136 



Inat the tendency' to A-ary is at times less marked m species near 

 tlie extreme limits of tins range, such as tlie violet M. Henrici, 

 the indigo M. horridula, the purple M, quintuplinervia, and the 

 red 31. punicea, than it is among species like M. aculeata, M, 

 rudis, M. Pmttii, M. Wallichii, where the normal colour is pale 

 blue. But Avith all this variation there is no instance, so far as 

 is at present known, of a species in which. the petals are some- 

 times yellow, at others blue or purple or red. * 



In 1896 it was pointed out that in M. primulina the stigma, 

 which is of the type with decurrent rays, is 2-partite, the lobes 

 being oblong and plano-convex, the outer convex surface being 

 2-rayed. It was further pointed out that in this species there 

 is a rudimentary disk composed of four papillae in two pairs, each 

 pair opposite to the stigmatic cleft. It is now known that this 

 solution of the stigma is not peculiar to M. jyrimulina. It is 

 sometimes, though not always, met with in ^le flowers of M. 

 discigera, where again there is a disk. In this case, however, the 

 solution may be complete, the individual stigmatic rays or loops 

 being quite discrete and stellately patent; as before, the disk- 

 lobes are alternate with tlie stigmatic rays or loops. There is, 

 however, no necessary connection between this tendency to solu- 



of that structure. A tendency to solution has been met with in 



M p.nlonomma from Kansu ; and in a specimen of M. speciosa, 



coilected by Forrest in 1914 in Yunnan (Forrest n. 13240), the 



stigmatic rays m some of the flowers are as completely separated 



as m the case of M. discigera, though, owing to the fact that the 



stigma m M. speciosa is considerably shorter than it is in M. 



discigera, the appearance which the discrete rays present is less 

 striking. "^ ^ 



In 1906 it was necessary to explain that of the alpine vegeta- 

 tion of the Himalayan region and of the Tibetan districts to the 

 north, between 89° and 990 e, practically nothing was then 

 known.* One of the sequels to the Abor Expedition of 1911-12t 

 was the ovganisntion of a systematic survey of the Abor country 

 proper and of the region to the north of the countiy inhabited by 

 the Abor tribes. This survey included in its scope the cxploratioh 

 of the ^catchment area of the Dibong, an affluent of the Brahma- 

 putra,., and while this was in progress two of the off^icers cm- 

 ployed Captam Morshead and Captain Bailev, were detached 

 from the mam party, and, travelling with ver/ light equipment, 



IT \ .^ ^'^n^'"'' }^^ ^'"^^^« "^ *^i« district of Chimdro which 

 ■separate the valley of the Diliong from that of the Dihong, the 



large river which debouches next to the west from the Hinmlava 

 ^Z^^t^'^''^ P^^'^^l, ^* '^^^^ probable that it was in the 

 ZX5" ••^T'^^^n'^* '^^-^^ ^^ "" M^^onopsis -from the Abor 

 Frnn!^' 'T!^ "'' Greenwich Park in 1914, were obtained. 

 mZi?r ^T J^^^^^^^-^^ich fluwered in June, came the 



matenal upon which the description provided below of the 

 striking speces J/, decora h<^been based. Subsequently the 



* Annah of Botany, vol. xx. p. .%30. 



t Kew Bunetin. ir>12, p. 1.59. 



+ Geographical Journal, vol. Ixii. p. 491 



