ORDER LEGUMINOS^E, OR PEA TRIBE. 425 



their stamens also are spreading and separate. A degree of 

 that irregularity in size, however, which is so striking among 

 the Papilionacece* is here still evident; some of the petals or 

 stamens being larger than the remainder. Few of this group 

 are ever seen in this country ; but in foreign climes they are 

 very abundant. 



590. The third division of this order comprehends those which 

 have flowers formed upon the plan of the last, that is, not 

 being papilionaceous, whilst the number of stamens is much 

 greater, being triple or quadruple that of the sepals. The flowers 

 are extremely minute, and grow in compact clusters ; the stamens 

 have very long, slender, and separate filaments. This is the 

 structure of the division which includes the Mimosas, amongst 

 which is the Sensitive Plant. Many of these have a very elegant 

 appearance, the clusters of flowers which they bear being 

 numerous, and often presenting gay colours. 



591. Besides the peculiar structure of the fruit, in which 

 these subdivisions all agree, the entire absence of a separate 

 albumen, so that the cotyledons are fleshy, and occupy with 

 the embryo the whole interior of the seed, is a character of 

 great importance, which prevails through the whole group, and 

 shows it to be a natural one. The number of species already 

 known is very considerable, certainly not less than 4000 ; and 

 many more must remain to be discovered. The order is diffused 

 over the whole habitable globe. Some species of it have a very 

 extensive range ; whilst others are restricted to particular coun- 

 tries. The Papilionaceous division with united stamens contains 

 most of the former ; but these appear to flourish best in tem- 

 perate and moderately warm latitudes, diminishing in number 

 towards the poles, and giving place to the Mimosa and Cassia 

 tribes nearer the equator. 



592. The properties of the different species of this order are 

 so various, that it is difficult to enumerate them all. Their 

 difference may also appear to invalidate the principles formerly 

 laid down, respecting the correspondence between structure and 

 properties (, 479) ; for, whilst some of them seem to be remark- 

 able for nothing but the large quantity of tasteless gum which 



