1s] THE GROUPS OF ANGIOSPERMS 343. 
The series of Leitneriales with the Leitneriacee is doubtful, 
sce we may have to do here with a reduced type. If such 
ree once shown, then this family would come next to the 
famamelidaceze, near the Rosales. 
The Juglandales occupy a higher position, nearly always 
msessing a perianth, which in the pistillate flowers is coalescent 
mihthe carpels, giving rise to apparent epigyny. The Juglan- 
alesare further distinguished from the Myricales by chalazo- 
my, but whether this character is constant must yet be deter- 
mined. 
The Fagales show the same stage of progress as the Juglan- 
Ss. Coordinate with the Fagales are the Urticales in which 
the Ulmaceze precede the Moracee and Ursticacee because of 
‘Se occurrence of hermaphrodite flowers. 
Ih the Proteales the perianth is-sometimes green, but in the 
“ority of cases it is petaloid, but a differentiation of calyx 
Hes not appear. This series is easily distinguished from the 
Satalales by its single free carpel with ventral placentation. 
The Santalales, comprising the Loranthacee, Myzoden- 
traceze, Santalacez, Grubbiacez, Opiliacez and Balanophoracee, 
“re their Syneceum made up of three (more rarely one or two) 
pels, and for each carpel but one embryo sac is developed. 
, In the Aristolochiales the petaloid segments of the perianth 
a more or less coalescent, and the carpels are provided 
| indefinite number of ovules and an inferior ovary. 
‘tage Polygonales form, in certain respects, the frantic? 
ge the Centrosperme, though the latter is characterized by 
Presence of perisperm. 
‘thol though itis mentioned above that certain series are mor- 
- further advanced than others, and that on the other 
“ttain series as the Fagales and Urticales occupy the 
ei “orphological stage of advancement, yet each of nese 
thik is be regarded as an independent plant-group, which 
‘i €nse be derived from any of the others. Nor is there 
Wien to suppose that any of these series constitute a point 
| & te for the following series. 
