1893.] natural sciences of philadelphia. 303 



Structure of Florets in Bidens bipinnata. 



"Watching closely, on August 28, the anthesis of Bidens bipin- 

 nata, precisely the same phenomena were observed as is recorded of 

 Heliopsis Icevis. The style pushes up the staminal tube, starting at 

 night-fall. During the next day the pollen is pressed through the 

 divisions of the incurved staminal appendages, the style branches 

 protrude the following nightfall, the staminal tube retires the next 

 day, followed the following day by the retreating style. 



In the arrangement of the floral parts there seems more definite- 

 ness than in many Composite. Normally there are five outer 

 "involucral bracts" with foliaceous scales, — and these spread, some 

 time before the rest of the flowers open. But sometimes only three 

 of this verticil spread, in which case the other two remain erect 

 and are not distinguishable from the inner series of involucral 

 scales. Cutting away these five carefully we find no sign of an 

 axillary bud at the base of these bracts. Then we have another series 

 of five (seven in appearance, where the two lower bracts have become 

 " scales "), at the base of these five we find axillary buds. By 

 examining a number of heads, it is soon seen that these axillary 

 buds develop to florets. The buds develop equally in the axis of 

 each scale.wben they rest, apparently for about a day, and growth 

 is resumed, when, strange to say, one leads off in rapidity of 

 growth, outstripping the rest. Occasionally another will nearly equal 

 it, but two at least exhibit a veiy puny growth. These form the 

 ligulate rays, and in consequence of the irregularity noticed, there 

 are often only two fully formed ray florets, the others remaining 

 beneath the outer series of scales in an independent state. Inside 

 of this series, another series of five scales appear, the axillary buds 

 of which develop to fully formed disc florets. The next series of 

 five, however, seldom all perfect, — usually only two. 



Taking the theoretical structure of the flower to be on the plan 

 of five, we look for indications of this numeral in the ovarium. I 

 have, however, never found traces of five awns. Four are usually 

 present ; but one or two so weak that unless searched for, two only 

 might be supposed present. The weakest is at the base of the scale 

 and the next weakest opposite. The two stronger are in the lateral 

 positions fronting the scales, and give eventually the bidentate char- 

 acter to the akene. At times there are only three teeth. We may 

 probably assume, in the theoretical conception of the floret, that 

 five "involucral scales" primarily formed the calyx which becom- 



