TABLE 10 - COLORS ASSOCIATED WITH THE UFO
Metallic	126 events		Orange	53 events
Red	90		Green	51
Blue	77		Yellow	42
White	69

In many cases, the colors were from luminous sources on the UFO, not the actual color of the UFO surface (except for metallic cases).

THE EM EFFECT

It was found convenient and appropriate to divide the effects on the vehicles' engines, electrical equipment, or steering into three separate categories. They are:

	1) "EM"--the cases of failure of the engine, lights, or electrical equipment such as the radio. An example would be the Levelland, Texas cases of 1957 (see catalogue).
	2) "Control"--those cases where the engine may or may not be affected, but the vehicle does not respond to the commands of the driver. It may float off the road, drive into a field, or execute other unplanned maneuvers (see the November 2, 1967 case from Ririe, Idaho).
	3) "Destruction"--those cases where the engine or electrical equipment are permanently damaged. Some examples are the melting of a battery, burning of wires and hoses, or malfunctioning of a Cruise Control (see the August 22, 1957 case from Cecil Naval Air Station).

Within the EM category, no attempt was made to subdivide the data into those instances where only the lights were affected, or only the motor, or combinations of these and other effects. Since the response to various forms of ambient energy differs for each system on a vehicle, and the direct cause of the EM events remains unknown, it was considered unwise to make any distinction. Conversely, because destruction of components is so damaging and does leave a physical remnant of the event, it was considered to be a separate category. Within this category, however, no attempt was made to further subdivide the data.  

          A fourth category labeled "Miscellaneous" groups all cases which did not fit into any easy classification scheme (see the February 2, 1969 case from Nuble, Chile). The data:

TABLE 11 - TYPE OF VEHICLE INTERFERENCE
EM	357 events	(82%)
Control	27	(6%)
Destruction	30	(7%)
Miscellaneous	20	(5%)

One additional effect was analyzed: those events where the witness reports that the engine of the vehicle started by itself without his turning the key or pressing on the gas pedal. Because of the uncertainties surrounding such instances, and their obvious remarkable nature, great care was taken when assigning a case to this category. The excitement of the moment, the possibility of not having turned the engine off (if it was not completely affected), the varied types of engine starting procedures, and the fact that in some cases the engine power level is only reduced, are all cause for caution. Still, twenty-one events were finally placed in this category, or about 5% of the total EM events. While rare, it is incontrovertible that these events do occur, even though the author's discussions with several auto mechanics has confirmed that such events should not occur.

MONTHLY AND YEARLY DISTRIBUTION OF EVENTS

As with other UFO events, the number of EM events per month or year has not been constant. Figure 4 presents the total number of events for each month of the year.

Figure 4. Number of Events by Month

Figure 5. Number of Events by Month-Trace Cases

With the exception of October and November, the variation by month does appear to be random. Since the UFO waves of 1954, 1957, and 1973 were all in October or November and included many EM events, this discrepancy can be explained, though it still does not explain why the waves occurred in only those months.

The distribution of EM events by month is closely matched by that for physical trace cases (see Figure 5), though with these, only the month of October deviates greatly from the norm. In Figure 6, the distribution of all UFOCAT cases by month, separated into the northern and southern hemispheres, is unlike the observed peak in the fall for all types of high-strangeness events. The July peak is usually explained by the fact that more people are outside in July, hence available to have a sighting. However, July is the middle of winter in the southern hemisphere, so this explanation seems lacking.

Figure 6. Distribution by Month

Figure 7. Number of Events by Year

The yearly distribution of reports since 1954 is displayed in Figure 7. It is clear that the distribution is not random; the number of events in the period 1966-79 is greater than the number for 1954-66 by over 40%. This difference is larger than can be attributed to chance (p < .01). The increase of EM events recently might be attributed to an actual increase, or perhaps more frequent reporting by witnesses, or an increased number of vehicles on the highway, or an increased number of UFO investigators, or any number of other factors. It is interesting that this increase parallels that of CE III cases, noted by Hendry (9). Figure 8 presents the yearly distribution of high-strangeness cases as well as all cases from UFOCAT since 1947. The data also parallels the EM data.

Figure 8. All UFOCAT Cases by Year

Saunders has investigated a possible "spatio-temporal invariant" of UFO sightings (10). He found that the distribution of certain UFO waves could be regularized if related to geographical location (longitude) and time period of separation. For negatively skewed waves, a period of 5.1 years was isolated, with a corresponding shift 30° to the east of the longitude describing the center of the wave area every period. Hendry has argued that this analysis was incomplete and did not account for all negatively skewed waves (11).

Given this past work, I examined the data under these conditions:

1. That a wave be defined as any month or months (consecutive) with more than ten events
   
2. That the events be located within one country, or two or three smaller ones such as in Europe
   
3. That the number of events decrease rapidly after the peak month, so that the distribution is negatively skewed

Only October 1954, November 1957, March-April 1967, and October 1973 meet these criteria. Thirty-seven months separate the first two waves, 112 months the second and third, and 79 months the third and fourth. The numbers are somewhat intriguing, in that 112 is almost exactly divisible by 37. On the other hand, 112 and 79 are not divisible by a large common factor; neither is 79 a factor of 37. Because of the failure to find any possible period with this first-order analysis, no spectral analysis was attempted. The period found by Saunders, 61 months, obviously does not conform to these results.

All of the waves occurred in the United States, except for the October 1954 wave in France, so no correlation exists with longitude.

Additional investigation will best be carried out after one more wave occurs. It can then be more readily observed whether any regularities exist. The EM waves correspond exactly with major UFO waves of all types. The longest period between both EM and all UFO waves in the post-1947 era has been slightly under nine and one-half years. We are nearing that same period today from the October 1973 wave.