43 o FAMILY CURCULIONIDAE 



time taken to boring horizontally instead of vertically. It may be possible 

 that these horizontal portions in the gallery are made just before one of the 

 moultings of its skin which take place periodically during the growth in size 

 of the grub. When the grub has reached full growth, and just before 

 pupating, the gallery is often curved at right angles to its previous direction, 

 and the larva then eats out the pupating-chamber in the sapwood and bark. 

 In fig. 23 a grub is shown in this position at the top of the stem. 



As has been said, the gallery increases in size with the development in 

 bulk of the insect, so that by the time the grub approaches full growth the 

 gallery is from three to four times the breadth of its occupant. The larval 

 gallery is always filled with loosely packed ejected wood-dust and w r ood 

 excreta containing short strands of ligneous wood fibres. Owing to its 

 irregular method of feeding the length and breadth of the larval galleries 

 vary very greatly, but galleries have been measured as long as 9 in. in 

 length, whilst others have wound about in an area of bark 6 in. to 8 in. 

 square. Fig. 23 and pi. xxxvii show the larval galleries in the cambium 

 . and sapwood. 



The position of pupation appears to vary according to the size of 

 the tree attacked and consequent thickness of the bark present. In the case 

 of old trees where the bark is very thick the insect eats out an elliptical 

 pupating-chamber which is made entirely in the inner bark of the trees 

 and does not groove the sapwood. This chamber or cradle is lined with 

 particles of bark fibres forming a rough loose cocoon in which the larva 

 pupates. In the case of young trees, on the other hand, the larva eats out an 

 elliptical depression in the sapwood about half to three-quarters of an inch 

 in length and having the appearance of a small kind of cradle. The other 

 half of this ellipse is eaten out in the cambium layer (see fig. id, in which a 

 newly matured beetle is shown in situ in the cradle) and bark, and the two 

 halves are then lined with a thick interwoven mass of stringy wood-fibres 

 which fits rather tightly into the cradle (pi. xxxvii). In this cocoon the larva 

 changes into the pupal stage. 



When the beetle is fully mature, and consequently ready to leave the 

 tree, it bites out a circular hole at one end of the cocoon (fig. xxxvii, the 

 lower of the two cocoons shown), and, crawling out of this, bores a large 

 circular hole through the thick bark of the tree, and thus escapes from it. 

 These holes have a considerable diameter, and are easily visible even in 

 the thick bark which this tree puts on (vide pi. xxxvii). Observations of 

 attacked trees have shown that the newly mature weevils do not always 

 get out of the tree in safety, owing to the outflow of resin which at times 

 catches the beetles whilst engaged in boring their way through the bark and 

 drowns them. Instances of this have been observed in situ in the trees. 



It is difficult at present to state with any certainty the number of 

 generations which this weevil passes through in the year, since the insect is 

 so often found in the tree in all three of its stages of larva, pupa, and beetle. 

 I am inclined to consider, however, that the insect has normally three genera- 



