It has repeatedly been said that there is no provision for discovery of documents for a summary trial in a magistrates' court and that section 97(1) should not be used as a disguised attempt to obtain discovery: see Reg. v. Skegness Magistrates' Court, Ex parte Cardy [1985] R.T.R. 49. The circuit diagrams would have been meaningless without explanation and Dr Williams, who disavowed any expertise in electronics, would not have been able to explain them. The proceedings would have had to be adjourned for the documents to be examined by an expert on behalf of the accused. Whether they would then have had any relevance is entirely speculative: the defence made no attempt to lay a foundation for their application by evidence of how a circuit diagram might demonstrate that, contrary to Dr Williams's empirical observations, the inaccuracy of the time reading did in fact have some effect on the breath analysis. I think that the justices were right to reject the application.

          The second question is whether, given Dr Williams's lack of expertise in electronics, his expert evidence was admissible on the question of the reliability of the device In my view Dr Williams was entitled to give this evidence by reason of his familiarity with the working of the Lion Intoximeter 3000. It is notorious that one needs no expertise in electronics to be able to know whether a computer is working properly. Dr Williams was qualified to say that he was familiar with Intoximeters which displayed the time incorrectly but nevertheless produced correct breath analyses. Sergeant O'Dell, who tested his Intoximeter to see whether it was calibrating correctly, was able to give evidence to the same effect. This was evidence on which the justices were entitled to make their finding that the statements produced by the device were accurate.

          This brings me to the chief question in Ms McKeown's case, which is whether the evidence of the Intoximeter's breath analysis satisfied the requirements of section 69(1) of the Police and Criminal Evidence Act 1984. If it did not, there was no other evidence of the alcohol content in her breath and her conviction cannot stand. It will be recalled that section 69(1) deals with the admissibility of "a statement in a document produced by a computer" as evidence of "any fact stated therein." In order for the statement to be admissible, it must be shown, in accordance with in subsection (1)(b):

"that at all material times the computer was operating properly, or of not, that any respect in which it was not operating properly or was out of operation was not such as to affect the production of the document or the accuracy of its contents."

          A "statement" has the same meaning as in section 10(1)(c) of Part I of the Civil Evidence Act 1968, where it includes "any . . . device in which . . . data . . . are embodied so as to be capable . . . of being reproduced therefrom." This would include the memory of a computer.

          I shall for the moment assume that the inaccuracy in the time display meant that "the computer . . . was not operating properly." The question is therefore whether that was "such as to affect the production of the document or the accuracy of its contents." If the words are read literally, it did. The document said that the first test had occurred at 23.00 GMT when it was in fact 00.13 BST. As to one hour, the discrepancy is merely as to the way in which the time was expressed. 23.000 GMT is the same time as 00.00 BST. But the remaining 13 minutes cannot, I think, be dismissed as de minimis. The inaccuracy of the time reading therefore affected the accuracy of a part of the contents of the document.

          In my view, however, the paragraph was not intended to be read in such a literal fashion. "[T]he production of the document or the accuracy of its contents" are very wide words. What if there was a software fault which caused the document to be printed in lower case when it was meant to be in upper case? The fault has certainly affected the production of the document. But a rule which excluded an otherwise accurate document on this ground would be quite irrational. To discover the legislative intent, it is necessary to consider the purpose of the rule.

          The first thing to notice is that section 69 is concerned solely with the proper operation and functioning of a computer. A computer is a device for storing, processing and retrieving information. It receives information from, for example, signals down a telephone line, strokes on a keyboard or (in this case) a device for chemical analysis of gas, and it stores and processes that information. If the information received by the computer was inaccurate (for example, if the operator keyed in the wrong name) then the information retrieved from the computer in the form of a statement will likewise be inaccurate. Computer experts have colourful phrases in which to express this axiom. But section 69 is not in the least concerned with the accuracy of the information supplied to the computer. If the gas analyser of the Intoximeter is not functioning properly and gives an inaccurate signal which the computer faithfully reproduces, section 69 does not affect the admissibility of the statement. The same is true if the operator keys in the wrong name. Neither of these errors is concerned with the proper operation or functioning of the computer.

          The purpose of section 69, therefore, is a relatively modest one. It does not require the prosecution to show that the statement is likely to be true. Whether it is likely to be true or not is a question of weight for the justices or jury. All that section 69 requires as a condition of the admissibility of a computer-generated statement is positive evidence that the computer has properly processed, stored and reproduced whatever information it received. It is concerned with the way in which the computer has dealt with the information to generate the statement which is being tendered as evidence of a fact which it states.

          The language of section 69(1) recognises that a computer may be malfunctioning in a way which is not relevant to the purpose of the exclusionary rule. It cannot therefore be argued that any malfunction is sufficient to cast doubt upon the capacity of the computer to process information correctly. The legislature clearly refused to accept so extreme a proposition. What, then, was contemplated as the distinction between a relevant and an irrelevant malfunction? It seems to me that there is only one possible answer to that question. A malfunction is relevant if it affects the way in which the computer processes, stores or retrieves the information used to generate the statement tendered in evidence. Other malfunctions do not matter. It follows that the words "not such as to affect the production of the document or the accuracy of its contents" must be read subject to the overall qualification that the paragraph is referring to those aspects of the document or its contents which are material to the accuracy of the statement tendered in evidence.

          Paragraph (a) of section 69(1), which deals with improper use of the computer, clearly has this meaning. The statement is inadmissible only if there are reasonable grounds for believing that the improper use has caused the statement tendered in evidence to be inaccurate. It was argued that because paragraph (b) uses different language and speaks of the "production of the document or the accuracy of its contents" rather than being concerned, as in paragraph (a), with the accuracy of "the statement," it must have a different meaning. I shall not speculate on the reasons why the draftsman thought it necessary to deal with improper use of the computer separately from the question of whether it was in proper working order. But there cannot have been any difference in the purpose of the two paragraphs: in both cases the legislature was concerned with the reliability of the statement tendered in evidence as a properly processed and reproduced piece of information. On the point now in issue I think it would be quite irrational if the effect of the two paragraphs was not the same.

          The justices had before them a certificate signed by Sergeant O'Dell under paragraph 8 of Schedule 3 stating that to the best of his knowledge and belief the requirements of section 69(1) had been complied with. In the absence of contrary evidence, they were entitled to accept this certificate as sufficient to satisfy section 69(1). The question is then whether they were obliged to regard the inaccuracy of the clock display as contrary evidence. The justices also had evidence from Dr Williams and Sergeant O'Dell, which they were entitled to accept, that the clock display was not affecting the proper functioning of the computer in processing the information from the breath analyser. Having accepted this evidence, there was in my view nothing to displace the effect of Sergeant O'Dell's certificate.

          I have considered the matter on the assumption that the error in the clock display showed that the computer was not operating properly. I should say, however, that I am not satisfied that this conclusion should have been drawn. Computer clocks, like any others, have to be set to the correct time and the most obvious explanation for the 15 minute discrepancy was that someone had made a mistake when he last set the clock. This would not have anything to do with the computer not operating properly. Furthermore, if the error lay in the clock mechanism itself, I doubt whether it would constitute part of "the computer" for the purposes of section 69(1). The section, as I have said, is concerned with the processing and storage of information and not with the accuracy of the information supplied. The clock, although no doubt physically in the same box as the computer, is something which supplies information to the computer rather than being part of the processing mechanism. But I do not explore this question any further because there was no evidence about why the time was inaccurate and I prefer to base my decision on the construction of section 69(1). In my view, there was admissible evidence upon which the justices were entitled to convict Ms McKeown and the Director's appeal in her case should be allowed.

          Mr. Christopher Jones was brought into Widnes Police Station about a week after Ms McKeown. He had driven his car onto a roundabout and appeared to be very drunk indeed. He provided one specimen of breath which registered 148; more than four times the prescribed limit. He then did not blow hard enough to provide a second specimen. The visual display and print out indicated that the test had aborted. The officer who operated the machine and gave evidence at the trial was Sergeant Draycott. Like Sergeant O'Dell, he noted that the time display was inaccurate by his watch; in his case, by an hour and 15 minutes, but the discrepancy of two minutes from the evidence in the earlier case is as likely to have been in the policemen's watches as an indication that the computer clock had fallen further behind.

          Mr. Jones was charged under section 7(6) with failing without reasonable excuse to provide a specimen of breath when required to do so in pursuance of section 7(1). At the trial he gave evidence that he had tried his best but was overcome by a fit of coughing. The justices rejected his explanation, found that he had deliberately refused to blow into the machine as instructed and convicted.

          It was submitted on behalf of Mr. Jones that he could not lawfully be required to provide breath for an Intoximeter with an inaccurate clock. But section 7(1) says only that a constable may require a person under investigation to "provide two specimens of breath for analysis by means of a device of a type approved by the Secretary of State." There is nothing about the approved device having to have an accurate clock. The Lion Intoximeter was an approved type of device and it seems to me impossible to argue that, by reason of the inaccuracy in its clock, the device at Widnes Police Station could no longer be described as a Lion Intoximeter. The point is reinforced by subsection 3(b), which says that a constable may not require a specimen of blood or urine to be provided at a police station except in specified circumstances, one of which is that ". . . a device or a reliable device of the type mentioned in subsection 1(a) is not available at the police station. . . ." This clearly contemplates, as one would expect, that a device may be "unreliable" and yet of the type approved by the Secretary of State.

          Secondly, it was argued that the inaccuracy in the clock was a "reasonable excuse" for Mr. Jones's failure to provide a specimen. The difficulty about this argument is that Mr. Jones never claimed to have the slightest notion that there was anything wrong with the time on the clock. His defence before the justices, on which he was disbelieved, was that the inadequacy of the specimen had not been his fault. It is therefore hard to see how the clock could have been an "excuse" for his failure to provide a specimen. I do not say that an excuse must be something which the accused had in mind when he failed or refused to provide the specimen. It would, for example, have been an excuse that he was unconscious when the request was made. But the concept of an excuse requires that it must have been a reason why the specimen was not provided. There must have been some causal connection between the excuse and the failure to provide the specimen. In this case, there was none.

          In any case, even if Mr. Jones had noticed that the clock was slow, I do not think that this would have been an excuse. If it was obvious that the device was in no state to analyse his breath correctly, he might reasonably have objected to going through a pointless exercise. But this would have been far from obvious and the evidence before the justices showed that it would have been wrong.

          Finally, it was submitted for Mr. Jones that the only admissible evidence that his second breath specimen was inadequate was the computer reading showing that the test had aborted. If this was inadmissible for failure to comply with section 69(1), he could not be convicted. The arguments against the admissibility of the computer evidence were the same as those which I have already rejected in the case of Ms McKeown. It follows that in my view this point also fails and the Director's appeal in Mr. Jones's case should also be allowed.

LORD CLYDE

My Lords,

          I have had the advantage of reading in draft the speech prepared by my noble and learned friend Lord Hoffmann. For the reasons which he gives I would allow the Director's appeals in both cases.