N,; CONTRIBUTIONS FROM NIK NATIONAL SERBARIUX. 



overlook them. Flowers were collected, and later Doctor Purpus 

 sent me fruit These -how clearly thai this plant is distinct from both 

 Dasylirion and Nblina, though much nearer the latter. 



A lx >ut Tehuacan tw<> remarkable species of Beaucarnea were discoi 

 ered, both of which seem to be undescribed. One of them has a most 

 singular trunk, at first nearly globular but afterwards sending up a 

 slender stem which becomes more <>r less branched. The swollen base 

 take- <>n a multitude of shapes, but isalwaj si ery large as compared with 

 the rest of the plant. One such measured at 0.6 meters above the base, 

 6.3 meters in circumference, and at 1.8 meters above the base contracted 

 abruptly into the slender stem proper. The basal expansion is made 

 up of very loose cellular tissue which when dead and dry is almost as 

 light as cork. This club-footed base suggests the specific name used 

 elsewhere in this paper for this species. The aspect of the plant is 

 shown in PL XXIII. facing page vv . 



Dasylirion quadrangidatum S. Wats., while perhaps not uncommon 

 in cultivation, is not often met with even in our larger herbaria This 

 specie> was found to he very common in eastern Hidalgo and the drier 

 parts of Queretaro. It forms a very distinct trunk L20 to L50 cm. 

 lone- below the crown of leaves and sends up a flowering stem 3.6 to 

 4.r> meters long. The leaves are very unlike those of all the other 

 species, being very thick and narrow and not prominently saw-toothed. 

 They are often 3 meters long. In some parts of Quere*taro the natives 

 use them as a thatch for their houses. They call the plant junquillo. 



Material of the small parasite Pilostyles (or Apodantho) of the 



Rafflesiaceae was cojlected at five localities, two near [xmiquilpan, one 

 in the desert of Queretaro, and two near Tehuacan. Whether more 



than one species was obtained I have not yet determined, hut there 



were two or three hosts, each being some species of Parosela. 



According to Doctor Robinson Pilostyles has been reported only 

 three times from North America, once by Geo. Thurber and twice by 

 Dr. ('. (J. Pringle. Material was collected, however, by Mr. Fred- 

 erick V. Coville in Texas in L904 and by Dr. E. Palmer in San Luis 

 Potosi in 1905. This genus, in North America at least, is always 

 parasitic on some genus of Viciaceae, generally on Parosela. The 

 plants are minute, reduced simply to flower parts, which may account 

 for the fact that the species have been so \^v\ rarely collected. 

 Unless one has seen specimens or is looking particularly for them he 

 will readily pass them by as secretions or insect work. 



All new species here described are based upon specimens in the 

 United States National Herbarium and when two or more collector- 

 are cited the type specimen is definitely stated. 



The line drawings are the work of Miss Juliet ( '. Patten, except 

 that plate 25 was made by the late Frederick A. Walpole and plate lo 

 bv Homer D. House, 



