CH. vii] CONFIRMATION BY PREDICTION 69 



rarity in New Zealand represented by the figure 4-3, i.e. 0-7 

 greater than the largest figure for any that reach the islands 

 (Kermadecs, 3-6), or a range of 84 miles less. This difference 

 between the two groups comes out in a very striking way if we 

 place the figures in columns by classes : 



miles less range tlian the first column. 



There are also 98 species that are endemic to New Zealand 

 and one or more of these island groups, but not found elsewhere 

 in the world. These have an a^-erage rarit}' in New Zealand 

 represented by 2-9, or in other words, they are a good deal more 

 widely ranging in New Zealand than tliose species which reach 

 Australia, etc. (enumerated above in the second column), but 

 do not reach these little islands. The difference of 1-4 in average 

 range represents 168 miles. Now here, still more than in the 

 previous case (p. 68), there is no conceivable reason why ranging 

 to these little groups of islands (and to auy one of them, though 

 they differ completely in climate and geology) should make these 

 endemic species more widespread in New Zealand than many 

 others whose distribution touches Aiistralia, etc., unless it be 

 simply that being older, they have had time to reach the islands, 

 and to range more widely in New Zealand itself. 



Another interesting point shows in the table given above, 

 which also indicates the greater age of these species, whether 

 wide or endemic, that reach the outlying islands. The wides that 

 reach them show 45 in class 1, whose range covers Stewart 

 Island, a separate island near to the south coast of New Zealand, 



