290 A FOTJETH NEW HONGKONG CALAMrS, 



Juxtu convallem Wongneichung ins. Hongkong, sub initio m. 

 Martii, 1875, detexit Dr. G. Dods. (Herb, propr. n. 18979.) 



Though I have examined a considerable number of seeds, I have 

 been unable — probably owing to their having been kept several 

 months— to satisfy myself perfectly as to their internal structure : the 

 albumen appeared to me in all deeply ruminated. If this is not an 

 error, it is a character hitherto supposed peculiar to the DcEmonoy^ops 

 group. 



The nearest ally of this pretty and graceful palm is no doubt C. 

 gracilis, Eoxb., a native of Chittagong. From Roxburgh's descrip- 

 tion and figure, copied by Griffith,* his plant differs from the Hong- 

 kong one by its lanceolate much more acuminate frond- segments with 

 aculcolate nerves ; the fruit is unfortunately unknown, or at all events 

 undescribed. C. radiatus, Thw., according to Dr. Thwaites's diag- 

 nosis, f has its 6-7 linear frond-segments all radiating from the ex- 

 tremity of the rachis and densely spiny vaginse. C. ^>«d?%s^emowMs, 

 Thw. ! may be readily distinguished by its fewer and longer frond- 

 segments, and more especially by its fruit — unknown when Dr. 

 Thwaites's character was published — which is somewhat larger, and 

 has the scales, which are straw-coloured with purple margins, dis- 

 posed in only 12 vertical series, each series consisting of about eight 

 scales. The numerical relations vertically and laterally of the fruit- 

 scales furnish a character seemingly of great constancy and the highest 

 importance, which has not been sufficiently attended to by writers, 

 with the exception of Elume. 



There is yet another character furnished by the loricae hitherto, so 

 far as I am aware, unnoticed by any author, and to which my 

 attention was first directed by my valued friend the Eev. Dr. Graves, 

 as distinguishing the fruit of C. melanoclmtes, Bl., from that of C. 

 Margaritce, mihi. These two species are very close allies, but in the 

 Malayan plant, as figured in Blume's double folio plate and also in his 

 detailed analysis,;}: the margins of the scales in the same spiral form a 

 continuous line ; that is to say, any given scale is not overlapped at 

 the upper limit of its anterior edges by the extremities of the scales 

 situated to the right and left ; in C. Margaritce, on the contrary, the 

 lateral scales in each diagonal series project f of a line beyond the 

 edge of the scale situated between and immediately beneath them,§ 

 so that the spiral, instead of describing an unbroken line, has a step- 

 like arrangement, or resembles the teeth of a saw. I have found no 

 exception to this in more than one hundred fruits I have examined. 

 The use of such a phrase as seriehus squamanuji spiralihus (or speiro- 

 stichis) plam's, or graclatis, as the case might be, would be a convenient 

 mode of indicating this difference, in diagnoses. All the Hongkong 

 Calami have speirostichi gradati. 



* Griffith, *' Palms of Brit. India," 64, t. 196. 



t Thwaites, " Enum. pi. Zeylan." 431. 



+ Blume, " Kiimphia,"ii., tt. 134, 137 B. f. 10. 



§ As shown in Blume's plate of Ceratolobus concolor (op. mox cit. , ii., t. 130. 

 f. 1. A). K f , y » 



