LIFE OF PLANTS. 26 1 



We thus perceive, that what are called the Monocotyledonous 

 plants have no proper cotyledons. As little do wc find polv- 

 cotyledonous plants, from the number of Mhicli the Pine 

 tribe, as we have remarked above, must be struck out. If, 

 in the case of dicotyledonous plants, the seed-lobes are di- 

 vided or cleft, as in Erodium, Canarlum^ and Lcp'ulhim, 

 tliey are fundamentally still but two, and all perfect plants, 

 reckoned upwards from the Polygoneae, must iluis be con- 

 sidered as dicotyledonous. 



387. 



In the study of nature, we must accustom ourselves to sus- 

 pect that in all bodies there are transition forms, and never 

 to believe that one and the same type remains without change. 

 There are thus transitions from cotyledons to leaves (in the 

 Liliaceae) ; transitions from albuminous substance to cotvle- 

 dons, and even to the root of the embryon. There are transi- 

 tions from leaf-stalks, and even from brandies, to leaves, as in 

 the Cereag and Acaciae, (181.) ; transitions from leafy appen- 

 dages to leaves in the Cistae ; from leaves to bractea? and to 

 the calyx ; from the calyx to the corolla in the Liliaceae ;. 

 from petals to filaments {Pancratium)^ and to nectaries {Con- 

 iortcE.) Nay, in the Canneae and Orchideae, the opposite 

 parts of fructification are so run into each other, that even 

 here transitions must be admitted. 



388. 

 When we are examining the evolution of the parts of 

 plants, we must further attend to the law of nature, which 

 was before stated (183.), namely, that simple forms always 

 j^recede those that are complex and subdivided. In imperfect 

 plants, where the embryon resembles a line or a point, and is 

 unevolved, the expansion of it takes place in pai-allel sur- 

 faces, when the plant unfolds itself, (290.) A more complex 

 expansion takes place when this complexity was prefigured in 

 the manifold subdivisions of the parts of the seed, and when 

 a more powerful crowding towards the first knot is evident, 

 i'rom ihis law (183.) that the more early forms are always 



