622 FIFTH REPORT OF THE ENTOMOLOGICAL COMMISSION. 



rather bright yellow, brownish near center. Average height of gall, 

 4"""; diameter, 3.3""". Cell as in P. c.mamma, the bottom wall at cen- 

 ter much less thick than at sides. 



P. celtidiscucurbitav'dr. '^.— GaW as seen on the upper side of leaf 

 either as in P. c.-ciwurbita, or more often as in P. c.-glohulus, i. e., the 

 impression is vertical and widening internally, but the upper rim is 

 either barely or not at all elevated. On the under side of the leaf 

 it is always sessile, not oval in outline, but depressed and button- 

 shaped ; the disk being greatly flattened and with a shallow depression 

 having a small central nipple and surrounded by two concentrical rims, 

 the inner one often obsolete, the outer one rarely so. The sides of the 

 gall have either faint longitudinal furrows or are irregularly rugose. 

 Size very variable ; average height, 1.75'"™; width, 3'"'". 



Whether or not this form is specifically distinct from the typical P. 

 c.-cucurbita can not yet be decided. Of the latter I have not many 

 specimens, all from Missouri; while of the variety I have numerous 

 specimens from Missouri and Texas. No specimens strictly inter- 

 mediate have yet been found. 



Besides the galls just described I possess single specimens which 

 apparently represent other species, l3ut their description is postponed 

 until more complete material can be obtained. 



The life-history of these Pachypsyllas varies somewhat with the spe- 

 cies, but the following summary from my notes gives the essential 

 facts : 



Most of the imagos issue in the fall of the year and hibernate in the 

 cracks or under the bark of the tree. The sexes pair in early spring, 

 and as soon as the young leaves put forth, the eggs are deposited singly 

 either on the upper or under sides of the leaf. The young larva settles 

 on the upper side of the leaf and inserts its beak in one of the pores. 

 The irritation from the puncture causes an abnormal growth of the 

 leaf substance, swells around the insect, so that this last appears to 

 sink into the leaf and is gradually carried with the growing gall to the 

 under side. The gall in all Pachypsyllas is due to the action of the 

 young larva, and not to the insertion of the egg. The gall itself soon 

 becomes quite perceptible, but the growth of the larva is very slow in 

 the early part of the season. After the month of July the larva de- 

 velopes more rapidly, and toward the end of September or in October 

 the full-grown jjupa with its abdominal spines saws an oval or slightly 

 crescent-shaped opening through one side of the roof of the gall, issues 

 therefrom, and changes to imago immediately afterwards. Many pupse 

 remain in the galls and fall to the ground with the leaves, where they 

 mostly perish, but some succeed in hibernating and change to imago 

 in early spring. 



OTHER HACKBERRY INSECTS. 



Some few other insects occur on Celtis, but they are not important 

 and I have so far found no time to work at them. 



