x] ZOSTER A AND HALOPHILA 129 



scope dans la maison d'un ami, a quelques centaines de metres de 

 la plante, et je resolus de n'en point partir que je n'eusse de- 

 couvert, si c'etait possible, le mode de reproduction de Zos- 

 tera....l\ m'importait de ne pas rester plus longtemps dans une 

 incertitude qui commen9ait a me peser." On a favourable day, 

 hot and absolutely still, he went out in a boat and examined 

 some flowering plants. The three conditions in which the in- 

 florescences were found proved that cross pollination is ensured 

 by protogyny. Some were still enclosed in the spathes, with 

 the anthers intact ; others showed stigmatic branches, ready for 

 pollination or recently pollinated, emerging from the spathe, 

 while the stamens were still enclosed and not completely ripe; 

 in others again the stigmatic lobes had all fallen, while the 

 anthers were exposed, and either all empty, or the lower ones 

 empty and the upper ones in the act of dehiscence. The anthers 

 were seen to open, and eject the thread-like pollen which formed 

 a floating cloud. In pollen-grains, which had just been expelled 

 from the anther, an outgrowth was observed at a little distance 

 from one end. When pollinated stigmas were examined, it was 

 noticed that these outgrowths, which were, in fact, young pollen- 

 tubes, were forcing their way into the stylar tissue, between the 

 cells whose walls were becoming mucilaginous and separating 

 from one another. The pollination of Zostera is scarcely possible 

 except in still water, as any movement would carry the pollen 

 completely away from the scene of operations. 



The best-known genus among the marine Hydrocharitaceae 

 is Halophila, three species having been investigated in detail 

 by Bayley Balfour 1 and Holm 2 . Bayley Balfour himself col- 

 lected his material of H. ovalis, (R. Br.) Hook. fil. (H. ovata, 

 Gaudich.) and of H. stifulacea^ (Forsk.) Asch. on the reefs 

 surrounding the island of Rodriguez east of Mauritius. 

 H. ovalis (Fig. 87) grows on spots just uncovered at full ebb 

 tide, while H. stipulacea prefers localities where it is always 

 submerged and subjected to a constant current. The rhizomes 

 are creeping, and produce numerous long filiform rootlets 

 1 Balfour, I. B. (1879). 2 Holm, T. (1885). 



