﻿MIMOSEiE 



855 



are also western or extratropical species, north and south, of limited areas and very 

 definite character, renmants of ancient races now evidently expiring. In the third extra- 

 Australian subgenus, the Grummifer (b ^ the circumstances have changed ; it is at once less 

 tropical and less American. Only 14 species are known in the New World to 47 in the 

 Old ; and those 14 are either Western, West-Indian, or extratropical north and south, 

 Not one (except the generally introduced A . farnesiana) is to be met with in tropical 

 South America east of the Andes, although some, within their areas, appear to be abun- 

 dant and variable. 



In Africa the ^ cacm constitute more tliaa half the whole nuraher of the MitnospoB of 

 that continent, and form a very promineat feature in the desert landscape of the dry 

 districts hoth within and south of the tropics, extending in some places northward to the 

 Mediterranean, and eastward, connected throu^'h Arabia with Western India. Thoy 

 belong to the two above-mentioned American and Asiatic subgenera Vnlgaret and 0«»»- 

 mifera;. Of the former, 22 species (to the American 42) include some forest races of 

 wide range, representative both of Asiatic and American ones. The African Gimmycra, 

 on the other hand, for surpass those of any other region, comprising at least 35 species, 

 often wide-spread, some very variable, and two at least apparently identical with West- 

 Asiatic ones, but others again local and definite. From the Mascarcnc islands I only 

 know of 2 species o£ Acacia ; one, belonging to the Vulgares, is rem;„\-able as being 

 closely representative of the BrazUian A. laceram, and much further from any Allican 

 or Asiatic species; the other. A. hete^-ophalla. is an extreme outlying species of the 

 AustraUan subgenus Fhyllodine^. and closely representative of the extreme outlying 

 species in an opposite direction, the A. Koa of the Sandwich Islands. Asia has 19 specie 

 of Acacu.. 9 each of the African and American subgenera Vulgare. ^nd^ Gumm.fera., and 

 some of them closely representative of species from either contmcnt ; the 



A. Mangiu^, one of the \'^y'">''ZH7^^i^<^ '^■io.<~i. allied to, if not identical with, 

 but if fiumphius's accounts ai. cor^ , ^\^^'^^J^^^^,^^^ ,„ [^^ ,.,„,^,, 1,1,,,,, 



a North-Austrahan ^^f . ^.IX The same Australian subgenus myiloaine., 

 presents us with 4 specie , all ^on "ig ^.^ extra-Australian 



and nearly allied to Nor ^^-^^^^t; ,Sier they are the persistent remnants 



Pk,Uodme<ehe.e -S-t^d from Aus^ , ^^^^ ^ ^ ^^^ ^^^^^^ ^^^ ^^ 



of the raoft in its ancient home, we ud-v 



unknown to recent botanists 



logy would lead t» the Wter conclusion ^^^ ^^.^^ ^^^ ^^^^^ development, bavin 

 It is in Australia itself that tne genu ,„,.„„..;„„ i„ number the 31mosc 



f^ 



iu M 111 ^^v.^^^^ rivaUing, or even surpassing in numoer iiie M,mu,u<, 



flourished beyond aU precedent, ^-^f^"^ ^ ^^^ ubiquitous A. fam^na, belong to 



of America. Of 393 species, 4 o° ^ ""=' = ^ ^ ^e„ local in North Au-^tralia; the 



the generaUy spread Gumm/er's but W ^^^ ^^_j^ subgenera purely endemic ; 



remainder are essentially Australia , ^anM^ and widely spread, 



and no less than 271 axe 2'^^l^ls bearing a .^nspicuous part in the arb<v 

 n+bpr« Innnl and defimte, the wlioie suuo A„«f«,iio and Tasmania. Oriiri- 



ahd fru 



,nd dehmte, .- "J^ ^ ^t of Australia and Tasmania. Ongi. 



tescent vegetation of ^"^ ' ^^7,^on, where it has left, as above men 

 ., . « nvTf, spa rene or Malayan TL^iui , .^ • .^ 



nating probably in some Mascarene ^ ^ ^ . ^^^,^ ^^.ith the greatest profusion, 

 .: ? L^^. ^o.v fftw traces, it has muitipuea 



tioned, such very few 



