J. Forens. Sci. Soc. (1975), 15, 257
Information in Forensic Science V. J. EMERSON Home Ofice Central Research Establishment, Aldermaston, Reading, Berks, RG7 4PN, England This paper reports the development of a central Home Ofice Forensic Science Information Service. The collection, collation and dissemination of literature, commercial and analytical information is also described as is the systematic collection and analysis of casework data and the monitoring of analytical methods.
Introduction I n the late 1930's when the Home Office first established its Regional Forensic Laboratories, the scientists working in those laboratories had precious little background information and only rudimentary apparatus. They were few in number and relied mainly on their own scientific expertise and their enquiring and agile minds to enable them to devise and develop ways in which this expertise could be applied to crime detection. I t was therefore not really surprising that each laboratory developed its own methods of examination and analysis and that these varied from laboratory to laboratory. Neither is it surprising that although the scientists generally examined most things submitted to them, they each had their own particular interest which sometimes meant that a greater degree of expertise in a particular subject was available a t one laboratory. There were particular examples of this, as in the case of document examination, which was carried out for the whole country at the laboratory in Cardiff, and in the case of ballistics most of which was carried out at the laboratory in Nottingham. Apart from a few standard works on the subject of Forensic Science (Jago, 1909; Robertson, 1949; Lucas, 1948; Smith, 1940) which described various methods used in the scientific investigation of crime, the only information which helped interpret the case findings was retained in the expertise of the senior staff. As the years passed this was passed on as the size of the service increased. I n the last ten years the Forensic Science Service in Great Britain has expanded at a faster rate than ever before. With this growth the interlaboratory communications and co-operation have also been greatly improved. During these ten years the service has seen the advent of the Home Office Central Research Establishment (HOCRE) whose task is basically to provide the necessary research and development for the United Kingdom Forensic Science Service. With the inception of HOCRE came the idea of a centralized information service and from a purely logical basis, apart from the obvious economic one, its implementation would enable a greater degree of effort to be put into this particular field, instead of each laboratory providing its own information when and where it was required. The information service of HOCRE started in 1967 and since then it has moved from a one man service to a flourishing division of 10 people. The original concept of the division as reported by Curry (1967; 1970) has changed to a broader based theme taking in all the various aspects of information. Also the acquisition of a mini-computer has facilitated the rapid processing of data. The five main projects of the division are: Information collection and presentation; Implementation of data banks; Systems analysis;
Communication links ; Quality control.
Information Collection and Presentation The numbers of scientific papers published every year have been variously reported, but suffice to say that it is many hundreds of thousands, and with Forensic Science covering such a broad spectrum of scientific interests, the task of screening them all for relevant articles is mammoth if not Herculean. The way in which this problem has been tackled falls into four main areas. Firstly, HOCRE itself takes 60 journals each of which is thoroughly checked, within days of its arrival, by a member of the division and another senior member of staff. The journal is then placed in the library for the benefit of the rest of the staff. During this checking procedure papers which are of immediate interest to the service as a whole are marked for "circulation". This means that a copy is taken and retained in the information room, and a t the end of each month further copies of these papers are taken and placed in folders under the headings "Biology", "Chemistry", "Toxicology" and "General & Law". These are then sent to all the Home Office Laboratories as a "current awareness" circulation. Papers that do not come into this category, but are none the less thought to be important, not only to the regional laboratories but also to the scientists a t HOCRE for research information are also copied and retained in the information room. Secondly, publications are covered by running tailored profiles on the United Kingdom Chemical Information Service (UKCIS) computer. This covers 13,000 journals contained in Chemical Abstracts condensates for the chemist and Biological Abstract Previews for the biologist. These produce about 100 references a month of which some 40% are obtained through the National Lending Library (NLL) for further scrutiny. As before all papers needed either for circulation or for general storage are placed in the information room. Thirdly, Current Contents (Life Sciences) are thoroughly checked each week. These cover approximately 1,100 journals, of which about 100 provide most of the "alerts" with occasional papers appearing in about another 200 journals. These "alerts" produce several hundred papers a year, some of which are in journals which are taken at HOCRE but not received until some time later. The papers in journals not taken are obtained in the usual way, either from local libraries or the NLL. Lastly, and by no means least, is the method which has been developed by co-operation with other Governmental Establishments. These Establishments now send to HOCRE, as a normal process, either a copy of each of their reports, as in the case of the Transport and Road Research Laboratory, or a list of their publications, as in the cases of the Microbiology Establishment and the Joint Fire Research Organisation. Lists of interesting papers which are held by the Ministry of Defence are also obtained. By all these means over 200 papers each month are collected of which some 15% are sent to the regional laboratories in the "current awareness" circulation. 14,000 papers have now been collected in this way. I t became abundantly clear in the early days of the information division that all of these papers would have to be stored and an easy means of retrieval would have to be developed. Advice was sought from the Organizations and Methods branch of the Home Office and H M Treasury. The recommendation was for 16mm cassetted microfilm to be used as the storage medium; in this way some 2,000 sheets of paper can be stored in a pocket size box. All regional laboratories were supplied with reader printers so that the films could be easily read and full size paper copies taken when the need arose. Filmed copies of all the papers collected and held at HOCRE are now supplied to all regional laboratories, and the "current awareness" circulation which initially is supplied in hardcopy is also filmed and indexed a t the end of each year. 258
The retrieval of papers has been achieved by the use of a keyword system using a mini-computer. Originally a time shared Burroughs computer in Brussels was used, linked to HOCRE Information Division by telex for the interrogation of the literature file. This service has now been discontinued and new programs have been written for the "in house" Hewlett-Packard 2100A computer. More details of these particular programs are yet to be published, but the basic system is as follows: Each paper is allocated a number which is the computer accession number, and then the contents of the paper are described by "keywords" which are contained in a thesaurus of some 3,500 words, synonyms, or dates. This thesaurus was drawn up by a group of scientists a t HOCRE who endeavoured to use the minimum number of words which could be capable of aptly describing any scientific paper likely to be of interest to the Forensic Science Service. (Inevitably there will come new techniques, chemicals and terms which will have to be added to the thesaurus from time to time, however it is most important that if this is done no previous record is still on file which should have that keyword on it, otherwise the paper will not be retrieved on interrogation.) The accession number together with all the keywords describing that paper are then fed onto the computer file. Retrieval of information in this way is achieved by interrogating the file by requesting a search to be made on a keyword or words which describe the problem of the enquirer. The file is then searched and the accession numbers of all the papers which have that keyword or combination of keywords is printed out in reverse order, the most recent paper number will head the list. Needless to say that if a common term is used a large number will be cited, and likewise if a collection of very unusual keywords are used then it is possible that no paper will be turned up. I t therefore rests very much with the interrogator as to how the file is searched. HOCRE personnel can take the accession numbers provided by the search, and then refer directly to the papers stored in the Information Room. Scientists in the regional laboratories are provided with a list of accession numbers relevant to their enquiry and can then go to their own microfilm copy of papers and check through to find the relevant information. Because the number of journals searched by various means is large there is always the need to obtain references, or copies of papers, or the loan of actual journals from outside sources. These sources are the libraries of the following establishments: The Atomic Weapons Research Establishment (AWRE), Aldermaston; The National Institute for Research in Dairying (NIRD), Shinfield; The Police College, Bramshill; The University, Reading; as well as the NLL a t Boston Spa. The Establishment owes a debt of gratitude to all these libraries. Regional laboratories from time to time have the need to follow a particular piece of short term research for a specific case; they may also have personnel with expertise in a particular field. I n these cases the scientists concerned regularly supply information to HOCRE and thus the service as a whole. Often this is at a prepublication stage and this co-operation is appreciated by all concerned. This paper so far has briefly outlined the standard means of collecting, collating and disseminating literature information, but there is another r61e which is performed by the information room and that is to follow up enquiries from regional scientists for commercial information. There are many cases which arise every year when an everyday product is involved and it is naturally a much easier task for identification or comparison if the formulation is known. This service is centralized for several reasons, the most important being that the Forensic Science Service relies on the co-operation of manufacturers in supplying this information, and if only one establishment is asking for assistance there is less chance of duplicating enquiries, and so an efficient service can be more easily maintained.
I n some cases the enquiry for this type of information will produce an answer which is concise, for example an enquiry into the constituent of a particular adhesive, and can be passed on quickly. I n other cases certain substances can be identified which lead a scientist to believe that it can be classified as say a soldering flux, but he wishes to identify the particular brand. Such enquiries are often lengthy and in some cases information is passed to a regional scientist, which requires him to complete the full investigation. However, with kind and full co-operation from colleagues in regional laboratories, information is fed back to HOCRE so that should this type of enquiry occur again, part if not all the information is already available.
Implementation of Data Banks Data banks are not by any means a recent innovation, all laboratories dealing with routine analysis problems have kept their own collection of reference data. I n the early days this was merely a collection of such details as melting points and Rf's on paper chromatograms, and then spread to UV, IR and TLC data, and perhaps more recently to mass spectra. Needless to say it is far from economic for each laboratory to put the effort and storage space into making its own comprehensive collection. This therefore started off the concept of centralized data banks a t HOCRE for the whole of the Forensic Science Service. The first bank of data to be set up with that of IR spectra and details of this together with a simple retrieval system using optical coincidence cards have been reported by Curry et al. (1969). In this system a collection of some 1,000 spectra were collected, coded and copied onto 16mm microfilm. The retrieval system was also reduced, copied and distributed so that each of the regional laboratories had their own copy of the retrieval system together with copies of the spectra to which they could refer. This system suffered from the problem of updating. Whereas the task of providing fiche copies of extra spectra was a simple one, continuously adding to the optical coincidence cards was far from simple. This was mainly due to the reduction and reproduction of the system for each laboratory which involved an expensive photographic process followed by accurate guillotining. When the mini computer came into use the whole problem was solved and it enabled all spectra to be put on file by a specified number and by the wavelengths of its six major peaks. This naturally meant that each laboratory did not possess its own retrieval system but had to resort to the use of HOCRE information room. However, six major peaks passed over a telephone line, followed by computer interrogation of the IR file, resulted in a few numbers being quoted to the enquirer which could very quickly be inspected on the microfilm. The final identification of a spectrum is always left to the regional scientist himself and the system merely acts as a means of reducing the number of possibilities. The bank of IR data has continued to grow and now contains more than 3,000 spectra which are all in the regional laboratories on microfilm and which are backed by the computerized retrieval. The infrared file is a good example of a bank of analytical data, but several others are also in existence; these include a collection of over 1,000 UV spectra of basic drugs in acid solutions, which although not computerized have been indexed and placed on microfilm in ascending order of absorption maxima. There is also a file on agricultural chemicals with analytical details. With the advent of mass spectrometry into forensic science and the fact that some of the regional laboratories will shortly be owning their own instrument there is the need for a collection to be made of the mass spectra of the more commonly encountered chemicals and drugs. The file for interpretation of mass spectra can be dealt with in two ways, which in fact for the Forensic Science Service will be complementary. The first will be by providing a library of normalized spectra of the substances most likely to be encountered in a regional laboratory. This will be prepared by the use of the computer to normalize the
spectra and plot them on the visual display unit, complete with major peaks and intensities. A hardcopy of these can then be taken and produced in a book for the regional laboratories. The library will be backed by a computerized retrieval system based on a n eight peak index. I n this way a service similar to that for IR spectra can be provided for the regional laboratories. The second way will be by placing on a computer file, by means of Wiswesser notation, a large number of substances which are less likely to be encountered and for which mass spectra will not have been generated. (Wiswesser Line Notation is a method of representing chemical formulae in a form compatible with a computer. The computer file can be searched for various structural fragments which have been identified by spectroscopic techniques.) I t is anticipated that this will be necessary to answer some of the more obscure problems. I n this way those laboratories starting out into the mass spectrometry field will have the benefit of both of these systems to assist with their interpretation. There is, of course, yet another way in which mass spectra can be interpreted and that is by means of the service provided by the Mass Spectrometry Data Centre via the Honeywell time share system. This service is commercially available and provides access by telephone to a collection of spectra stored on a large computer. I t can be considered as a final reference for the identification of difficult spectra. However it is not only in the analytical field that there is a need for data banks. There are several cases where a bank of information on a specific topic is necessary for the correct interpretation of evidence found at scenes of crime. To cover this side of data collection, contracts are placed with various commercial bodies for the supply of specific collections. I n this way collections of photomicrographs of hard and soft woods have been prepared and supplied. A collection of all the sole and heel patterns produced by British shoe manufacturers has also been obtained. A collection of tyre patterns found in the United Kingdom is at present being assembled under contract. With the kind co-operation of manufacturers, a collection of headlamp lenses found on cars in the United Kingdom has been obtained as well as a collection of pharmaceuticals. Obviously the number of collections that could be made are legion, but time and staff do not permit us to undertake them all. Those mentioned have been considered to be the most important. None of these collections is claimed to be complete, but great efforts are made to achieve as complete a collection as possible.
Systems Analysis The expertise of the individual in assessing the significance of evidence is obviously passed on from the more senior members of staff, but unfortunately the imparting of such knowledge is not always complete before the retirement of such senior staff. The knowledge therefore retires with them and in some way this knowledge must be stored for the benefit of the whole service, so that new staff can have access to the past experiences of the service. Currently more than 100,000 case records are completed annually by the nine forensic science laboratories in England and Wales. Each laboratory has its own filing system, usually by means of sequential numbering. I t is therefore very difficult for information contained in these files to be readily extracted. Occasionally when a particular piece of information is required, the laborious task of going through these case files is carried out. The problems that then arise are, was the sample a random one, was it truly representative, was it in fact large enough? Clearly it is desirable to collect data continuously so that up to date information is always available. Already some information is being collected, because on a general basis the Home Office requires each laboratory to make half yearly returns on the number of cases and exhibits received, and these are split into 18 classifications. For several years now, the toxicologists have been keeping data on the drug 261
levels in blood and tissue received a t the laboratories and these have been converted at HOCRE into the "Register of Human Toxicology". Such information has been found invaluable on several occasions in assisting the pathologist with the interpretation of the results submitted to him. Also for several years all laboratories have been recording the origin and physical properties of control glass samples submitted in case work. This data has been collected centrally at H O C R E where the results have been analysed and tables of probabilities have been drawn up which are designed to help the regional scientist interpret the significance of his results for the courts. All of these have proved useful, and it has become obvious to most of the scientists who have used such data that it is desirable for this type of data collection to be carried out across the board of forensic case work. Coleman and Walls (1974) have written that the interpretation of scientific evidence should be more objective, and that surveys should be carried out to produce the data for such an objective assessment. There is an immense problem both from a time and labour aspect to obtain a truly random selection of articles for sampling and to ensure data collection. However, if all laboratories collect all the relevant information from case work exhibits on a similar basis, a tremendous amount of useful information can be acquired. I t is with this in mind that the systems analysis project was started at HOCRE over 18 months ago with the major objectives as follows: The provision of statistical information on control and suspect samples in specific areas (including those systems already in existence) ; The provision of information for planning research policy at HOCRE. Such a project clearly will take time before it is fully operational. There are several reasons for this but one most important aspect is the design of the information collection forms. I t is essential that the regional scientist, who will be responsible for their completion, should feel the need not only of the whole scheme but also for the specific points of data which he is expected to supply. I t is because of this that to date, through the various specialist committees drawn from the staff of regional laboratories, forms for the collection of data in eight specific areas have been designed and evaluated. Pilot trials have been run and in fact in at least one area continuous data collection is being carried out. Already scientists have been able to back their intuitive thoughts with figures which may be applicable either regionally or nationally. The project is continuing.
Communication Links The whole information division and the service which it provides have grown and the latter now extends throughout the British Isles. I t has been said that one of the problems which beset the dinosaur was that its brain became too far away from its belly. Not wishing to meet the same fate as such a noble beast, HOCRE Information Division has been looking very closely into the communications system within the service in an attempt to see that the interchange of information between two points can be carried out with the maximum speed and with the minimum delay. The general modes of information transfer in the past have been telephone and postal services. Each of these suffer drawbacks, and each of these involve a personal interface. I t has been suggested that removal of such an interface may increase the number of enquiries by as much as tenfold, and it is clearly desirable that all laboratories should have easy and ready access to all the data banks established a t HOCRE. I n an attempt to establish the preferred form of inter-laboratory communication the following systems are under consideration and pilot trials are in the process of being carried out: Telex links between two laboratories and H O C R E ; Facsimile links between two laboratories and HOCRE ; A Computer terminal in a laboratory connected to the HOCRE computer.
The pilot trials a t present in operation indicate that telex is a very favourable form of communication because both sender and receiver have a printed record of the original request and the replies sent. I t also makes the enquirer frame a concise and specific request which can be handled more easily. Names and figures also are not misunderstood or inaccurately recorded and transmissions are fast and can be sent whether there is anyone at the receiving end or not. The main disadvantage appears to be the necessity for a competent operator rather than each individual scientist performing a one finger exercise, whether he be cutting a tape or communicating direct. The use of facsimile machines has proved far more specialized and does involve a n operator at each end at least at the beginning and end of the transmission. However, its main advantage over any other means of communication is its ability to transmit graphical data very quickly. A photograph, photomicrograph or infrared spectrum can be transmitted in as little as four minutes and received in a legible form. There have been occasions when it has been very important to transmit such data quickly and the facsimile machine has, under these circumstances, proved invaluable. However, it is extremely difficult under such circumstances to try and calculate the cost effectiveness of such a system. The third pilot trial is about to be started. A Hewlett-Packard 2100A computer with only 16K core does not provide an adequate true time sharing capability. However, by means of two acoustic couplers and a portable teletype it has been possible to interrogate the computer files from a remote station via a standard telephone line. While it is accepted that there is still the need for a person to place the telephone receiver into the acoustic coupler and for the computer to be switched on, once the connection has been made the remote station has complete access to all the files for the time allocated to it by the information room. This is designed to try to establish the usage when direct access is preferred as opposed to the access via a human interface. Inevitably this will necessitate a trained operator in the laboratory used for the experiment, and should the usage and facilities be extended to all the regional laboratories, then trained operators would be required in each laboratory. Not wishing to anticipate the results of the experiment there is little that can be said at this stage. However, with all the regional laboratories growing in size and in information needs, the situation is fast approaching where an officer in each laboratory could well be employed to deal with the interchange of information with HOCRE. Some laboratories have in fact started this and a greater degree of efficiency has been achieved in these cases. We look forward perhaps optimistically to the time when all the regional forensic science laboratories have their own computer terminal, connected, perhaps by means of' a private line, to a large central computer where all the data collected and processed by HOCRE can be accessed individually by them.
Quality Control This is a process used by industry to ensure that the quality of the goods, allowed from a particular firm on to the market, is at the desired level. I t is a process used by biochemists to ensure that results produced in various laboratories are accurate as well as precise and are thus capable of interpretation by any hospital. I t is a process which is in use in the forensic science service to ensure that the results obtained at all the regional laboratories are accurate as well as precise and that these are not affected by the fact that apparatus, techniques and operators vary between laboratories. This quality control is carried out through various specialist committees who decide to look a t a particular analysis or analytical technique used throughout the service. I t is then the function of Information Division of HOCRE to send out samples to each laboratory and collect and analyse the results obtained. The results are
then discussed by the appropriate committees and in some cases suggestions for the improvement of techniques are put forward. This is used for two main purposes, either the checking of improvements to existing methods, or the testing and evaluating of new methods before they are used in routine casework, and both of these are vitally important so that all regional scientists know that the service they are supplying is of high quality and also standard. I n all these ways the information service works to assist the scientists in all the regional laboratories. The train of information starts with providing the necessary literature which will furnish the scientist with new and relevant methods and techniques. I t passes on to the data banks which help identify the extracted material or the source of the material and the systems project enables the scientist to determine the frequency of occurrence and hence interpret the significance of the evidence to the courts. The most efficient and accurate means of communication are being evaluated so that this information can be imparted to the regions in the optimum way, and finally the techniques used in laboratories are monitored so that a uniform level of precision and accuracy is maintained. References COLEMAN, R. F. and WALLS,H. J., 1974, Crim. Law. Review, p. 276. CURRY, A. S., 1967, Chemistry in Britain, 5, 501. CURRY, A. S., 1970, J. Forens. Sci. Soc., 10, 151. C., 1969, J. Pharm. Pharmac., 21, 224. CURRY,A. S., READ,J. F. and BROWN, JAGO,W., 1909, Forensic Chemistry and Chemical Evidence, Stevens & Haynes, London. LUCAS,A., 1948, Forensic Chemistry and ScientiJic Criminal Investigation, 4th Ed, Arnold, London. W. G. A., 1949, Aids to Forensic Medicine and Toxicology, (ed RyEel, ROBERTSON, J. H.), 12th Ed, Baillitre, Tindall & Cox, London. SMITH,S., 1940, Forensic Medicine, 7th Ed, Churchill, London.
Information in Forensic Science V. J. EMERSON Home Ofice Central Research Establishment, Aldermaston, Reading, Berks, RG7 4PN, England This paper reports the development of a central Home Ofice Forensic Science Information Service. The collection, collation and dissemination of literature, commercial and analytical information is also described as is the systematic collection and analysis of casework data and the monitoring of analytical methods.
Introduction I n the late 1930's when the Home Office first established its Regional Forensic Laboratories, the scientists working in those laboratories had precious little background information and only rudimentary apparatus. They were few in number and relied mainly on their own scientific expertise and their enquiring and agile minds to enable them to devise and develop ways in which this expertise could be applied to crime detection. I t was therefore not really surprising that each laboratory developed its own methods of examination and analysis and that these varied from laboratory to laboratory. Neither is it surprising that although the scientists generally examined most things submitted to them, they each had their own particular interest which sometimes meant that a greater degree of expertise in a particular subject was available a t one laboratory. There were particular examples of this, as in the case of document examination, which was carried out for the whole country at the laboratory in Cardiff, and in the case of ballistics most of which was carried out at the laboratory in Nottingham. Apart from a few standard works on the subject of Forensic Science (Jago, 1909; Robertson, 1949; Lucas, 1948; Smith, 1940) which described various methods used in the scientific investigation of crime, the only information which helped interpret the case findings was retained in the expertise of the senior staff. As the years passed this was passed on as the size of the service increased. I n the last ten years the Forensic Science Service in Great Britain has expanded at a faster rate than ever before. With this growth the interlaboratory communications and co-operation have also been greatly improved. During these ten years the service has seen the advent of the Home Office Central Research Establishment (HOCRE) whose task is basically to provide the necessary research and development for the United Kingdom Forensic Science Service. With the inception of HOCRE came the idea of a centralized information service and from a purely logical basis, apart from the obvious economic one, its implementation would enable a greater degree of effort to be put into this particular field, instead of each laboratory providing its own information when and where it was required. The information service of HOCRE started in 1967 and since then it has moved from a one man service to a flourishing division of 10 people. The original concept of the division as reported by Curry (1967; 1970) has changed to a broader based theme taking in all the various aspects of information. Also the acquisition of a mini-computer has facilitated the rapid processing of data. The five main projects of the division are: Information collection and presentation; Implementation of data banks; Systems analysis;
Communication links ; Quality control.
Information Collection and Presentation The numbers of scientific papers published every year have been variously reported, but suffice to say that it is many hundreds of thousands, and with Forensic Science covering such a broad spectrum of scientific interests, the task of screening them all for relevant articles is mammoth if not Herculean. The way in which this problem has been tackled falls into four main areas. Firstly, HOCRE itself takes 60 journals each of which is thoroughly checked, within days of its arrival, by a member of the division and another senior member of staff. The journal is then placed in the library for the benefit of the rest of the staff. During this checking procedure papers which are of immediate interest to the service as a whole are marked for "circulation". This means that a copy is taken and retained in the information room, and a t the end of each month further copies of these papers are taken and placed in folders under the headings "Biology", "Chemistry", "Toxicology" and "General & Law". These are then sent to all the Home Office Laboratories as a "current awareness" circulation. Papers that do not come into this category, but are none the less thought to be important, not only to the regional laboratories but also to the scientists a t HOCRE for research information are also copied and retained in the information room. Secondly, publications are covered by running tailored profiles on the United Kingdom Chemical Information Service (UKCIS) computer. This covers 13,000 journals contained in Chemical Abstracts condensates for the chemist and Biological Abstract Previews for the biologist. These produce about 100 references a month of which some 40% are obtained through the National Lending Library (NLL) for further scrutiny. As before all papers needed either for circulation or for general storage are placed in the information room. Thirdly, Current Contents (Life Sciences) are thoroughly checked each week. These cover approximately 1,100 journals, of which about 100 provide most of the "alerts" with occasional papers appearing in about another 200 journals. These "alerts" produce several hundred papers a year, some of which are in journals which are taken at HOCRE but not received until some time later. The papers in journals not taken are obtained in the usual way, either from local libraries or the NLL. Lastly, and by no means least, is the method which has been developed by co-operation with other Governmental Establishments. These Establishments now send to HOCRE, as a normal process, either a copy of each of their reports, as in the case of the Transport and Road Research Laboratory, or a list of their publications, as in the cases of the Microbiology Establishment and the Joint Fire Research Organisation. Lists of interesting papers which are held by the Ministry of Defence are also obtained. By all these means over 200 papers each month are collected of which some 15% are sent to the regional laboratories in the "current awareness" circulation. 14,000 papers have now been collected in this way. I t became abundantly clear in the early days of the information division that all of these papers would have to be stored and an easy means of retrieval would have to be developed. Advice was sought from the Organizations and Methods branch of the Home Office and H M Treasury. The recommendation was for 16mm cassetted microfilm to be used as the storage medium; in this way some 2,000 sheets of paper can be stored in a pocket size box. All regional laboratories were supplied with reader printers so that the films could be easily read and full size paper copies taken when the need arose. Filmed copies of all the papers collected and held at HOCRE are now supplied to all regional laboratories, and the "current awareness" circulation which initially is supplied in hardcopy is also filmed and indexed a t the end of each year. 258
The retrieval of papers has been achieved by the use of a keyword system using a mini-computer. Originally a time shared Burroughs computer in Brussels was used, linked to HOCRE Information Division by telex for the interrogation of the literature file. This service has now been discontinued and new programs have been written for the "in house" Hewlett-Packard 2100A computer. More details of these particular programs are yet to be published, but the basic system is as follows: Each paper is allocated a number which is the computer accession number, and then the contents of the paper are described by "keywords" which are contained in a thesaurus of some 3,500 words, synonyms, or dates. This thesaurus was drawn up by a group of scientists a t HOCRE who endeavoured to use the minimum number of words which could be capable of aptly describing any scientific paper likely to be of interest to the Forensic Science Service. (Inevitably there will come new techniques, chemicals and terms which will have to be added to the thesaurus from time to time, however it is most important that if this is done no previous record is still on file which should have that keyword on it, otherwise the paper will not be retrieved on interrogation.) The accession number together with all the keywords describing that paper are then fed onto the computer file. Retrieval of information in this way is achieved by interrogating the file by requesting a search to be made on a keyword or words which describe the problem of the enquirer. The file is then searched and the accession numbers of all the papers which have that keyword or combination of keywords is printed out in reverse order, the most recent paper number will head the list. Needless to say that if a common term is used a large number will be cited, and likewise if a collection of very unusual keywords are used then it is possible that no paper will be turned up. I t therefore rests very much with the interrogator as to how the file is searched. HOCRE personnel can take the accession numbers provided by the search, and then refer directly to the papers stored in the Information Room. Scientists in the regional laboratories are provided with a list of accession numbers relevant to their enquiry and can then go to their own microfilm copy of papers and check through to find the relevant information. Because the number of journals searched by various means is large there is always the need to obtain references, or copies of papers, or the loan of actual journals from outside sources. These sources are the libraries of the following establishments: The Atomic Weapons Research Establishment (AWRE), Aldermaston; The National Institute for Research in Dairying (NIRD), Shinfield; The Police College, Bramshill; The University, Reading; as well as the NLL a t Boston Spa. The Establishment owes a debt of gratitude to all these libraries. Regional laboratories from time to time have the need to follow a particular piece of short term research for a specific case; they may also have personnel with expertise in a particular field. I n these cases the scientists concerned regularly supply information to HOCRE and thus the service as a whole. Often this is at a prepublication stage and this co-operation is appreciated by all concerned. This paper so far has briefly outlined the standard means of collecting, collating and disseminating literature information, but there is another r61e which is performed by the information room and that is to follow up enquiries from regional scientists for commercial information. There are many cases which arise every year when an everyday product is involved and it is naturally a much easier task for identification or comparison if the formulation is known. This service is centralized for several reasons, the most important being that the Forensic Science Service relies on the co-operation of manufacturers in supplying this information, and if only one establishment is asking for assistance there is less chance of duplicating enquiries, and so an efficient service can be more easily maintained.
I n some cases the enquiry for this type of information will produce an answer which is concise, for example an enquiry into the constituent of a particular adhesive, and can be passed on quickly. I n other cases certain substances can be identified which lead a scientist to believe that it can be classified as say a soldering flux, but he wishes to identify the particular brand. Such enquiries are often lengthy and in some cases information is passed to a regional scientist, which requires him to complete the full investigation. However, with kind and full co-operation from colleagues in regional laboratories, information is fed back to HOCRE so that should this type of enquiry occur again, part if not all the information is already available.
Implementation of Data Banks Data banks are not by any means a recent innovation, all laboratories dealing with routine analysis problems have kept their own collection of reference data. I n the early days this was merely a collection of such details as melting points and Rf's on paper chromatograms, and then spread to UV, IR and TLC data, and perhaps more recently to mass spectra. Needless to say it is far from economic for each laboratory to put the effort and storage space into making its own comprehensive collection. This therefore started off the concept of centralized data banks a t HOCRE for the whole of the Forensic Science Service. The first bank of data to be set up with that of IR spectra and details of this together with a simple retrieval system using optical coincidence cards have been reported by Curry et al. (1969). In this system a collection of some 1,000 spectra were collected, coded and copied onto 16mm microfilm. The retrieval system was also reduced, copied and distributed so that each of the regional laboratories had their own copy of the retrieval system together with copies of the spectra to which they could refer. This system suffered from the problem of updating. Whereas the task of providing fiche copies of extra spectra was a simple one, continuously adding to the optical coincidence cards was far from simple. This was mainly due to the reduction and reproduction of the system for each laboratory which involved an expensive photographic process followed by accurate guillotining. When the mini computer came into use the whole problem was solved and it enabled all spectra to be put on file by a specified number and by the wavelengths of its six major peaks. This naturally meant that each laboratory did not possess its own retrieval system but had to resort to the use of HOCRE information room. However, six major peaks passed over a telephone line, followed by computer interrogation of the IR file, resulted in a few numbers being quoted to the enquirer which could very quickly be inspected on the microfilm. The final identification of a spectrum is always left to the regional scientist himself and the system merely acts as a means of reducing the number of possibilities. The bank of IR data has continued to grow and now contains more than 3,000 spectra which are all in the regional laboratories on microfilm and which are backed by the computerized retrieval. The infrared file is a good example of a bank of analytical data, but several others are also in existence; these include a collection of over 1,000 UV spectra of basic drugs in acid solutions, which although not computerized have been indexed and placed on microfilm in ascending order of absorption maxima. There is also a file on agricultural chemicals with analytical details. With the advent of mass spectrometry into forensic science and the fact that some of the regional laboratories will shortly be owning their own instrument there is the need for a collection to be made of the mass spectra of the more commonly encountered chemicals and drugs. The file for interpretation of mass spectra can be dealt with in two ways, which in fact for the Forensic Science Service will be complementary. The first will be by providing a library of normalized spectra of the substances most likely to be encountered in a regional laboratory. This will be prepared by the use of the computer to normalize the
spectra and plot them on the visual display unit, complete with major peaks and intensities. A hardcopy of these can then be taken and produced in a book for the regional laboratories. The library will be backed by a computerized retrieval system based on a n eight peak index. I n this way a service similar to that for IR spectra can be provided for the regional laboratories. The second way will be by placing on a computer file, by means of Wiswesser notation, a large number of substances which are less likely to be encountered and for which mass spectra will not have been generated. (Wiswesser Line Notation is a method of representing chemical formulae in a form compatible with a computer. The computer file can be searched for various structural fragments which have been identified by spectroscopic techniques.) I t is anticipated that this will be necessary to answer some of the more obscure problems. I n this way those laboratories starting out into the mass spectrometry field will have the benefit of both of these systems to assist with their interpretation. There is, of course, yet another way in which mass spectra can be interpreted and that is by means of the service provided by the Mass Spectrometry Data Centre via the Honeywell time share system. This service is commercially available and provides access by telephone to a collection of spectra stored on a large computer. I t can be considered as a final reference for the identification of difficult spectra. However it is not only in the analytical field that there is a need for data banks. There are several cases where a bank of information on a specific topic is necessary for the correct interpretation of evidence found at scenes of crime. To cover this side of data collection, contracts are placed with various commercial bodies for the supply of specific collections. I n this way collections of photomicrographs of hard and soft woods have been prepared and supplied. A collection of all the sole and heel patterns produced by British shoe manufacturers has also been obtained. A collection of tyre patterns found in the United Kingdom is at present being assembled under contract. With the kind co-operation of manufacturers, a collection of headlamp lenses found on cars in the United Kingdom has been obtained as well as a collection of pharmaceuticals. Obviously the number of collections that could be made are legion, but time and staff do not permit us to undertake them all. Those mentioned have been considered to be the most important. None of these collections is claimed to be complete, but great efforts are made to achieve as complete a collection as possible.
Systems Analysis The expertise of the individual in assessing the significance of evidence is obviously passed on from the more senior members of staff, but unfortunately the imparting of such knowledge is not always complete before the retirement of such senior staff. The knowledge therefore retires with them and in some way this knowledge must be stored for the benefit of the whole service, so that new staff can have access to the past experiences of the service. Currently more than 100,000 case records are completed annually by the nine forensic science laboratories in England and Wales. Each laboratory has its own filing system, usually by means of sequential numbering. I t is therefore very difficult for information contained in these files to be readily extracted. Occasionally when a particular piece of information is required, the laborious task of going through these case files is carried out. The problems that then arise are, was the sample a random one, was it truly representative, was it in fact large enough? Clearly it is desirable to collect data continuously so that up to date information is always available. Already some information is being collected, because on a general basis the Home Office requires each laboratory to make half yearly returns on the number of cases and exhibits received, and these are split into 18 classifications. For several years now, the toxicologists have been keeping data on the drug 261
levels in blood and tissue received a t the laboratories and these have been converted at HOCRE into the "Register of Human Toxicology". Such information has been found invaluable on several occasions in assisting the pathologist with the interpretation of the results submitted to him. Also for several years all laboratories have been recording the origin and physical properties of control glass samples submitted in case work. This data has been collected centrally at H O C R E where the results have been analysed and tables of probabilities have been drawn up which are designed to help the regional scientist interpret the significance of his results for the courts. All of these have proved useful, and it has become obvious to most of the scientists who have used such data that it is desirable for this type of data collection to be carried out across the board of forensic case work. Coleman and Walls (1974) have written that the interpretation of scientific evidence should be more objective, and that surveys should be carried out to produce the data for such an objective assessment. There is an immense problem both from a time and labour aspect to obtain a truly random selection of articles for sampling and to ensure data collection. However, if all laboratories collect all the relevant information from case work exhibits on a similar basis, a tremendous amount of useful information can be acquired. I t is with this in mind that the systems analysis project was started at HOCRE over 18 months ago with the major objectives as follows: The provision of statistical information on control and suspect samples in specific areas (including those systems already in existence) ; The provision of information for planning research policy at HOCRE. Such a project clearly will take time before it is fully operational. There are several reasons for this but one most important aspect is the design of the information collection forms. I t is essential that the regional scientist, who will be responsible for their completion, should feel the need not only of the whole scheme but also for the specific points of data which he is expected to supply. I t is because of this that to date, through the various specialist committees drawn from the staff of regional laboratories, forms for the collection of data in eight specific areas have been designed and evaluated. Pilot trials have been run and in fact in at least one area continuous data collection is being carried out. Already scientists have been able to back their intuitive thoughts with figures which may be applicable either regionally or nationally. The project is continuing.
Communication Links The whole information division and the service which it provides have grown and the latter now extends throughout the British Isles. I t has been said that one of the problems which beset the dinosaur was that its brain became too far away from its belly. Not wishing to meet the same fate as such a noble beast, HOCRE Information Division has been looking very closely into the communications system within the service in an attempt to see that the interchange of information between two points can be carried out with the maximum speed and with the minimum delay. The general modes of information transfer in the past have been telephone and postal services. Each of these suffer drawbacks, and each of these involve a personal interface. I t has been suggested that removal of such an interface may increase the number of enquiries by as much as tenfold, and it is clearly desirable that all laboratories should have easy and ready access to all the data banks established a t HOCRE. I n an attempt to establish the preferred form of inter-laboratory communication the following systems are under consideration and pilot trials are in the process of being carried out: Telex links between two laboratories and H O C R E ; Facsimile links between two laboratories and HOCRE ; A Computer terminal in a laboratory connected to the HOCRE computer.
The pilot trials a t present in operation indicate that telex is a very favourable form of communication because both sender and receiver have a printed record of the original request and the replies sent. I t also makes the enquirer frame a concise and specific request which can be handled more easily. Names and figures also are not misunderstood or inaccurately recorded and transmissions are fast and can be sent whether there is anyone at the receiving end or not. The main disadvantage appears to be the necessity for a competent operator rather than each individual scientist performing a one finger exercise, whether he be cutting a tape or communicating direct. The use of facsimile machines has proved far more specialized and does involve a n operator at each end at least at the beginning and end of the transmission. However, its main advantage over any other means of communication is its ability to transmit graphical data very quickly. A photograph, photomicrograph or infrared spectrum can be transmitted in as little as four minutes and received in a legible form. There have been occasions when it has been very important to transmit such data quickly and the facsimile machine has, under these circumstances, proved invaluable. However, it is extremely difficult under such circumstances to try and calculate the cost effectiveness of such a system. The third pilot trial is about to be started. A Hewlett-Packard 2100A computer with only 16K core does not provide an adequate true time sharing capability. However, by means of two acoustic couplers and a portable teletype it has been possible to interrogate the computer files from a remote station via a standard telephone line. While it is accepted that there is still the need for a person to place the telephone receiver into the acoustic coupler and for the computer to be switched on, once the connection has been made the remote station has complete access to all the files for the time allocated to it by the information room. This is designed to try to establish the usage when direct access is preferred as opposed to the access via a human interface. Inevitably this will necessitate a trained operator in the laboratory used for the experiment, and should the usage and facilities be extended to all the regional laboratories, then trained operators would be required in each laboratory. Not wishing to anticipate the results of the experiment there is little that can be said at this stage. However, with all the regional laboratories growing in size and in information needs, the situation is fast approaching where an officer in each laboratory could well be employed to deal with the interchange of information with HOCRE. Some laboratories have in fact started this and a greater degree of efficiency has been achieved in these cases. We look forward perhaps optimistically to the time when all the regional forensic science laboratories have their own computer terminal, connected, perhaps by means of' a private line, to a large central computer where all the data collected and processed by HOCRE can be accessed individually by them.
Quality Control This is a process used by industry to ensure that the quality of the goods, allowed from a particular firm on to the market, is at the desired level. I t is a process used by biochemists to ensure that results produced in various laboratories are accurate as well as precise and are thus capable of interpretation by any hospital. I t is a process which is in use in the forensic science service to ensure that the results obtained at all the regional laboratories are accurate as well as precise and that these are not affected by the fact that apparatus, techniques and operators vary between laboratories. This quality control is carried out through various specialist committees who decide to look a t a particular analysis or analytical technique used throughout the service. I t is then the function of Information Division of HOCRE to send out samples to each laboratory and collect and analyse the results obtained. The results are
then discussed by the appropriate committees and in some cases suggestions for the improvement of techniques are put forward. This is used for two main purposes, either the checking of improvements to existing methods, or the testing and evaluating of new methods before they are used in routine casework, and both of these are vitally important so that all regional scientists know that the service they are supplying is of high quality and also standard. I n all these ways the information service works to assist the scientists in all the regional laboratories. The train of information starts with providing the necessary literature which will furnish the scientist with new and relevant methods and techniques. I t passes on to the data banks which help identify the extracted material or the source of the material and the systems project enables the scientist to determine the frequency of occurrence and hence interpret the significance of the evidence to the courts. The most efficient and accurate means of communication are being evaluated so that this information can be imparted to the regions in the optimum way, and finally the techniques used in laboratories are monitored so that a uniform level of precision and accuracy is maintained. References COLEMAN, R. F. and WALLS,H. J., 1974, Crim. Law. Review, p. 276. CURRY, A. S., 1967, Chemistry in Britain, 5, 501. CURRY, A. S., 1970, J. Forens. Sci. Soc., 10, 151. C., 1969, J. Pharm. Pharmac., 21, 224. CURRY,A. S., READ,J. F. and BROWN, JAGO,W., 1909, Forensic Chemistry and Chemical Evidence, Stevens & Haynes, London. LUCAS,A., 1948, Forensic Chemistry and ScientiJic Criminal Investigation, 4th Ed, Arnold, London. W. G. A., 1949, Aids to Forensic Medicine and Toxicology, (ed RyEel, ROBERTSON, J. H.), 12th Ed, Baillitre, Tindall & Cox, London. SMITH,S., 1940, Forensic Medicine, 7th Ed, Churchill, London.
