1890.] NATURAL SCIKXCES OF PIIILADELrnrA. 273 



foreign pollen. The large i)urple autliei's of the Cai).siciun are 

 drawn closely together and discharge pollen copiously over the 

 stigma. 



Lycopus Virginlcus L. — It cannot be said that there is any ar- 

 rangement of the ])arts of the flower suggestive of special ada])tatiou 

 to self-pollination, but the later flowers of the season, borne on the 

 stolons, are often pushed into the earth in the bud and expand 

 wholly beneath the surface, ])roducing seeds as abundantly as those 

 on the erect stems. No one will doubt, on carefully observing the 

 plants that the flowers self- fertilize. Dr. Gray seems to limit 

 the tuberiferous character of this species to a depauperate form 

 growing from Lake Superior northwards. The tubers are freely 

 borne by the typical form around Philadelphia, and they are often 

 as large as Kidney Beans. 



Humamtlis Virginiana L. — The arrangements for self-pollination 

 here are among the most interesting known to me. It is well-known 

 that the anther-cell opens by an operculate lid or valve. What 

 might be termed the hinge to this lid is on the upper part of the 

 stamen. When ready to expose the pollen, the lid previously point- 

 ing downward, becomes perfectly erect. The fiice which, previous 

 to opening, faced one point of the horizon, is turned to the opposite, 

 and the inside face becomes the outside. The pollen adheres to the 

 inside face of the valve, and so complete is this adherence, that on 

 the erection of the valve, every partof the pollen is withdrawn from 

 the cell. The erection of this valve is, however, accomplished in- 

 stantaneously, and the sudden jerk dislodges some of the pollen, 

 which is scattered around. Some falls back into the cell, some 

 over the adjacent stigmas, but a large portion still adheres to the 

 face of the lid, falling away gradually as moved by the wind or other 

 agency. 



It may be well to note here the order in which the parts of the 

 flower successively mature. As the four strap-like petals expand, 

 neither anthers nor stigmas have reached maturity, which seems to 

 be the work of the second day. As the stigmas become viscid, a 

 pair of anthers only expose the pollen. One anther, and one cell 

 only of each anther, bursts at a time, the other follows some time 

 during the same day. Before evening of that day, the four cells ex- 

 pose the inner pollen-clothed surfaceof the lid. The pair of stamens 

 that open first are those that are in front of the pair of styles — that 

 is, the pair which alternates with them. The next day the pair on 



19 



