cii. XV] ENDEMISM AND DISTRIBUTION: SPECIES 157 



endemic species that covers all New Zealand, and reaches the 

 Chathams, and the other endemics occupy smaller and smaller 

 areas within this. The figures by zones show: 



2235556665 



a result exactly similar to that for Ranunculus. The presence of 

 wides does not seem in any way necessary, nor to cause the 

 species of a genus to behave in any way differently. 



If we go to Ceylon, and take a few species of the pan-tropical 

 genus Eugenia, of which Ceylon has 29 species endemic to the 

 island and 14 found elsewhere (6 only in southern India), we 

 find that E. cyclophylla occurs only on Adam's Peak, E. lu^ida 

 on several peaks close together, E. sclerophylla on a number of 

 peaks and in the plains between, E. assimilis throughout the 

 mountains and in the moist plains, E. hemispherica in all this 

 and also in South India, and E. operculata in these regions, and 

 also in Burma, Malaya, and China. And many other genera 

 show the same type of dispersal, which, in fact, a little study 

 soon shows to be the usual type. If one go to the state of Rio 

 de Janeiro in South Brazil, which has an area about equal to 

 Ceylon, one finds 52 Eugenias endemic to the state (which is 

 very mountainous), and 6 going beyond it, 3 only into Minas, 

 the next state, the other three as far as the states of Alagoas 

 (1000 miles north along the coastal plain), Rio Grande do Sul 

 (the same south) and Goyaz (600 miles inland, across the moun- 

 tains). And one may find Eugenia behaving in the same manner 

 in many other places. In Brazil it has many endemic species in 

 Minas, the next state to Rio, but on the other and dripr side 

 of the mountains that fringe the coast. 



This general type of distribution shows very clearly in the 

 case of very many genera, whether they be endemic genera with 

 all their species in a confined area, like Doona in Ceylon, or 

 whether they be genera of wide distribution that have developed 

 many endemic species within a certain small area, like Ranun- 

 culus in New Zealand. In such cases they do not seem as yet to 

 have encountered any barriers of a very serious kind. But one 

 may also find a great* number of genera, or sections of genera, in 

 which the same thing is displayed over a very much larger area 

 than what would entitle the contained species to be considered 

 endemic. In Callitris, for example, C. glauca occupies the whole 

 range of the genus over Australia and Tasmania (130); two 

 others range from New South Wales to Tasmania and to West 



