OF PICUS loXBDWHU 



tbe fluid from the cavity during maturation, it is probablp ,1 > .. 



tag exit of the fig-insects by hearing the gnawing of the »J« * ''*" ^ t »1™-* 



by tactile sensation of the vibration of tissue connect,, \ZtT?*i ««M« nl 



there in waiting, and, as soon as the fig-insects begin to em»T ♦ an, ' ,hey "" 



and carry them off. neerino' rlnwn int« +i,„ i . . " ' ,u '"'' 1"" to mi. 



and carry them off, peering down into the canal of exit and ' ""* '"'" Ceed to "■"• 

 it to secure insects which have not yet fairly emerged Sol re *' l " ng d w " ilrt 

 maturing receptacles is not excessive in relation to the number of ^ """ ,K!r 



are contented to carry off their prey to their nests on other treeTf ''"*" ' * '*"* 

 circumstances, there are none on F. Roxburgh*, due, no doubt mlv • t" A '"" 



nient strength and resistance presented by the leaves rendering them U !I 

 late, but mainly to the fact that the tree is not liable to be i„f«"! t ." 

 scale-insects, the presence of which is the ordinary determinant of the l,Li!l ' 



nests.^ The nests, during a great part of the year at all events, are 



i • t 



lucre cow Immimw 



it is only during the rainy season that young are to be found in 

 however, an excess of receptacles mature simultaneously, the ants enstru ,"','„ , .,' ',',' 



managing in a wonderful fashion to bend the large, stiff leave, and nmn . r i' 



by the usual tough, papery web used in nest building elsewhere, ami ,, { to 

 accumulate large numbers of corpses in these local larders 



is : )t 



110 



Sima rufonigra also utilises the fig-insects as a source of food-supply, but 

 - so constant or methodical in its attendance as the previous species is, due 



doubt, to the great abundance of the latter, and the ferocity with which any inter! rem 



with its rights is resented. In fact, as a rule, ants of this species are only found on tree 

 not visited by (Ecophylla, or only after the latter has left the receptacles on tl 





cessation of emergence of fig-insects from them. In the latter case they frequently enter 

 the receptacular cavity to pick up the bodies of insects which have died without emenrinir 



Conclusion. 



It remains now to consider certain points regarding the relation which the presence of 

 the fig-insect holds to the fertilization of the receptacles of F. Eoxburghii. There can be 



little room for doubt that the phenomena indicate that, while the development of embryos 

 in the female receptacles of the tree is essentially connected with the access of the 

 insects to the receptacular cavity, it is yet normally independent of the introduction 

 of pollen by their agency. The fact that the access of a single insect or of a pair of 

 them only is sufficient to determine the development of ten or twelve thousand embryc , 

 is in itself almost conclusive against the occurrence of any ordinary process of pollination. 



The obstacles through which a passage has to be forced ere the receptacular cavity I 

 reached are of such nature and amount as to render it almost inconceivable that j >llen 

 should be introduced in sufficient quantity (Plate III, figs. 1, 2), and there is at the 



3 time an absolute want of evidence to show that such introduction takes place. 



sam 



have carefully examined very many receptacles at various periods shortly after 

 access of insects to the cavities, and have never been able to detect any evidence of 

 general distribution of pollen over the stigmatic surface. Examination of individual 

 flowers has jriven like results: in most cases it has been impossible t> find any pollen 



within the receptacle or cavity, and in tbe few cases in which any was found it 

 was represented by one or two shrivelled grains adherent to the corpses of msec . 



must be borne in mind, too, that if .we accept the hypothesis that tiie develop- 



It must be borne in mind, too, that if .we accept the hypotnesi* »» — — ^ 

 ment of the embryos is due to ordinary processes of pollination, we must assume no 



