PITTIER — TREATMENT OF THE GENUS CASTILLA. 259 



achenes. In ('. daguensis and 0. panamensis the cup-like receptacle 

 remains rather deep and closed, while it is quite shallow and open in 

 0. fallax, 0. lactiflua, G. costaricana, C. guatemalensis, and C. nico- 

 yensis. The counterpart of the closed receptacles of 0. ulei is found 

 in C. elastica, where these organs are more or less distorted and re- 

 flexed so as to embrace the twig on which they grow. As to size, they 

 are exceptionally large in this last species, as also in ('. costaricana. 



No data are available as to the seeds of C. australis, C. daguensis, 

 and C. lactiflua. Castillo, fallax has the smallest seeds, C. pana- 

 mensis and C. ulei the largest ones; they are roundish and also small in 

 O. nicoyensis, ovoid-elongate and rather large in C. ulei, distinctly 

 ovoid in the remaining species. In all cases they are more or less 

 flattened by being pressed one against another. The integuments con- 

 sist of a white, translucent testa, covering a brownish or purplish 

 arillus. There is no albumen and the cotyledons are large. 



PRELIMINARY SURVEY OF SPECIES. 



The writer does not make any claims to finalty in the status here 

 assigned to the forms described. As the fruit of tiis own experience 

 in the field, supplemented by the observations kindly supplied by his 

 fellow-workers, Messrs. O. F. Cook and (J. X. Collins, and of his 

 careful study of a large amount of well preserved materials, he has 

 been able to satisfy himself that most if not all of these forms offer 

 constant characters to be found in all specimens proceeding from a 

 certain area and to be easily recognized by any skilled botanist. At 

 the same time it must be borne in mind that, owing to the uniformity 

 in the general aspect of the representatives of the genus, it really 

 needs an exercised eye to distinguish them on sight. Microscopic 

 examination of complete specimens, including branchlets with leaves 

 and terminal buds, clusters of both male and female inflorescences, and 

 ripe fruits, is necessary to insure correct identification. 



The ten forms here described have not. probably, the same tax- 

 onomic value: the distinctive characters may prove more pronounced 

 and weighty in some, and not so sharply defined or of less signifi- 

 cance in others. To the former a definite specific status will be read- 

 ily conceded, while the latter may remain in abeyance or be reduced 

 to the rank of subspecies or even of mere forms. Neither have we 

 necessarily recorded the whole number of specific representatives of 

 the genus ; we know almost nothing of several of the rubber-producing 

 trees of South America, especially of those of Ecuador, Venezuela, 

 and Colombia, and it is even probable that Castilla elastica may not 

 be the only form peculiar to Central Mexico. The present attempt at 

 a monograph of the genus does not, therefore, pretend to be either 

 complete or final. Among the species the status of which can not be 



