1898] THE GROUPS OF ANGIOSPERMS 349 
The Contortae are by no means to be sharply distinguished 
from the next comprehensive series, Tubiflore. It is simply a 
question of differing tendencies of development. One sub- 
_ sties, the Oleineze, with the families Oleaceze and Salvadoracee, 
sometimes still possesses distinct petals. In the second sub- 
vries, the Gentianineze, belong the Loganiacea, Gentianacee, 
Apocynacee, and Asclepiadacee, the first of which comprises 
genera which show characters common also to the three other 
_ families, but also contains genera which lean more toward the 
Tubiflore and Rubiales. The Loganicee may thus represent an 
older type, from which the other families of Gentianinee, and 
_ perhaps the Rubiales, have branched off. 
The largest and most difficult series is the Tubiflorz, within 
_ Which numerous families are separated from one another only 
by very poorly defined characters. Advancement takes place 
_ ftom those families with actinomorphous flowers and several- 
_ wuled carpels to those with zygomorphous flowers and few- 
/ wuled carpels. But ina part of the Tubiflore the tendency to 
“aw-formation especially prevails, and in these the carpels con- 
—-Sinonly a few, often only two ovules. In close connection with 
' these Tubiflore are the Hydrophyllacee and Convolvulacee. 
Therefore, I distinguish first the subseries Convolvulinez, with 
the Convolvulacese and the Polemoniacee. The Borragininee 
: ‘oMstitute the second subseries, of which the Hydrophyllacex 
4 “a capsules and may be regarded as the starting point for 
iy Borraginacewe which possess claw-fruit. The subseries 
_Stbenine joins on to the Convolvulacee by having in the 
: ie esacece and the Labiate the ovule with the micropyle 
“ted downward. Then come the numerous fami 
: x themselves about the Solanacee and Scrophulariacee in 
Mich placente with numerous ovules mostly occur. These | 
Between 
p "* gtouped together under the subseries Solaninee. 
“thin and the Convolvulinee are the Nolanacee. The om 
ce belong near the Solaninex, yet owing to the great num 
oe of peculiarities of the Acanthacee, and in spite a 
“have i common with the Bignoniacee and Scrophulariacex, I 
ntured to separate them as a subsection. 
