Forensic firearm examination


Forensic firearm examination is the forensic process of examining the characteristics of firearms or bullets left behind at a crime scene. Specialists in this field are tasked with linking bullets to weapons and weapons to individuals. Obliterated serial numbers can be raised and recorded in an attempt to find the registered owner of the weapon. Nitric acid (HNO3) is the most common reagent used for this. Examiners can also look for fingerprints on the weapon and cartridges. Fingerprints are key pieces of evidence. If crime scene investigators find prints at a scene, they will be dusted, photographed, collected, and analyzed both by hand (using comparison microscopes) as well as compared to databases for potential references. By examining unique striations, scratches left behind on the bullet and weapon, individual fired rounds can be, but not always are, linked back to a specific weapon. These striations are due to the rifling inside the barrels of firearms. Rifling spins the bullet when it is fired out of the barrel to improve precision. Although striations are individualized evidence and will not match any other bullet or weapon, microscopic striations in the barrel of the weapon will change about every three to five shots. This is important because if attorneys wish to present ballistics evidence in court, it would be hard to prove beyond a reasonable doubt that one specific bullet would match one specific weapon. Forensic ballistics examiners may not fire more than five shots at most from a weapon found at a scene for this exact reason. Known exemplars taken from a seized weapon can be directly compared to samples recovered from the scene using a comparison microscope as well as newer 3-D imaging technology. Striation images can also be uploaded to any existing national databases. Furthermore, these markings can be compared to other images in an attempt to link one weapon to multiple crime scenes. Like all forensic specialties, forensic firearm examiners are subject to being called to testify in court as expert witnesses.


The ability to compare ammunition is a direct result of the invention of rifling around the turn of the 16th century. By forcing the bullet to spin as it travels down the Gun barrel, barrel of the weapon the precision is greatly increased. At the same time, the rifling leaves marks on the bullet that are indicative of that particular barrel. Prior to mass production of firearms, each barrel and bullet mold was hand made by gunsmiths making them unique. The first successful documented case of forensic firearm examination occurred in 1835 when a member of the Bow Street Runners in London matched a recovered bullet from a murder victim to a specific mold in a suspect's home confirming that he made the bullet; this gave further evidence that the bullet maker was the perpetrator and he was convicted. As manufacturing and automation replaced hand tools, the ability to compare bullets became impossible due to the standardization of molds within a specific company. However, experts in the field postulated that there were microscopic differences on each barrel left during the manufacturing process. These differences were a result of wear on the machines and since each new weapon caused a tiny amount of wear, each barrel would be slightly different from every other barrel produced by that company. Also, each bullet fired from a specific barrel would be printed with the same marks, allowing investigators to identify the weapon that fired a specific bullet. One of the first uses of this knowledge was in 1915 to exonerate Charles Stielow of the murder of his neighbors. Stielow was sentenced to death and appealed to Charles S. Whitman, the Governor of New York, who was not convinced by the evidence used to convict Stielow. Whitman halted the execution until an inquiry could be conducted and after further examination it was shown that Stielow's firearm could not have fired the bullets recovered from the victims. The invention of the comparison microscope by Calvin Goddard (ballistics), Calvin Goddard and Phillip O. Gravelle in 1925 modernized the forensic examination of firearms. Simultaneous comparison of two different objects at the same time allowed to closely examine striations for matches and therefore make a more definitive statement as to whether or not they matched. One of the first true tests of this new technology was in the aftermath of the Saint Valentine's Day Massacre in 1929. During the Prohibition in the United States, Prohibition Era, competing gang members were fighting over Rum-running, bootlegging operations within the city of Chicago. Members of the Chicago Outfit and the Egan's Rats led by Al Capone attempted to remove all competition from Chicago by eliminating the North Side Gang leader Bugs Moran. The massacre missed Moran, who was not present, but killed seven members of the North Side Gang. The murderers attempted to cover up their crime by posing as police officers, even dressing in police uniforms. Witnesses saw two "officers" leaving the scene, which implicated the Chicago police department as the perpetrators of the massacre. High levels of police corruption during that time period made it seem likely that the police department committed the killings. The investigation stalled until December 1929 when Fred Burke, a member of the Egan's Rats, shot and killed a police officer in St. Joseph, Michigan. Officers searching for Burke were led to a home in nearby Stevensville, Michigan, Stevensville. While Burke was not there, inside officers found an arsenal of weapons including two Thompson submachine guns. The Chicago police department was contacted and the weapons were brought back to Chicago for testing. Goddard was asked to compare the weapons to collected evidence found at the massacre using his new "ballistic-forensics" technique. After test firing the guns, Goddard proved that the weapons were those used to kill the members of the North Side Gang, absolving the Chicago police department of all involvement. The successful use of Goddard's technique resulted in the solidification of his place as the father of forensic firearm examination.

Examination of the firearm

Any firearm collected during the course of an investigation could yield viable evidence if examined. For forensic firearm examination specific evidence that can be recovered include weapon serial numbers and potentially fingerprints left on the weapon's surface.

Fingerprint recovery

Fingerprint recovery from the surface of firearms is done with cyanoacrylate (more commonly known as superglue) fuming. Firearms are placed in a specially designed fume hood designed to evenly distribute fumes instead of removing them. Liquid superglue is placed in a container and heated until it is in a gaseous state. The circulating fumes adhere to the oils left behind by the fingerprint, turning the print white. The resulting white print can be enhanced with fingerprint powder to increase the contrast of the white print against the weapon's finish. While using the fuming technique on recovered guns is commonplace, the recovery of fingerprints from the surfaces of a firearm is challenging due to the textured grip and the general condition of recovered weapons. If fingerprints are recovered, they can be processed through fingerprint databases such as the Integrated Automated Fingerprint Identification System (IAFIS). Various parts of the recovered weapon can also be tested for touch DNA left by whomever handled it. However, the low levels of DNA that can be recovered presents numerous issues such as contamination and analysis anomalies such as allele drop-out and drop-in.

Serial number recovery

Serial numbers became commonplace after the United States passed the Gun Control Act of 1968. This law mandated that all guns manufactured in or imported into the country have a serial number. Prior to 1968, many firearms either did not have a serial number or the serial numbers were not unique and were reused by a manufacturer on multiple firearms. If a recovered weapon has had the serial numbers altered or destroyed, examiners can attempt to recover the original numbers. The two main methods for the restoration of serial numbers are magnetic particle inspection and chemical restoration. It is recommended that magnetic particle inspection be performed first due to the Nondestructive testing, nondestructive nature of the method. If magnetic particle inspection fails, chemical restoration is the next step in the forensic analysis. If the serial number is successfully restored it can be used to help investigators track the weapon's history, as well as potentially determine who owns the weapon. Firearm databases such as the National Crime Information Center of the United States and INTERPOL, INTERPOL's Firearm Reference Table can be used by investigators to track weapons that have been lost, stolen, or used previously in other crimes.

Magnetic particle inspection

Originally developed as a method to detect flaws or irregularities in Ferromagnetism, magnetic materials, magnetic particle inspection can be used on firearms to visualize the serial number underneath the obliterated area. When performing this technique, examiners place the weapon in a magnetic field. The irregularities in the metal, in this case the serial number, cause the field to deform. When a solution of ferrous particles is added to the weapon's magnetized surface they will be attracted to the area where the magnetic field has deformed and will build up in the area. If Fluorescence, fluorescent particles are added to the ferrous solution, ultraviolet light can be used to make it easier to visualize any recovered serial number.

Chemical restoration

Chemical restoration is a type of chemical milling. Typically, chemical milling is used to slowly remove material to create a desired shape. In serial number restoration, small amounts of metal are removed until variations in the metal corresponding to the serial number are visible. This is possible because stamping (metalworking), stamping the numbers distorts the grain boundary structure underneath the surface of the metal. However, chemical restoration is limited to that depth and is only successful when the obliteration of the serial number is superficial. Examiners performing a restoration first sand the area where the serial number used to be. This removes any debris from the area left when the serial number was obliterated. The examiner then chooses a chemical, usually an acid, that will be used to slowly bring the number back to the surface. The type of chemical that is used depends on the material the weapon is made of. These acids can range from Fry's Reagent for a magnetic metal, which is a mixture of hydrochloric acid, cupric chloride, and distilled water, to an acidic ferric chloride solution for a non-magnetic, non-aluminum material.

Examination of cartridges

Spent cartridges found at a scene can be examined for physical evidence such as fingerprints or compared to samples that match them to a weapon. The examination of the cartridge relies on the unique tool marks left by the various parts of the weapon including the firing pin and the ejector in semi and fully automatic firearms. These markings can be compared and matched to known exemplars fired from the same weapon using the same parts. The examination of the marks left on the cartridge is done using a comparison microscope. Examiners view the questioned cartridge and the known exemplar simultaneously, looking for similar microscopic marks left during the firing process. Cartridges are also routinely examined for fingerprints as the act of loading the ammunition into the magazine, or chamber, leaves recoverable impressions. These fingerprints can survive the firing processes and, while a rare occurrence, fingerprints have been obtained from cartridges recovered from the scene. Cartridges are subjected to cyanoacrylate fuming and examined for any usable prints. Usable prints are photographed and can be uploaded to fingerprint databases such as IAFIS for comparison with known exemplars. Cartridges can also be swabbed for trace DNA left by the individual who loaded the magazine. The extremely low levels of recoverable DNA present the same issues as swabbing a firearm for DNA. Advancements in microscopic stamping have led to a push for the inclusion of firing pin microstamping. The microstamp is etched onto the firing pin and is transferred to the cartridge during the firing process. Each firing pin would have a unique serial number allowing investigators to trace casings found at a crime scene to a known firearm. The practice is not in use , although California has enacted legislation that requires microstamping on all newly sold firearms. The law, and microstamping in general, has received significant opposition from gun manufacturers due to increased costs associated with introducing the microstamps into the manufacturing lines.

Examination of bullets

Class characteristics

Preliminary examination of the bullet can exclude a large number of weapons by examining the general characteristics of a recovered bullet. By determining general aspects of the fired ammunition, a number of weapons can be immediately excluded as being incapable of firing that type of bullet. The make and model of the weapon can also be inferred from the combination of different class characteristics that are common to specific manufactures. The three main class characteristics of all bullets are the lands and grooves, the caliber of the bullet, and the rifling twist. All three can be tied directly to the type of barrel that was used to fire the bullet. The lands and grooves of barrel are the bumps and valleys created when the rifling is created. The caliber is the diameter of the barrel. The twist is the direction of the striations left by the barrel's rifling, clockwise (right-handed) or counterclockwise (left-handed). Most barrels will have a right-handed twist with the exception of weapons created by the Colt's Manufacturing Company which uses left-handed twists. Weapon barrels that match the class characteristics of recovered bullets can be examined further for individual characteristics to determine if the bullet came from that particular weapon.

Individual characteristics

In order to compare individual striations, examiners must obtain a known sample using the seized weapon. For slower-traveling bullets, such as pistols or revolvers, known bullet exemplars are created by firing the weapon into a water tank. The spent bullet can be recovered, intact, as the water slows down the bullet before it can reach the tank walls. For faster traveling bullets, such as those fired from high-powered rifles and military style weapons, water tanks cannot be used as the tank will not provide enough stopping power for the projectiles. To examine these weapons, investigators must fire them at a target at a shooting range, controlled range with enough backing to stop the bullet and collect the spent round after it has been fired. Once a known exemplar is produced, the evidence sample can be compared to the known by examining both at the same time with a comparison microscope. Striations that line up are examined more closely, looking for multiple consecutive matches. There is no set number of consecutive matches that equates to a match declaration, and examiners are trained to use the phrase "sufficient agreement" when testifying. The degree to which an examiner can make that determination is based on their training and expertise. All findings by examiners are subject to questioning by both sides, prosecution and defense, during testimony in court.

Striation databasing

Bullets and casings found at a scene require a known example to compare to in order to match them to a weapon. Without a weapon, the striation pattern can be uploaded to a database such as the National Integrated Ballistic Identification Network (NIBIN) maintained by the Bureau of Alcohol, Tobacco, Firearms and Explosives, ATF or the United Kingdom's National Ballistics Intelligence Service (NABIS). Information uploaded to these databases can be used to track gun crimes and to link crimes together. Maintainers of these databases recommend that every recovered firearm be test fired and the resulting known exemplar be uploaded into the database. In the 1990s, there were two databases that were formed for storage of pictures of shell casings and bullets in gun crimes. The first was the Drugfire system which was used by the FBI. The second, the Integrated Ballistics Identification System, IBIS (Integrated Ballistic Identification System) was created by Forensic Technology, Inc. and eventually bought by the Alcohol Tobacco and Firearms (ATF) in 1993. The FBI and ATF realized that their systems would not work together, and they needed to find a way to share information between them. The NIBIN board was created in 1997, in hopes of creating one imaging system.  A year after the creation of the NIBIN board, both the ATF and FBI decided to put their resources together toward one of the systems, and created the National Integrated Ballistics Information Network, with IBIS as the system.


Firearm examiners have attempted to determine the shooter's position by the location of spent bullet casings. The use of ejection pattern studies were originally part of incident reconstruction and methods for determining shooter location continue to be explained in major crime scene examination books. However, the validity of ejection pattern analysis has been brought into question by multiple studies that look at the reproducibility and end determination of shooter position by qualified examiners. Studies have shown that over 25% of spent casings land somewhere other than to the right and rear of the shooter. This is the most commonly accepted location for where spent cartridge casings should fall, and the large percentage of casings that end up somewhere else raises concerns for the validity of the examination technique. Investigators should only present a location gained from an ejection pattern study as a tentative estimate when using the information in a courtroom setting. Prior to September 2005, comparative bullet-lead analysis was performed on bullets found at a scene that were too destroyed for striation comparison. The technique would attempt to determine the unique elemental breakdown of the bullet and compare it to seized bullets possessed by a suspect. Review of the method found that the breakdown of elements found in bullets could be significantly different enough to potentially allow for two bullets from separate sources to be correlated to each other. However, there are not enough differences to definitely match a bullet from a crime scene to one taken from a suspect's possession. An additional report in 2004 from the National Academy of Sciences (NAS) found that the testimony given regarding comparative bullet-lead analysis was overstated and potentially "misleading under the federal rules of evidence". In 2005, the Federal Bureau of Investigation indicated that they would no longer be performing this type of analysis. Further criticism came from the 2009 NAS report on the current state of various forensic fields in the United States. The report's section on firearm examination focused on the lack of defined requirements that are necessary in order to determine "matches" between known and unknown striations. The NAS stated that, "sufficient studies have not been done to understand the reliability and repeatability of the methods." Without defined procedures on what is and what isn't considered "sufficient agreement" the report states that forensic firearm examination contains fundamental problems that need to be addressed by the forensic community through a set of repeatable scientific studies that outline standard operating procedures that should be adopted by all firearm examiners. Another report issued in 2016 by the United States President's Council of Advisors on Science and Technology confirmed the NAS's findings, finding only one appropriately designed study that examined the rate of false positives and reliability amongst firearm examiners. In fall 2020 a paper was written by Itiel E. Dror and Nicholas Scurich and presented in their paper, ''The (Mis)Use of Scientific Measurements in Forensic Science''. The paper looked at the validity of ballistic forensic experts when attempting to make an identification of a shell or bullet. In their study they found that while some experts would come to the conclusion that the bullets were a definite match, another expert looking at the same evidence would determine it inconclusive. Dror and Scurich argue an “inconclusive” determination affects the error rate for the study, and provides very little confidence in the overall findings of the scientists. According to Dror and Scurich, the error rate, which was zero to one percent, could be higher. Their reasoning behind this is that if an “answer” was marked as inconclusive, it must count as a correct answer which decreases the error rate making it lower than it probably should be. They wondered how different the error rate would be if inconclusive was not an option. In addition, Dror and Scruich noted that the scientists seemed to come up with a more conclusive decision on the evidence if there was the added part of a human life hanging in the balance. In 2021 Alex Biederman and Kyriakos N. Kotsoglou responded to the Dror and Scurich paper and raised issues. Some of the issues raised by Biederman and Kotsoglou included: a paradox in which examiners' results agreed with ground truth but would be considered "error" via Dror and Scurich's proposals. Biederman and Kotsoglou also pointed out that Dror and Scurich's proposals would set false incentives where examiners would be directed to be "diving what the mythical forensic wisdom of the consensus opinion might (and hence enshrine the false belief in the existence of such wisdom), be rather than the ground truth". Biederman and Kotsoglou concluded "In all, our analysis does not leave much intact from recent attempts to label 'inconclusives as errors."

See also

*Edward O. Heinrich *Ballistics *Crime reconstruction *Evidence packaging


{{reflist, 2, refs= {{cite book , title=Science for Lawyers , chapter=Ballistics , chapter-url= , last=Steele , first=Lisa , publisher=American Bar Association , date=2008 , access-date=January 19, 2016 {{cite journal , last=Hamby , first=James , title=The History of Firearm and Toolmark Identification , journal=Association of Firearm and Tool Mark Examiners Journal , volume=31 , number=3 , date=Summer 1999 , access-date=January 16, 2016 , url= {{cite book , title=Convicting the Innocent: Errors of Criminal Justice , last=Borchard , first=Edwin , year=1932 , chapter-url= , chapter=Stielow and Green , access-date=January 20, 2016 , publisher=New Haven Yale University Press {{cite web , title=Comparison Microscopy , publisher=National Forensic Science Technology Center , access-date=June 25, 2016 , url= {{cite web , title=The St. Valentine's Day Massacre , last=O'Brien , first=John , date=February 14, 2014 , access-date=June 25, 2016 , url= , publisher=The Chicago Tribune , archive-url=, archive-date=March 5, 2016 , url-status=live {{cite web , title=St. Valentine's Day Massacre: Tale of two guns , last=Ashcroft , first=Brent , url= , access-date=June 25, 2016 , publisher=WZZM13 {{cite web , title=Baltimore native helped solve 1929 St. Valentine's Day Massacre , last=Rasmussen , first=Frederick N. , date=February 12, 2011 , access-date=June 25, 2016 , url= , publisher=The Baltimore Sun {{cite web , title=In the Public Eye: Finding Fingerprints on Firearms is Actually Very Rare , date=September 2, 2015 , access-date=May 31, 2016 , url= , publisher=Forensic Magazine {{cite web , title=CYANOACRYLATE (SUPERGLUE) FUMING , access-date=May 31, 2016 , url= , publisher=Minnesota Bureau of Criminal Apprehension {{cite web , title=Public Law 90-618: To amend title 18, United States Code, to provide for better control of the interstate traffic in firearms , date=October 22, 1968 , access-date=May 31, 2016 , url= , pages=1213–1236 {{cite web , title=Firearm Identification in the Forensic Science Laboratory , last=Thompson , first=Robert , publisher=National District Attorneys Association , date=2010 , access-date=January 19, 2016 , url= {{cite journal , last=Gulick , first=Gary , journal=Evidence Technology Magazine , volume=6 , issue=3 , date=May–June 2008 , access-date=January 9, 2016 , url= , title=Lifting Latent Fingerprints from Difficult Surfaces {{cite web , date=November 2011 , access-date=January 9, 2016 , title=Firearms Tracing Guide , publisher=Bureau of Alcohol, Tobacco, Firearms, and Explosives , url= {{cite book , title=Cartridges and Firearm Identification , last=Walker , first=Robert E. , publisher=CRC Press , date=2013, url=, isbn=978-1-4665-8881-3, page=573 {{cite web , title=Magnetic Particle Recovery of Serial Numbers , last1=Utrata , first1=Dave , last2=Johnson , first2=Marcus , publisher=Midwest Forensics Resource Center , date=October 2003 , access-date=July 4, 2016 , url= {{cite web , title=Technical Procedure for Serial Number Restoration , publisher=North Carolina State Crime Laboratory , date=September 5, 2014 , access-date=January 10, 2016 , url= {{cite web , title=Serial Number Restoration , access-date=July 2, 2016 , url= , publisher=Minnesota Bureau of Criminal Apprehension {{cite web , url= , publisher=Iowa Division of Criminal Investigation , access-date=July 2, 2016 , title=Iowa Division of Criminal Investigation Criminalistics Laboratory Firearm & Toolmark Section Restoration of Obliterated Serial Numbers {{cite web , title=FEU08 – SOP for Serial Number Restoration of Obliterated Stampings in Various Metal Surfaces , date=November 26, 2013 , access-date=December 17, 2016 , url= , publisher=District of Columbia Department of Forensic Sciences , page=2 {{cite web , title=INTERPOL Firearms Reference Table (IFRT) , access-date=January 12, 2016 , url= , website=INTERPOL {{cite web , title=NCIC Files , access-date=January 12, 2016 , website=Federal Bureau of Investigation , url= , archive-url= , archive-date=February 20, 2016 , url-status=dead {{cite web , work=The Guardian , title=Forensics: Fingerprints can be recovered from fired bullet casings , last=Randerson , first=James , date=June 3, 2008 , access-date=June 18, 2016 , url= {{cite journal , doi=10.1016/j.fsigen.2009.02.007 , pmid=19647709 , title=Development of STR profiles from firearms and fired cartridge cases , journal=Forensic Science International: Genetics , volume=3 , issue=4 , date=September 2009 , pages=242–250 , last1 = Horsman-Hall , first1 = Katie M. {{cite report , publisher=The Education Fund to Stop Gun Violence , title=Cracking the Case: The Crime-Solving Promise of Ballistic Identification , date=June 2004 , access-date=June 18, 2016 , url= , archive-url= , archive-date=April 14, 2008 {{cite act , title=Crime Gun Identification , date=2007 , number=1471 , url= , type=Bill , access-date=June 18, 2016 {{cite web , title=Smith & Wesson says it won't follow California 'microstamping' law , last=Mather , first=Kate , date=January 23, 2014 , access-date=June 18, 2016 , url= , work=Los Angeles Times {{cite book , title=Gunshot Wounds: Practical Aspects of Firearms, Ballistics, and Forensic Techniques , edition=3rd , last=DiMaio , first=Vincent J.M. , publisher=CRC Press , date=2016 , url=, isbn=978-1-4987-2570-5, page=1 {{cite web , title=Firearms & Tool Mark , website=North Carolina Department of Justice , url= , access-date=June 4, 2016 {{cite web , title=A high-tech approach to tracking gun crimes , date=February 25, 2016 , access-date=June 20, 2016 , publisher=Southwest Journal , last=Thomas , first=Dylan , url= {{cite web , title=Intelligence service links 350 guns to crimes , date=January 11, 2010 , access-date=June 20, 2016 , url= , work=BBC News {{cite web , title=Bullets, Casings, and You , publisher=Bureau of Alcohol, Tobacco, Firearms and Explosives , access-date=June 21, 2016 , url= {{cite book , title=Practical Crime Scene Processing and Investigation , edition=2nd , last=Gardner , first=Ross , pages=300–301 , url= , publisher=CRC Press , date=2012, isbn=978-1-4398-9778-2 {{cite journal , title=Firearms and Toolmarks in the FBI Laboratory , date=April 2000 , volume=2 , number=2 , journal=Forensic Science Communications , access-date=June 5, 2016 , url= , archive-url= , archive-date=September 20, 2015 , url-status=dead {{cite book , title=Introduction to Criminalistics: The Foundation of Forensic Science , last1=Fisher , first1=Barry A.J. , last2=Tilstone , first2=William J. , last3=Woytowicz , first3=Catherine , url=, publisher=Elsevier Academic Press , date=2009 , page=39, isbn=9780080916750 {{cite journal , title=Fired Cartridge Case Ejection Patterns From Semi-Automatic Firearms , journal=Investigative Sciences Journal , volume=2 , number=3 , date=November 2010 , access-date=May 28, 2016 , url= , last1=Lewinski , first1=William , last2=Hudson , first2=William , last3=Karwoski , first3=David , last4=Redmann , first4=Christa {{cite web , title=FBI's Forensic Test Full of Holes , last=Solomon , first=John , date=November 18, 2007 , access-date=December 28, 2016 , url= , website=The Washington Post {{cite journal , title=A Metallurgical Review of the Interpretation of Bullet Lead Compositional Analysis , last1=Randich , first1=Erik , last2=Duerfeldt , first2=Wayne , last3=McLendon , first3=Wade , last4=Tobin , first4=William , journal=Forensic Science International , date=July 17, 2002 , volume=127 , issue=3 , pages=174–191 , doi=10.1016/S0379-0738(02)00118-4 , pmid=12175947 {{cite press release, title=FBI Laboratory Announces Discontinuation of Bullet Lead Examinations, url=, publisher=FBI National Press Office, date=September 1, 2005 , access-date=December 28, 2016 {{cite book , title=Strengthening Forensic Science in the United States: A Path Forward , author=National Research Council , publisher=National Academies Press , date=2009 , isbn=978-0-309-13131-5 , url= , access-date=June 12, 2016 {{cite report , title=Forensic Science in Criminal Courts: Ensuring Scientific Validity of Feature-Comparison Methods , date=September 2016 , access-date=December 31, 2016 , url= , pages=104–114 , url-status=live , archive-url= , via=NARA, National Archives , work=Office of Science and Technology Policy , archive-date=January 20, 2017 Forensic techniques