1922] JURICA—UMBELLIFERAE 299 
families have a number of general characters in common. He also 
states that not a single anatomical character permits an absolute 
separation of the two families. After describing the details of this 
anatomical relationship, he concludes that the Umbelliferae, with 
the Araliaceae, form a continued morphological series and a very 
natural group. VIGUIER next shows that the Cornaceae are closely 
related to the Araliaceae, claiming that even good systematists at 
times take a single genus out of one of the families and put it 
into another. WANGERIN (80) and WARMING (81), however, would 
separate the Cornaceae from the Umbellales on account of the 
nature of the ovule. No less apparently conclusive is the account 
of WETTSTEIN (83) in regard to the inter-relationship of the three 
families forming the Umbellales. 
With this inter-relationship established, we can safely return 
to VIGUIER (78), who compares the Pittosporaceae with the Umbel- 
liferae and Araliaceae on the basis of the presence of secreting canals, 
for which reason VAN TreGHEM (77) has added this as a fourth 
family of the order Umbellales, and his studies upon the ovule of 
the Pittosporaceae confirm his view. 
This interpretation seems plausible, for the ovary of the 
Pittosporaceae is likewise bicarpellate, although the ovules them- 
selves are free and generally parietally arranged in two rows. 
The single integument, and the fact, as WETTSTEIN (83) has it, 
that the corolla is very often somewhat sympetalous (“‘Korolle 
manchmal etwas sympetal”), prove that at any rate the Pitto- 
sporaceae are out of place among the Rosales, and no doubt belong 
among the Sympetalae. The inclusion of the Pittosporaceae in 
the Umbellales by VAN TrecHemM likewise demands the transfer 
of the entire Umbellales to the sympetalous dicotyledons. Accord- 
ingly they are not to be viewed as an order closing up the Archi- 
chlamydeae line, the next of kin to the Myrtales. This is in spite 
of the fact that the Umbellales have some characters in common 
with some of the Myrtales, especially with Hippurus of the 
Haloragidaceae, such as epigyny, single integument, etc., for they 
stand too sharply apart from the rest. The position of families 
forming the order Myrtales certainly deserves a reconsideration, 
and this has already been done in part by SCHINDLER (61), who 
