210 PItOCEEDINGS OF THE NATIONAL MUSEUM. vol.xxi. 



a ^eat trunk. It is similar in mineral character to No. 0, and the leaf 

 scars and other organs agree well witb the upper i)arts of that si)ecimen. 

 The fractures are downward, but follow the i)lane of the petioles, which 

 are here erect. In No. 15, however, two large and nearly equal branches, 

 whose axes were nearly at right angles to each other, are represented. 

 Viewed from the broken sides the two axes are ch*arly seen in contact, 

 having a gnarly appearance, such as is normally produced at the junc- 

 tion or crotch between two branches." 



This branching character, as I was well aware, does not belong to the 

 large perfect specimen, but having no others, I thought it possible th^t 

 some of the small secondary axes might in other cases become primary 

 branches. But after seeing so many other specimens of C. colossalis, 

 all agreeing in this respect, and also a large number of the present 

 species also all agreeing and exhibiting no tendency.to vary in the 

 direction of the other si)ecies, it became obvious that the branching 

 forms aJl belonged to one species and the simple ones to another. The 

 specimen No. lo clearly belongs to the branching species, and now it is 

 easy to see other specific differences. 



CYCADEOIDEA FURCATA, new species. 



Trunks hirge, forking above, or sometimes with a third branch, sini])h' 

 below, laterally compressed, eccentric, light colored, soft and of low spe 

 cific gravity, 35 to 45 cm. high, 25 to 30 by 35 to 40 cm. in diameter, 90 to 

 110 cm. in girth ; organs of the armor mainly horizontal ; leaf scars sub 

 rhombic or somewhat triangular, the vertical angles generally rounded, 

 the lateral acute, variable in size, averaging 15 by 25 mm., those on the 

 branches smaller, or sometimes nearly as large, empty f ramentaceous 

 walls variable, usually thin, 1 to 5 mm., much thicker in the angles, firm 

 in texture, grooved or divided into two or three plates; reproductive 

 organs few, large,*elliptical, 4 to 7 by 7 to 10 cm. in diameter, either set 

 in depressions or somewhat elevated, surrounded by bract scars, either 

 cavitous in tlie center or solid, the larger ones simulating small branches ; 

 armor 4 to 7 cm. tliick; cortical parenchyma 7 cm., clearly distinguish- 

 able from the darker zonfeof wood cm. in thickness; medulla elliptical, 

 9 to 11 cm. in diameter. 



This species is thus far represented by only two specimens, namely, 

 Nos. 18 and CO of the Yale collection, the latter of which is in such a 

 complete state of preservation that little can be known of its internal 

 structure. It is distinguished from all other trunks known to me by a 

 true dichotomy, consisting of a simple trunk with two nearly equal erect 

 branches and a natural Junction or crotch at their joint of separation. 

 The axis is far to one side and the trunk is flattened on that side, the 

 entire true base being lateral and the trunk standing on a false base 

 belonging to the armor, but naturally flattened in transverse direction. 

 These peculiarities were doubtless the result of the position in which 

 the trunk originally grew among rocks. Besides this striking charac- 



