Traditional plant breeding involves crossing varieties of the same species in ways they could cross naturally.For example，disease－resistant varieties of wheat have been crossed with high－yield wheat to combine these properties.This type of natural gene exchange is safe and fairly predictable. Genetic engineering（GE）involves exchanging genes between unrelated species that cannot naturally exchange genes with each other.GE can involve the exchange of genes between vastly different species， e.g.putting scorpion toxin genes into maize or fish antifreeze genes into tomatoes.It is possible that a scorpion toxin gene，even when it is in maize DNA，will still get the organism to produce scorpion toxin， but what other effects may it have in this alien environment？We are already seeing this problem——adding human growth hormone genes to pigs certainly makes them grow——but it also gives them arthritis and makes them cross－eyed，which was entirely unpredictable. It will be obvious，for example，that the gene for human intelligence will not have the same effect if inserted into cabbage DNA as it had in human DNA——but what side－effect would it have？In other words，is GE food safe to eat？The answer is that nobody knows because long－term tests have not been carried out. Companies wanting a GE product approved in the UK or USA are required to provide regulatory bodies with results of their own safety tests.Monsanto's soya beans were apparently fed to fish for 10 weeks before being approved.There was no requirement for independent testing，for long－term testing，for testing on humans or testing for specific dangers to children or allergic people. The current position of the UK Government is thatThere is no evidence of long－term dangers from GE foods.In the US，the American Food and Drug Administration is currently being prosecuted for covering up research that suggested possible risks from GE foods.