DERIVATION OF THE FLORA AND FAUNA 277 



Rhipsalis is Gondwanic it is likely, to say the least, that the archetypes of the family 

 itself stem from the Afroantarctic triangle (pp. 364, 365). 



Much could be said to this, but I shall confine myself to remarking that the 

 claims of Rhipsalis outside America to true citizenship are still under suspicion and 

 that the existence of this specialized type of angiosperms in the Gondwana flora 

 needs proofs of some sort. But Croi/.at docs not stop here. On p. 523 we read: 



Considering that the New World is uniformly at the receiving end of angiospermous 

 tracks throughout the first epoch of migration we are drawn to conclude that none of 

 the primary angiosperms ever originated in the Americas. 



But who will tell us where the first angiosperms were evolved? We have to 

 go back to times when "the Americas", as we know them, did not exist and where, 

 just as everywhere else, all centres of origin are secondary. I cannot see that his 

 opinion that the primary centre must be looked for in an "Afroantarctic triangle" 

 rests on a stable foundation, but everybody will agree when he formulates the 

 following recommendation to the "phytogeographers of the Academic school": 

 "Look, and keep silent awhile ere you speak." 



Phrygilantlius and Saiitalum; compare what will be said below about Coprosnia. 



Dysopsis. Antarctic according to Croizat p. 51 ; I regarded it as neotropical. 



"A large part of the Ericaceae originated with the Empetraceae in the Afro- 

 antarctic Triangle", Croizat writes p. 381. That the Antarctic has been used as 

 a migration route of the Gaultherioideae seems certain, let alone where the family 

 or complex of families had its beginning. Croizat finds that the occurrence of a 

 species of Pernettya in Juan Fernandez and of another in Galapagos is 



one of the most interesting aspects of the Gaultherioideae. ... a track of this nature is 

 nothing unusual, of course, because dispersals of this extent and nature are commonplace 

 between the vicinity of Juan Fernandez and Hawaii, not to speak of the Galapagos and 

 the Revilla Gigedos (pp. 167, 168). 



We have examples of a floristic "contact" between Juan Fernandez and Hawaii, 

 but they are very few and to call this type of dispersal commonplace is an inadmis- 

 sible exaggeration of facts. The American area of Pernetlya extends from Fuegia- 

 Falkland to Mexico, and its presence on Juan Fernandez and Galapagos, outlying 

 stations west of the continent, is not surprising; we need not construct a special east 

 Pacific track to explain them, nor assume that "the South Pacific is a fundamental 

 center of the Ericaceae". 



I have mentioned Evipetriim and the boreal centre of Empetraceae, also that 

 Croizat wants them to have followed the same stream as Ericaceae: "they begin 

 their visible dispersal only westward from the Cape region" (p. 349), "with the 

 present Atlantic as the axis of their distribution" (p. 353). Empetrmn used to pass 

 as a classical example of bipolarism, but to those who still believe that bipolarity 

 exists Croizot says: "bipolarism, as we know, is a bugaboo of academic phyto- 

 geography" — I am .afraid that some of us didn't know that. 



The case of Cumijiia (for which CROIZAT uses the fancy name Johoivia), whether 

 or not to be linked with the Old World and Hawaiian Prasioideae, is embarrassing: 



