70 



with some rapidity to the length of a few feet. In the Jour- 

 nal of Botany it is said to inhabit " old dykes about Buenos 

 Ayres, and to have green flowers which are remarkably fra- 

 grant, particularly in damp evenings." They have indeed 

 much the smell of Pergularia odoratissima, and almost their 

 colour, except that the centre is occupied by a white angular 

 coronet. They are about an inch in diameter. What makes 

 this rare plant a most interesting species is the distinctness 

 with which it exhibits the curious formation of pollen tubes, 

 and the singular phsenomena connected with fertilization, in 

 Asclepiadaceous plants. Before the flowers expand the hol- 

 low formed by the closed up coronet is dry, and all the parts 

 are in the ordinary condition ; but with the expansion of the 

 corolla appears an abundant secretion of watery matter, 

 which bathes and lubricates all the parts connected with the 

 column. At the same time the surface of the sides of the 

 column, over the cells of the anthers, becomes gradually 

 tumid ; shortly afterwards a substance like tow is seen pro- 

 truding from beneath the membranous apex of the anthers ; 

 it increases in quantity, covers over the apex of the stigma, 

 and eventually makes that organ, which in reality is quite 

 smooth, appear as if woolly. This tow-like matter consists 

 entirely of pollen tubes, which quit the pollen-bags on their 

 outer edge near their point of attachment to the arms of the 

 gland, distend the valves of the anther, and follow the course 

 of the membranous apex of that organ, which directs them 

 with unerring certainty to the stigma. I am not aware that 

 this modification of the plan of fertilization in Asclepiadaceae 

 has been before noticed; it has been stated, in such cases as 

 bave been previously examined, that, in this order, the 

 pollen tubes do not direct themselves to the apparent stigma, 

 but to the lower and under part of its discoid head ; here 

 however they manifestly pass upwards in the direction of 



that part through which fertilization occurs in ordinary 

 plants. "^ 



If we take CynancJium vincetoxicum as the type of the 

 genus Cynanchum, as seems most convenient, it will not be 

 possible to allow this remarkable plant to form a part of it. 

 Its convex two-lobed stigma is extremely difl^'erent from the 

 flat or concave angular one of Cynanchum ; and its tubular 

 coronet, which converges in suclf a way as completely to 



I 



\ 



