SIPHON AN TH.E. 



Ill 



Mr. Maximowicz in 1877 (Mel. Biol, x, 82) diagnosed P. rhinanthoides from P. labelL 



as follows 



filaments glabrous, beak shorter than the 2-dentate galea, whole plant 



a 



span high p> rh mHthoide . 



anterior filaments hirsute, beak longer than the edentulous gale* . . P. labellata; 



adding to the diagnosis in 1888 (Mel. Bwl xii, 786), as regards P. rhmanthoida, 

 "very like P. labellata. 11 The original plant is not at Calcutta, but gpecinu ns from 

 Sairam, Turkestan, collected by Mr. A. Kegel ard named by Mr. Maximowj z have 

 been examined. In these the anterior stamens are hirsute, as MM. Fischer and Meyer 

 say they found them in Schrenk's specimens and not, as Bunge tells us ho found 

 them {I.e. 276), 1 all glabrous. In Griffith's specimens, (,I.»ur. 10GO), the ant. rioi 

 stamens are hirsute, and not, as M. Boissier (I.e.) indicates, all jrlabnujs. Tho baits 

 on the filaments of var. ft/pica are not numerous, but appear to be usually prcsenl , 

 I have not found them once absent in 23 individual corollas analysed. Thus the 

 principal diagnostic character fails. Nor do any of the other charaetn* hold; 

 Himalayan specimens of "P. labellata" vary a great deal in size, but have the beak 

 very uniformly twice the length of the hood, while specimens from Phari (of -which 

 I have examined 11), though not exceeding 12 cm. in height, have the beak 

 thrice the length of the hood. On the other hand, the gathering from Iiamsor 

 (Duthie 576) consists of dwarf specimens with the short beak of var. typia, but 

 with the densely hairy anterior filaments of "P. labellata 11 Among the 48 specimens 

 of "P. labellata 1 '' in the herbaria of Calcutta and Saharanpur examined by me, 43 



have edentulous corollas, 4 have the throat 2-auricnlate (a flower from one of 



is shown in Plate 1, fig. 8), and one from Zalang Karpo in Karnag collected by 



Dr. Stoliczka has the throat on both sides very distinctly toothed, the U h being 

 much longer than in any specimen of P. siphomntha, a species in which this character 

 is a normal one. It is not therefore possible to consider P. labellata specifically separable 



from P. rhinanthoides 



Dr. Griffith's Hadii-gak plant, referred by Mr. Bentham to P. Elcph 



Boiss 



in 



1846, is by Boissier himself, so recently as 1879, only doubtfully placed there. And 



this cannot be its position, since Boissier describes P. Eleplms as bavmg th<- ;,pcx <,f 

 the beak 2-fid, whereas the apex of beak in Griffith's plant, which agrees completely 

 with Fischer and Meyer's description of P. rhinanthoides, and is preci-Hy the tame a* 

 A. Regel's Sairam specimens, is entire. Genuine P. Elephas, on the other hand, cannot 



be satisfactorily diagnosed from P. siphonantha. But another Griffithian spec.men referred 

 by Bentham I.e. to P. labellata is not that plant, but is P. vphonamha var. hretduba ; 

 consequently P. labellata Boiss. (I.e. 489), which depends on this specmen alone » not 

 the same as P. labellata Jacq., but is=P ,iphonantka Don. Sir J. D. Hooker, U. 313 



reduces P. labellata to P. siphonantha; the following are the d.aguostic characters of 

 th9 two species: 



calyx 5-toothed, margin of lip ecibate, lobes not emarginate, basal 



porti 



of hood inclined forward, beak entire at apex 



P 



calyx 3-toothed, margin of lip ciliate, midlobe emarginate, basal 



portion of hood erect, beak 2-fid at apex . . 



P 



i t • v„i, *h» Utter* Ic are employed the reference is to the ■ynonymy that follows the specific 



'In every ease mwh,e^ the 1 tter Ic *™ W* ^ . of the telt 





