270 



Leguminosae are represented by only one species (Sophora 

 tetraptera). Quensel, who was with Skottsberg, found the 

 islands were composed, as far as he could observe, solely of 

 volcanic material, of uncertain age, but probably belonging- to 

 early Tertiary times. Skottsberg considers that colonization from 

 existing* neighbouring lands, Chile, Polynesia, New r Zealand, 

 etc., will not alone account for the existing flora, since the 

 relatively short time which has passed since the present islands 

 appeared would not be sufficient for the development of such 

 isolated types as Lactoris and Robinsonia. He believes that the 

 old element did not originate in the Juan Fernandez group, but 

 is older than these islands and once spread over a large area in 

 Polynesia, Chile, etc., subsequent changes in the Quaternary 

 Period, however, altered the floras of these countries and isolated 

 that of Juan Fernandez. The existence of more land in the 

 Pacific in Tertiary times may have given developmental centres for 

 a fauna and flora, whose scattered remains now especially inhabit 

 the oceanic islands. Many things suggest that in spite of their 

 recent volcanic nature the islands are the remains of a greater 

 island or group of islands, but nothing definitely suggests that 

 they formed in earlier times part of a transoceanic land-bridge or 

 of a great Pacific continent. 



The forest consists of evergreens, except Berber is corymbosa 9 

 which was leafless in July and August. It is a typical rain- 

 forest, but is not sub-tropical, since it lacks obligatory epiphytes 

 and lianes. The presence of two climbing ferns (Arthropteris 

 altescandens and Blechnum Schotii) is interesting, and in general 

 appearance the Yaldivian forest is strongly recalled by the 

 presence of such types as Myrtaceae, Drirays, Sophora, EscaUonia, 

 Pernettya, Rhaphithamnus, and Chusquea. On the other hand 

 the lack of species of Nothofagus, so common in Chile, is striking, 

 as is also the presence of snch types as Fagara, Psychotria 9 

 Boehmeria and Santahnn, which are absent from the' Chilean 

 rain-forest. These types are tropical or subtropical, but they do 

 not change the mainly temperate character of the rain-forest. 

 The absence of Coniferae is common to Juan Fernandez and many 

 other oceanic islands. The Composite trees (Rhetinodendron, 

 Robinsonia, Dendroseris, etc.) and other "feather bushes" give 

 a characteristic facies to the vegetation. 



A graphic account is given of a visit to the last remaining tree 



of the native sandle-wood (Santalum). The tree had only one 



living blanch, and the reader is made to share with the author 



the intense emotions of a botanist witnessing the death of a 

 >pecies. 



America south of 41° S. (4, 6, 11, 18 ? 22. 23.) 



The results of the Swedish expeditions in the southern part of 

 South America are extensive and important from the several 

 aspects of systematic, geographical and ecological botany. In 

 the lo-t published work here dealt with, Skottsberg has brought 

 together much of the information obtained during the four or 

 five years he spent in the Southern Hemisphere, and has further 

 modified some of the views he published in previous papers. 



