Species crosses in Rats. 103 



After we had noticed tho frequent appearance of novel characters 

 in the descendence of hybrid plants, especially at the grounds of the 

 firm de Vilmorin, we planned a series of crossbreding experiments with 

 animals, designed to test the possibility that here, as in plants, cross- 

 breeding species would i-esult in new characters. When wc did not 

 succeed in obtaining hybrids between the tame mice and related species, 

 we turned our eyes to rats of the Mus rattus group. From the ex- 

 periments of Morgan we knew that hybrids between the roofrat and 

 the houserat were fertile and the material seemed promising. We were 

 fortunate enough to be able to buy Dr. Bonhnte's rats when he had 

 to leave for the Orient, and this gave us our start. 



Experimental. 



The experiments to be described here are crossbreding experiments 

 with rats of the Mus rnttvs group. Mus raftus is a cosmopolitan group 

 of species, common in the old world, in Europe, Africa and Asia. In 

 colder climates the rats live in houses and warehouses and in such 

 localities there exists only one species, a houserat, which is exceedingly 

 multiform as to size and colour, but which is always found in colonies 

 which are remarkably uniform. Lloyd has studied this phenomenon in 

 British India, and has come to the conclusion that these small groups 

 are essentially small species. The rats used in our work belong to 

 seven species: . 



Mus fedorum, the North African treerat, 



Mus alexandrinmn, the North African houserat, 



Mus rattus, the West European houserat, 



The Javanese houserat, 



The Javanese treerat, 



The Javanese feldrat, 



The fieldrat from East Sumatra. 



Origin of the first novelties. 



We began our work with a colony of rats bred by Dr. Lewis 

 Bonhote in England. He had crossed Mus tectorum and alexandrinum. 

 The formei- is a typical whitebellied , active treerat, rufous red, with 

 long tail, alexandrinum is a graybellied larger rat whose mode of life 

 corresponds entirely with that of the European Mus rattus. The hybrids 

 were whitebellied, and among the descendants of these hybrids a minority 

 of greybellied rats were obtained. Dr. Bonhote also obtained a few 



