58 P. Day

many reasons why they will be cautious. For example, at the present time
each product of transformation using the Agrobacterium system is poten-
tially different from every other. This is because as yet there appears to be
no control over the sites of integration of introduced DNA or the number
of copies introduced. In a plant breeding context, therefore, the effect of
integration site will have to be examined. Even if the molecular biologist
has appeared to 'gild the lily' by adding one or more new characters to an
established and widely-grown crop plant variety, the new line will need to
be tested for stability of the introduced characters) over several years and
many trial sites, to make quite sure that transformation has not introduced
an unanticipated defect which could render it valueless under certain cir-
cumstances. Plant breeders are used to uncovering the Achilles' heels in the
most cherished products of their selection programmes.
 Perhaps one of the most important scientific and social issues concerns
framing sensible legislation to allow the introduction of products of genetic
engineering into agriculture outside the containment of the laboratory and
glasshouse, which is at present mandatory for all such materials (see
Halvorsen et al., 1985; Teich et al., 1985). The difficulties encountered in
arranging field tests of Ice-bacteria and of corn seeds treated with Pseudo-
monas fluorescens engineered to produce the delta endotoxin of Bacillus
thuringiensis kurstaki are well known. The companies interested in com-
mercializing these two microorganisms have spent millions of dollars, and
thousands of man-hours, attempting to proceed through a thickening tan-
gle of regulations and lawsuits. There can be no doubt that society must
insist on safeguards to protect the environment. However, such safeguards
have to be tempered by the knowledge that no-one can guarantee, with
complete certainty, that any given released organism, whether or not ge-
netically engineered, will have no adverse effects whatsoever. Answers
must be provided to certain questions before field tests and environmental
release are allowed. In my view, the requirement to produce, in an appli-
cation for an experimental use permit, more than a thousand pages of
evidence to show that an engineered organism is safe and harmless, would
be absurd if it is to become the norm. Unless procedures are simplified,
and sensible tests established, we will deny ourselves important benefits in
a technology that may well greatly reduce our dependence on agrochemi-
cals, which in the long run could be far more damaging both to us and the
environment than products of genetic engineering.