1886.] G. King— 'New Species of Ficus from New Guinea. 399 



ovarial cavities of the egg of an insect. On the other hand, a nnmber 

 of perfectly formed females, all well adapted for the reception of pollen, 

 are shut up together in a receptacle which contains neither male nor gall 

 flowers, and to which it must from a very early stage be very difficult for 

 pollen-bearing insects to get access. Yet each of the females situated 

 in such apparently disadvantageous circumstances bears a well-formed 

 embryo. No doubt the Blastopliaga or other Hymenopterous insect deve- 

 loped in the gall flowers, in some way, conveys the pollen of the males to 

 the perfect females imprisoned in the neighbouring receptacles. But al- 

 though one can understand that it is to the advantage of the insect to 

 enter the receptacle containing the gall flowers, since these afford it 

 such a suitable nidus for its egg, and that the mature insect in escaping 

 from these receptacles may, inadvertently carry along with it some of 

 the pollen which the anthers are then shedding, yet it is difficult to under- 

 stand how the pollen so removed is conveyed into the interior of the 

 receptacle containing the perfect female flowers, and how these females 

 are so universally fertilised by it. 



" This arrangement, by which the receptacles are practically diee- 

 cious, obtains, as I have said, in a large proportion of the species of 

 Ficus. There is, however, a group of species ( TJrostigma) in which it 

 does not obtain, and in which male, fertile female, and gall flowers are 

 contained in the same receptacle. In this group the difference in struc- 

 ture in the early stages between gall and fertile female flowers is very 

 slight, and in some cases I could find no difference whatever. And even 

 the lipe achenes of the fertile females are in many cases undistinguish- 

 able externally from the ovaries containing far advanced pupae, and it 

 is only by cutting them open that they can be recognised. As regards 

 the relation in this group of TJrostigma of the male flowers to the fertile 

 female and gall flowers, there are two types of arrangement. In one set 

 of species (of which F. Bengalensis and tomentosa are good examples) 

 the male flowers are comparatively few in number, and are confined to 

 a zone at the apex of the receptacle, just under the ostiolar scales ; while 

 in another set the male flowers are intermixed with the fertile female 

 and gall flowers over the whole surface of the interior of the recep- 

 tacle. 



" A third small group (Syncecia) has neuter flowers mixed with the 

 fertile females in one set of receptacles ; while the other set of recep- 

 tacles contain only male and gall flowers. And a fourth group (which 

 I have named Falceomorphe) has male flowers which, in addition to an 

 anther, contain an insect-attacked or gall pistil. These pseudo-herma- 

 phrodite flowers are confined to the sub-ostiolar zone, the remainder of 

 the receptacle being occupied by gall flowers : while perfect female 



