P3 to P2 conditions, because the difference between these levels 
of containment is large. Additionally, the substitution of 
biological for physical containment measures, as embodied in 
the proposed revisions of the guidelines, has not been demonstrated 
to be valid or effective. 
(b) In concert with further study of genetic transfer 
mechanisms , several measures can be taken immediately to guard 
against bacterial contamination of recombinant DNA experiments. 
These additions to § II-D might include: (i) the mandatory 
incorporation of control plates or broth cultures into recombinant 
DNA experiments to insure the identity of hosts and vectors and 
(ii) the physical isolation of experiments requiring different 
levels of biological containment. These measures are easily 
implemented in the laboratory, and they would be an effective 
barrier to the inadvertent contamination of cultures, a not 
uncommon occurrence in microbiology laboratories. 
F. Experiments with Viral DNA 
The reason for the containment levels which are currently 
mandated for experiments with viral DNA is the possibility that 
a harmful virus or viral product would gain an extended host 
range when cloned in a novel host organism. The previously 
mentioned results of experiments carried out by Drs . Rowe and 
Martin indicate that this concern is justified for the cloning 
of complete viral genomes and hybrid genomes in bacteria. 
The formation of viral hybrids is an important model for 
[A-195] 
