412 William Russel Dtidley 



accession and quickened movement of the earlier waves of the 

 flood. The anther opens near its lower obtuse end, rips up- 

 ward along the depressed median line with a quick movement, 

 and exposes the numerous pollen filaments, lying parallel and 

 obliquely placed in each pollen-sac. The masses are white, 

 silky, and appear somewhat spirally twisted. The slow ad- 

 vance and recession and the sudden lashing of the swell carry 

 the long locks of sea-grass with them, throwing the spadices 

 back and forth, and dragging them roughly on one another 

 and on the leaves. 



The pollens average i millimeter in length by .004 to .005 

 of a millimeter in breadth. They are slightly flattened and 

 broadened at the extremity (Fig. D., pn.), and some are enlarged 

 toward the middle. Each pollen filament when first exposed is 

 protected by a layer of air, and a cluster of them loosened from 

 the pollen sac springs immediately to the surface of the water, 

 while the filaments repel one another sufficiently to form at once 

 a silvery arachnoid film, perhaps a centimeter across. These 

 are never abundant, but they float hither and thither with the 

 water, and among the doubling and swinging pistillate plants. 

 At the lowest stages of the tide the films of pollen could easily 

 be thrown upon the half-exposed pistillate spadices, and would 

 adhere to the protruded stigmas, as they were observed to do 

 in the aquarium. Unquestionably this is one mode of pollin- 

 ation. 



When the pollen has been exposed for hours in the open 

 sac, it does not necessarily rise to the surface but floats in the 

 water where it can be more readily carried to the usually 

 submerged stigmas. Clavand, (Actes d. 1. S. Linn. d. Bor- 

 deaux, T. II, 1878,) describes this as the mode of transference 

 in Zostera marina, and mentions no others ; but the young 

 pollens of the Pacific coast Zostera spring to the surface, 

 exactly as do those of Phyllo.spadix. I have examined mi- 

 croscopically both the floating and the submerged pollens of 

 Phyllospadix, taken in the conditions above described, and 

 found in both the natural streaming of the protoplasm. Both 

 would presumably be capable of effecting impregnation at the 

 time. The protoplasm was observed streaming in a filament 



