484 ISLAND LIFE 



to a very distinct family (Discoglossida^), confined to 

 the False arctic region. 



Of the fresli-water fishes we need only say here, that 

 none belong to peculiar Australian types, but are related 

 to those of temperate South America or of Asia. 



The Invertebrate classes are comparatively little known, 

 and their modes of dispersal are so varied and exceptional 

 that the facts presented by their distribution can add little 

 weight to those already adduced. We will, therefore, now 

 proceed to the conclusions which can fairly be drawn from 

 the general facts of New Zealand natural history already 

 known to us. 



Deductions from the Peeulicirities of the JVew Zealand 

 Fauna. — The total absence (or extreme scarcity) of 

 mammals in New Zealand obliges us to place its union 

 with North Australia and New Guinea at a very remote 

 epoch. We must either go back to a time when Australia 

 itself had not yet received the ancestral forms of its 

 present marsupials and monotremes, or we must suppose 

 that the portion of Australia with which New Zealand 

 was connected was then itself isolated from the mainland, 

 and was thus without a mammalian jiopulation. We shall 

 see in our next chapter that there are certain facts in the 

 distribution of plants, no less than in the geological struc- 

 ture of the country, which favour the latter view. But 

 we must on any supposition place the union very far back, 

 to account for the total want of identity between the 

 winged birds of New Zealand and those peculiar to 

 Australia, and a similar want of accordance in the 

 lizards, the fresh-water fishes, and the more important 

 insect-groui3s of the two countries. From what we know 

 of the long geological duration of the generic types of 

 these groups we must certainly go back to the earlier 

 portion of the Tertiary period at least, in order that there 

 should be such a complete disseverance as exists between 

 the characteristic animals of the two countries; and we 

 must further suppose that, since their separation, there 

 has been no subsequent union or sufficiently near approach 

 to allow of any important intermigration, even of winged 

 birds, between them. It seems probable, therefore, that 



