Jan. IS, 191S Agrilus Bilineatus 289 



The eggs were hatched in the laboratory in from 10 to 13 days. It was 

 found that the outer membrane became dry and shriveled and even 

 cracked in from 3 to 6 days, while the inner membrane became brown 

 and the embryo seemed to develop ajt one side of the egg, which became 

 plumper than the other side (PI. XXXVIII, fig. 3). 



THE LARViE 



When the larvse hatched, they broke through the egg membrane on 

 the side toward the bark and immediately began to burrow. As a result, 

 they were not exposed to view, and the eggshells were found filled with 

 the frass which had been burrowed out at the start. In a few cases the 

 eggs had been entirely loosened from the bark for examination, and none 

 of the larvse from these eggs succeeded in getting a burrow started, except 

 in one case where the egg was artificially fastened to the bark before the 

 lan,'a left it. Since the eggs are stuck to the bark in the depths of cracks, 

 they offer a certain resistance to the lar\'a's efforts, which makes it pos- 

 sible for its mandibles to get hold of the bark and start the burrow. 



The newly hatched lar\'ae (PI. XXXVUI, fig. 4) measure only from i 

 to 1^-2 mm. in length, but one was found capable of reaching the cambium 

 layer in 24 hours by burrowing for 2}^ mm. The fact that the eggs were 

 laid in the depths of cracks made it possible for the larvse to reach the 

 soft cambium layer by penetrating but a few millimeters of hard bark 

 tissue. Having reached the cambium layer they started off in any con- 

 venient direction. Obsen^ations show that burrows made during the 

 first instar often go obliquely across the grain of the wood or with the 

 grain, the larvae being indifferent as to whether they go up or down the 

 tree. If care is taken in removing the bark when green, the tiny burrows 

 can be traced to a widening of the burrow which marks the end of the 

 first instar. The burrows measured showed that the larvae had bur- 

 rowed for a distance of 60 to 135 mm. when the first molt took place. 



In most cases the widenings in the burrow occurred when the lar\-ae 

 had gone into the wood, less often into the bark, and were again returning 

 to the cambium layer. It is evident that these points marked the limits 

 of instars and that the molting took place in these excavations in the 

 wood, which were often two or three times the length of the larvae. 

 Places were found, however, where the burrows showed that the molt 

 had been made in the cambium layer. The longest larva found in a 

 burrow of the first instar measured 4.6 mm. in length, and the average 

 width of the burrows was 270/i. 



The burrows made during the second instar measured about 900/i in 

 width and took about the same course through the cambium layer, but 

 they were about twice as long. At the beginning of the third instar 

 quite a different course was usually found, especially in green bark on the 

 trunks of trees, where the burrows were almost always transverse to the 

 grain of the wood. The burrows of the fourth instar were about 2 mm. 

 in width and often attained the length of 500 or 600 mm. Where the 



