Australian Flowering Plants. 221 



once apparent from a consideration of the doctrine of permuta- 

 tions and combinations. Let it be assumed, however, that a 

 family may have only 1,000 genera, and furthermore let it 

 be assumed that each genus may have only 2,000 species. Then 

 because the South American and Australian continents have 

 few points in common beyond the fact that each is in the 

 southern hemisphere, the chance that a particular family of 

 flowering plants should arise independently in each country 

 is somewhat similar to the chance that a die with 250 faces 

 and thrown fairly into the air, should fall down with any par- 

 ticular face uppermost. But there are 250 ways in which this 

 may be done, therefore the chance is only 1 in 250. Simi- 

 larly the chance that a particular genus should arise inde- 

 pendently in each country in such a particular family would 

 be similar to the chance of two particular faces on two dice 

 falling uppermost if the dice were thrown fairly, the one with 

 250, the other with 1,000 faces. But the odds are 250 to 1 

 against one particular face turning up on the one die if thrown 

 by itself and the odds are 1,000 to 1 against a particular face 

 turning up on the other die, therefore the odds against both 

 of these particular faces falling uppermost when the two dice 

 are thrown together is 250,000 to 1. Similarly the chance 

 that a particular species should arise in two countries inde- 

 pendently, all other things being equal, is somewhat similar to 

 the chance that a particular face on each of three dice, one 

 of 250 faces, one of 1,000 faces, and one of 2,000 faces, should 

 fall uppermost if thrown fairly into the air. But the odds 

 against this are 250 x 1,000 x 2,000 or 500,000,000 to 1, and 

 the chance that two identical species should be evolved 

 independently in each of two isolated countries . would be 

 only 1 in 250,000 billions. And the chance that 80 genera and 

 60 species should be evolved independently would be infinitely 

 more remote than the vanishing chance already mentioned. 



On the other hand, it is well known that plants are con- 

 tinually being transported by winds, sea currents, by birds, and 

 by man, from various lands to other lands, and, moreover, 

 geology teaches us that, in the past, various land masses have 

 been directly connected to each other. Identity of families, 

 genera, or species, therefore, are more simply explained by con- 

 sanguinity of origin than by the assumption of multiple origins. 

 The present distribution of the plants and animals in Aus- 

 tralasia, New Zealand, South Africa, South America, and the 

 Antarctic Islands, is, therefore, concerned rather with the rela- 

 tive merit of certain hypotheses of transport or of migration, 



