BIDENTATyE AND A.NODo: IVE. 



67 



This is best shown by diagram 10, which exhibits the distribut 



shows how unlike it is to that of 



any other 



of the section 



highest actual number also) of Bidentate species occurs in tl 



The highest percentage (as it happ 



md 

 the 



The decrease of the percentag 



th 



Siberio-Turkestan prov 



Caucasus appears to indicate that the section was 



evolved on the Siberian meridian during the last period of retreat 

 paratively recently. Another indication that this type of corolla ori 



d 



at, too. com 



that th 



(p 



only alternate-leaved Bidentate species 



ginatcd in Siberia is 



Chin 



a 



dolichorrhiza) 



d the WYsti 



II 



meridian 



an excellent example of overflow from one to 



in 



This section contains the greater number of 



temperate-intermediate flora 



and that are characterised by a zonal distrib 



ilayi 



r icent 

 individuals that belong to the 



ion. In 



discussing species of that class, we have already seen that these particul 



probably evolved under the temperate conditions they now prefer; and this as we t 



8]>eci< - w- re 



could only happen at a time when climatic 



had ceased to be progr< sis 



„ ^-^^.v, wi*ug<.B xi«U. W.l.^U u» uv profit ive, 



species, must have been just as the point of furthest retreat was 



which time, for these 



being reached. But diagram 10 shows further that this character of distribution is not 



confined to these few species, but is exhibited by the section as a whole 



for tl 



higher percentage in each of the four provinces border 



on the Circumpol 



than 



w a 



in 



any province more remote, and also than in the Circum polar province it-elf. It M»ems 

 therefore impossible to come to any other conclusion than this: — that the Bn untax 

 were evolved under temperate rather than under Arctic or Alpine cond 



they s 



and 



t 



vince 



appeared at a comparatively recent period. 

 The occurrence of Bidentate in the most easterly division of the Iliniala) 



at first sight to invalidate this reasoning, but in reality it aiWdi 



pro 



app ears 



it tl 



ie 



strongest confirmation 



1 



first, the Himalayan B 



are 



Indo-Cl 



not Emodic — they thus characterise the div 

 temperate rather than Alpine. 



whe 



Then, the complete ab 



the physical 



dit 



and 

 are 



of Bi dentate from CI 



(the 



gle 



pt 



of P 



dolichorrhiza, as h 



b 



said, is an example of overflow) 



indicates an independent appearance of Bidentate in Indo-China. And final I) 



a 



11 



Indo-Chinese Bidentam have opposite leaves. Now we have already seen that 

 appearance of even the most highly specialised of the Rhyncholopm in all prob- 



thes 



the 



ability preceded the occurrence of opposite leaves 



any 



section 



It b 



theref 



almost certain that in Indo-China. as in Siberia, the appearance of Bidentate was 



subsequent, and probably 

 the Rhyncholoph^e 



in 



vjuiuw > — — ' 



both provinces was long subsequent, to the appearan 



f 



The distribution of the Anodont* shown in diagram 11 alongside of that of the 



e evidence advanced already 



Bidentate require 



brief notice 



It confirms tl 



from other sources that the Anodontous is the most archaic type of 



f 



d that 



the AtfODONTiE form the oldest section 



The 



Bidentate enables us to see that the apparent 

 hypothesis, for the Siberian meridian now disapp 

 with the curious 



planation of the distribution of the 



tradiction of this, on the migration 



We 



however, left face 



face 



fact that the Caucasus province, which 



the equatorial termination of 



the Siberian meridian, exhibits a percentage of arch 



forms higher than that of any 



higher 

 planation of th 



province 



itself 



The 



is 



than that of the Circumpolar province 

 that the Caucasus has been a centre of preservat 



poss 



ible 



for forms of 



archaic type 



But this explanation appears perfectly sufficient 



for when th 



thward 



gration into the Caucasus province first took plac 





it did so owin 



>^ 



to 



ctic cond it 



prevailing on the pi 



of Turkestan 



But owing to the presence of an Aralo-Casj 



Ann. Hoy. L'ot. Gaud. Calcutta, Vol. III. 



