270 TAKING UP OF POLLEN BY INSECTS. 



up over a hollow labellum. Near the apex of the column is the anther, and 

 lower down the rostellum, whilst below the rostellum the column is deeply 

 excavated. The edges of this pit are fleshy, and are prolonged into two curious 

 processes resembling a pair of horns. These processes are curved and tapering, 

 and are inclined obliquely forwards and downwards. In most species, including 

 Gatasetum tridentatum (here represented), the horns (or antennae, as they are 

 sometimes called) cross one another diagonally (see fig. 275'). Each horn, 

 originally a ribbon-like lobe, is rolled up lengthwise so as to form a tapering tube. 

 The substance of both horns passes without any definite line of demarcation 

 into the tissue of the rostellum above. Although this tissue has been examined 

 with the greatest care, nothing special has been found in it to account for the 

 extraordinary irritability which it exhibits. It has been ascertained by experi- 

 ment that any pressure on the lower extremity of the horn acts as a stimulus, and 

 that this stimulus is at once transmitted upwards through the cells of the tissue to 

 the part of the rostellum which forms the viscid disc. The slightest touch applied 

 to the tip of one of the horns is instantly followed by the rupture of the tissue 

 which has hitherto retained the viscid disc in position, and by the consequent 

 liberation of that portion of the rostellum. The viscid disc had, however, in its 

 turn served to keep a curved elastic band which attaches the disc to the pollinia on 

 the stretch, and in its proper position (fig. 275 '), so that when the disc is set free 

 the band flies up and straightens itself up with a jerk. The viscid disc and the 

 pollinia are torn from their recesses by the recoil of the band, and are carried with 

 it in an ample curve away from the column, which till then has served as their 

 common base (fig. 275 *). During its flight the viscid disc goes first, and it naturally, 

 by sticking to some object in the way, brings the pollen-masses to a standstill. 

 From the time of its being shot ofl^, the band connecting the disc and the pollinia is 

 quite straight (see figs. 275 ^ and 275 ^). 



The expulsive apparatus exhibited by most species of Dendrobium is altogether 

 difierent. In the species selected for illustration — viz. Dendrobium fimbriatum 

 (figs. 275^ and 275^) — the column is capped by an anther in shape like- a bell. 

 The anther is septate, and contains in its loculi pollen-masses, which since they 

 are unconnected with any viscid disc, are therefore liable to fall out of the anther 

 in certain positions of the latter. The anther is borne by a slender subulate 

 filament, to the extremity of which it is articulated in such a manner that a 

 gentle push is sufficient to set it rocking. When the flower first opens, and before 

 it has yet been exposed to any contact, the bell-shaped anther rests mouth down- 

 wards on a notch at the top of the column, and is held in that position by two 

 tooth-like processes to the right and left of the notch (see fig. 275^°). A push 

 administered from the front displaces the anther and causes it to fly back, whilst 

 the pollinia contained in it are simultaneously expelled (see fig. 275 ^^). The 

 pollinia being unfurnished with viscid discs (fig. 275^^), it is not quite evident 

 how the insects which supply the necessary stimulus on their visits to the flowers 

 get loaded with the pollen, There is, however, every probability that the expulsion 



