Ammdbroma, a New Genus of Plants. 55 



undetermined. Kunth (1. c.) expressed no opinion as to the 

 place of Corallophyllum in the Natural System, and he seems 

 to suspect that the plant he examined was in an abnormal con- 

 dition. It is left by him among his "genera incertce sedis." 

 Endlicher also places it with " genera dubice sedis." 



Sir "W. Hooker, who first described Pholisma, from Nuttall's 

 specimen, and gave a good figure of the plant in his Icones 

 Plantarum (1. c), regards it as nearly related to Corallophyllum, 

 and refers it to Orobanchacese, though he thinks it will yet 

 form a distinct group near that order, but with a very different 

 fruit. "Walpers follows Hooker without comment. Lindley* 

 places both genera, with a mark of doubt, at the end of Mono- 

 tropacea?, which they certainly resemble much more than they 

 do Orobanchaceae. Like the former they are parasitical on 

 roots ; and in the spiked inflorescence of Pholisma there is an 

 approach to Hypopithys. Most of the genera of Monotropacece 

 are gamopetalous, and in half of them the anthers open by lon- 

 gitudinal slits.f The pollen, also, is simple and spherical. 



On the other hand, Corallodendron and Pholisma, as well as 



* Vegetable Kingdom, p. 462. 



f Viz. In Pterospora, Allotropa, Torr. and Gray (in Bot. Wilkes's Expl. Exped. 

 ined.), and Hemitomes, Gray, in Newberry's Bot. of Williamson, and Abbot's 

 Pacif. Railroad Expl. An examination of good specimens of Hemitomes, collected 

 in Washington Territory by George Gibbs, Esq., shows that the anthers are dis- 

 tinctly 2-celled ; but they open and discharge their pollen even before the flower 

 is expanded. The lines of dehiscence are near the connective. After opening, 

 the broader portion of the cells is rolled backwards till each nearly meets its fel- 

 low, forming a large and spurious cell. A narrow portion of each proper cell is 

 left. These also incline towards each other, so that another smaller, spurious, and 

 apparently abortive cell is formed. Hence, after flowering the anther might 

 easily be regarded as only one-celled by abortion. An examination of an unex- 

 panded flower shows the true structure of the anther ; and proves that in the 

 withered state the spurious cells are at right angles to the normal ones. Hence 

 the name Hemitomes is quite inapplicable, and I propose that it be changed to 

 Newberrya, in honor of the first discoverer of the plant, who has distinguished 

 himself by investigating the recent and fossil botany of the Western and Pacific 

 States. 



