iV£OTT7£J5— EPIPACTIS 



79 



Wheldon and Travis' diagnosis says : "Rhizome slender, far creeping". I searched 

 carefully for this but could never find it. I once thought I had done so, but the 

 supposed slender creeping rhizome turned out to be an underground shoot of grass 

 which had pierced right through the middle of the Epipactis stem! Later Mr Travis 

 wrote to me that he was now convinced that the plant had no creeping rhizome. 



E. dunensis differs from E. leptochila as follows : 



E. leptochila E. dunensis 



Root-system strongly developed. Very poor. 



Rhizome large, knotted, descending. A small irregular mass, very deeply buried 



at the base of the stem. 



Roots very numerous, rather tliick, fleshy. Few, slender, wiry, hard. 



Flowers larger, wide open; sepals and petals Sepals and petals short, broad, narrowing 



long, tapering gradually to an acute point. suddenly to an obtuse point. Flowers not 



opening wide. 



Epichile long, narrow, pointing straight Short, as broad as long, with recurved tip. 



forward. 



No pollen-globules faU into the flower. Globules of poUen fall into the flower. 



Seed-testa long and narrow, 8-10 times as Testa broader, only half as long, cells not 



long as broad ; cells narrow, compressed compressed, cell-walls very thin. 



sideways, cell-walls thick. 



E. dunensis resembles E. leptochila in the upper parts of the polhnia coming over 

 the upper edge of the stigma on to its viscous surface, and was placed under it as 

 a variety by Drs T. and T. A. Stephenson (J.B. p. 205 (1921)), before it was raised 

 to specific rank. 



Fertilisation. Mostly, if not entirely, self-fertilised. The pollinia are at first friable, 

 and httle globules of poUen fall on the sepals, petals, lip, and no doubt also on the 

 stigma, even in bud, effecting self-pollination. The main method, however, is as 

 follows. The pollinia emerge from the anther and lie in the cUnandrium or shallow 

 cup at the apex of the column beneath the anther. Here they swell and disintegrate, 

 and appear to become matted together by pollen-tubes, so that they cease to be friable. 

 The upper parts of the pollinia come over the sloping edge of the stigma on either 

 side, probably pushed over by the expansion of the pollen, and thus reach its frontal 

 viscid surface, resulting in fertilisation in the ordinary way. Little masses of pollen 

 become rooted to the upper part of the stigma by pollen-tubes, for on attempting 

 to remove them with a needle, it is found that they have to be moved backwards 

 and forwards several times before they become detached. As a rule there is no 

 rosteUum, but occasionally it is developed, and in one case, on touching the 

 rosteUum with a pencil, the pollinia came away, spreading themselves like a little 

 mat over the point of the pencil. There is therefore a possibility of cross-fertilisation 

 occasionally through attacliment of the pollinia to an insect visiting a flower for the 



