HEYERDAHL'S KON-TIEl THEORY — ^JONKER 539 



nesia from Melanesia or from Micronesia — to say nothing of the pos- 

 sibility that the Polynesians reached their present living area by one 

 of these latter routes. The oldest known finds in southeastern Asia 

 indicate the occurrence of the bottle gourd in China sometime before 

 the beginning of our era. For the sake of completeness, I must also 

 add that Heyerdahl further adduces evidence that both in Peru and 

 in some Polynesian islands gourds were used to make flutes. 



Merrill (1954) believed it most probable that bottle gourds reached 

 the South American coast from Africa in a floating state. Pie men- 

 tioned floating experiments showing that these gourds could stand 

 floating in salt water for nearly 2 years and still contain viable seeds. 

 He also was unable to provide any other explanation. 



A third very important crop occurring already in pre-Columbian 

 times both on the American mainland and on the Pacific islands, and 

 in fact in nearly all tropical regions, is the coconut, Cocos nucifera L. 

 Formerly this distribution was ascribed to the ability of the coconut 

 to float for a long period. However, floating experiments have shown 

 that coconuts drifting in sea water rather quickly lose their powers 

 of germination, and that fruits floating more than 110 days are no 

 longer viable. Heyerdahl's voyage on the Kon-Tiki raft supplied an 

 important contribution, as the crossing lasted 101 days ; but of course 

 smaller objects drift somewhat more slowly. It seems out of the ques- 

 tion, consequently, that drifting coconuts can cross this part of the 

 Pacific within the critical period of 110 days. Moreover, the coconuts 

 that floated on the raft in sea water decayed, and then the pelagic 

 fauna rapidly completed their deterioration. However, the other coco- 

 nuts on the raft that were not subjected to the sea water maintained 

 their viability. I agree with Heyerdahl in concluding that the dis- 

 tribution of Cocos in pre-Columbian times was possible only with 

 the help of man. 



But there is no unanimity among botanists concerning the region of 

 the origin of the species. A number of botanists regard South 

 America as such, and especially either Colombia or Panama, since 

 related genera and species occur in those countries. Also, old travel 

 stories reported rich coconut vegetation in regions not previously 

 visited by European travelers. Other competent botanists, e.g., Merrill 

 (1937, 1954), are convinced, following Alphonse de Candolle (1883), 

 that the species originated in the Old World Tropics. It is not neces- 

 sary in this paper to review Merrill's arguments, but I shall only 

 remark that his mention of fossil species, probably belonging to the 

 genus Gocos^ in both New Zealand and in India, in my opinion is 

 irrelevant. Distribution by man must be connected with the recent 

 area of distribution of the species. A large number of genera and 

 species in Tertiary times occupied an area covering two or three con- 



