In the vexed case of John McInnes, exhumed from Stonehouse Cemetery in the bitter winter of 1996 and promptly pilloried by the media as the legendary serial killer Bible John, the magic bullet failed to find its mark.

Despite the best efforts of Strathclyde Police Forensic laboratory, then latterly of the Cambridge University team who proved 80 years on that the bodies in the Russian woods were those of the Czar and his family, the DNA from McInnes's remains could not be matched with the body fluids left behind on the underwear of murder victim, Helen Puttock.

The police, the forensic scientists, and the Crown were left ruefully to contemplate the egg on their collective chins while the media moved on to the next sensation.

However, the biggest loser was probably DNA testing itself, because public confidence was undoubtedly shaken in a process which has become confused in the wider public mind as a universal answer for all crime investigation.

Any detective will tell you that that is not the case. Before DNA matching comes into its own as one more corroborative tool in the Crown's armoury, the police have first of all to track down and collar a prime suspect and the key to that is twofold - sound detective work and an expertise at handling the media, the better to evoke public response.

The misunderstanding of DNA sampling does not stop there, but is much wider. In a Scottish case, which ended only last week in an acquittal, DNA samples were introduced by the prosecution and would have undoubtedly influenced the jury but for a defence submission - successful in the event - that there was nothing to indicate the age of the samples or when they were left at the scene of the crime.

It was precisely the same situation that would have faced the police and the Crown had the McInnes DNA taken from the Stonehouse grave proved a match; it would have indicated that McInnes had had a sexual encounter with the unfortunate murder victim at some point, but it would most emphatically not have proved that McInnes had killed her.

Realistically, it would have allowed the police to say they were not looking for anyone else.

Widening out the argument, it could have been argued that the victim just might have borrowed clothes from a friend - a friend who had had a sexual encounter with McInnes, whose semen had somehow survived washing. The arguments can be tested to destruction.

There have been many challenges to DNA's alleged infallibility, both in UK courts and in the US, and the challenges continue even in the face of a rapidly advancing technology. Many in the past have centred on incidental contamination, or degradation of the specimen, or even invasion of a person's DNA by outside chemical agents.

At any rate, there is continuing concern by defence lawyers and by judges over the overpowering manner in which DNA evidence is presented to juries when it should, like all forensic science, carry commonsense health warnings about the fallibility of technique.

While a positive DNA sample will undoubtedly persuade a proportion of guilty accused to throw up their hands and confess, a case pointing to a reasonable doubt existing in many other instances can undoubtedly be made.

Perhaps these points should be uppermost when any consideration is given to the idea of a national DNA data base covering the entire UK population, because, attractive thought the idea is to the police, a whole raft of doubts exist over the incidence of accidental mismatches (which have led to proven miscarriages of justice both here and in the United States) and over the dubious misuse of statistics by prosecutors.

Aside from the scientific

misgivings, the notion is unthinkable in terms of civil

liberty arguments.

Professor Alan Miller of the Scottish Human Rights Centre, points out: ''The danger of

scientific development like

this is that change is determined by policing interest and that it is only after the passage of years that we realise it should have been examined much more fully.

DNA samples were introduced into an unprepared criminal court in the early 1980s when it had not been validated and that led to a number of clear miscarriages of justice. ''The lesson is that we should allow the police to contribute their arguments as to what benefits their interest but that it is a matter for all of society to decide.''

Professor Miller says any move towards a whole-population data base would undoubtedly lead to challenge in the European Court on the grounds of invasion of privacy and that the judges would almost certainly take the view that it was disproportionate to the threat it allegedly met or the benefits

it conferred.

There are other, more sinister, DNA applications in the background which are also provoking fears, including advances in technique which would allegedly allow the scientists to take a sample from a crime scene and extrapolate strands from it which indicate - allegedly - colour of hair, colour of eyes, gender, and ethnic origin.

This in turn leads to one of the types of evidence, identification - whether by eye witnesses or, increasingly, from CCTV - which often leads to the courts of appeal.

As the technology advances it becomes of greater and greater interest to the insurance industry, or to employers, eager to isolate people with the notional criminal gene.

Professor Miller says: ''It is time for government to step back and evaluate all the different factors in the development of DNA technology.''