PITTIER TREATMENT OF THE GENUS CASTILLA. 265 



Golfo de Osa district; it was found only once, however, in condition to be col- 

 lected. According to Terraba Indians, used to traveling between their village 

 and David in the Republic of Panama, there are several individuals or groves 

 of the same species along the mule-path between Rio Chiriqm Viejo and the 

 mentioned town of David. These localities, as well as those where specimens 

 of the tree have been collected, are on the Pacific side of Central America. A 

 Frenchman who had spent several years in and about Bluefields, in Nicaragua, 

 and whom the writer met in Fort Union in 1002, informed him that the "hole 

 macho,"' called also "hide Colorado " or " gutta-percha " by the native Spaniards 

 and " tunu " by the MIsquito Indians, is rather common on the Atlantic coast 

 of Nicaragua. This seems to be confirmed by Koschny. There are also indi- 

 cations, collected in Guatemala, of the presence of such a tree in eastern 

 Spanish Honduras; but, in the absence of herbarium specimens, it is not pos- 

 sible to decide whether the last-mentioned data refer to C. fallax, or to the 

 enigmatic C. tunu of British Honduras. The only definite character heard of 

 is the absence from the latex of merchantable rubber. 



This species and the tunu of British Honduras are distinguished from the 

 other Central American species of the genus by the very important fact that 

 they do not produce commercial rubber. The milk coagulates in a hard, sticky, 

 unelastic mass, erroneously called "gutta-percha" by the natives, the price 

 of which is so low as to render its collection unprofitable. It is therefore of 

 urgent necessity to publish accurate and detailed descriptions of these useless 

 species, so as to avoid the mistake of planting or distributing their seeds for 

 agricultural purposes. Unfortunately, however, the attempts hitherto made at 

 formulating their distinctive characters have only resulted in confusion. 



As far back as 1883, Sir Darnel Morris mentioned "tunu" in his book on 

 British Honduras, but looked upon it as the local name of the true rubber tree. 

 Nevertheless, his description of the fruit is more suggestive of C. falla.r than of 

 the Mexican C. elastica: "The fruit, of a brownish green color when ripe, has 

 very much the appearance of a raspberry flattened or depressed, about an inch 

 in diameter, the numerous seeds being massed together and enclosed In papery 

 capsules, covered with a brown tomentum." « This account does not contain 

 any allusion to the red pulp that characterizes the fruit of the true C. elastica 

 and seems to point to the presence in British Honduras of a species the nutlets 

 of which are free and dry. At the same time, it is known from other sources 

 that the name "tunu " applies to a species that does not produce rubber, and is 

 therefore distinct from C. elastica. 



Sir Joseph D. Hooker does not mention Morris's work in the memoir he 

 presented in 18Sr> to the Linnean Society on "Castilloa elastica and some allied 

 plants." In Hooker's paper is found a short description of fruits assumed to 

 belong to the tunu of British Honduras, illustrated by 3 drawings; & but neither 

 the description nor the figures of these fruits indicate any relation to 0. fallax. 

 Some of the fruits ascribed on the same plate to €. elastica (tigs. 1-0) appear 

 even more similar to those of the former species than those that represent the 

 tunu, and these bear a striking resemblance to V. costaricana. 



Writing in 1898,° Mr. Hemsley rectifies the error of Morris in applying the 

 name "tunu" to C. elastica, accepting Hooker's statement of the existence in 

 British Honduras of a second species of Castilla. As a further proof, he men- 

 tions that Mr. Rowland W. Cater has sent leaves of what is known in that coun- 



a Morris, D. The Colony of British Honduras, p. 74. (1883.) 



5 Trans. Finn. Soc. Bot. ser. 2, vol. 2, p. 212, pi. 28, figs. 7-9. (1885.) 



c Kew Bull. Misc. Inform., FS9S, p. 141. 



