distinguish the inks, Brunelle objectively measures the separation of the dyes and the width of the dye columns, and again comparing these measurements with those he has made of his ink library samples, he makes his identification.
Perhaps the most basic flaw in Brunelle's procedure is that his "ink library" was so incomplete that it was impossible for him to exclude a great number of inks which he never had examined. He conceded that he had no foreign ink library; that he had no Japanese inks,3(a) nor those of a number of other countries. (N.T. 556-7). Obviously, a procedure which relies entirely on comparison and elimination cannot be used to make a positive identification when an unknown number of inks cannot be excluded.
Furthermore, Brunelle's results are affected to an unknown extent by a great number of variables. A trier of fact would have to speculate about the importance of these variables, especially when the witness could not recall and could not document how these factors could have altered his results. First, there is no way to measure the quantity of ink that is removed from the questioned signature (N.T. 652). Therefore, each "plug" from the questioned document contained an unknown quantity of ink and Brunelle admitted that different concentrations of ink on paper were compared. (N.T. 761). Furthermore, Brunelle had no idea how many plugs he took from the document and put into solution, (N.T. 653), and he did not know the strength of the solution (N.T. 654). Although different solvent systems can give different separations of the dye components (N.T. 783), comparisons were made of dyes in different solvents (N.T. 787). In putting the plug into solution, the paper necessarily is mixed in (N.T. 654), but Brunelle ran no chromatographic test on the paper itself. (N.T. 658). He also conceded that scribble samples of a known ink were taken from a type of paper different from the questioned document (N.T. 676). Brunelle testified that the distance that a dye will travel on a chromatogram will not be uniform because of the nature of the coating on a chromatogram plate, (N.T. 659), and that a chromatogram result can be affected by the manner in which the ink is placed on the chromatogram (N.T. 805). Finally, Brunelle knew that ultraviolet light causes the dyes to fade and can even change the relative intensities of the dyes on the chromatogram, (N.T. 838) but he had no idea as to how long the questioned document had been exposed to ultraviolet light. Perhaps even more basic than these variables is the fact that none of the manufacturers who testified stated that the production of inks is so uniform that two batches of the same ink would give uniform chromatographic results. In fact, the representative of the Scripto Company testified that the actual composition of its inks will change over a period of time (N.T. 419). Brunelle conceded that the same dye in inks manufactured by different companies would not necessarily be identical (N.T. 774).
As if to demonstrate the unpredictable effects of all these variables, Brunelle showed us a slide in which three chromatograms of the same ink showed entirely different results (N.T. 805). He conceded that he could not tell on the basis of these chromatograms that they were all prepared using the same ink (N.T. 812).
We are convinced that the state of the art in this field of ink identification is not yet sufficiently advanced to be reasonably scientifically certain that an ink of unknown composition is the same as a known ink. The most that can be said is that an unknown ink appears to be similar to or different from a known ink. The technique used by Brunelle was first developed in 1967 (N.T. 1259), and Brunelle has been using this procedure only three years (N.T. 542). It is most interesting that the scientist who developed and taught Brunelle the solvent system for using this chromatographic analysis (N.T. 907) testified that because the "variables haven't been investigated thoroughly" (N.T. 1289), and because the questioned ink has no unique pattern, the state of the art is such that it cannot be positively identified (N.T. 1290). This view is shared by all of the experts in chromatographic identification who testified, other than Brunelle (N.T. 1053, 1248, 1366).
In Frye v. United States, 54 App. D.C. 46, 293 F. 1013, 1014 (1923), the court, in deciding that the results of polygraph tests are not admissible, made this statement as to scientific evidence:
"Just when a scientific principle or discovery crosses the line between the experimental and demonstrable stages is difficult to define. Somewhere in this twilight zone the evidential force of the principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs."
We decided, therefore, that we could not allow a jury to speculate as to the guilt or innocence of Levin and Coopersmith based entirely on this ink-identification evidence which has not achieved this level of acceptance.
2. INADEQUACY OF BRUNELLE'S METHOD
Apart from this inherent inadequacy in the identification technique, we had grave doubts as to the reliability of Brunelle's methodology in conducting these tests. It is widely recognized that the party offering the results of laboratory tests must not only produce an expert to attest to the scientific reliability of the test, but also, he must vouch for its correct administration in the particular case. See McCormick, Handbook of the Law of Evidence p. 357 (1954).
Contrary to the accepted practice in every laboratory, Brunelle kept no detailed log concerning the tests he was performing. (N.T. 575, 616-18, 643, 675). Not only did this make cross-examination difficult, but it also made it impossible for a jury to evaluate his findings. For example, it could not be determined whether any experiments had given negative or conflicting results. He permitted the solution in the chromatogram procedures to "run" only 30 minutes, which fails to permit the more definitive separation a longer period would provide. (N.T. 769). Although he said that he had experimented with longer runs, he had no such results and had no notes concerning them. Finally, he testified that the original comparison chromatograms were made on glass plates which quickly faded, but that no photographic slides were made of these plates. (N.T. 669-70). Consequently, there were no exhibits of any chromatographic plates or slides which displayed the ink from the questioned document alongside the known inks which were believed to be similar (N.T. 764-5). Since Brunelle's examination of these plates gave him reason to single out Scripto 3852 as the questioned ink, the destruction of this crucial evidence, without photographing the plates, made it impossible for the jury to see the basis for his conclusion, and left the defendants with no opportunity to cross-examine him on this vital point.
Considering all of the above, it is clear that the Government failed to establish a prima facie case and failed to meet the minimum standards of proof. The only defect which the Government offers to cure in a trial of Bruno on the conspiracy charge is the completeness of Brunelle's ink library. Subsequent to the first trial, Brunelle purportedly was able to secure a large number of foreign ink samples from Interpol in Switzerland. However, the Government concedes that the library is not complete as to all European inks and the added samples still do not include the Japanese inks. More importantly, the mere addition of more inks to Brunelle's library does not cure the inherent defects in the identification process nor in the particular procedures followed by Brunelle which in this case amount to a denial of procedural due process because the very procedures he followed assured the destruction of the defendants' right to meaningful cross-examination. We hold, therefore, that none of Brunelle's identification testimony would be admissible in a future trial of Bruno, and that since the Government concedes it has no other way in which to challenge the authenticity of the questioned document, Bruno's motion to dismiss the indictment as to the conspiracy count must be granted.
B. COLLATERAL ESTOPPEL
But for the severance which defendant Bruno had not requested, it is clear that our decision in the first trial would have resulted in his acquittal on the conspiracy charge. In these coincidental circumstances, the Government asks for a second chance and three more weeks of the court's time to challenge the authenticity of the questioned document. Apart from the evidentiary defects of the Government's case, the policy reasons for the application of the doctrine of collateral estoppel are fully applicable:
"Relitigation results in waste of public funds and the valuable time of the court with already crowded dockets, and delays the opportunity of other citizens to gain initial adjudication of their claims. * * * A single criminal suit is likely to involve a greater expenditure of public funds since it involves the additional expense of prosecution. A policy forbidding reexamination of issues encourages care in the preparation and presentation of the prosecution's initial case."
We must decide whether there is any reason why the Government should not be collaterally estopped from relitigating the authenticity of the questioned document.
In the recent case of Ashe v. Swenson, 397 U.S. 436, 90 S. Ct. 1189, 25 L. Ed. 2d 469 (1970) the Supreme Court reinforced the view generally recognized in the federal courts that collateral estoppel is available as a defense in criminal trials:
"'Collateral estoppel' is an awkward phrase, but it stands for an extremely important principle in our adversary system of justice. It means simply that when an issue of ultimate fact has once been determined by a valid and final judgment, that issue cannot again be litigated between the same parties in any future lawsuit. Although first developed in civil litigation, collateral estoppel has been an established rule of federal criminal law at least since this Court's decision more than 50 years ago in United States v. Oppenheimer, 242 U.S. 85, [37 S. Ct. 68, 61 L. Ed. 161]. As Mr. Justice Holmes put the matter in that case, 'It cannot be that the safeguards of the person, so often and so rightly mentioned with solemn reverence, are less than those that protect from a liability in debt.' 242 U.S., at 87 [37 S. Ct. 68, 61 L. Ed. 161]. As a rule of federal law, therefore, '[it] is much too late to suggest that this principle is not fully applicable to a former judgment in a criminal case, either because of lack of "mutuality" or because the judgment may reflect only a belief that the Government had not met the higher burden of proof exacted in such cases for the Government's evidence as a whole although not necessarily as to every link in the chain.' United States v. Kramer, [2 Cir., 289] F.2d 909, 913."