150 



tlie plant again among rocks in Mupine, at 14,000 ft., in July, 

 1908 {Wilson 3030). In 1913, in his work on Western China, 

 Mr. Wilson, while maintaining the view that M. Prattii is 

 distinct from M. rudis, has accepted the Fflanzenreich arrange- 

 ment and has referred both the Ta-chien-lu and Mupine M. 

 Prattu (vol. i. p. 138), and the form of M. horndula collected by 

 him on the higher alps of Western Szcchuan (vol. ii. p. 9) to 

 M. racemosa, Maxim. 



_ Since 1908 both M. Prattii and M. rudis have been in cultiva- 

 tion m European gardens; observation of the two in the living 

 state has fully confirmed the justice of Mr. Wilson's criticism. 

 Though nearly allied, the two differ considerably in the consis- 

 tence of their leaves, those of M. Prattii beincr softer in 

 *^'''i/^®i?^'^ °^ ^ diferent shade of green ; in their prickles, which 

 i^J ^^^'Z ''^"" ^^I'c weaker and are always pale in colour, those of 

 M. rudis usually being purple, at least at the base ; in the anthers, 

 which are whitish or buff-coloured in M. Prattii, yellow in M. 

 rudis; and m the stigma, which in M. Prattii is greenish white, 

 but m 31 rudis IS orange. The style, too, in .1/. Prattii is more 

 slender, while the stigma is smaller than in M. rudis. Another 

 ditierence between M. Prattii and M. rudis has impressed cultiva- 

 tors 01 these two species. In the former jolant all the pedicels 

 except the terminal one which supports the flower that opens first 

 are short or very short, thus imparting a relatively compact 

 appearance to the inflorescence ; in the latter all the pedicels are 

 long or very long and the inflorescence is therefore more open and 

 lax. mt this difference, though sufficientlv strikinn- to deserve 

 notice, IS not one on which too great stress should be laid. Within 

 M. aculeata we find the same striking difference in ap])earance, 

 yet m tliat species no great account has been, or can be, taken of 

 the character. Moreover, there is evidence that the same varia- 



1 -i-i • • f* , -I . ; I — "^ '->iviciacc iiiuL ine same vuiiti- 



b hty IS manifestecl m ¥. Prattii. The specimens collected by 

 ::'• X v^.'? at Sien-wha-shan, in West Kansu (Purdom 736), 



, pt T ^^l^^°g^^s^atl« f rom the original type of ^l Prattu from 

 t .n .^ ?r ^ '"^ T,^«t Szechuan. But tlie plant met with by 

 tr. Zl ' ".1 °^ ^^''- ^"V^- ^"^''8' ^^'^^« ^^1 +^'e stem pedicels 



^Jr^'-T' *^^ f. "' *^" corresponding condition of M. 

 acul^eata, Eoyle, and has, moreover, what is rather rare m M. 

 ^^^^eata, a considerable number of long, sbmder, 1-flowcred, 

 simple radical scapes surrounding the base of the central stem. 

 In this instance, too, the character would apnear to be a fixed 

 ?r;/T '° ^ plant presented to Kew by the late Mr. R. Wood- 

 T>p[f.'J p'""' r ^^^dj^isefl i^ ^^t Arley Castle from seed of tbis 

 .cronfr... -^^ ?^' }^^^ character of long stem-pedicels with 



o^S^^'""-^ '^^^'''"^ ^'"^^^^ ^^^P^^ ^^ ^^ i^^^'r'^-e^l a^ in the 

 ougmai specimen. 



Meconops 



^^. ..^vvv.tupsis luuis, jrTain in JJot Man- f S'lHS n<^\\X\ 

 M. racemosa, Franch ir^ "Rull Qrl -n V i?^' ^- .^^'"^ ^f'^^^^ 

 n.S8a\ At iT, T^i -n Y -l^^/i- ^00. Bot. Fr., vol. xxxiii. p. 



ar Ll-l ?• P^^P^^y-' P- 41 (1889) ; nee Maxim. M. horrid u.u, 

 rSUnsqr^ l7 ^^ J«^™-.A«- Soc. Beng. vol. Ixiv. pars 2, 



M. sinnata, var. Prattu excl. (1906): Fedde 1 c r> ''^iH svn 

 M. stnuata, var. Fm«n excl. (1909). ^' ^ 



