614 



SCIENCE 



[N. S. Vol. 5LVII. No. 1225 



West Africa and tropical America. This has 

 long been known to botanists and often dis- 

 cussed, as, for example, by Schimper and 

 Engler. The latter is inclined to insist that 

 the present distribution demands a land con- 

 nection, and this hypothesis is heightened by 

 the usual assumption of such a land bridge 

 by paleogeographers based on the distribution 

 of littoral faunas, the absence of marine Ter- 

 tiary deposits on the facing coasts, the evidence 

 of the foundering of earth blocks in certain 

 areas, and the absence of Mediterranean ele- 

 ments in the early Tertiary marine fauna of 

 Patagonia. 



If the community of floras on the two sides 

 of the tropical Atlantic was entirely confined 

 to strand types the conclusion would be in- 

 evitable that this was due entirely to the action 

 of ocean currents, but such is not the case, 

 nor are these features confined to the Eecent 

 floras, for in the Tertiary floras of oiir Gulf 

 states a number of genera common to "West 

 Africa and tropical America had already made 

 their appearance. 



An important conclusion of Guppy's is that 

 the sphere of influence of the ocean current in 

 determining plant distribution between the 

 Oriental and Occidental tropics is limited, and 

 that their action leaves the main facts of gen- 

 eral distribution imconfused. In the com- 

 parison of the floras of the West African and 

 American tropical littoral the most direct 

 journey, namely from the former to the latter 

 via the Main Equatorial current, requires that 

 a fruit or seed will float unharmed for from 

 two to three months. The restdts for 53 spe- 

 cies are embodied in a table showing that 6 

 mangrove, 7 estuary and riverside, and 19 

 beach plants, or 60 per cent., are found in both 

 regions, and of these 88 per cent, show experi- 

 mentally that they possess the required buoy- 

 ancy and resistance to enable them to make 

 such a journey. 



While the evidence is only negative it may 

 be noted that the oldest known fossil occur- 

 rences of the genera Rhizophora, Avicennia, 

 Laguncularia, Oanavalia, Bodoncea, Chrysoba- 

 lanus, Conocarpus and Carapa are American. 

 Guppy's discussion is eminently open-minded 



and shows conclusively that most of the com- 

 mon species have the necessary powers of re- 

 sistance, while only 24 per cent, of the strand 

 plants confined to the American tropics possess 

 fruits or seeds capable of making the trans- 

 Atlantic trip. I am, however, inclined to dis- 

 sent from his conclusion, at least as regards 

 its application during the Tertiary, namely, 

 that since the Main Equatorial current offers 

 the easiest avenue of distribution, this dis- 

 tribution has been from east to west. The 

 more general conclusions derived from a study 

 of geological distribution point in many cases 

 to an original distribution from west to east. 

 The fact that genera like Moschoxylon (Meli- 

 aceee) with about three score existing species 

 in tropical America and West Africa has fossil 

 species in the lower Miocene of Chile and 

 Colombia indicates that other explanations 

 than dispersal by ocean currents are involved, 

 and the distribution of the family Humiriacese 

 is another illustration of the same kind. Many 

 such instances are given in a recent discussion 

 of fossil floras.'^ 



This is not the place, however, to elaborate 

 this thesis since the distribution of the recent 

 floras of the two regions is fairly well known 

 and Guppy's work concerns only the littoral 

 species and in discussing these he is never 

 dogmatic. 



The chapter devoted to the current con- 

 nections in the Southern Hemisphere, based 

 largely on records of bottles and wreckage, is 

 also exceedingly interesting, but the time in- 

 volved in such journeys between these out- 

 lying land masses is too great to be a large 

 factor in the curious similarities of elements 

 in antipodean floras, which are really results 

 of geological history. Moreover there is rather 

 definite evidence of land connection, especially 

 the early Tertiary connection of Graham Land 

 and Patagonia, at which time Fagus and Notho- 

 fagus, as well as other northern derived types 

 appear to have invaded Patagonia and Chile 

 from the Antarctic. 



For shorter distances currents may have 

 been important factors, e. g., tropical Aiistralia 



1 U. S. Geol. Survey Professional Paper 91, pp. 

 72-140, 1916. 



