The important thing is simply this; that on the basis of that fragmentary print, it was not possible to determine even the type of pattern that the impression was. Accordingly, we had to compare it with all types of fingerprint patterns, of which there are really four basic types--the arch, tented arch, loop, and whorl. And we are still making comparisons in that case, and we have not been able to identify these few points. Now, that means simply this--that the theory that we are going on an assumption that people do not have the same fingerprints--and we find it not necessary to compare, say for example, a loop pattern with a whorl pattern, and as there is a possibility that, it is contended by some of these so-called authorities, that maybe the points that you find in a loop may be found in the same arrangement in a whorl--is not true. I think that that case, a practical case we have actually worked on, disproves that theory so strongly in my mind that I am convinced that no two people can possibly have the same fingerprints. Mr. EISENBERG. That is, you had a print with seven points, and these same seven points appeared in none of the millions---- Mr. LATONA. Of the millions that we actually compared over a period--well, since 1937. You may recall the case. It was the Matson kidnapping case out in Tacoma, Wash. That is one of only three major kidnapping cases the FBI has not yet solved. Mr. EISENBERG. Are palmprints as unique as fingerprints? Mr. LATONA. Yes; palmprints are. They are not as useful for purposes of setting up a file in order to conduct searches, for the simple reason that there are not as many variations of patterns occurring with any frequency in the palms as occur on the tips of the fingers. That is primarily why the fingertips are used--because you have 10 digits, and there is a possibility of finding variations of the four basic pattern types which can be additionally subdivided by utilizing certain focal points which occur in those particular patterns, which enable us to actually subdivide our files into millions of groups. Accordingly, when you make a search in the fingerprint file, it can be reduced actually to a matter of minutes, whereas to attempt to set up a palmprint file to the extent of the size of the fingerprint file we have in the FBI would be a practical impossibility, much less a waste of time. The CHAIRMAN. Approximately how many fingerprints do you have these days? Mr. LATONA. At the present time, we have the fingerprints of more than 77 million people, and they are subdivided in this fashion: we have two main files; we have the criminal files and we have what are referred to as civil files.