30 SCIENCE PROGRESS. 



Islands, as well as of the Galapagos, are confined to 

 single islands, and often, we are assured, by competent 

 explorers, to very restricted areas and easily counted 

 individuals. This affords a reason why the botanists of 

 Cook's voyages failed to obtain nearly complete collec- 

 tions of the plants of Tahiti, though much time was spent 

 there. Subsequent English collectors have made fresh dis- 

 coveries, yet they have by no means exhausted the flora. 

 French collections, as might be expected, are much richer, 

 especially those made by persons actually resident in the 

 islands ; but they want a number of species existing in the 

 English collections. This excessive rarity of individuals 

 of the endemic element is not satisfactorily accounted for 

 by the destructive effects of cultivation and the struggle 

 with introduced plants of excessive reproductive powers 

 combined, because many of them would appear to inhabit 

 localities where they are neither exposed to direct de- 

 struction nor to competition. It is one of the problems 

 to be studied on the spot. Reduced reproductive power 

 may explain the phenomenon ; but local investigation 

 should reveal the real cause. What is certain, some of 

 the most highly differentiated forms of vegetable develop- 

 ment still linger on in remotely distant localities. Lepinia, 

 a genus of Apocynaceae, 1 having one of the most remark- 

 ably constructed fruits in the whole range of the vegetable 

 kingdom, was discovered in Tahiti by the French col- 

 lectors Lepine and Vesco in 1847, an d by Nadeaud, a 

 writer on the botany of the island, in 1856; but, as far 

 as can be ascertained, it exists in no English collection 

 from the island; yet, in 1890, the Rev. R. B. Comins 

 collected specimens of the very same species in the 

 Solomon Islands, some 3000 miles to the westward. No 

 intermediate locality is known ; and, to all appearance, 

 it is one of the least likely of plants to be conveyed a 

 distance by any of the ordinary means of dispersal. In 

 the flowering stage the carpels of the ovary of this plant 

 are small and apparently consolidated, but having the four 



1 A?inales des Sciences Naturelles, 3 me serie. xii. p. 194, t. 9. 



