1962] 
Roth and Stay — Cockroaches 
197 
in the brood sac, and the mature oocytes degenerated. In addition to 
the fifteen females shown in table 4, two females had their nerve cords 
severed prior to ovulation and oviposited normally; 24 and 25 days 
later their oocytes had grown considerably and were 2.18 mm. and 
2.72 mm. respectively. The nerve cord may be severed at any site 
between the second and sixth segments to eliminate the inhibition of 
the corpora allata during pregnancy. Two females had their nerve 
cords severed between the second and third abdominal segments 4 
days after oviposition; 29 days later their oocytes were 2.75=1=0.01 
mm. Six females had their cords severed between the third and fourth, 
fourth and fifth or fifth and sixth abdominal segments, 4 days after 
oviposition; 22 days later their oocytes were 1.8911=0.24 mm. long. 
Six pregnant females taken from cultures (histories unknown) had 
their nerve cords severed between the fourth and fifth, or fifth and 
sixth segments; 20 days later their oocytes measured 2.28=1=0.33 mm. 
Unmated females of the parthenogenetic strain oviposited normally 
in 98 percent of 248 individuals examined (Roth and Willis, 1961). 
Twenty-two females had their nerve cords severed prior to oviposition. 
Of these, 15 (68%) ovulated in the normal period of time and de- 
posited eggs in the uterus ; 8 oviposited all their eggs and had normal 
oothecae but the other 7 had small abnormally shaped oothecae and 
from 1 to 12 mature oocytes remained in their ovaries. The remaining 
7 of the 22 females operated upon failed to retract the ootheca into the 
brood sac; in 4 of these one or more mature oocytes remained in the 
ovaries but in the others all the eggs were laid. Of 9 females that 
were sham-operated when 1 to 4 days old, all oviposited normally and 
no mature oocytes remained in their ovaries. Apparently an intact 
nerve cord is necessary for normal deposition of mature oocytes and 
for normal formation and retraction of the egg case in some females 
of P. surinamensis (parthenogenetic strain). Some center, possibly in 
the brain, may be involved in this behavior. 
Virgin females of the bisexual strain almost invariably fail to retract 
their oothecae into the uterus (99% of 138 females, Roth and Willis, 
1961) and drop the incompletely formed ootheca. Thirteen virgin 
females had their nerve cords severed when 1 to 9 days old. All ovi- 
posited abnormally, which is the typical behavior of virgins of the 
bisexual strain; 10 dropped their oothecae and all had mature oocytes 
left in their ovaries. The other 3 carried their oothecae extruding 
from the abdomen but failed to retract them; 2 had some mature 
oocytes left in the ovaries but the third had none. Virgin females of 
the bisexual strain with severed nerve cords behaved like unoperated 
virgin females in oviposition and deposition of mature oocytes. 
