1878. ] On the Natural Succession of the Dicotyledons. 727 
Again, there are cases in which the lobes of the nominally 
gamopetalous corolla are so deeply parted that it becomes diffi- 
cult to determine whether they are at all united, and in which the 
union sometimes actually ceases to exist. Of this class might be 
mentioned at random Symplocos, Statice, Naumbergia, Anagallis, 
Chionanthus, Fraxinus. The last named genus is of special inter- 
est in consequence of the close general resemblance between the ash 
and maple families, which are, however, widely separated in the 
present system. ‘That beautiful climber, Boussingaultia baselloides, 
commonly known as the Madeira vine, is botanically related to 
‘the Portulacacee, but the petals are barely united at the base, 
for which reason it has been removed from its natural association 
and placed in the monopetalous division. 
Thus we find that while some botanists have preferred to main- 
tain well established orders intact by allowing them to embrace 
genera and species whose corolla would require them to be 
placed in different divisions, others have chosen rather to remove 
such anomalous genera to their appropriate divisions and if neces- 
sary to create new orders for them. Some, for example, leave Paro- 
nychia, Anychia, etc., in the Caryophyllacee, with which they are 
clearly allied, while others place them in the apetalous division 
near the Polygonacee with which they doubtless are also allied. 
But there are many cases which cannot be thus easily disposed 
of, as the Agutfoliacee, Euphorbiacee, Asclepiadacee, etc., in which 
cases the entire order is changed about from one division to | 
another, according as the author may think the preponderance of 
characters requires. 
Not only does it frequently occur that an order which cannot 
be divided contains genera representing two of the general 
types of corolla upon which the divisions are founded, but some 
orders, as the Primulacee, actually embrace all three of these 
types. In the order just named we find Maumbergia which is 
frequently polypetalous, and G/aux which is always apetalous, 
while most of the genera are monopetalous. 
These few examples, which might easily be extended, certainly 
show that the so-called divisions of the Dicotyle do not form a 
natural series. They rather indicate that they represent three 
parallel and co-ordinate series, in any one of which orders closely | 
corresponding may be named in one or both the others. Thus the = 
Caryophyllacee may be compared with the Polygonacee,the Acer- 
