No. 2367. FOSSIL PLANTS FROM COSTA RICA— BERRY. 171 



and Bolivia, where I have observed it, and probably elsewhere through 

 out northern South America. Heliconia has not previously been 

 recognized in the fossil state, but I have a species as yet unpublished 

 from the late Tertiary of eastern Bolivia, and Musophyllum elegans 

 described by Engelhardt from the Tertiary of Santa Ana, Colombia, 

 is present in material collected by C. F. Bowen at Betijaque, Vene- 

 zuela. The genus Musophyllum was founded by Goeppert in 1854 

 for fossil Musaceae from the Island of Java, and a number of European 

 and American species have subsequently been described. Fossil 

 forms are liable to be confused with the genera Canna and Geonoma 

 and their allies, but undoubtedly the bulk of the fossil species actually 

 represent the banana. There is no evidence that the existing culti- 

 vated species which flourish so prodigiosuly in the American Tropics 

 were ever indigenous in the Western Hemisphere, and I can see no 

 reason for not substituting Heliconia for Musophyllum in the Ter- 

 tiary record of tropical America. 



Class DICOTYLEDONAE. 

 Order PIPERALES. 

 Family PIPERACEAE. 



Genus PIPERITES Goeppert. 



PIPERITES CORDATUS, new species. 



Plate 22, fig. 1. 



Description. — Leaves of medium size, approximately equilateral, 

 cordate in general outline, with an acuminate tip, and a not deeply 

 cordate base. Margins entire, full, and evenly rounded. Texture 

 subcoriaceous. Length, about 10 cm. Maximum width, in the 

 lower half of the leaf, about 8 cm. Petiole stout, expanded proxi- 

 mad, about 4.5 cm. long. Primaries seven, from the top of the peti- 

 ole, diverging at acute angles, all curved including the midrib, stout, 

 prominent on the lower surface of the leaf, aerodrome. Secondaries 

 thin but well marked, arching along the margins and internally 

 mostly transversely percurrent. 



This is an exceediugly well-marked species of Piperaceae which 

 finds many similar forms among existing tropical American species of 

 Piper and related genera. Since its generic affinity can not be posi- 

 tively demonstrated, it is referred to the genus Pipe-rites proposed 

 by Goeppert for fossil leaves of the plants of this family. 



This ancient and specifically abundant family has hitherto fur- 

 nished but few fossil species, its past rarity being thought to be a 

 matter of lack of preservation or discovery, since its extensive 

 modern distribution would seem to indicate that its ancient history 



