- 6 - 
Lapp& 
scientific accountability was heavily circumscribed and respon- 
sibility for unforeseen, adverse consequences downplayed. But 
some of the risks with recombinant molecules are foreseeable, in 
the broadest sense. ^We can review the safety record of Fort 
Oetrick in Frederick, Maryland, where most of our biological 
warfare research was conducted. There we readily find instances 
of laboratory accidents which were not predicted by the containment 
precautions which were taken. Researchers were almost routinely 
exposed to sub-clinical doses of the agents with which they worked, 
and like the present-day workers with SV40 and Shope papilloma 
viruses, developed the tell-tale antibodies indicating exposure to 
potentially lethal agents. Biological excursions in which aerosols 
of tularemia or other pathogenic organisms escaped from the laboratory 
provide ample precedent for the unpredictability of much of yet to 
be experienced hazards. Research at Detrick, in fact, provides the 
only real basis against which to equate health risks; yet it is not 
mentioned once in either document. 
VIhat does an actual risk estimation in the report look like? 
As a thought exercise, the environmental impact statement walks the 
reader through a complex series of hypothetical events which would 
have to happen simultaneously to create a biohazard. The genetically 
engineered organism which possesses a potentially harmful gene must 
first escape from the experimental situation (risk: 1 in 100) ; it 
must survive after escape (risk: 1 in 10,000); it must grow and 
Appendix K — 15 
