82 NATURAL HISTORY OF PLANTS. 



The Pohjgalnccw have been considered so analogous to the 

 Tremandracea'. by some authors that they have treated the former 

 as only an irregular form of the latter. This is true especially as 

 regards the gynsficeum, which has in both groups almost always the 

 same number of cells, with an ovule whose various parts have the 

 same direction ; but the perianth is very different in the Treman- 

 dracea'^ and more analogous to that of certain Mahjacece [Lasiopetalece) 

 with which they have also been connected. It appears to us that 

 tlie Tremandracece and Poli/rjalaceai should be ranged side by side. 

 The latter show an affinity with the Geraniacecc and Lhiacea'.^ 

 from which they are distinguished particularly by the anterior keel 

 of their flower and by the organization of the ondroceum. They 

 have also, like the Linacew^ the ovules and seeds ofiheUiiphordtacew, 

 and in certain cases the same gynseceum. We shall obtain a good 

 idea of these relationships by comparing the Foli/e/alaeece, plants 

 with milky juice, with the irregular flowers of Pedilanthus on the 

 one hand, and on the other with irregular types of Chailletiea}^ such 

 as Tapura, whose resemblance to Moutahea is considerable. It has 

 also been said, and with reason, that there is a great analogy between 

 the Pohjgalacea' and the Sapindaeece ; the latter, however, are easily 

 distinguished by the situation of their disk, exterior to the ancboceum, 

 the symmetrical form of the latter differing from that of the Polyga- 

 lacew^ and by the du'cction of the different parts of the ovules when 

 they are indefinite in number. The Viohccea' could only be confounded 

 with the Xanl//op/i///lea' among the Polygaiaceo', they alone having 

 parietal placentas ; but the Violacece have an isostemonous andro- 

 ceum, dissimilar stamens when the flower is as irregular as in the 

 Xaiifhojti/Uetr, and, in a fi-uit which is often capsular, ai-illate seeds. 

 Finally the Krameriew by their resupinate flowers have some 

 analogy to the Legiiminosece with which they have often been con- 

 nected ; but the relationship is more in appearance than reality, and 

 we have seen that the Kramerias have a di-earpellary gyna^ceum 

 which renders all assimilation between the two groups impossible. 



of the climbinf^ species from warm regions. ridnca volubilis and of a Comesperma [Catncoma 



But few researches have been made in this way. Iiwidii). See Done, in Arch. Miis. (1839), i. 205, 



Gruegeu {in Bill. /<//. [1S.50], IGl), has indi- t. 10. — Ohiv. Sleiii. Dicot. 6. fievcraX Pnli/i/aliiccrc 



cated the peculiarities of the structure of Syeii- are cited as not havinc; medullary rays. 



