92 ME. F. ^Y. KEEBLE ON 



flowers of these two species are more regular than those of most other Cingalese 

 Loranths. 



Lorantlms Qardneri, Thw. : flower when open is regular, the stamens form a ring, the 

 corolla splits to 5 lobes, recurved about middle (fig. 6, PI. X.). 



L. neelgherrensiSi Wight & Arn. (fig. 1 1?, PL X.) : the corolla-lobes are very long, 

 " lobes much longer than tube, reflexed above the middle," and " one or two of the lobes 

 at times combined for half their length " *. By this fusion of several lobes, tbe flower, on 

 opening, has a slightly zygomorphic appearance, inasmuch as, in the reflexion of the 

 lobes, the space between the two which are not fused forms a sort of throat (fig. 1, PI. X.), 

 all the lobes being bent back, away from this throat. A further result of this mode of 

 opening out of the flower is that the stamens, from forming a ring round the style, 

 come to form a row^ behind it. 



In many of the Cingalese species a slit, similar in origin to that in the flower of the 

 last species, occius in the corolla-tube, w^hereby, at the time of opening, the upper part 

 of the tube by growth of its inner surface opens out laterally, so that all the five lobes, 

 whose inner surfaces also at the same time grow more rapidly than their outer, come to 

 stand in a row, and the stamens also w hich arise from the bases of tlie lobes similarly 

 stand side by side. 



The slit is well marked in Lorantlms Sciirrula^ Linn., L. biiddleoides, Thw., L. tomen- 

 tosus, Ileyne (fig. 1, PI. XI.), L. cuneatus, Heyne (fig. 3, PI. XI.), X. suborblcidaris, Thw., 

 L. longiflorus., Desr. (fig. 11, PL X.), L. lonchiphjllus, Thw^, L. sclerophyllus, Thw. 



The effect of the slit is, as is illustrated in the figures [cf. fig. 1 5, PL XI.), to turn the 

 regular flower into what might perhaps be called a physiologically zygomorphic one, by 

 enabling the upper intact part of the tube to flatten itself out in the way just 

 described. 



A similar opening out of an originally tubular corolla occurs, as is well know^n, in 

 some species of AnigosantJios, LabilL, a genus of Rgemodoraceae. 



In enquiring into the significance of these slits it must be remembered that, as other 

 observers have already shown, these tube-flowered Loranths are bird-fertilized f. My 

 own observations confirm this, for in Ceylon the common honey-sucker, a species of 

 Nectarinia, is always to be found, especially in the early morning, visiting these 

 fl.owers. 



I shot some of these birds which were busy in a Lorantlms bush and found their beaks 

 covered with pollen. Whether other birds also act as carriers of Lorantlms pollen I 

 could not determine. Now birds are less precise in their methods than butterflies, and 

 the pollen-carriers— their beaks — are much larger and by no means symmetrical. By 

 the spreading slit or throat a bird's beak has ample space to reach the nectar w^hich fills 

 the bottom of the tube. Thus the natural slit saves the flower to some extent, but not 

 wholly, from being torn. Further, the arrangement of the stamens side by side, rendered 

 possible by the opening out of the part of the corolla-tube above the slit, has the 



* Hooker, Fl. lirit. India, vol. v. p. 216. 



t Cf' Wallace, 'Darwinism:' '* In Australia and New Zealand . . . Loranthus {&c.) . . . are cross-fertilized by 

 birds " (p. 320). 



