INTRODUCTION TO THE REPORTS ON INSULAR FLORAS. 41 



As already stated, the size of the flowers or flower-heads of the St Helena indigenous 

 plants is on the whole rather above than below the average of those of their allies in other 

 parts of the world. When we come to colour, however, the equality fails altogether, red 

 being almost entirely wanting in the insular plants and blue unknown ; for the slight tinge 

 of blue in Melliss's figure of WcMenbergia angustifolla seems to have been added to show 

 the flowers on the white ground. Burchell describes (MSS. in Bibl. Kew) the flowers of 

 all the St Helena species of Wahlenbergia as white. The species of Melhania, like those 

 of many allied plants, change in colour after expansion. At first they are pure white, 

 becoming pink or brown red according to Melliss, who, however, represents them with a 

 clear, rosy red. Of no other oceanic island are we able to give such full particulars on 

 this point ; but it may be asserted that brilliantly coloured flowers are very rare in such 

 situations. 1 Brigham (Proceedings of the Boston Natural History Society, xii. p. 6) says : 

 •' There are but few show-flowers, and still fewer fragrant ones, in the Hawaiian Flora. In 

 the colouring white or greenish white is predominant, and yellow and pink follow at a 

 respectful distance. There are very few blue flowers. Strongylodon lucldum is a rich 

 crimson, and some other leguminous plants are violet, but the various and brilliant 

 colouring of the Californian plants is wholly absent." 



The flora of Juan Fernandez partakes, however, in some measure of the generally 

 more brilliant colouring of the flowers of the south temperate zone. Indeed, Philippi 

 states (Botanische Zeitung, 1856, p. 635) that the Juan Fernandez species of Chilian 

 genera are handsomer, and especially larger-flowered, than their continental congeners. 



Returning to the question of the existence of structural peculiarities in insular plants, 

 we may safely say that no general ones exist, and the very distinct genera that occur are 

 not disproportionate. Not a single order, or suborder, or tribe, is endemic in the smaller 

 oceanic islands ; and, as more fully set forth elsewhere, the greatest isolation of an order 

 is that of the Balanopseae in New Caledonia. The endemic genera of St Helena present 

 no anomalies, nor indeed any specially striking differentiations in structure ; for it is 

 habit rather than structure that characterises the Compositse. In the Tristan da Cunha 

 gr'oup, and in St Paul and Amsterdam, the endemic element includes no higher rank than 

 species. The other islands in the Southern Indian Ocean which have essentially the same 

 flora (see Part II., p. 251) possess two very distinct, though by no means anomalous genera, 

 namely, Pringlea and Lyallia. In the Juan Fernandez Lactoris we have an example of 

 a truly anomalous genus, the position of which in the natural system is not evident. 



The most highly differentiated endemic genera of the Sandwich Islands flora are : 

 Schiedea and Alsinidendron, shrubby Caryophyllese exhibiting important structural 



1 Livrelation to the fertilisation of plants by insects, the absence of native butterflies in St Helena is note- 

 worthy. On this point Melliss (St Helena, p. 180) says: "The varied hues of the St Helena landscape need 

 little to add to their brilliancy [brilliant with exotic plants], or the almost entire absence of butterflies would 

 be more striking. There are but four species in the island, and they have all been imported. With moths it 

 is different : they abound." 



