2 . 
The difficulties of estimating by theoretical considerations the potential 
hazards of DMA recombinant experiments appear insurmountable. A comparison 
might be made to the dangers involved in using radioisotopes in biological 
research. With radioisotopes the direct dangers to lab personnel , dangers of 
disposal into the water supply or into landfill, etc. can be quantified. It is 
known that exposures to certain levels of radiation is dangerous, whereas exposure 
to certain lower levels is not. It is thus possible to say Unit "(his level of 
safety (procedure) is adequate" while "that level of safety is inadequate." The 
situation is fundamentally different in the case of the recombinant DNA experiments. 
Here the dangers can not be quantified; they can only be guessed at over an 
extremely broad range of probabilities (from high to low) for any single, given 
experiment. It is thus very difficult, indeed somewhat unrealistic, to say "this 
level of safety (procedure) is adequate" while "this level is inadequate" as the 
Asilomar and Woods Hole Guidelines have attempted. Rather an inevitable, prudent 
philosophical approach for this type of hazard seems one which says "we will do 
those recombinant DMA experiments judged to be worthwhile, in as safe a way as 
possible." This approach is justified because the potential hazardous consequences 
of the DNA recombinant work are extremely serious, and because of the all-important 
distinction between quantifiable (e.g. radioisotopes) and non-quantifiable (e.g. 
recombinant DNA) hazard probabilities. 
The detailed analysis in the preceding paragraph, concluding that worthwhile 
DNA recombinant experiments should be conducted as safely as possible, leads us 
to propose a conservative set of guidelines. 
A. first of all, in overview of the specific proposals below, we believe 
consensus is possible in the research community that the following experiments not 
be done at this time. (Certain detailed reasons and hazards are elaborated in 
the individual categories listed later). 
[ 353 ] 
