It's nearing the end of a science lesson at Dr Challoner's Grammar School in Amersham when the assistant headmaster enters, marches to the front of the classroom and raises her hands to speak. "We have made it clear to you that we do not want drugs in this school. And that from time to time we may bring in some dogs. Well," Mrs Horrocks tells the class of boys aged 13 and 14, "they are here. If anyone is frightened of dogs put up your hand." One boy's arm hovers upwards but quickly comes down again as excitement reaches a crescendo among his classmates.
The children are told to leave their bags and queue up outside while the dogs, trained to sniff out illegal drugs - from cannabis to crack cocaine - get down to business. A spaniel leashed to its handler darts around the empty classroom, sniffing the children's pencil cases, blazers and bags. As the boys file back in grinning, some of them a little nervously, each is smelled by the second dog, a black labrador. None of the bags or pupils has been "indicated" by the dogs, whose powerful sense of smell can detect traces of cannabis or cannabis smoke up to a month old.
In fact, during the four dog visits to the school - undertaken by private firm Grosvenor International Services (GIS), which takes its canines into 30 schools - no pupil has been caught with, or smelling of, drugs. This may not be surprising - official figures estimate that only two in 10 secondary children have taken an illegal drug in the past year.
So why, along with scores of other state schools and despite a recommendation by the Government's drug expert panel this month against the use of sniffer dogs and random testing, is Dr Challoner's bringing in the canine drug squad?
"There are enough pressures on young people in society these days and we want them to be able to be free from the pressure of illegal drugs," says the headmaster Mark Fenton. "I wanted to introduce the drug dogs as a deterrent. I have to put the interests of the wider school community before individual pupils."
Mr Fenton's motives seem to come from a genuine concern for his pupils' welfare. And supporters of the tactic say pupils and parents have overwhelmingly supported sniffer dog and drug-testing schemes, that it gives pupils an excuse not to succumb to peer pressure and is a way of spotting children in need of help. Although protocols differ, most schools will counsel pupils suspected of using drugs, while those caught with class-A drugs face expulsion or arrest.
But the sniffer dog solution, along with the more high-tech tactic of drug testing pupils pioneered in Kent - both endorsed by the Prime Minister - has come under fire for being a violation of children's civil rights and a heavy-handed, costly, error-prone, counter-productive measure that forces drug-use underground.
There is also scepticism about whether dogs and tests help children who need it most. The majority of schools using drug dogs are in relatively affluent areas - it has yet to be tried in inner London schools, for example. Therefore, critics say their use is a cynical PR exercise in branding a school "a drugs free zone" rather than a serious attempt at preventing and tackling drug use. If a pupil does have a drug problem, detractors say, would the fear of being detected by a dog or a random test drive them to increased truancy and make them less likely to confide in a teacher?
Yet these strategies - borne on the back of the increased use of sniffer dogs and drug testing in adult society, and media scare stories about drugs flooding playgrounds and dealers at the school gates - are being adopted by more and more schools. In the past four years, the number of state schools using sniffer dogs has rocketed from virtually nothing to at least 150.
Police have found themselves in the interesting position of having to dissuade headteachers from adopting an array of measures that would breach government guidelines on the use of sniffer dogs in their eagerness to jump on the drug-dog bandwagon.
"I get some teachers calling me up and ranting about what they want us to do with the dogs - some want to try to scare the kids; some want to con their pupils by searching their bags while they are in assemblies," says Inspector Bill Stupples, schools liaison officer at Merseyside Police. "I tell them using dogs is not that easy and that it is fraught with danger." He says it is only a matter of time before a parent sues a school for incriminating their child.
Sniffer-dog firms admit their four-legged employees can show heightened interest in girls who are menstruating or carrying the contraceptive pill, pupils who have been prescribed the attention deficit disorder drug Ritalin, and in the smell of food, other dogs and cannabis smoke from, for example, the school bus or even parents. Teachers also have to be aware. On occasion, they have been indicated by dogs and have had to explain themselves to their headteacher like their pupils.
The Association of Chief Police Officers earlier this year warned forces to refrain from the "covert" use of sniffer dogs in schools after Kent Police were criticised for secretly searching pupils' bags in classrooms while children were attending special "fun" drug-dog assemblies.
The body of evidence to support the effectiveness of using dogs or drug tests in schools is flimsy. The only study into the use of sniffer dogs in England, carried out by the John Grieve Centre for Policing and Community Safety among six schools in Buckinghamshire between 2003-4 - of which Dr Challoner's was one - declared the project a success on the basis that the dogs were a "hit" with children. It said the scheme should be taken up in all schools. But the findings have to be taken with a pinch of salt when the study's small print reveals only two per cent of pupils gave feedback.
Encouragingly, Sharafat Ali, the young people's commissioner for Buckinghamshire drug and alcohol team, says the next raft of 10 schools being piloted in the county will more accurately reflect the emphasis on consulting young people contained in the Government's Youth Matters document: "We want to make sure the scheme is focused on outcomes for young people, rather than schools. There was no recognition of that within the original pilot." Now children's views on the sniffer dogs will be gauged before and after they come into schools.
A report into drug testing in American schools last year by the Joseph Rowntree Foundation stated: "It is a matter of concern that student drug-testing has been widely developed in the US and may conceivably be so in the UK on the basis of the slimmest available research evidence."
Former education secretary Ruth Kelly praised the country's first school drug-testing scheme, initially funded by the News of the World at the Abbey School in Faversham, Kent - to be expanded to six schools in the county in January, as a "hugely effective" way of tackling substance misuse - despite the fact the school based its success on an improvement in exam results and behaviour. Peter Walker, Abbey's former headteacher who introduced the tests, is now helping the Department for Education and Skills produce guidelines on drug testing in schools.
Although Buckinghamshire drug and alcohol team is pleased with GIS' involvement in its pilots and police forces have been given guidance on using dogs in schools, headteachers need to tread carefully. "We are only one of a handful of firms out there doing it effectively," says John Franklin-Webb, the director of GIS. "It is not an easy thing to get right. There are a lot of ex-security guards who get themselves and a dog trained up and offer their services to schools. It frightens the life out of me."