102 ME. F. W. KEEBLE ON 



end of the liypoeotyl to penetrate the host. Be this as it may, the character of producing 

 aerial roots at an early stage is often of great use to seedling Lot^anthus loniceroides 

 developing on unpropitious branches, since the aerial roots, rapidly elongating and 

 sending out suckers at intervals, are guided by negative heliotropism to the older part of 

 the tree, and so stand a good chance of happening upon better nourished branches. 



Some interesting cases in which the seed was not so normally and conveniently placed 

 are given in woodcut 4 a, b. It may here be mentioned that unattached hypocotyls 

 become irregularly curved, and the heads, instead of becoming funnel-shaped, send out 

 from their distal surface irregular white projections. 



Loranthus capitellattis. — Thwaites suggested that this is probably only a variety of L. 

 loniceroides. Sir J. D. Hooker * supports this view. The fruit differs in some respects 

 from that of X. loniceroides (cf. fig. 11, PL XI., and fig. 4, PI. XL) ; but if the seed and 

 mode of germination may be used as evidence, there can be little doubt that Thwaites's 

 view is correct. The germinating seeds of these two species are so alike that it is almost 

 impossible to distinguish them {cf. fig. 12, Pi. XL, with fig. 6 d, PI. XL). 



Fig. 4. 



^^®- * "> •* IB March 3 



A 



L. Gardneri resembles L. RooJcerianus in that it does not throw out a hypocotyl. 

 The viscin is mainly attached to the head end (hypocotyl), and so this, and this end only, 

 is firmly fixed to the host — hence the need for a long hypocotyl is avoided (fig. 7, 

 PI. X.). Fig. 8, PL X., shows a seedling germinating on a leaf of wild Durian — interesting, 

 inasmuch as in this unfavourable position a very long aerial root has been thrown 

 out and is growing along the surface of the leaf. 



L. neelgherrensis. — On germinating, the seed, which becomes firmly attached by its 

 viscid layer (especially towards its base), rapidly protrudes its minutely lobulated 

 hypocotyl, which at once curves (fig. 2, PL X.). There is in this species no distinct 

 terminal suctorial disc, though, in the rounded free end, the green colour is replaced by 

 white, marking the place whence the *' root " will issue. 



As in other tropical Loranths, the seed commences to germinate directly the fruit is ripe, 

 and the hypocotyl often bursts through the apex of the fruit-coat, e. g. figs. 3 and 4, PI. X. 



If the free end of the hypocotyl soon comes in contact with a branch, this end 



* Hooker, Fl. Brit. India, vol. v. p. 221. 



