Defendant Ernest Williams appeals from a capital murder conviction for the shooting death of Renee Williams (no relation) in her home January 8,1981. That evening appellant and his brother Darnell Williams went to the victim’s home, where appellant confronted the victim’s brother Kirk Mischeaux about an alleged debt Mischeaux owed appellant. An argument ensued culminating in Mischeaux being shot three times with a .45 caliber weapon. Mischeaux testified that after he fell from the shots he saw the defendant still holding the gun and going to another part of the house. Ms. Williams was shot in her bedroom while talking on the telephone to the Berkeley Police Department. Appellant denied shooting the victim, claiming that he dropped his gun while wrestling with Mischeaux, and that his brother Darnell shot the victim with appellant’s gun. The jury returned a verdict of guilty of capital murder, from which this appeal is taken. Since appellant was sentenced to life imprisonment, this court has jurisdiction.
The first point on appeal is whether the trial court erred in admitting the testimony of Dr. Robert Briner, director of the crime laboratory at Southeast Missouri State University, regarding the results of a gunshot residue test performed on samples taken from the hands of both Darnell and the appellant. Dr. Briner testified that the test data were consistent with appellant “having fired a .45 caliber weapon,” and with Darnell “just having handled the weapon and not necessarily firing it.” Appellant objects to the admission of this testimony on the ground that the particular test employed here has not been generally accepted by the scientific community as reliable.
“[T]he determination of admissa-bility of expert testimony vests largely in the discretion of the trial court, and the exercise of that discretion will not be disturbed on appeal unless an abuse of such discretion is shown.” State v. Sager, 600 S.W.2d 541, 572 (Mo.App.1980). Expert opinions based on scientific tests are admissible if the scientific principle involved is generally considered by the scientific community as reliable. State v. Johnson, 539 S.W.2d 493 (Mo.App.1976). Appellant cites no authority suggesting that the scientific principle upon which this test wasi based is not generally accepted as reliable. On the contrary, appellant’s brief admits that the basic principle of chemically analyzing samples taken from suspects’ hands to determine the presence of trace metals emitted by the discharge of a firearm, has been generally accepted in this jurisdiction since 1975. State v. Ross, 523 S.W.2d 841 (Mo. App.1975). Ross held that “firearm residue testing by neutron activation analysis has *311crossed a line between the experimental and demonstrable stages and the evidential force of the principle must be recognized.” State v. Ross, 523 S.W.2d at 845.
The test at issue is based on the same scientific principle as the neutron activation analysis. That is, it involves chemical analysis of samples from suspects’ hands to determine the presence of trace metals. It differs from the Ross test in that is uses a different laboratory process for the chemical identification of the trace elements. Dr. Briner testified that this process, known as electroplating or stripping voltammetry is commonly used in environmental industries to detect the presence of metallic materials, and that the procedure has been applied to “the analysis of metallic residues off of gunshot swabs” for over three years at the crime laboratory under Dr. Briner’s direction.
Dr. Briner testified to his own credentials, including a B.S. in Chemistry and a PhD in Biochemistry. He also testified that he had prepared and chemically analyzed “several hundred [gunshot residue kits] over the past eight to nine years we have been doing gunshot residues.” Dr. Briner testified that he was co-authoring an article at the time of the trial “correlating ... the electroplating procedure with other procedures that have been used in gunshot residue [testing] in the past.” See Briner and Couchoiy, An Inexpensive Approach to Inorganic Gunshot Residue Analysis Using Anodic Stripping Voltammetry, (1982).
The record clearly supports the trial court’s admission of the gunshot residue evidence as expert testimony based on a scientific principle generally accepted as reliable.1 “Objections to the manner in which the analysis was conducted go to the weight rather than the admissibility of the evidence.” State v. Johnson, 539 S.W.2d at 501.
The record also supports the reliability of the samples from which the test results were produced. Sargeant Hoffmeis-ter of the Berkeley police testified about the circumstances in which the appellant’s and Darnell’s hands were swabbed to obtain the test sample. He stated that the swabs were taken according to step-by-step instructions on the kit, and to his knowledge neither appellant nor Darnell had an opportunity to wash his hands or to do anything else prior to the taking of the samples that would have interfered with the results. Dr. Briner testified concerning the control built into the sample collection kits used by the investigating officers.
“Yes, they are all very similar in that they have vials marked for poth palms, right and left palm of the hand area, the area to be swabbed. And they also are for both right and left back of the hand.
And then there is also a control sample in there for the officer to provide us with a sample of the liquid, or nitric acid that he uses to swab the hands, so that we can find out whether or not he has any contamination in his acid or not.”
The test employed here is subject to the same limitations as neutron activation analysis in that:
“While there seems to be little doubt that neutron activation analysis can accurately detect the presence and amount of certain chemical elements, it is doubtful that it can conclusively establish that those elements were present as the result of firing a weapon....
“However, the presence of these chemicals together in the amounts found on this defendant’s hands could reasonably provide some evidence that defendant had fired a gun....”
State v. Ross, 523 S.W.2d 841, 845-46 (Mo.App.1975), quoting State v. Spencer, 298 Minn. 456, 216 N.W.2d 131, 134-35 (1974).
On this record we hold that it was within the trial court’s discretion to admit the results of the gunshot residue test into evidence.
*312Appellant’s second point is that the trial court erred in refusing to exclude testimony of officer Hunt concerning the defacement of the serial number on the gun used to kill Renee Williams. Appellant contends that this testimony was irrelevant and constituted evidence of a separate crime, since possession of a weapon without a serial number violates § 571.130 RSMo.2 Although evidence concerning the defacement had been admitted earlier without objection or motion to strike, appellant contends that the prosecution attempted to prejudice appellant before the jury with this evidence after it was admitted, and seeks review for plain error.
The plain error rule is to be used sparingly and is limited to cases with a clear showing of manifest injustice. State v. Sammons, 640 S.W.2d 488, 490 (Mo.App.1982). The plain error rule is inapplicable here. It is unlikely the jury was prejudiced, since the defacement was never referred to as a crime and since it is doubtful that any jurors were aware of the criminality of mere possession of such a defaced weapon. Certainly there is no evidence of manifest injustice.
Had the point been properly preserved for appeal it would have been denied. The general rule that evidence of a separate crime is not admissible has several well established exceptions, including “multiple crimes so related to each other that proof of one tends to establish proof of the other.” State v. Frazier, 550 S.W.2d 590, 596 (Mo.App.1977). The defaced serial number was a material fact relevant to the identification of the murder weapon as well as to other elements of the crime, such as preparation and consciousness of guilt. Therefore, the court did not err in admitting evidence of it.
Appellant’s third point alleges improper and prejudicial remarks by the prosecutor in closing argument. During the first part of the State’s argument the trial court overruled appellant’s objection to the following statement: “By the way, I cannot force Darnell Williams to testify, if you were expecting that.” This incident arose while the prosecutor was commenting upon the unlikelihood of Darnell Williams using appellant’s gun to shoot the victim when Darnell had his own gun.
During the closing half of his argument, the prosecutor stated
“The defendant’s attorney argued that the State had not a shred of evidence, and that if the State had any more evidence, they would put it on. Both parties are entitled to put on evidence.
The only person you heard from was the defendant Ernest Williams. I could not make Darnell Williams testify. The defense did not produce Darnell Williams.”
Appellant’s objection to this argument was sustained and the jury was instructed to disregard the remark. However, appellant’s motion for a mistrial was overruled.
Appellant argues here that the first statement constituted an improper statement of the law and an improper argument of the law by the prosecutor. In support of his argument he cites to State v. Phillips, 511 S.W.2d 841 (Mo.1974) and State v. Wright, 571 S.W.2d 734 (Mo.App.1978). These cases merely hold that it is not improper to permit a witness to invoke the privilege against self-incrimination in the presence of a jury.3 Here, the prosecutor did not say he could not subpoena Darnell, only that he could not force him to testify. The authorities relied upon by appellant do not indicate such a statement to be an improper statement of law. Nor do we find the statement to amount to an improper argument of law by the prosecutor. As a general rule it is not the prerogative of counsel to inform the jury as to the substantive law of the case. State v. Jordan, 646 S.W.2d 747, 751 (Mo.banc 1983). But the prosecutor’s comment that he could not force Darnell to testify falls far short of the usurpation of the court’s function of in*313structing the jury condemned in such cases as State v. Shelby, 634 S.W.2d 481 (Mo. 1982) and State v. Holzwarth, 520 S.W.2d 17 (Mo.banc 1975). The statement, made as an aside in the midst of a legitimate argument regarding the improbability of defendant’s testimony, was neither definitional nor instructional. Although the legal privilege against self-incrimination may be said to support the accuracy of the statement, it was nothing more than a statement of fact.
Appellant’s complaint regarding the repetition of the statement in the closing half of the prosecutor’s argument is meritless, not only by reason of what has been said above, but in addition because this argument was clearly retaliatory. Thus, even if technically improper, which it was not, the statement would not constitute grounds for reversal. State v. Sims, 639 S.W.2d 105,110 (Mo.App.1982). Nor do we find merit in appellant’s contention that the argument was an improper attempt to cast an adverse inference upon his failure to call his own brother as a witness. Even if appellant’s attempt to bathe his brother in the light of a cold-blooded murderer could be said to alter the general rule that failure to call a member of one’s family who is in a position to have relevant knowledge justifies an adverse inference, State v. Moore, 620 S.W.2d 370, 373 (Mo.banc 1982), there is no ground for reversal here. The case relied upon by appellant, State v. Hustead, 615 S.W.2d 556 (Mo.App.1981) is inapposite. There a prosecutor after being prohibited from calling two co-defendants for the purpose of having them envoke the privilege against self-incrimination in the presence of the jury, was permitted, over objection to argue at great length that defendant’s failure to call these witnesses justified an inference their testimony would be adverse to the defense. The seven words of the prosecutor here are not comparable to the deliberate prosecutorial and ethical misconduct reflected in Hustead. Furthermore, in Hu-stead, the trial court overruled the objection to the argument, thereby placing its imprimatur upon the comments of the prosecutor. Here, the trial court immediately sustained the objection and instructed the jury to disregard the remark. The denial of appellant’s motion for a mistrial, “a drastic remedy [to] be exercised only in extraordinary circumstances where the prejudicial effect can be removed in no other way,” State v. Raspberry, 452 S.W.2d 169, 173 (Mo.1970), was well within the sound discretion of the trial court.
Appellant’s final two points were conceded at oral argument. First, the record clearly shows that the state did not occasion excessive pre-trial delays calling for dismissal. Second, felony murder is not a lesser included offense where capital murder is charged, State v. Baker, 636 S.W.2d 902 (Mo.banc 1982); State v. Clark, 652 S.W.2d 123 (Mo.banc 1983).
The judgment of the trial court is affirmed.
SNYDER, P.J., and DOWD, C.J., concur.. See Peterson and Wong, Fundamentals of Stripping Voltammetry, American Lab 116 (November 1981); Liu, Lin, and Taylor, The Application of Anodic Stripping Voltammetry to Forensic Science, 16 Forensic Science International 43 (1980); Liu, Lin, and Nicol, Anodic Stripping Voltammetry Analysis of Gunshot Residues, 16 Forensic Science International 53 (1980).
. Now § 571.050, RSMo 1981.
. But see the later case of State v, Wright, 582 S.W.2d 275 (Mo.banc 1979), which prohibits this practice when the State is aware beforehand the privilege will be claimed by the witness.