I2 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 134 
plant at the University of Florida. This filter had been in operation 
for five years but had never been flooded during that time to control 
filter flies. The filter was finally flooded because cockroaches had been 
seen collecting on the outside walls at night. When the flooding water 
was within 4 inches of the top, the filter medium was covered with a 
mass of cockroaches so thick that the rocks were completely hidden 
(see frontispiece, pl. 1). When the insects were killed by insecticide, 
those removed occupied a minimum volume of over 24 cubic yards; 
several times that quantity died in the adjacent woods. Contamina- 
tion of cockroaches in trickling filters can probably be assumed. Fair 
(in Rosenau, 1940) stated that trickling filters remove go to 95 per- 
cent of the sewage bacteria. Obviously, more pathogenic bacteria are 
present in the influent sewage than are present in the effluent. Silver- 
man and Griffiths (1955) state that the percolating filter is not a 
serious obstacle to the passage of helminth ova in sewage, and they 
cite experiments by others to prove this point. Hence, helminth eggs 
could also be a source of contamination to cockroaches residing within 
the filter bed. 
Nelson (1952) reported that cockroaches were the most prevalent 
filth-carrying insects of San Tomé, Venezuela. They were present 
in practically every building under natural conditions of no control, 
and they were numerous in sewer lines as well: “Literally millions 
of roaches, crickets, and centipedes—not to mention rats—live in 
sewer lines and invade homes through connecting lines or overland. 
. . . Cockroaches overran the kitchen and dining hall; they hid under 
and in sugar bowls; they were sometimes found in soup and in food. 
The first treatment of the messhalls and other kitchen installations 
yielded an approximate half barrel of dead bodies, and it was im- 
possible to walk across a floor without crushing hundreds of bodies.” 
Schoof and Siverly (1954) surveyed 22 sewer manholes in Phoenix, 
Ariz., for a 7-week period and found a weekly average of 92 to 143 
specimens of Periplaneta americana per manhole. Jackson and Maier 
(1955) surveyed sewer manholes in the same city and found 300 to 
400 P. americana in certain manholes. 
Eads et al. (1954) reported that about 40 percent of 762 sewer 
manholes in Tyler, Tex., were found to be infested with P. americana, 
Four other species of cockroaches were also captured in the manholes: 
Periplaneta fuliginosa in 3 manholes, Blatta orientalis in 10 manholes, 
Parcoblatta bolliana (Saussure and Zehntner) in 1 manhole and Par- 
coblatta pensylvanica (DeGeer) in I manhole. Dr. R. B. Eads (p.c.) 
stated that P. fuliginosa has become a very common domiciliary 
cockroach over much of the State. 
