DICOTYLEDONS 217 



SUMMAEY 



Owing to the enormous number of Dicotyledons the 

 task of arranging them systematically is a formidable 

 one, and it is unlikely that any arrangement yet pro- 

 posed can be considered final. Very much more infor- 

 mation is needed in regard to the development of both 

 floral and vegetative parts, as well as in regard to the 

 embryology, especially in the obscure and doubtful types, 

 before we shall have the data necessary for a satisfactory 

 classification. Their relation to the Monocotyledons is 

 also very uncertain, and a thorough comparison of the 

 lower types of Dicotyledons with these is very much 

 needed. 



It is generally admitted that the apocarpous Choripet- 

 alse, i.e. those with entirely distinct carpels, are the most 

 primitive. The simpler RauunculaceEe offer many resem- 

 blances to the apocarpous Monocotyledons, and it is possi- 

 ble that here we have a point of contact between the two 

 groups. It is, however, not at all impossible that the 

 Dicotyledons have had a multiple origin, and the possi- 

 bility of the derivation of the Piperinese, and possibly the 

 lower Centrospermse from monocotyledonous types like 

 the aroids, is not improbable. This would imply that 

 some, at least, of the apetalous Dicotyledons are prim- 

 itive types, and not reduced from petaloideous forms. 

 The occurrence of numerous apetalous types among the 

 oldest fossil remains of Dicotyledons also strongly sug- 

 gests their primitive character. If this view is correct, 

 it is quite possible that the Amentaceae and some other 

 Apetalae constitute a line of development entirely dis- 

 tinct from that of the petaloideous forms. 



