Dr. Schleiden on the Structure of the Ovule in Plants. 165 



dually together, that the difference alluded to is absolutely 

 incapable of being employed as a ground of division at the 

 time of flowering ; when the seeds are ripe it then indeed af- 

 fords a well-defined distinctive character. But since we have 

 genera which cannot be divided {Euonymus) in which this 

 double form occurs, such a character can in no case be made 

 use of to establish and justify a division, unless nature evi- 

 dently indicates it otherwise ; and indeed the less so, when, as 

 in Ranuncnlace(£, nature has set no value on the structure it- 

 self of the ovule, and when peculiarities otherwise most con- 

 stant within the limits of family are found to be among the 

 most variable. Of this nature is the number of integuments 

 of the ovule, which in Ranunculacece vary even in the same 

 genus. 



With an integumentum simplex there are, Thalictrum, Ane- 

 mone, Hepatica, Ranunculus, Ficaria, Caltha, Helleborus, Del- 

 phinium tricorne and chinense, and the Podophyllece, 



With an integumentum duplex there are. Clematis, Adonis, 

 Trollius, Isopyrum, Aquilegia, Aconitum, Prnonia, Delphinium 

 fissum, elatum, bicolor, consolida, Ajacis, and the Magno- 

 liacece. 



So great is the difficulty of examining most plants of this 

 family with reference to the original structure of their ovule, 

 which in general is no longer to be recognized even in the de- 

 veloped bud, that I will not assert that some error may not have 

 crept into the preceding enumeration (perhaps in Delphinium). 

 But if, as I trust, the greater part is correct, then the con- 

 clusion is justified — that the number of integuments, which is 

 of fixed constancy in most other families, here appears as a 

 totally variable and consequently secondary character, ac- 

 cording to which alone the family can neither be restricted 

 nor extended. 



An example of similar anomalies also occurs in the family 

 of the Aroidece. Here there is nothing constant in the forma- 

 tion of the ovule, but the integumentum duplex pertaining to 

 all Monocotyledons. We find moreover in this family ovula 

 erecta [Arum), pendula {Pothos), atropa [Sauromatum), he- 

 mianatropa [Meconostigma), anatropa {Calla), and even hy- 

 per tropa {Orontium aquaticum), Robert Brown united Typha- 



