INSECTS AFFECTING PARK AND WOODLAND TREES 42 I 



middle of Juikj in MassachuscLls, lu llic early pari of Jul\- in norlhcrn 

 Maine, the larvae feediny on the leaves late in June and in July and early 

 Auij^iist. The yoiuii^ are nearly full yrown by the last of July or first week 

 in Auorust, according to the latitude. Still a few occur on trees in Massa- 

 chusetts, as late as the last week of August or early September. 



l)r PackartI states that it is very iloublful if lher(; are two broods. lie 



obser\cd at llrunswick Me., that all the eggs had hatched by June 23 or 



28. They are placed in two rows, alternatc-ly, not exactly parallel, one 



being placed a little in advance; of the other, and are inserted at the base 



of the fresh young partl\- tUneloped leaves of the new shoots, which 



are usually at this time only about i or i'^ inches in length. Ihe 



presence of the eggs causes a deformation of the; shoots, which curl over, 



the incisions being in all cases observed on the innc^r sidct of the shoot. 



1 )r Packard has described the operations of v.^^o; laying as follows: 



The sawing the slit in which the eggs are placed rccpiires about 

 3 minutes. The tw^o sets of serrated blades of the ovipositor were thrust 

 oblicfuc^ly into the shoot by a sawing movement, the lower set of blades 

 being the most active, and sliding in and out alternately. After the 

 incision is sufhcientl)' dee]), the egg is expelled through the inner blades of 

 the ovipositor. 



Dr Packartl states that though the slit is at first closed, as soon as the 



embryo increases in si/e the twigs swell where they have f)een cut and the 



slit enlarges and gaps more or less, and thus affords ready egress for the 



newly hatched larva, which rarc-ly eats the terminal shoots, but crawls on 



the lea\es of the whorls next thereto. It first nibbles one side of the 



needle or leaf, leaving it half eaten and rough, serrate, and partly 



withered along the edge. The presence of the injured leaves is of 



great aid in detecting the }(ning sawfly larvae, which collect on the 



verticils of the larches after llu\' have shed their tirst skin and almost 



invariabl)- begin to eat the needles one after another. In this way one 



verticil after another is devounil ami when the larvae are half grown, they 



occasionally collect around the main stem of the twig in singular clusters, 



with the hinder |)art of the bod\- curled over the back. Owing to the 



oblique |)osition in reference one to another, the\' look much like a ball 



