258 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 



The style varies in length, thickness, etc., according to the species 

 thus affording good means of identification. It is long (3 mm. and 

 over), slender, and smooth in C. daguensis and C. panamcnsls; long 

 and sparsely hairy in ('. fallax; slender, smooth, and about 2 mm. long 

 in C. costaricana. In C. lactifma it is thick and short (1.5 to 2 mm.) 

 with spinescent bristles at the base only, while in C. guatcmalensis the 

 hairs, longer and not so stiff, extend around the base and on one side 

 up to the stigmas. Castillo, nicoyensis has a very short (1 mm.), 

 thick, and quite smooth style, but in C. elaslica this organ varies in 

 length from 0.8 to 1.5 mm., and is also thick and sparsely covered 

 with short hairs. The style of C. australis is the most reduced and is 

 also hairy. 



The stigmas are normally two to each style, but very often three or 

 five are found in C. elastica, and not seldom three in C. guaternalensis. 

 They are short and cushion-shaped in ('. australis, rather slender and 

 moderately developed in C. daguensis, (J. fallax, C. costaricana, and 

 C. panamensis, and long, broad, and conspicuously twisted in the other 

 Central-American species. In the Yucatan specimens of C. guate- 

 rnalensis they are so prominently long and otherwise peculiarly 

 shaped that we are in doubt as to whether these specimens do not 

 constitute a separate type. In most of the species the stigmatic 

 surface is more or less pupillose-verruculose, but in C daguensis the 

 papillae, instead of being covered with a sugary nectar, are provided 

 with an indument of long, flagellate hairs. In 0. australis this same 

 surface is also hairy, if we may judge from Hemsley's plate. 



I S FRUCTESCENCE. 



Directly after the pollinating process is over, the perianth usually 

 begins to thicken and it then turns gradually into an orange red pulp. 

 As with the growth of the ovules, now changed to seeds, more space 

 is needed, the receptacle increases its surface, all its parts, but mainly 

 the outer scales, altering their original size, shape, and texture. As 

 these changes are gradual and it is often difficult to decide whether a 

 given receptacle is still an inflorescence or has become an infructes- 

 cence, the characters derived therefrom are always more or less in- 

 definite and should be used with caution in the identification of the 

 several species. 



An exception to the thickening and coalescence of the perianths is 

 found in C. fallax, in which they remain free, green, and hairy, with 

 more or less adherence to the seeds. In the other species the ripe pulp 

 is sweetish and eagerly sought for by certain birds; in G. ulei it is 

 even a favorite delicacy of the native Indians. In this species the 

 whole infructescence has a quite characteristic appearance, being 

 globose and entirely covered with scales and containing only 3 to 5 



