Sapindales.] 



POLYGALACE^. 



377 



pules, of some supposed difference in the relative position of the lai'gest petal, the 

 anthers openuig longitudinally, and the presence of some fleshy glands between the ovary 

 and stamens ; I cannot, however, concede anything like ordinal importance to these 

 circumstances. Trigonia may be regarded either as an approach on the part of Milk- 

 worts to the Sapmdaceous structui-e, as is indicated by the longitudinal dehiscence of 

 its antliers, the greater symmetry of its flowers, and its 3-valved fruit ; or as actually a 

 member of the Soapworts, approaching Milkworts. The supposed relation between it and 

 Spindle trees or Leguminous plants, which M. Cambessedes suggests, appears to be a very 

 slight indication of analogy. — See Fl. Bras. Mer. v. 2. p. 112. 



Krameria has much higher claims to separation. The whorls of its flowers are so 

 completely dislocated that it is difficult to determine 

 the relative position of the parts ; there is no 

 trace of the quasipapilionaceous structure generally 

 characteristic of Milkworts, its ovary is imperfectly 

 2-celled, and it is said that no albumen exists in its 

 seeds. Certainly these ai'e points of moment. 

 Nevertheless, its definite hypogynous stamens, 

 porous anthers, misymmetrical flowers, definite 

 pendulous o"\^les, bur-like fruit, which resembles 

 that of Salomonia, and in some degree its habit, are 

 conformable to the ^lilkwort structm-e ; and in the 

 absence of all trace of the existence of other genera 

 approacliing this kind of organisation, it seems 

 expedient to regard it as a mere exception to the 

 usual structvu'e of an Order whose general condition 

 is in many respects very anomalous. It, too, may 

 be regarded as assisting to bring into contact the 

 Milkworts and Soapworts, for Kramei'ia cytisoides 

 has ternate leaves. 



Soulamea is another instance of the elevation of 

 a solitai'y genus into a Natural Order. This is a 

 Molucca plant, also without albumen in its seeds, 

 and having a regular trimerous flower with 2-celled 

 anthers. It may perhaps be considered as an 

 instance of the usual irregular flower of Milkworts 

 assuming a regular type. 



Lastly, of the genus Moutabea, promoted by Endhcher, who stationsitnear Storaxw^orts, 

 regarding it as monopetalous, it may be said mth tolerable confidence that it has not a 

 single featm'e that can justify its separation from Milkworts. Like Xanthophyllum its 

 petals are equal-sized, and as for their adhesion into a tube, that is no more than what 

 occm's in all the Polygalas, whose stamens hold together parts which under ordinary 

 cu'cumstances are distmct. So entirely, indeed, does Moutabea agree with Polygala, that 

 it even has its eight 1 -celled anthers opening at the point, in combmation with a 5-petalled 

 corolla. The ben-ies of Moutabea longifoUa are said to be eatable ; so are those of 

 Mundia spinosa. A tubular calyx exists in Moutabea, but that will hardly be insisted 

 upon as a ground for forming it into a Natural Order. 



Most of the genera are hmited to one or two of the five parts of the globe ; thus Sa- 

 lomonia is only found in Asia, Soulamea in the Moluccas, Muraltia at the Cape of Good 

 Hope, and Monnina and Badiera in South America. Comesperma is found both in 

 Brazil and New HoUand, and, what is very remarkable, there is in the former country 

 a species of the Cape genus Mundia. Polygala itself occurs in four of the five parts ; 

 under the torrid zone and in temperate chmates, at Cayenne, and on the mountains of 

 Switzerland ; it is, however, very miequally distributed. This genus inhabits almost 

 every description of station — dry plains, deep morasses, woods, mountains, cultivated 

 and barren soils. Comesperma is only known in Brazil and Australia. Monnina and 

 Krameria inhabit open places in the temperate parts of South America. 



Milkworts offer, as has been stated, considerable diversities of structure, and there- 

 fore, as might have been anticipated, the pm-poses to which they are applicable, are by 

 no means uniform. The greater part are bitter, and in their roots milky. To this cate- 

 gory may be referred the following cases. Polygala amara is a European perennial, all 

 the parts of which are extremely bitter; it is much extolled in pulmonary complaints 

 and spitting of blood. P. vulgaris and major have similar uses, but are inferior in energy, 

 A strong bitter taste pervades all the parts of Polygala rubeUa, a North American 



Fig. CCLXIV. 



Fig. CCLXIV.— Krameria cistoidea.— Hooker. 1. an expanded flower; 2. a diagram, showing the 

 relative position of the parts ; 3. a stamen ; 4. a perpendicular section of an ovary. 



