150 
and animal cells as hosts. So biological containment is assured by the ex- 
treme fastidious nature of animal cells in culture. So all you have to deal 
with is physical containment now. It seems to us the closer the donor DNA 
species is to the natural host of the virus you are using as vector, the 
lower the containment need be, and the further that difference the higher 
the containment. So if you are putting mouse DNA into polyoma, which is a 
mouse virus which picks it up anyway, and putting that in mice, PI would 
seem enough. If, on the other hand, you are putting a lower eukaryotic DNA 
into that system where you have less reason for saying that the thing can't 
have any harm at all, then you would have a slightly higher level of contain 
ment, or a much higher level of containment. But it must reflect, I think, 
the relationship between the donor and the natural host species of the virus 
Those are the crucial relationships. And the species of the host cell that 
you are putting the recombinant in when you have got it. You should base 
any classification on that inverse set of relationships. 
DR. FREDRICKSON: The suggestion that Dr. Tooze referred to in his tele 
gram about a meeting on viruses is in accord with one made by other American 
commentators, and with our own feelings here after reviewing extensively 
the comments. You heard Dr. Barkley describe this morning some of the 
efforts toward a comity of containment definitions, which I think was con- 
siderably helped by a meeting which EMBO and the NIH sponsored last summer, 
and we have been making some preliminary overtures today to decide whether — 
not really whether, but how we might be able to have a similar meeting 
between Americans and Europeans on this matter of virology. We are going to 
have an SV40 polyoma conference here sponsored by the Fogarty International 
Center in February, and we decided that that wouldn't be quite the group 
that would be most useful here. Dr. Tooze, I am looking forward to getting 
together with you during this next day or so and seeing if we can't work 
out something, because we would be very much interested in doing the same. 
Whether it would be possible to have it in time to affect the revisions 
of these Guidelines depends on our ambitions for the revision, but we would 
do our best. 
Professor Rosenblith. 
PROFESSOR ROSENBLITH: I can't help but be impressed with the discus- 
sion that has just taken place, in the sense that those of us who are not 
experts in these fields need to understand the range of scientific debate or 
uncertainty before we superimpose upon that a range of policy uncertainties. 
And it was really that that I was asking for before, and I think we have 
been illuminated on that a good deal, and I think that I for one, perhaps 
parochially, am also impressed with the fact that the way in which numbers 
are being used in the area, even the taxonomy of containment, if I may say 
so, lends itself to significant misunderstandings. And I think at interna- 
tional meetings, if no other benefit were to be derived, and I think a great 
deal from what we have heard will be derived, ought to really deal with this 
question of the taxonomies of containment, and especially this — I will say 
beautifully produced, colorful chart, where I at least was no longer sure as 
to what the trade-offs between the various taxonomies were. 
[ 354 ] 
