Flowers of Cassia fistula — ^Neal 
85 
bent inward and are borne on filaments to 
10 mm. long. Because of their nature and 
position they serve chiefly as footholds for 
the wood bee. Next to them are the four 
starch-bearing, sterile stamens with filaments 
10 mm. long and anthers 5 mm. long. Three 
fertile stamens arise from the flower’s base, 
the filament basally being directed upward, 
then after 5 mm. suddenly turning backward, 
and finally bending in a large curve out of the 
flower, so that the 5X2 mm. anthers are to 
25 mm. from the flower base. 
Tischler states that he has nothing to add 
to this description and that like Knuth he 
saw the pollinating wood bee "milk” the 
starch-bearing anthers (my Nos. 4, 6, 8, 10) 
for food and get pollinated by the fertile an- 
thers (my Nos. 5, 7, 9), the pollen of which 
it then carried to stigmas of other flowers. 
The position of the style may be between 
stamens Nos. 5 and 7 or between Nos. 7 
and 9. 
ABERRANT FLOWERS 
The cause of the aberrations in tree B is not 
known. The four trees in our garden^ A, B, 
C, D, are about 12 years old. The 22 other 
trees examined are probably older. The gen- 
eral aspect of tree B is like that of the com- 
mon form of C. fistula. But the leaves are 
slightly shorter and the leaflets smaller than 
those of A, C, D; none, however, was mal- 
formed. According to Ridgway’s Color Chart, 
the "wax yellow” flowers of tree B are a duller 
yellow than the "lemon chrome” flowers of 
tree A, for example; this comparison, on be- 
ing analyzed, shows that the flowers of both 
B and A have the same hue, orange-yellow- 
yellow (25 per cent orange, 75 per cent 
yellow), but that "wax yellow” is grayer, a 
stage nearer neutral. Besides other differences 
of the flowers, some racemes though at first 
as large soon become small and rounded, with 
flowers clustered at the tip only, instead of 
remaining long and more or less obconical, 
as in the common form; other racemes are 
small and rounded from the beginning. The 
seeds look like those of tree A, for example; 
but the pods are far fewer, some are shorter, 
and others have constrictions. 
All these slight differences might well be 
within the range of the species. Hybridism 
does not explain the differences, since the 
regular shaped sepals, petals, and stamens 
found among the irregular shaped ones of 
tree B were much like those of the common 
form of flower of trees A, C, and D. 
Fig. 4. Stamens of common form of flower of Cassia fistula, front and back views. Upper row: mature stamens 
from open flower, anther cells open. Lower row: stamens from bud, anther cells not open. A and C, No. 2; F and 
D, No. 8; E and G, No. 1 {E shows No. 1 from two different flowers); F and H, No. 9. 
