.8 



GEOGRAPHICAL DISTRIBUTION. 



atural 



Sea to the north-west and the barrier of an inclement mountain chain on the south-east, 

 the path of immigration into this province was comparatively narrow, and accommodated 

 comparatively few species. Then, during the period of stability that followed, the 



conditions that made the shores of the Mediterranean Ocean a centre of pre- 

 ■ervtttfoa for the present flora of the Mediterranean basin, prevented the wide extension 

 of such species as had gained access to the arctic and sub-arctic portions of the 



And finally, when retreat began, temperate conditions were early established 

 the territory through which these invading forms had reached the Caucasus. Their 

 retreat being thus cut off, they could only extend in the Alpine parts of the Caucasus 

 province itself. But as their advance had been greatly checked, the forms thus cut off 



province 



were 



few in number, and in place of finding the area subject to natural cond 



suitable for their existence steadily curtailed, they found it continue ample. Internal 

 pressure thus being minimised, and there being practically no external pressure, there 



little stimulus in this province, compared with any other, to variation at all, 

 h variations as naturally appeared deviating far less than variations elsewhere did 

 from the original archaic types. 1 



was 



M. 



Messrs 



Hemsley {Jour. Linn Son. xxri, 204 -220 ; April 12, 1890) while this paper has been passing through the press. They are 

 .ntained in a list of 94 Chinese Pediculares, 10 of which are advanced as new. The following species in the list call 

 b* remark —16 P COnifera Maxim. A new species from Hupeh, forming a monotypic series with a corolla-structure 

 that corroborates the evidence on which (p. 18) the tribe VERTICILLAT2E has been suppressed. -27.^ P. filicifoHa 

 Hemsl A new species from Hupeh, a member of the series Axillares. 

 in Southern China.-34. P. hirtella Franchet. A new species from Yunnan of great interest as being the first alternate- 

 leaved Bidentata reported from the Chinese meridian of distribution. In the absence of specimens and in the absence of a 



30. P. gracilis Wall 



full 



integrifolia 



Hook. f. is now reported from Yunnan.-41. P. leiandra Franchet. This name is proposed for a plant from Yunnan that 



Mel 



Wall. As Mr. Maximowicz has been able (cf. Mel 



Wallichian specimen, while M 



M. Jacquemont and named (in Herb. Paris) by the late Mr. Bentham, the proposed name is not at present acceptable. There 

 is nothing in its description to separate the Yunnan plant from the Sikkim form of P. porrecta (plate 20, B) with fleshy 



-L.J »l:. a:i.i.:~ ~i„„«- :„ ~±-in*i*i+ „~<- «,,™,>;fl™iiTr ^^lin^ fmm t.hft t.rnft P. vorrecta of the N.-W. Himalaya. -48. 



V. macilenta Franchet. A new species 



50. P melampyrifl* 



ame is substituted for P. melampyrifolia Franchet {Mel. Biol, xii, 895 : 1888* which has priority. But « la publication 



est un fait que l'auteur ne peut pas annuler," {A. Be Cand , Lois, 61) and M. franchet is respc 



so, for the earlier name.--66. P. oxycarpa Franchet is now reported from Szetschuan.— 71. P. 



Franc A 



Fragiles, but on account of 



79. P. strobilacea 



An interesting new species from Yunnan placed, on account of its habit, near the group 



its inadequately described corolla kept apart and situated between that group and the group 



Franchet. A new Yunnan species apparently a Siphonantha and probably a Longijlora ; unfortunately the staminal insertion 



is not given in the description.— 88. P vagans Hemsl. An interesting new Rhyncholopha from Szetschuan which helps 



to connect the group Excelsa with the rest of the section.— 92. P. Viali Franchet. A still more interesting new species 



from Yunnan, connecting the American series Surrectse Maxim 



Mi 



Fragi 



Yuan 



•93. P. villOSula Franchet. A new Obthoebhyncha from 



(excluding 



Three (P 



if era, fil 



The total for China is now (cf. p. 25) raised from 49 to 54, the endemic factor is increased from 35 to 38, and the percentage 



decreased 

 101 to 107. 

 from 90 tc 



The other six are endemic in Yunnan, and raise the Himalo- Yunnan total (cf. p. 26) from 



oxycarpa 



to 94 and the endemic percentage falls from 89* 1 to 87*8. For Yunnan (cf. p. 35) the total is now 47 (endemic 

 33), so that the percentage changes from 70*0 to 701 ; for Szetschuan (cf. p. 37) the total is now 30 (endemic 9), so 



the percentage changes from 32*1 to HQ0 

 The extension o£ P. gracilis to Hupeh 



appearance 



as had (p. 44) to be discounted. If 

 annexation (p. 37) of Yunnan to the 



Himalaya, this is more than counterbalanced by the occurrence of P. integrifolia in Yunnan as well as in Chumbi, 



