1922] JURICA—UMBELLIFERAE 301 
On the other hand, Hatter (30) would have a common 
“Umbellifloren” stock reaching from the Terebinthaceae, and giving 
rise to Cornaceae, and through these to two branches, namely, to 
Umbelliferae on one side and to Rubiales on the other. From this 
it is evident that, although these systematists disagree as to details, 
they agree upon the fact that the Umbellales are related to Rubiales. 
Recognizing ENGLER’s scheme of classification as the one more 
commonly and more widely accepted, and in view of the numerous 
facts, both from the taxonomic and morphological field, I suggest 
that the Umbellales be placed among the Sympetalae parallel with 
the Rubiales. Even if the viewpoint of WANGERIN (80) and 
WarMiInG (81) in regard to the Cornaceae should be proved to be 
correct and universally accepted, this would not affect the remaining 
families of the Umbellales, for then the Cornaceae would be trans- 
ferred to a position in close association with the Caprifoliaceae. 
Summary 
1. The floral development shows an acropetal succession of 
floral cycles, namely, sepals, petals, stamens, and carpels. 
2. In the genera in which the sepals are represented by mere 
calyx teeth or are obsolete, the calyx primordia also make their 
appearance, but fail to develop any further. 
3. The carpels are two in number and later fuse to form the 
ovary. 
4. Two anatropous ovules begin to Geysop 3 in each cavity, but 
usually the lower one reaches maturity. 
5. The hanging anatropous ovule has a single integument. 
6. The nucellus is very small, and the hypodermal archesporial 
cell is easily recognizable. 
7. The megaspore mother cell produces a perfect linear tetrad, 
as a result of two successive divisions. 
8. The embryo sac develops from the innermost megaspore, 
the three others aborting. 
9. A regular eight-nucleate and subsequent seven-nucleate 
embryo sac results from three successive divisions of the megaspore, 
followed 2 the fusion of the polar nuclei. 
