- 10 - 
one step (from P3 + EK2 to P2 + EK2 or P3 + EK1). (ii) Containment levels 
for plant virus DNAs in EK hosts were not changed (P3 + EK1 or P2 + EK2). 
(iii) In both cases the levels are lower by one step compared to the P3 + 
EK2 level required for plant pathogenic procaryotes. These, although not 
known at present to exchange chromosomal DNA with £. col i , do exchange 
plasmid and/or viral DNAs, whereas no natural genetic exchange of any kind 
has been shown or is likely to occur between fungi or plant viruses with E. 
col i . 
Since I basically agree with the approach which has been followed in 
dealing with lower eucaryotes, I suggest that a similar approach be used, 
when applicable, for procaryotes as well. Furthermore, containment levels 
for experiments with procaryotic DNA in EK hosts should remain one step 
lower than those for various types of lower eucaryotic or viral plant path- 
ogic DNAs involving a similar potential biohazard, since the former are much 
more likely to be introduced into these hosts than the latter by standard 
in vivo manipulations. 
I wish to make a final general comment: if the Guidelines are to survive, 
they should take into account, among other things, (i) the comparative risks 
of in vivo (or "natural") and in vitro ("novel") methods of accomplishing a 
particular experimental goal (e.g. the construction of a particular strain), 
and (ii) the possible indirect consdquences of over-regulating one experimental 
method ( in vitro ) over another ( in vivo ) which, although "natural", is not 
necessarily all that biohazard-free. It is clear that the basic revisions 
proposed are meaningful and overdue and that the original stringency has 
already caused much unnecessary delay in certain areas of research. Stringency 
and overregulation, however appealing to some, can indeed have 
undesirable side-effects, since classical genetic research is not entirely 
[Appendix A — 119] 
