404 



THE AMERICAN NATURALIST 



[Vol. LI 



ing. As a rule nothing happens in particular when we 

 introduce a NorA^egicus to a Eattus rat. But if we put in 

 one cage two Eattus rats which are strangers to each 

 other, they almost always start a fight, and generally it is 

 a matter of life and death. 



Mus norvegicus is a real water-rat, sewer-rat, field-rat, 

 and in some parts of Java, where the poor houses have 

 no floor, and where there are many covered sewers, as in 

 Solo and Sourabaya, it becomes a house-rat in a certain 

 sense. But it takes extreme negligence of the inhabitants 

 of a house to make it shelter this rat. It happens that in 

 houses where the bedstead is never moved from its place, 

 and where the space below it is used as a place to dump 

 the garbage, that this rat establishes itself there, exca- 

 vating numerous large burrows. 



It is very remarkable that this rat, which is extremely 

 uniform all through its range in Europe, is very rare in 

 this island, where the geographical distribution shows it 

 to be a recent immigrant, which has come in by way of 

 the big rivers. The skull, the shape of the parietal ridges, 

 the relative size of the bullae, the relative size of the 

 molars, the relative position of the choane, characters 

 which are very constant in Europe, become very variable 

 here. In size it varies very much, the biggest individuals 

 weighing nearly twice as much as the biggest European 

 animals. The color, which hardly varies in Europe, 

 varies between very light gray agouti to a silvery blackish 

 dark gray, with dark belly there. It looks to us very 

 probable that this great variability may be the result of 

 crossbreeding between this species and species of Gu- 

 nomys or Bandicota. The variation of the species in Java 

 is certainly towards the characters of these rats, which 

 have a somewhat similar mode of life as Mus norvegicus. 



As yet it has not been possible to make Bandicoots 

 breed in cages, although we have tried to make them do 

 so in very quiet concrete rooms. Whenever it will be 

 possible to breed these rats it will be very interesting to 

 observe whether Mus decmnamis will mate with Bandi- 



