CH. XIV] PALAEOBOTANICAL STANDPOINT 139 



In the paper on the "Sources and Distribution of the New 

 Zealand Flora, pp„ 354-362, the conclusion is drawn that there 

 were two main plant-invasions by which New Zealand was popu- 

 lated, a northern and a southern. In a subsequent paper (134) 

 this conclusion is amplified. The northern invasion is split into 

 three a principal one from the north, and two subsidiary ones, 

 called the Ivermadec and western invasions respectively ' 



It will at once be seen that we have here postulated three 

 invasions (northern, western, and Kermadec), which in their 

 general direction are poleward, and one invasion (southern) 

 winch IS equatorward. Bearing in mind the conclusions we have 

 reached as to the relationship between direction of migration 

 and change of climate, it would appear that the three poleward 

 mvasions must have occurred whilst the climate of all the regions 

 involved, or possibly only that of New Zealand, was becomina 

 warmer: the southern in^'asion, equatorward, must have oc" 

 curred whilst the climate of the regions involved was becoming 

 colder. '^ 



It will be sufficient for our argument if we consider onlv the 

 two main divisions, the northern and southern. 



Dr Willis brings forward strong evidence (132, and cf p 81) 

 to show that of the two, the northern was much the older We 

 have,- therefore, to consider one very old migration polewards 

 whilst the climate was warming, and one newer, equatorwards, 

 whilst the climate was cooling. 



For the migration oi floras (plants in ecolooical association) 

 as opposed to the casual transport of individuals, Dr AVillis 

 rightly insists that land connection, complete or all but com- 

 plete Avith the source of dispersal, at the time of dispersal is 

 necessary. It is inconceivable that associated assemblages could 

 travel 111 one definite stream except bv land. A sea-passage 

 must have sifted out species with inferior powers of dispersal 

 across Avater in a Avay that is not found to have occurred. 



We have now traced the conditions necessary for these two 

 main invasions as postulated. For the northern", a very ancient 

 land connection between New Zealand and Indo-iAIala\-a, Avith 

 a climate increasing in temperature, certainly in New Zealand, 

 and probably over the whole of these regions." For the southern,' 

 a very much later land connection southwards, at least as far 

 as the Campbells and Aucklands, Avith the climate of these 

 regions becoming colder. It will readily be seen that for the 

 western and Kermadec iuA-asions, conditions A'cry similar to 



