support the fact that an unusual aerial phenomenon occurred, 2) the circumstances of the actual recovery of the specimen are reported, 3) there is data to suggest that the specimen is in fact linked to the observed aerial object, and 4) physical analysis has been performed by a competent laboratory of known reliability. In several cases the sample is available for continuing study by independent scientists. In the present paper we will try to establish the frequency of such cases and the type of analysis they suggest. In conclusion we will examine hypotheses that may deserve further testing.

     In an excellent catalogue compiled by Mr. Larry Hatch³ and made available to researchers and to the general public one finds 15,181 unexplained aerial phenomena reports that have been tabulated in computer-readable form. We have broken down these cases according to the classification system used by this author (Vallee, 1990) in order to bring out the distribution of incidents across various situations. Under this classification, inspired from Hyneks definition of close encounters (Hynek, 1972), each case is given a type and a category. Hynek used a single digit representing the "kind" or type of incident, ranging from "1" for a simple sighting and "2" for physical effects to "3" for report of a lifeform or living entity. We have extended this typology using "4" in cases when witnesses experienced a transformation of their sense of reality (often corresponding to the popular characterization of the incident as an "abduction) and "5" in cases of lasting physiological impact, such as serious injury or death.

     The categories to which the typology is applied range from "CE" for close encounters and "MA" for maneuvers (trajectory discontinuity) to "FB" for fly-by (no observed discontinuity in flight) and "AN" for simple anomalies in which no UFO was reported: unusual lights or unexplained entities fall into this last category.

     Using this classification we would speak of a particular case as a CE-3 incident, or a MA-2 incident, etc., leading to the simple matrix of Table 1, which provides a convenient way for establishing a baseline in comparing reports from various countries or from various epochs.

     ³Larry Hatch's computer catalogue is available from him in diskette format under the DOS operating system running on standard IBM-compatible PCs. His address is 142 Jeter street, Redwood City, California 96062.