TROPICAL NATURE 



purple or blue, three -lilac, and two red or pink, showing a 

 very similar proportion of white and yellow flowers to what 

 obtains farther south. 



We have, however, a remarkable flora in the southern 

 hemisphere, which affords a crucial test of the theory of greater 

 intensity of light being the direct cause of brilliantly-coloured 

 flowers. The Auckland and Campbell's islands, south of New 

 Zealand, are in the same latitude as the middle and the south 

 of England, and the summer days are therefore no longer 

 than with us. The climate, though cold, is very uniform, 

 and the weather "very rainy and stormy." It is evident, 

 then, that there can be no excess of sunshine above what we 

 possess, yet in a very limited flora there are a number of 

 flowers which Sir Joseph Hooker states are equal in 

 brilliancy to those of the arctic flora. These consist of 

 brilliant gentians, handsome veronicas, large and magnificent 

 Composite with purple flowers, bright ranunculi, showy 

 Umbelliferse, and the golden -flowered Chrysobactron Eossii, 

 one of the finest of the Asphodeleee. l All these fine plants, 

 it must be remembered, are peculiar to these islands, and 

 have therefore been developed under the climatal conditions 

 that prevail there ; and as we have no reason to suppose that 

 these conditions have undergone any recent change, we may 

 be quite sure that an excess of light has had nothing to do 

 with the development of these exceptionally bright and hand- 

 some flowers. Unfortunately we have no information as to 

 the insects of these islands, but from their scarcity in New 

 Zealand we can hardly expect them to be otherwise than very 

 scarce. There are, however, two species of honey -sucking 

 birds (Prosthemadera and Anthornis), as well as a small 

 warbler (Myiomoira); and we may be pretty sure that the 

 former at least visit these large and handsome flowers, and 

 so effect their fertilisation. The most abundant tree on the 

 islands is a species of Metrosideros, and we know that trees 

 of this genus are common in the Pacific islands, where they 

 are almost certainly fertilised by the same family of Meli- 

 phagidse or honey-sucking birds. 



I have now concluded this sketch of the general pheno- 

 1 Coloured figures of all these plants are given in the Flora Antarctica, vol. L 



