464 



CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 



Alphonse de Oandolle's description of his Lucumat capiri is founded on the ''Caiques 

 et Dessins" and partakes of their inaccuracies. As the author himself acknowledges 

 in his "Phytographie," " "although these drawings are generally exact," they are 

 not quite satisfactory as types of new species. Nevertheless, taken us a whole, the 

 description agrees wonderfully, except for the fruit, with the specimens described 

 here. 



With due allowance for the "personal equation," the tree described by Hemsley 

 corresponds satisfactorily with the materials in the U. S. National Herbarium, except 

 for the lesser dimensions of the leaves, the 10 to 12 instead of 14 or 1-3 primary veins, 

 and the hirsute ovary. It may be inferred from the fact that the fuzzy pubescence 

 on the back of the corolla lobes in their earlier stage is given as a character that the 

 Kew specimens are undeveloped ones; this would also explain the discrepancy in 

 the size of the leaves and the number of primary veins; the former are "molliter 

 ferrugiueo-hirsuta" only when young, the thick pubescence then hiding the terminal 



venation. The ;i ovarium hirsu- 

 tum," however, is not so easily 

 explained in the absence of the 

 type specimens. 



The only serious discrepancy 

 between Gray's description and 

 the specimens at hand, which be- 

 long to the same numbers of the 

 same set as those on which he es- 

 tablished his species, is in the 

 length of the leaves, this being 

 underestimated. 



The specimens of the Mexico 

 National Museum come from the 

 mountains of the State of Michoa- 

 oan, on the western slope of Mexico, 

 where the tree was first noticed. 

 Those collected by Palmer grew 

 farther north in the neighboring 

 State of Jalisco, in the same alti- 

 tudinal zone. The exact origin of 

 the Kew specimens is unknown, 

 but it is known that Jiirgensen, 

 in the employ of Galeotti, collected principally in the mountains of Oaxaca. 



The name capiri, already given to the tree in time of Mociiio, again appears on the 

 labels of the specimens collected by Leon, and this is also a good incidental confirma- 

 tion of the view expressed at the beginning as to the definitive reduction of Lucuma 

 capiri, Sideroxylon mexicanum, and S. pctiolare to the primitive Achras capiri Moc. & 

 Seas., which now becomes Sideroxylon capiri. 



Fig. 90. — Diptwlis minutifiora, floral details, a, Four se- 

 pals, that at tlie left exterior, that at the right interior 

 (fourth sepal wanting); b, two exterior corolla lobes; 

 c, interior corolla lobes; d, staminodes; e, stamens; /, 

 pistii; g, cross section of ovary. Scale 6. 



A NEW SPECIES OF DIPHOLIS AND ONE OF MXMUSOPS. 



Dipholis minutifiora Pittier, sp. nov. Figure 90. 



A small tree with dry bark and reddish, very hard wood. Twigs terete or sub- 

 angulose, the newest ones whitish-tomentose, the oldest substriate and lenticelloee. 

 Puds ferruginose-pubescent. 



Petioles 1.5 to 2 cm. long, canaliculate, minutely pubescent. Leaf blades coria- 

 ceous, obovate or oblong, 4 to 8 cm. long, 2 to 4 cm. broad, acute-rounded at base, 

 obtuse, or broadly rounded and sometimes emarginate at the tip, light green above, 

 paler or brownish beneath; main rib impressed above, prominent and obscurely silky- 



" Page 141. 



