62 
other tested mammals, and more recently it has been shown that it does not 
even become pathogenic when containing certin plasmids that normally confer 
pathogenicity to more robust strains of coli. 
In addition, it has recently been shown that recombinant DNAs consist- 
ing of lambda phage containing any one of thousands of different eukaryotic 
DNA segments, in this case from yeast, never grow as well as the parental 
lambda vector. That is, they are selected against in a growth race against 
the normal DNA. 
Finally, experiments in our laboratory and in John Carbon's labora- 
tory indicate that the probability that the eukaryotic genes are expressed 
in IS. coli is quite small — much less, for example, than is the probability 
of expression by prokaryotic genes in _E. coli . 
Why, in the face then of this evidence, were the containment levels 
continously increased? I suggest that one of the important factors has 
been a preoccupation and speculation as to the conceivable hazards of these 
experiments without a corresponding concern about their benefits. In con- 
sidering this balance between benefits and hazards, I should like to dis- 
tinguish between areas of fact and of speculation. Some of the benefits 
result from a fact that can be simply stated. By the use of these shot- 
gun experiments, one can isolate individual genes and examine in molecular 
detail their structure, function, and arrangement in the complex chromosomes 
of higher organisms. Without that use, one cannot. 
This is an important fact, because such an examination is critical to 
a determination of how these genes are reproduced and of how they are turned 
on and off in different cells. This determination is, in turn, of critical 
importance to our understanding of cell development, both normal and 
pathological. 
By contrast, we enter the speculative arena when we consider either the 
directly-applied benefits or the hazards of such experimentation. I do not 
wish to belittle either area of speculation. The applied benefits that we 
can imagine are indeed considerable, ranging from the economical production 
of medically useful proteins, such as specific antibodies, to significant 
increases in our food supply. 
The hazards that one can imagine are certainly not trivial, or we 
should not be here today. 
However, one's evaluation of each involves guesswork, and whatever the 
balance obtained between them, the result must be stacked up against the 
hard fact that I referred to before, namely that these experiments allow a 
dramatic increase in our knowledge about one of the most important charac- 
teristics of life — that is, its genetic structure and function. 
[ 203 ] 
