October 2, 1903.] 



SCIENCE. 



437 



the many plants for the study of which the 

 usual dried and shriveled herbarium material 

 is nearly worthless, but in deference to cur- 

 rent botanical opinion the public has con- 

 tinued to infer that rubber planting is equally 

 practicable in all localities in which wild trees 

 are found, and that all trees are equally suit- 

 able for cultivation, except the elusive ' tunu ' 

 or ' ule macho' (male rubber), which Hems- 

 ley has recently attempted to separate as a 

 distinct species. 



Until the intervening territory has been 

 thoroughly explored it can not be known with 

 certainty whether two supposed species from 

 distinct localities intergrade or not, but for 

 agricultural purposes this is a matter of little 

 importance. There are at least two different 

 kinds of Castilla in cultivation in Mexico and 

 two in Costa Rica, and the indications are 

 that these four are distinct species. The 

 Castilla introduced from Panama to Ceylon 

 and other British colonies represents a fifth 

 type, while three others are of merely botanical 

 interest, as yet, since they are not known to 

 produce commercial rubber. 



Before touching upon the characters which 

 distinguish the species, it may be well to ex- 

 plain that Castilla is partially dioecious, some- 

 what after the manner of the edible fig. The 

 inflorescence consists, as in the fig, of a 

 fleshy receptacle which bears either stamens 

 or pistils; the pistillate inflorescence is tur- 

 binate, the staminate more or less flabellate 

 or funnel-shaped. There are trees which bear 

 only male inflorescences, or at least there are 

 some which bear crops of males without fe- 

 males, but along with the females there are 

 also male inflorescences, smaller and gener- 

 ally of a different shape from those of the 

 male trees. The primary male inflorescences 

 arise normally in groups of four, but of the 

 secondary or complemental male inflorescences, 

 those which subtend the females, there are 

 never more than two, as though the female 

 inflorescence were the equivalent of two male 

 clusters. 



The original description of Castilla. elastica 

 (the name Castilloa being an unwarranted 

 emendation) was not accompanied by any in- 

 dication of a definite locality, but there is 



every probability that Cervantes had refer- 

 ence to the Castilla of eastern Mexico, which 

 seems to differ from all of its relatives in its 

 robust habit and in the large size of its ripe 

 fruits, which also have numerous and very 

 distinct vertical grooves. Cervantes' plate 

 shows, in addition, long, slender complemental 

 male inflorescences. 



The CasliUa of the Soconuseo district of the 

 state of Chiapas (C. lactiflua) is peculiar in 

 having the complemental inflorescence flat- 

 tened and with a broad mouth; it is very sim- 

 ilar to the primary except in the smaller size. 

 The specific name alludes to the fact that the 

 milk of the tree flows freel.v when the bark is 

 cut, so that it can be collected in quantity and 

 coagulated by improved ' creaming ' methods 

 instead of the rubber being harvested wholly 

 or partly by pulling the ' scrap ' (burueha) 

 from the gashes in which it has dried. Large 

 yields of scrap rubber are sometimes reported 

 from wild trees, but the tapping to which they 

 are subjected is very severe, and the removal 

 of the rubber from the wounds delays healing 

 and exposes the tree to the attacks of insects, 

 so that the cultural production of scrap rub- 

 ber is not likely to be profitable. 



On the peninsula of Nicoya, which extends 

 into the Pacific Ocean from the northern part 

 of Costa Rica, is a rubber tree easily recog- 

 nized by the dark olive color of its inflores- 

 cences of both sexes, and by the deeply bi- 

 lobed, long-stalked primary male inflorescence; 

 the complemental inflorescences are also long 

 and slender and are usually grown together at 

 the base. The branches, leaves and floral or- 

 gans are also much more densely hairy than 

 those of the Castilla or of the more humid 

 eastern slope of Costa Rica (C. costaricana 

 Liebmann). In C. nicoyensis the individual 

 fruits are very prominent, as in the two 

 Mexican species, but in C. costaricana they 

 are separate only at the broadly rounded or 

 flattened apex, leaving no characters by which 

 Hemsley's Castilla tunu, from British Hon- 

 duras, can be distinguished, except the re- 

 puted deficiency of rubber, which is by no 

 means lacking in C. costaricana. It seems cer- 

 tain, however, the ' tunu ' tree which Hems- 

 ley has recently figured and described from 



