ; the other three with ordinary scaly rat-like tails. From the ranges of the genera given above it will be seen that all the first group, the Mures, are confined to the Old World, and that of the Sigmodontes four genera are found in Madagascar and the rest in America, thus giving us a very remarkable instance of the peculiar affinity that the fauna of Madagascar has with that of the New World. This affinity is usually explained by the fact that those animals which show it belong as a rule to groups formerly distributed over both the Old and New Worlds, and that since the isolation of Madagascar these, owing to the competition of more highly-organized forms, have been exterminated or strongly modified throughout the continents of the eastern hemisphere, while in the western they have been preserved to the present time. Thus in the present case it seems probable that the original ances tors of the Murinse, if not indeed of the whole family Muridse, were Sigmodontes having molars with their cusps biserially arranged, 1 and that these, being less powerful in the struggle for existence, as is shown by the manner in which roving members of the Mures rapidly multiply at the expense of the indigenous Sigmodontes of any place they may be introduced into, have gradually succumbed to the more recently developed Mures wherever the latter were able to penetrate, Madagascar having previously become an island, and therefore inaccessible to them. Other groups, however, also probably descendants of Sigmodont Muridae, have become so strongly modified either as to structure or habits as to have been able to avoid the rivalry of the Mures, and thus to exist side by side with the latter ; such probably are the Hamsters (Cricetus) and the Voles (Arvicolci), both of which have modifications of the biserial arrangement of the molars. As to the Murines from Australia a region isolated from the rest of the world far earlier than Madagascar with their very various degrees of specialization, it seems prob able, as Mr Wallace has suggested, 2 that from very early times individual rats and mice have drifted on floating trees and other objects from island to island along the Indian archipelago down to Australia, and that the de scendants of the earliest arrivals have become the most modified, and that others have been continually joining them, until we get the present state of affairs, namely, one or two genera very markedly different from Mus, others but slightly different, and finally numerous species not generically separable from the European and Asiatic rats and mice. (o. T.) MOUSE-BIRD (Dutch Muisvoyel), the name by which in Cape Colony and Natal the members of the genus Colius 3 of Brisson are known partly, it would seem, from their general coloration, but more probably from their singular habit of creeping along the boughs of trees with the whole tarsus applied to the branch. By the earlier systematists, who had few opportunities of examining the internal structure of exotic forms, Colius was placed among the Fringillides ; but nearly all travellers who had seen one or another species of it in life demurred to that view. Still its position was doubtful till Dr Murie, in an elabo rate treatise on its osteology (Ibis, 1872, pp. 262-280), showed that it was no Passerine, and subsequently (Ibis, 1873, p. 190) proposed Pamprodactylx, as the name of the group of which the Family Coliid& is the sole type this word being coined to indicate the obvious character of all the toes being ordinarily directed forwards, but by no means the only peculiar character these birds possess. A few years later most of Dr Murie s views were confirmed 1 The teeth of the extinct genus Cricetodon from tlie Miocene of France and Germany are in their essential structure quite similar to those of Hesperomys. 2 Australasia, p. 53, 1879. 3 Some other generic divisions have been suggested, but on grounds so slender as hardly to merit consideration.
