254 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 



others each female receptacle is flanked by two staminate receptacles, 

 smaller than and somewhat distinct in shape from the flowers of the 

 male specimens. In other words, there are strictly unisexual indi- 

 viduals bearing only male flowers, and other individuals that are 

 truly monoecious, bearing together pistillate and staminate flowers. 

 It has been noticed that the younger trees, blooming for the first time, 

 bear invariably male flowers, but it is by no means certain that there 

 are trees permanently male and others permanently moncecious, nor 

 do we know the governing cause of the explained sexual conditions. 

 The stamens are fertile in both kinds of male flowers and, while the 

 wind seems to be the main agent of pollen transportation, a direct 

 pollination by means of insects between the male and female flowers of 

 the same clusters, or the transportation of pollen from tree to tree by 

 the same agency, is not out of question. The male flowers are usually 

 crowded with small thrips that are also found, but not so frequently, 

 on the female flowers. 



Mr. Cook has called " primary " the larger staminate inflorescences 

 found on the exclusively male trees, and " secondary " or " comple- 

 mental " the ones on the monoecious trees. These expressions are used 

 in the same sense in the course of this paper. 



PRIMARY MALE INFLORESCENCE. 



The inflorescences of this class appear in pairs in the axils of the 

 leaves, or in defoliate axils, and have the general appearance of 

 a flattened or depressed cone, or of a fan, more or less emarginate at 

 the base. This fan opens longitudinally, sometimes only by a narrow 

 slit at the top (G. fallax), sometimes with the lobes spreading out into 

 a flat disk (G. nicoyensis). In other cases the inflorescence is more 

 or less distinctly 3-winged {('. costaHcana) , or the lobes are diversely 

 lobulate or distorted (G. elastica). As to the number of receptacles 

 in ouch axil, it seems to vary from 2 to 8, but to be almost constant in 

 one species or iti one group of species. There are no data as to G. ulei 

 and C. australis. In G. daguensis, I noted 1 to 3 receptacles in each 

 axil in the only specimen at hand, but this may be either an error of 

 observation or an anomaly, as geminate receptacles seem to be the 

 rule; the same may apply to G. fallax, of which only dry, brittle speci- 

 mens have been handled. From 2 to 8 receptacles in each cluster 

 were ascertained to exist in G. lactifkia and G. guatemalensis, while the 

 number seems to be limited to 2 or 4 in G. costaricana, G. panamensis, 

 G. nicoyensis, and G, elastica. 



The receptacles are sessile or subsessile in C. falla,/>, but distinctly 

 stipitate, as far as known, in the other i) species. Except perhaps in 

 C. ffuatemalensis, the stipe is not a pedicel in the strict sense of the 

 word, but simply the basal attenuation of the body of the receptacle, 



