14 



also very close to Ambrosia, but the beaks are 2 or more. The 

 number of beaks corresponds also to the number of cavities in 

 the bur. Each cavity and beak contains usually only one pistil 

 but sometimes two. These species are closely related to the 

 original Gaertneria. In all these species the spines are rather 

 few, seldom 30, and either short and without any hooks at the 

 end, or else more or less flattened or channeled on the upper side. 

 The original species of Franseria on the contrary has numerous 

 spines and numerous series, the number of spines being over 100. 

 They are long and slender and hooked at the end, and the whole 

 fruit in structure agreeing with Xanthiiim. 



The only one who has tried to make segregates in the genus is 

 Delpino, who proposed the genera Xanthidium, Hemixanthidium 

 and Hemiamhrosia, but his arrangements cannot be followed, 

 because he included in Xanthidium the original supposed Fran- 

 seria and Gaertneria and applied the name Franseria to the i- 

 beaked species most closely related to Ambrosia. Besides the 

 name Xanthidium is preoccupied . Hemixanthidium was proposed 

 on a species which Delpino claimed had two kinds of pistillate 

 heads, the one kind described as the ordinary one, the other form 

 as found occasionally but as evidently caused by some disease. 

 His Hemiambrosia is based on the species which would be included 

 in Ambrosia. 



There are two species of Franseria, however, that are very 

 peculiar in their structure, namely, F. eriocentra and F. Bryanti. 

 Both have a single beak which is scarcely oblique and with 

 several teeth. 



The former has only one pistil, but the spines are in several 

 series and the plant is of quite different habit, otherwise the 

 plants could be included in Ambrosia. The most peculiar of all 

 is F. Bryanti, which also has a single beak, and the spines are 

 practically in a single series. According to these characters the 

 plant should be included in Ambrosia, but the bur contains 

 several pistils and is several-celled, although the beak is single 

 and the spines are enormously elongated, sometimes 2-3 cm. 

 long. If none of the other species of Franseria are regarded as 

 generic types, this one should. It is more distinct from Fran- 



