216 



b5' coming- into contact with tlie pollinateil l)ack of a passing Antbopbora 

 in searcli of lioney. There does not seem to lie any provision for the self- 

 fertilization of P. Poitstciiioii. 



The longer of the dimorphic pistils of iJerteiisiu Vinjuiica (L) D.C. are 

 of the same length as the stamens and may be self-fertilized by contact 

 with the dehiscing anthers. The shortest of the other form do not reach 

 beyond the end of the narrow tube, and are fertilized by honey-bees. 

 Honey is secreted at the liase of a tube 2." mm. long and is further pro- 

 tected by a puliescent ring 2 mm. aljove the receptacle. No insect was 

 found on the flowers that could reach the honey in a legitimate way. but 

 a liig bumble-bee was seeu on the corolla making slits in the tulje just 

 al)ove the pultescent ring. Through the opening the tongtie of the bee 

 Avas inserted and the honey reiuoved, with ease, as it passed rapidly from 

 one flower to another. 



A calendine poppy. Stiiloplioniin aiiilnjlhiin (INIich.) Xutt.. under ctilti- 

 vatiou came into bloom April 23. early in the forenoon. At 3:40 p. m. 

 the petals began closing and by sundown were completely folded over the 

 stamens. Although it was raining the next day the petals under obser- 

 vation again opened in all their golden splendor. It was not clearly evi- 

 dent that the stamens of this plant were proterandruus. though the stigma 

 gi-enlly iuei-eased in size after the bud had opened. Usually the flowers 

 did not wither under two days. Small bees Avere noticed crawling on the 

 flowers, a single honey-bee was seeu collecting pollen, and it is prol)- 

 able cio-!s-fertilization was the result of ibeir movements. Flowers pro- 

 tected by a net from insect visitors produced capsules of the normal size, 

 well filled with seeds. 



In July it was noticed that wliile the calendine poppy was ijroducing 

 an abundance of seeds none could be found on the ground under the 

 plant. The seeds of a dehiscing capsule, which were placed in a heap 

 on a small stone, all disappeared l)y next morning. When it was recalled 

 lliat ants are known to carry small seeds into their nests they were sus- 

 pected of carrying them away. This inference seemed probal>le. as the 

 seeds were provided with a fleshy crest on one edge whicli an ant could 

 grasp. At last a common Iflack ant. about 6 mm. long, was seen with a 

 seed in its mouth and watched until it disappeared in a round hole. Later 

 ; i ant was followed to anotlier hole. Tne moutli to these holes was level 

 with the surface of the ground and not through tlie usual hillock of sand 



