DKRIVATION OF THE FLORA AND FAUNA 373 



In this assumption he may be quite right, but it finds, as already pointed 

 out by Rkiciie {20J. 269), httle support in the flora, for the two groups have 

 only a single species in common {2^1), and the endemic genera and species are 

 very different. Joiiow's belief in their {permanent isolation from the continent 

 remained unchanged. 



In his tables, Continjente A and H, pp. 218-220, is a column indicating 

 (not always with sufficient accuracy) tlic nature of the fruits or seeds, and another 

 stating the probable dispersal agent. Of the 143 vascular plants known at that 

 time, all the ferns and 34 phanerogams were supposed to have been wind-borne, 

 61 had been transported by birds, either inside or adhering to their feet or 

 plumage; only one, Sophora ''tctrapterd\ with winged pods, had drifted with 

 the current. Five {Eryjigiioti hupleuroidcs and sarcophyllmn, Apiuui fernajidczianuni, 

 Collctia spartoidcs and Fagara (Zanthoxylum) niayu] offered too serious difficul- 

 ties to make their presence in the islands ex|3licable, but I cannot find that they 

 are more "impossible" than many of the others. In an earlier paper {2j_^) I sur- 

 veyed the nature of the diaspores; I shall not return to this question here but 

 only repeat that almost one third of the flowering plants show no special adap- 

 tation to any particular mode of dispersal across the water, and I shall add a 

 few remarks on some knotty cases. Among the island Compositae, a family known 

 to be well adapted to wind dispersal, are several species which lose their pappus 

 when the achenes are still enclosed in the involucre, or where it is reduced to 

 uselessness. Only one species of SopJiora, called tetraptera, was recognized by 

 JOH(3W and recorded from Chile, Juan Fernandez, Easter Island and New Zealand. 

 Genuine tetraptera is restricted to New Zealand, but this is of less importance, 

 the pods have narrow wings (poorly developed in some forms) which v^ere sup- 

 posed to help the pods to keep afloat, but as they open on the tree and drop 

 their rather heavy seeds to the ground, the wings serve no purpose. Halorrhagis 

 is a somewhat similar case. Joiiow, and others before and after him, identified 

 the plant found on Masatierra with H. erecta (alata), a New Zealand species with 

 four narrow wings on the fruit, and he did not hesitate to regard it as wind- 

 borne. His erecta is, however, an endemic species; there are two more on Masa- 

 fuera (see 22g, with illustrations), and their fruits are quite or almost unwinged. 

 There can be no question of an adaptation to wind carriage. The same applies 

 to Selkirkia, another of Joiiow's anemochorous plants. Edible fruits, if not too 

 big, will be swallowed by birds, but the indigestable stones and hard-coated 

 seeds will be dropped before the islands are reached, and the diet of the wide- 

 ranging and fast-flying sea birds is another. Epizoic transport is possible in a 

 small number of cases, but otherwise the dispersal mechanisms, many of them 

 quite wonderful, serve to maintain the population within its range, and not to 

 stock distant islands. 



Other advocates of the permanent isolation of Juan Fernandez are e.g. PLATE 

 (/pp), GouRLAY [301), Burger (^/), and Goetsch [124). Plate drew his con- 

 clusions from a comparison between the littoral faunas of Juan Fernandez and 

 the opposite coast: 



