14 MEMOIRS FROM THE DEPARTMENT OF BOTANY OF COLUMBIA UNIVERSITY 
The form of the style is somewhat various. In most species this organ is very slen- 
der and filiform ; in many, however, it 1s more or less thickened below by a glandular 
tissue. In P. arguta and its relatives it is thickened near the middle and also glandular. 
In P. fruticosa (Plate 101) and a few Asiatic species, it is club-shaped, that is glandular 
at the top and in that species the stigma is unusually prominent and slightly 4-lobed. 
In all the others it is rather inconspicuous. 
In the tribe the style is more or less constricted at the point of attachment to the 
ovary and usually deciduous sooner or later. The point of attachment is very variable 
and evidently of great importance. In all species of Horkelia, Ivesia and most included 
in Potentilla, the style is inserted nearly at the apex of the ovary. (Plate 1, Fig. 6.) 
In Sibbaldia, Fragaria, Comarum, Potentilla Anserina (Plate 1, Fig. 7), P. Egedu, P. 
fruticosa and P. tridentata the style is lateral, and in Chamaerhodos and P. arguta and its 
allies (Plate 1, Fig. 5), it is attached to the ovary very near its base. In all cases where 
the style is terminal the seed is also pendulous and anatropous (Fig. 6); where the style 
is lateral the seed is ascending and amphitropous (Fig. 7); and where the style is nearly 
basal the seed is ascending and orthotropous (lig. 5). 
The number of pistils is very variable. In most of the members of the tribe, the 
pistils are more than 20 ( Plate 1, Figs. 9, 10, 13 and 14), in some very numerous, but in 
Sibbaldia, Chamaerhodos (Fig. 11), Ivesia (Fig. 8) and one or two Potentillae, they are 
from 3-15, in I. depauperata (Fig. 12) and I. sabulosa 2, and in T. santolinoides (Fig. 15) 
only 1 
RELATIONSHIP OF THE TRIBE. 
As stated before, the so-called gamosepalous part of the calyx is to be regarded as an 
expanded flower-axis. Having this fact in mind, it is easy to see that the flower of Po- 
tentilla does not differ very much from that of Ranunculus. The main difference is 
that in the latter, where the flower-axis is not expanded laterally, the stamens, petals 
and sepals are borne right under the pistils, while in the Rosacrar they are peri- 
gynous or epigynous. The organs of vege tation and the other floral characters of 
Ranunculus and Potentilla are so similar that an ordinary observer easily mistakes one 
for the other or regards them as belonging to the same genus. It is evident that the 
RaNuNCULACEAR and RosacraAn are very nearly related, and of the genera belonging to 
the latter, it is questionable which should be regarded as standing nearest RANUNCU- 
LAckAR, Potentilla or Geum, especially the group containing G. Rossii (R. Br.) Ser. and 
G. turbinatum Rydb. 
Through Waldsteinia, the tribe connects with the Dryapbear, through Fragaria and 
