164 OEIGINAL ARTICLES. 



Tropical American forms is, according to De Candolle, but a section 

 of Tournefortia, including in one subsection the Canarian plant, while 

 other subsections include two African and an Indian species. 



Of the genus Persea, of which one species (P. indicd) is a domi- 

 nant tree in the Canaries, Madeira, and Azores, two species grow in 

 the Southern States, while others are found in New Grenada, Peru, 

 and as far South as Chili. Commelyna agraria, Euphorhia tenella, and 

 Bidens pilosa cannot be accepted as indigenous to the Atlantic 

 Islands. 



We may"gather from the above paragi'aphs that a close and very 

 peculiar analogy subsists between the Flora of Tertiary Central Eu- 

 rope and the recent Eloras of the American States and of the 

 Japanese region ; an analogy much closer and more intimate than is 

 to be traced between the Tertiary and recent Floras of Europe. We 

 find the Tertiary element of the Old World to be intensified towards 

 its extreme eastern margin, if not in numerical preponderance of 

 genera, yet in features which especially gave a character to the Fossil 

 Flora. I have taken occasion to show, in the above notices, that 

 this accession of the Tertiary element is rather gradual and not 

 abruptly assumed in the Japan islands only. Although it there 

 attains a maximum, we may trace it from the Mediterranean, Levant, 

 Caucasus, and Persia, in Chamcerops, Platanus, Liquidamhar, Ptero- 

 carya, Juglans, ^c, then along the Himalaya and through China ; 

 the Eastern Himalaya and China, indeed, forming with Japan one 

 great botanical region. The table given at p. 175 shows that about 120 

 Tertiary genera are represented in Europe and Asia, including Japan, 

 taken together, while, as stated already, but 88 are represented in the 

 Southern American States. We learn also that during the Tertiary 

 epoch, counterparts of Central European Miocene genera certainly 

 grew in North-west America, amongst them, one marked genus now 

 limited to the Japanese region (Salisburia). We note, further, that 

 the present Atlantic Islands Flora affords no substantial evidence of 

 a former direct communication with the main land of the New World, 

 though the cu'cumstance of an extraordinary predominance in it of 

 the Mediterranean element tends to countenance the probability of 

 the hy[)othesis of E. Forbes and others that a coimection formerly 

 existed between these Islands and some part of Western Europe. 

 The consideration of these facts leads me to the opinion that 

 botanical evidence does not fiivour the hypothesis of an Atlantis. On 

 the other hand, it strongly favom's the view that at some period of 

 the Tertiary epoch. North-eastern Asia was united to North-western 

 America, perhaps by the line where the Aleutian chain of islands 

 now extends, since there is sufficient ground to belie\'e that the 

 temperature in that latitude was liigh enough to allow the migra- 

 tion of types, which at the present period, are characteristic of lower 



