NOTES ON MASCARIONE OECfllDOLOaY. 291 



Anjracum eburneum and su'perham by outgrowths fi'om the labellum 

 itself, Avhich appear in the first-named in the form of two hiteral 

 tubcrck^s, one on each side, and in the other two of a single median 

 lanceolate fleshy plate. AVorthy of notice, too, is the space between 

 the lobes of the rostellum (fig. 2), seen best in flowers which 

 ure provided with a long slender curved spur ; the curve or curves 

 being often very sharp and trending forwards, an insect, in order to 

 get to the bottom of the spur, would probably be forced to move its 

 head against the edge of the anther-cap, detaching the latter and 

 getting the viscid gland or glands aittached to its head. The cone-like 

 rostellum is found in a few rare cases where the labellum is wound 

 round the column ; as it is very strong and curved upwards, it seems 

 admirably adapted to resist pressure from above, as of an insect 

 struggling to force its way down to the nectar contained in the short 

 spur. On the other hand, a forward projecting rostellum is not required 

 here, the space between labellum and column being very slight, and 

 totally filled by the insect's body during sipping of the nectar. I can 

 call to mind only two p'ants with this structure, Lidrostdclujs poly- 

 stachys, I'tchb., found in Madagascar, Bourbon, and Mauritius; and a 

 iiew species from lloJiiguez, closely allied to it, and I venture to 

 think tiiat they should be separated genericaily. The liguliform ros- 

 tellum is usually found where the entrance to the spur is wide; over 

 this it projects, often extending a considerable way beyond the spur on 

 to the limb of the labellum. 



(b.) In Listrostachys Pescatoriana. The flower of this plant (fig. 5) 

 wouid appear to be fertilised in a very unusual manner. Tae labellum 

 is clawed, and at the junction of limb and claw is the small orifice of 

 the spur. In consequence of the presence of this claw the anther is 

 evidently placed too far behind the spur for an insect to remove the 

 pollinia while sipping nectar ; but the small flowers are much crowded, 

 so that it seems possible for an insect, while sipping from one flower, 

 to draw out pollinia from another, and deposit them on the stigma of 

 a third. 



{c.) In Cynorchis (figs. 6 and 7). — The most striking point in 

 this Ophrydeous genus is the large, erect, fleshy rostellum (r), either 

 contiguous to the arms of the anther (fig. Q) or distant from them 

 (fig. 7) ; the anther-arms are very long, and the anther-case is 

 placed far back on the column, so that the caudicles are very 

 much elongated. The only use that one can see for this curious 

 rostellum is that it serves to keep straight the released but not yet 

 firmly-fixed pollinia, which, were it not for this provision, might run 

 the risk of coming in contact with foreign parts, and of being rubbed 

 off upon them and so lost, an accident to which, considering the great 

 length of the eaudicles, one might suppose these pollinia to be some- 

 what liable. Further, in C. fastiyiata each of the laterally-placed 

 etigmatic lobes (fig. 6, s.) may perhaps help in the same way as the 

 rostellum, each caudicle finding at its pollen-end an external wall, 

 and at its giand-end an internal one. On the other hand C. piirpuras- 

 cens (fig. 7) has its stigmatic lobe placed below the arms of the 

 anther, so that its eaudicles have only one wall, an internal one, 

 g.uarding their central portion from straying, A third species of this 

 i^enus from Mauritius ( C. Buryana, Ldl.) I know only from 



u 2 



