1894- THE GALL-MAKING INSECTS OF AUSTRALIA. 113 



In Opisthoscelis the adult female has the segmental divisions well 

 defined, and though both the fore and middle legs are wanting there 

 is a well-developed hind pair, in which the tarsal claw is produced 

 into a long thread-like joint sometimes much longer than the whole 

 coccid. 



Another remarkable genus, Cylindrococcus, forms scaly fruit- 

 like galls (which are frequently mistaken for seeds or cones) upon the 

 " she-oaks " (Casiiarina), while the members of the genera Spharococcus 

 and Frenchia form several curious galls very distinct from each other. 

 It is doubtful if the last three genera can be included in the sub- 

 family Brachyscelinae, but they are mentioned here on account of the 

 gall-making habits of most of their species. 



From their minute size it is difficult to observe the larvae of many 

 of these coccids when upon the plants, but taking one as a type, Ascelis 

 pramollis, Schrader, a species rather common about Sydney, found 

 upon the " bloodwood " [Eucalyptus corymhosa), we may illustrate how 

 the larvae of Brachyscelinae attack the foliage. Upon leaving the 

 mother gall they scatter all over the leaves, but only those fortunate 

 enough to come across the younger growth towards the tips of the 

 branches gain a foothold. They attach themselves to the epidermis, 

 simply sinking into the leaf, the larvae disappearing in a very short 

 time, leaving the apical orifice, like a pin prick, and a minute blister 

 as the only indications of their presence. 



In the case of a parent insect (as in the Cynipidae) forming an 

 incision in a growing woody tissue, it is quite possible that she might 

 inject some acid or irritant secretion that would have a good deal to 

 do with the ultimate form of the gall ; but where the larva itself 

 directly attacks the plant it is quite evident that there can be no 

 irritant matter of any consequence introduced with it, and the final 

 shape of the v/oody excrescence must be altogether influenced by the 

 mode of feeding or working of the insignificant-looking little coccid. 



It is, therefore, very remarkable that a number of white grub- 

 like coccids, with aborted appendages, differing so little from each 

 other in specific characters, should each have the power of forming a 

 distinct and well-defined form of gall, which never varies in size or 

 shape unless attacked by either outward or interior inquilines. The 

 question then arises. What first causes a gall to be oval in one 

 species, and in another species to put forth branching horns ? 



Their growth is very rapid, and the form of the gall is determined 

 and apparent at a very early stage of formation on the tree. It is 

 quite evident, in ihegenwsBrachyscelis, at any rate, that the after-shape 

 of the gall must be anticipated, because the inner skin, or wall of the 

 gall, in contact with the female coccid, is smooth, hard, and sapless 

 some time before its full development. Where the larva is a free- 

 moving creature, it can no doubt materially direct the accumulation 

 of vegetable matter either at the base, sides, or apex of the excrescence 

 by feeding upon or neglecting one portion more than other, and 



