366 NATURAL SCIENCE. May. 



to walk on five or four legs, holding the prey beneath her by one or 

 both legs of the intermediate pair. All her maternal care did not ensure 

 the safety of her offspring from the attacks of a large ichneumon fly, 

 Ephialtes albicincius, which was found in the tunnel, and whose grub 

 and pupa were found in the cocoons of the Crabro. The pupa (the 

 limbs of which are, as is usual in hymenoptera, free) can turn itself 

 over when disturbed. 



In the paper in Natural Science already mentioned, I also 

 referred to Dr. Sharp's researches upon stridulation in ants. M. Ch. 

 Janet has ('3) been working at the same subject, but, like the former 

 observers mentioned by Dr. Sharp, he does not distinguish between 

 the true stridulating organs and the ordinary sculpture of the 

 segments. He mentions that a number of red ants {Myrmica) 

 imprisoned between two pieces of glass fixed by cement, made a noise 

 like the gentle boiling of water in a closed vessel, when their prison 

 was held to the ear. The noise was specially audible when the ants 

 were irritated by the observer blowing between the two pieces of 

 glass. 



In Natural Science for November last (vol. iii., p. 353), Mr. 

 Rothera gave us an interesting paper on gall-producing insects, in which 

 he brought forward arguments against the view that the formation of a 

 gall is due to poison deposited with the egg in the plant by the 

 parent insect. Dr. F. Heim, however, in a recent memoir (4), defends 

 this view. His observations were not made upon any of the gall- 

 flies (Cynipidae), but upon a species of saw-fly {Nematns salicis, 

 Jur,'), whose grubs pass their early stage within galls produced 

 upon the leaves of willows. The female fly appears in May and 

 makes a triangular cut close to one of the veins of the leaf; 

 in this the eggs and poison are laid, and the gall is formed before 

 the grub is hatched. It is evident, therefore, that the malformation 

 cannot be due to any mechanical irritation by the larva, but from Mr. 

 Rothera's article it appears that in such a case he would attribute it 

 to some chemical irritant in the egg. The gall takes three or four 

 weeks to attain its full size, by which time it appears as an ellipsoidal 

 swelling, i cm. long, 5 mm. wide, and 5 mm. thick, of a paler green than 

 the normal hue of the leaf, and with a somewhat purplish tint. Trans- 

 verse sections showed the gall to have an epidermis with thickened 

 cuticle, almost destitute of stomata ; beneath this were cells contain- 

 ing a red colouring matter, shown by chemical tests to be a derivative 

 of tannin, like the pigments found in fruits; the internal layer con- 

 sisted of small cells, very rich in chlorophyll. The grub, when 

 hatched, is of the usual saw-fly type and colourless, but becomes 

 greenish by feeding. It undergoes no moult within the gall. After 

 two or three weeks it makes a round opening in its habitation, by 



1 This insect is not Newatus salicis, L., whose grubs do not hve in galls, but spend 

 all their lives free on the leaves. I have not been able to trace the synonymy. 



