Case 3.


Haneda Air Force Base, Japan, August 5–6, 1952

Brief summary: USAF tower operators at Haneda AFB observed an unusually bright bluish-white light to their NE,
alerted the GCI radar unit at Shiroi, which then called for a scramble of an F94 interceptor after getting radar returns
in same general area. GCI ground radar vectored the F94 to an orbiting unknown target, which the F94 picked up on
its airborne radar. The target then accelerated out of the F94’s radar range after 90 seconds of pursuit that was
followed also on the Shiroi GCI radar.

The visual and radar sightings at Haneda AFB, Japan, on August 5–6, 1952, represent an example of a longpuzzling
case, still carried as an unidentified case by Project Bluebook, at my latest check, and chosen for analysis
in the Condon Report. In the latter, is putatively explained in terms of a combination of diffraction and mirage
distortion of the star Capella, as far as the visual parts are concerned, while the radar portions are attributed to
anomalous propagation. I find very serious difficulties with those “explanations” and regard them as typical of a
number of rather casually advanced explanations of long-standing UFO cases that appear in the Condon Report.
Because this case has been discussed in such books as those of Ruppelt, Keyhoe, and Hall, it is of particular interest
to carefully examine case-details on it and then to examine the basis of the Condon Report’s explanation of it, as
example of how the Condon Report disposed of old “classic cases.”

Haneda AFB, active during the Korean War, lay about midway between central Tokyo and central Yokohama,
adjacent to Tokyo International Airport. The 1952 UFO incident began with visual sightings of a brilliant object in
the northeastern sky, as seen by two control tower operators going on duty at 2330 LST (all times hereafter will be
LST). It will serve brevity to introduce some coded name designations for these men and for several officers
involved, since neither the Condon Report, nor my copies of the original Bluebook case-file show names (excised
from latter copies in accordance with Bluebook practice on non-release of witness names in UFO cases):

Coded Designation


Airman A

One of two Haneda tower operators who first sighted light; rank was A/3c.

Airman B

Second Haneda tower operator to first sight light; A/1c.

Lt. A

Controller on duty at Shiroi GCI unit up to 2400, August 5; 1st Lt.

Lt. B

Controller at Shiroi after 0000, August 6; 1st Lt.

Lt. P

Pilot of scrambled F-94; 1st Lt.

Lt. R

Radar officer in F-94; 1st Lt.


Shiroi GCI Station, manned by the 528th AC & W (Aircraft Control and Warning) Group, lay approximately 20
miles NE of Haneda (specifically at 35 deg. 49’ N, 140 deg. 2’ E) and had a CPS-1 10-cm search radar plus a CPS
10-cm height-finding radar. Two other USAF facilities figure in the incident, Tachikawa AFB, lying just over 20
miles WNW of Haneda, and Johnson AFB, almost 30 miles NW of Waneda. The main radar incidents center over
the north extremity of Tokyo Bay, roughly midway from central Tokyo to Chiba across the Bay.

The Bluebook case-file on this incident contains 25 pages, and since the incident predates promulgation of
AFR200-2, the strictures on time-reporting, etc., are not here so bothersome as in the Lakenheath case of 1956,
discussed above. Nevertheless, the same kind of disturbing internal inconsistencies are present here as one finds in
most Bluebook case reports; in particular, there is a bothersome variation in times given for specific events in
different portions of the case-file. One of these, stressed in the Condon Report, will be discussed explicitly below;
but for the rest, I shall use those times which appear to yield the greatest over-all internal consistency. This will
introduce no serious errors, since the uncertainties are mostly only 1 or 2 minutes and, except for the cited instance,
do not alter any important implications regardless of which cited time is used. The overall duration of the visual and
radar sightings is about 50 minutes. The items of main interest occurred between 2330 and 0020, approximately.
Although this case involves both visual and radar observations of unidentified objects, careful examination does
not support the view that the same object was ever assuredly seen visually and on radar at the same time, with the
possible exception of the very first radar detection just after 2330. Thus it is not a “radar-visual” case, in the more
significant sense of concurrent two-channel observations of an unknown object. This point will be discussed further
in Section 5.

Visual Observations

A. First visual detection. At 2330, Airmen A and B, while walking across the ramp at Haneda AFB to go on the midnight shift at the airfield control tower, noticed an “exceptionally bright light” in their northeastern sky. They went immediately to the control tower to alert two other on-duty controllers to it and to examine it more carefully with the aid of the 7x50 binoculars available in the tower. The Bluebook case-file notes that the two controllers already on tower-duty “had not previously noticed it because the operating load had been keeping their attention elsewhere.”

B. Independent visual detection at Tachikawa AFB. About ten minutes later, according to the August 12, 1952, Air Intelligence Information Report (IR-35-52) in the Bluebook case-file; Haneda was queried about an unusually bright light by controllers at Tachikawa AFB, 21 miles to their WNW. IR-35-52 states: “The control tower at Tachikawa Air Force Base called Haneda tower at approximately 2350 to bring their attention to a brilliant white light over Tokyo Bay. The tower replied that it had been in view for some time and that it was being checked.”

This feature of the report is significant in two respects: 1) It indicates that the luminous source was of
sufficiently unusual brilliance to cause two separate groups of Air Force controllers at two airfields to respond independently and to take alert-actions; and 2) More significantly, the fact that the Tachikawa controllers saw the source in a direction “over Tokyo Bay” implies a line-of-sight distinctly south of east. From Tachikawa, even the north end of the Bay lies to the ESE. Thus the intersection of the two lines of sight fell somewhere in the northern half of the Bay, it would appear. As will be seen later, this is where the most significant parts of the radar tracking occurred subsequently.

C. Direction, intensity, and configuration of the luminous source. IR-35-52 contains a signed statement by Air man A, a sketch of the way the luminous source looked through 7-power binoculars, and summary comments by Capt. Charles J. Malven, the FEAF intelligence officer preparing the report for transmission to Bluebook.

Airman A’s own statement gives the bearing of the source as NNE; Malven summary specifies only NE.
Presumably the witness’ statement is the more reliable, and it also seems to be given a greater degree of precision, whence a line-of-sight azimuth somewhere in the range of 25 to 35 deg. east of north appears to be involved in the Haneda sightings. By contrast, the Tachikawa sighting-azimuth was in excess of 90 deg. from north, and probably beyond 100 deg., considering the geography involved, a point I shall return to later.


Several different items in the report indicate the high intensity of the source. Airman A’s signed statement refers to it as “the intense bright light over the Bay.” The annotated sketch speaks of “constant brilliance across the entire area” of the (extended) source, and remarks on “the blinding effect from the brilliant light.” Malven’s summary even points out that “Observers stated that their eyes would fatigue rapidly when they attempted to concentrate their vision on the object,” and elsewhere speaks of “the brilliant blue-white light of the object.” Most of these indications of brightness are omitted from the Condon Report, yet bear on the Capella hypothesis in terms of which that Report seeks to dispose of these visual sightings.

Airman A’s filed statement includes the remark that “I know it wasn’t a star, weather balloon or Venus, because I compared it with all three.” This calls for two comments. First, Venus is referred to elsewhere in the case-file, but this is certainly a matter of confusion, inasmuch as Venus had set that night before about 2000 LST. Since elsewhere in the report reference is made to Venus lying in the East, and since the only noticeable celestial object in that sector at that time would have been Jupiter, I would infer that where “Venus” is cited in the case-file, one should read “Jupiter.” Jupiter would have risen near 2300, almost due east, with apparent magnitude –2.0. Thus Airman A’s assertion that the object was brighter than “Venus” may probably be taken to imply something of the order of magnitude –3.0 or brighter. Indeed, since it is most unlikely that any observer would speak of a –3.0 magnitude source as “blinding” or “fatiguing” to look at, I would suggest that the actual luminosity, at its periods of peak value (see below) must have exceeded even magnitude –3 by a substantial margin.

Airman A’s allusion to the intensity as compared with a “weather balloon” refers to the comparisons (elaborated below) with the light suspended from a pilot balloon released near the tower at 2400 that night and observed by the tower controllers to scale the size and brightness. This is a very fortunate scaling comparison, because the small battery-operated lights long used in meteorological practice have a known luminosity of about 1.5 candle. Since a 1-candle source at 1 kilometer yields apparent magnitude 0.8, inverse-square scaling for the here known balloon distance of 2000 feet (see below) implies an apparent magnitude of about –0.5 for the balloon-light as viewed at time of launch. Capt. Malven’s summary states, in discussing this quite helpful comparison, “The balloon’s light was described as extremely dim and yellow, when compared to the brilliant blue white light of the object.” Hereagain, I believe one can safely infer an apparent luminosity of the object well beyond Jupiter’s –2.0. Thus, we have here a number of compatible indications of apparent brightness well beyond that of any star, which will later be seen
to contradict explanations proposed in the Condon Report for the visual portions of the Haneda sightings.

Of further interest relative to any stellar source hypothesis are the descriptions of the configuration of the object as seen with 7-power binoculars from the Haneda tower, and its approximate angular diameter. Fortunately, the latter seems to have been adjudged in direct comparison with an object of determinate angular subtense that was in view in the middle of the roughly 50-minute sighting. At 2400, a small weather balloon was released from a point at a known distance of 2000 ft from the control tower. Its diameter at release was approximately 24 inches. (IR-35-52 refers to it as a “ceiling balloon,” but the cloud-cover data contained therein is such that no ceiling balloon would have been called for. Furthermore, the specified balloon mass, 30 grams, and diameter, 2 ft, are precisely those of a standard pilot balloon for upper-wind measurement. And finally, the time [2400 LST = 1500Z] was the standard time for a pilot balloon run, back in that period.) A balloon of 2-ft diameter at 2000-ft range would subtend 1 milliradian, or just over 3 minutes of arc, and this was used by the tower observers to scale the apparent angular
subtense of the luminous source.


As IR-35-52 puts it:

“Three of the operators indicated the size of the light, when closest to the tower, was approximately the same as the small ceiling balloons (30 grams, appearing 24 inches in diameter) when launched from the weather station, located at about 2000 ft from the tower. This would make the size of the central light about 50 ft in diameter, when at the 10 miles distance tracked by GCI.... A lighted weather balloon was launched at 2400 hours...”

Thus, it would appear that an apparent angular subtense close to 3 minutes of arc is a reasonably reliable estimate for the light as seen by naked eye from Haneda. This is almost twice the average resolution-limit of the human eye, quite large enough to match the reported impressions that it had discernible extent, i.e., was not merely a point source.

But the latter is very much more clearly spelled out, in any event, for IR-35-52 gives a fairly detailed description of the object’s appearance through 7-power binoculars. It is to be noted that, if the naked-eye diameter were about 3 minutes, its apparent subtense when viewed through 7X-binoculars would be about 20 minutes, or two-thirds the naked-eye angular diameter of the full moon — quite large enough to permit recognition of the finer details cited in IR-35-52, as follows: “The light was described as circular in shape, with brilliance appearing to be constant across the face. The light appeared to be a portion of a large round dark shape which was about four times the diameter ofthe light. When the object was close enough for details to be seen, a smaller, less brilliant light could be seen at the lower left hand edge, with two or three more dim lights running in a curved line along the rest of the lower edge of the dark shape. Only the lower portion of the darker shape could be determined, due to the lighter sky which was believed to have blended with the upper side of the object. No rotation was noticed. No sound was heard.”

Keeping in mind that those details are, in effect, described for an image corresponding in apparent angular size to over half a lunar diameter, the detail is by no means beyond the indiscernible limit. The sketch included with IR-35-52 matches the foregoing description, indicating a central disc of “constant brilliance across entire area (not due to a point source of light),” an annular dark area of overall diameter 3–4 times that of the central luminary, and having four distinct lights on the lower periphery, “light at lower left, small and fairly bright, other lights dimmer and possibly smaller.” Finally, supportive comment thereon is contained in the signed statement of Airman A. He comments:

“After we got in the tower I started looking at it with binoculars, which made the object much clearer.

Around the bright white light in the middle, there was a darker object which stood out against the sky, having little white lights along the outer edge, and a glare around the whole thing.”

All of these configurational details, like the indications of a quite un-starlike brilliance, will be seen below to be almost entirely unexplainable on the Capella hypothesis with which the Condon Report seeks to settle the Haneda visual sightings. Further questions ultimately arise from examination of reported apparent motions of the luminous source, which will be considered next.

D. Reported descriptions of apparent motions of the luminous source. Here we meet the single most important ambiguity in the Haneda case-file, though the weight of the evidence indicates that the luminous object exhibited definite movements. The ambiguity arises chiefly from the way Capt. Malven summarized the matter in his IR-35-52 report a week after the incident; “The object faded twice to the East, then returned. Observers were uncertain whether disappearance was due to a dimming of the lights, rotation of object, or to the object moving away at terrific speed, since at times of fading the object was difficult to follow closely, except as a small light. Observers did agree that when close, the object did appear.

In contrast to the above form in which Malven summarized the reported motions, the way Airman A described them in his own statement seems to refer to distinct motions, including transverse components:

“I watched it disappear twice through the glasses. It seemed to travel to the East and gaining altitude at a very fast speed, much faster than any jet. Every time it disappeared it returned again, except for the last time when the jets were around. It seemed to know they were there. As for an estimate of the size of the object — I couldn’t even guess.”

Recalling that elsewhere in that same signed statement this tower controller had given the observed direction to the object as NNE, his specification that the object “seemed to travel to the East” seems quite clearly to imply a non radial motion, since, if only an impression of the latter were involved, one would presume he would have spoken of it in some such terms as “climbing out rapidly to the NNE.” Since greater weight is presumably to be placed on direct-witness testimony than on another’s summary thereof, it appears necessary to assume that not mere radial recession but also transverse components of recession. upwards and towards the East, were observed.

That the luminous source varied substantially in angular subtense is made very clear at several points in the casefile:

One passage already cited discusses the “size of the light, when closest to the tower...,” while, by contrast, another says that: “At the greatest distance, the size of the light appeared slightly larger than Venus, approximately due East of Haneda, and slightly brighter.” (For “Venus” read “Jupiter” as noted above. Jupiter was then near quadrature with angular diameter of around 40 seconds of arc. Since the naked eye is a poor judge of comparative angular diameters that far below the resolution limit, little more can safely be read into that statement than the conclusion that the object’s luminous disc diminished quite noticeably and its apparent brightness fell to a level comparable to or a bit greater than Jupiter’s when at greatest perceived distance. By virtue of the latter, it should be noted, one has another basis for concluding that when at peak brilliance it must have been considerably brighter than Jupiter’s –2.0, a conclusion already reached by other arguments above.

In addition to exhibiting what seems to imply recession, eastward motion, and climb to disappearance, the source also disappeared for at least one other period far too long to be attributed to any scintillation or other such meteorological optical effect: “When we were about half way across the ramp (Airman A stated), it disappeared for the first time and returned to approximately the same spot about 15 seconds later.” There were scattered clouds over Haneda at around 15–16,000 ft, and a very few isolated clouds lower down, yet it was full moon that night and, if patches of clouds had drifted very near the controllers’ line-of-sight to the object, they could be expected to have seen the clouds. (The upper deck was evidently thin, for Capt. Malven notes in his report that “The F94 crew reported exceptional visibility and stated that the upper cloud layer did not appreciably affect the brilliancy of the moonlight.”) A thin cloud interposed between observer and a distant luminous source would yield an impression of dimming and enhanced effective angular diameter, not dimming and reduced apparent size, as reported here. I believe the described “disappearances” cannot, in view of these several considerations, reasonably be attributed to
cloud effects.

I have now summarized the essential features of the Haneda report dealing with just the visual observations of some bright luminous source that initiated the alert and that led to the ground-radar and air borne-radar observations yet to be described. Before turning to those, which comprise, in fact, the more significant portion of the over-all sighting, it will be best to turn next to a critique of the Blue book and the Condon Report attempts to give an explanation of the visual portions of the sighting.

Bluebook Critique of the Visual Sightings
In IR-35-52. Capt, Malven offers only one hypothesis, and that in only passing manner: He speculates briefly on
whether “reflections off the water (of the Bay, I presume) were...sufficient to form secondary reflections off the
lower clouds,” and by the latter he refers to “isolated patches of thin clouds reported by the F-94 crew as being at
approximately 4000 feet...” He adds that “these clouds were not reported to be visible by the control tower
personnel,” which, in view of the 60-mile visibility cited elsewhere in the case-file and in view of the full moon then
near the local meridian, suggests that those lower clouds must have been exceedingly widely scattered to escape
detection by the controllers.

What Malven seems to offer there, as an hypothesis for the observed visual source, is cloud-reflection of
moonlight — and in manner all too typical of many other curious physical explanations one finds scattered through
Bluebook case-files, he brings in a consideration that reveals lack of appreciation of what is central to the issue. If he
wants to talk about cloud-reflected moonlight, why render a poor argument even weaker by invoking not direct
moonlight but moonlight secondarily reflected off the surface of Tokyo Bay? Without even considering further that
odd twist in his tentative hypothesis, it is sufficient to note that even direct moonlight striking a patch of cloud is not
“reflected” in any ordinary sense of that term. It is scattered from the cloud droplets and thereby serves not to create
any image of a discrete light source of blinding intensity that fatigues observers’ eyes and does the other things
reported by the Haneda observers, but rather serves merely to palely illuminate a passing patch of cloud material. A
very poor hypothesis.

Malven drops that hypothesis without putting any real stress on it (with judgment that is not always found where
equally absurd “explanations” have been advanced in innumerable other Bluebook case-files by reporting officers or
by Bluebook staff members). He does add that there was some thunderstorm activity reported that night off to the
northwest of Tokyo, but mentions that there was no reported electrical activity therein. Since the direction is
opposite to the line of sight and since the reported visual phenomena bear no relation to lightning effects, this carried
the matter no further, and the report drops that point there.

Finally, Malven mentions very casually an idea that I have encountered repeatedly in Bluebook files yet nowhere
else in my studies of atmospheric physics, namely, “reflections off ionized portions of the atmosphere.” He states:
“Although many sightings might be attributed to visual and electrical reflections off ionized areas in the atmosphere,
the near-perfect visibility on the night of the sighting, together with the circular orbit of the object would tend to
disprove this theory.” Evidently he rejects the “ionized areas” hypothesis on the ground that presence of such areas
is probably ruled out in view of the unusually good visibility reported that night. I trust that, for most readers of this
discussion, I would only be belaboring the obvious to remark that Bluebook mythology about radar and visual
“reflections” off “ionized regions” in the clear atmosphere (which mythology I have recently managed to trace back
even to pre-1950 Air Force documents on UFO reports) has no known basis in fact, but is just one more of the all
too numerous measures of how little scientific critique the Air Force has managed to bring to bear on its UFO
problems over the years.

Although the final Bluebook evaluation of this entire case, including the visual portions, was and is
“Unidentified,” indicating that none of the above was regarded as an adequate explanation of even the visual
features of the report, one cannot overlook extremely serious deficiencies in the basic reporting and the interrogation
and follow-up here. This incident occurred in that period which my own studies lead me to describe as sort of a
highwater mark for Project Bluebook. Capt. Edward J. Ruppelt was then Bluebook Officer at Wright-Patterson
AFB, and both he and his superiors were then taking the UFO problem more seriously than it was taken by USAF at
any other time in the past 22 years. Neither before nor after 1952–3 were there as many efforts made to assemble
case-information, to go out and actually check in the field on sightings, etc. Yet it should be uncomfortably apparent
already at this point in this discussion of the Haneda case that quite basic points were not run to ground and pinned
down. Ruppelt, in his 1956 book, speaks of this Haneda case as if it were regarded as one of the most completely
reported cases they’d received as of mid-1952. He mentioned that his office sent a query to FEAF offices about a
few points of confusion, and that the replies came back with impressive promptness, etc. If one needed some
specific clue to the regrettably low scientific level of the operation of Bluebook even during this period of
comparatively energetic case-investigation, one can find it in study of the Haneda report. Even so simple a matter as
checking whether Venus was actually in the East was obviously left undone; and numerous cross-questions and
followup queries on motions, angles, times, etc., not even thought of. That, I stress, is what any scientist who studies
the Bluebook files as I have done will find all through 22 years of Air Force handling of the UFO problem.

Incompetence and superficiality — even at the 1952 highwater mark under Ruppelt’s relatively vigorous [missing].
And in the final paragraph discussing this case, the Condon Report merely rounds it off to: “In summary, it
appears that the most probable causes of this UFO report are an optical effect on a bright light source that produced
the visual sighting...” (and goes on to a remark on the radar portions we have yet to examine here) .
There are some very serious difficulties with the more specific parts of the suggested explanation, and the
vagueness of the other parts is sufficiently self-evident to need little comment.

First, nothing in the literature of meteorological optics discusses any diffraction-produced coronae with a dark
annular space extending out to three or four diameters of the central luminary, such as is postulated in the above
Condon Report explanation. The radial intensity pattern of a corona may be roughly described as a damped
oscillatory radial variation of luminosity, with zero intensity minima (for the simple case of a monochromatic
luminary) at roughly equal intervals, and no broad light-free annulus comparable to that described in detail by the
Haneda controllers. Thus, lack of understanding of the nature of coronae is revealed at the outset in attempting to fit
the Haneda observations to such a phenomenon.

Second, droplets certainly do not have to be “spaced at regular intervals” to yield a corona, and Minnaert’s book
makes no such suggestion, another measure of misunderstanding of the meteorological optics here concerned. Nor is
there any physical mechanism operating in clouds capable of yielding any such regular droplet spacing. Both
Minnaert and cloud physics are misunderstood in that passage.

Third, one quickly finds, by some trial calculations, using the familiar optical relation (Exner equation) for the
radial positions of the minima of the classical corona pattern, that the cited drop diameter of 0.2 mm = 200 microns
was obtained in the Condon Report by back-calculating from a tacit requirement that the first-order minimum lay
close to 3 milliradians, for these are the values that satisfy the Exner equation for an assumed wavelength of about
0.5 microns for visible light. This discloses even more thorough misunderstanding of corona optics, for that firstorder
minimum marks not some outer edge of a broad dark annulus as described and sketched by the Haneda tower
operators, but the outer edge of the innermost annulus of high intensity of diffracted light. This clearly identifies
basic misunderstanding of the matters at hand.

Fourth, the just-cited computation yielded a droplet diameter of 200 microns, which is so large as to be found
only in drizzling or raining clouds and never in thin scattered clouds of the sort here reported, clouds that scarcely
attenuated the full moon’s light. That is, the suggestion that “patches of fog or mist” collected under an hypothesized
inversion could grow droplets of that large size is meteorologically out of the question. If isolated patches of clouds
interposed themselves on an observer’s line of sight to some distant luminary, under conditions of the sort prevailing
at Haneda that night, drop diameters down in the range of 10–20 microns would be the largest one could expect, and
the corona-size would be some 10 to 20 times greater than the 3 milliradians which was plugged into the Exner
equation in the above, b), and Fifth, the vague suggestion that “Raman brightening” or other “interference effects associated with propagation within and near the top of an inversion” is involved here makes the same serious error that is made in attempted optical explanations of other cases in the Condon Report. Here we are asked to consider that light from Capella, whose altitude was about 8 deg. above the NE horizon (a value that I confirm) near the time of the Haneda
observations, was subjected to Raman brightening or its equivalent; yet one of the strict requirements of all such
interference effects is that the ray paths impinge on the inversion surface at grazing angles of incidence of only a
small fraction of a degree. No ground observer viewing Capella at 8 deg. elevation angle could possibly see
anything like Raman brightening, for the pertinent angular limits would be exceeded by one or two orders of
magnitude. Added to this measure of misunderstanding of the optics of such interference phenomena in this
attempted explanation is the further difficulty that, for any such situation as is hypothesized in the Condon Report
explanation, the observer’s eye must be physically located at or directly under the index-discontinuity, which would
here mean up in the air at the altitude of the hypothesized inversion. But all of the Haneda observations were made
from the ground level. Negation of Raman brightening leaves one more serious gap in the Capella hypothesis, since
its magnitude of 0.2 lies at a brightness level well below that of Jupiter, yet the Haneda observers seem to have been
comparing the object’s luminosity to Jupiter’s and finding it far brighter, not dimmer.

Sixth, the Condon Report mentions the independent sighting from Tachikawa AFB, but fails to bring out that the
line of sight from that observing site (luminary described as lying over Tokyo Bay, as seen from Tachikawa) pointed
more than 45 deg. away from Capella, a circumstance fatal to fitting the Capella hypothesis to both sightings. Jupiter
lay due East, not “over Tokyo Bay” from Tachikawa, and it had been rising in the eastern sky for many days, so it is,
in any event, unlikely to have suddenly triggered an independent response at Tachikawa that night. And, conversely,
the area intersection of the reported lines of sight from Haneda and Tachikawa falls in just the North Bay area where
Shiroi GCI first got radar returns and where all the subsequent radar activity was localized.

Seventh, nothing in the proffered explanations in the Condon Report confronts the reported movements and
disappearances of the luminous object that are described in the Bluebook case-file on Haneda. If, for the several
reasons offered above, we conclude that not only apparent radial motions, but also lateral and climbing motions
were observed, neither diffraction nor Raman effects can conceivably fit them.

Eighth, the overall configuration as seen through 7X binoculars, particularly with four smaller lights perceived
on the lower edge of the dark annulus, is not in any sense explained by the ideas qualitatively advanced.
Ninth, the Condon Report puts emphasis on the point that, whereas Haneda and Tachikawa observers saw the
light, airmen at the Shiroi GCI site went outside and looked in vain for the light when the plotted radar position
showed one or more targets to their south or south-southeast. This is correct. But we are quite familiar with both
highly directional and semi-directional light sources on our own technological devices, so the failure to detect a light
from the Shiroi side does not very greatly strengthen the hypothesis that Capella was the luminary in the Haneda
visual sightings. The same can be said for lack of visual observations from the F-94, which got only radar returns as
it closed on its target.

I believe that it is necessary to conclude that the “explanation” proposed in the Condon Report for the visual
portions of the Haneda case are almost wholly unacceptable. And I remark that my analysis of many other
explanations in the Condon Report finds them to be about equally weak in their level of scientific argumentation.
We were supposed to get in the Condon Report a level of critique distinctly better than that which had come from
Bluebook for many years; but much of the critique in that Report is little less tendentious and ill-based than that
which is so dismaying in 22 years of Air Force discussions of UFO cases. The above stands as only one illustration
of the point I make there; many more could be cited.

Next we must examine the radar aspects of the 8/5–6/52 Haneda case.

Radar Observations
Shortly after the initial visual sighting at Haneda, the tower controllers alerted the Shiroi GCI radar unit (located
about 15 miles NE of central Tokyo), asking them to look for a target somewhere NE of Haneda at an altitude which
they estimated (obviously on weak grounds) to be somewhere between 1500 and 5000 feet, both those figures
appearing in the Bluebook case-file. Both a CPS-1 search radar and a CPS-4 height-finder radar were available at
Shiroi, but only the first of those picked up the target, ground clutter interference precluding useful CPS-4 returns.
The CPS-1 radar was a 10-cm, 2-beam set with peak power of 1 megawatt, PRF of 400/sec, antenna tilt 3 deg., and
scan-rate operated that night at 4 rpm. I find no indication that it was equipped with MTI, but this point is not

It may help to keep the main sequence of events in better time order if I first put down the principal events that
bear on the radar sightings from ground and air, and the times at which these events occurred. In some instances a
1–2 minute range of times will be given because the case-file contains more than a single time for that event as
described in separate sections of the report. I indicate 0015-16 LST (all times still LST) as the time of first airborne
radar contact by the F-94, and discuss that matter in more detail later, since the Condon Report suggests a quite
different time.


Time (LST)



Tower controllers at Haneda see bright light to NE, call Shiroi GCI within a few minutes.


Lt. A, Shiroi radar controller on evening watch, looks for returns, finds 3-4 stationary blips to NE of Haneda on low beam of CPS-1.


Lt. B comes on duty for midwatch at Shiroi; he and Lt. A discuss possible interceptor scramble.


Lt. A calls Johnson AFB, asks for F-94 scramble. Fuel system trouble causes delay of 5-10 min.


Lt. B has unknown in right orbit at varying speeds over north Tokyo Bay, 8 miles NE of Haneda. Loses contact again.


F-94 airborne out of Johnson AFB, Lt. P as pilot, Lt. R, radarman.


Shiroi alerts F-94 to airborne target to its starboard as it heads down Tokyo Bay, and Lt. P visually identifies target as C-54 in pattern to land at Haneda. Lt. B instructs Lt. P to begin search over north Bay area at flight altitude of 5000 ft.


Shiroi regains CPS-1 contact on unknown target in right orbit over same general area seen before, target splits into three separate targets, and Lt. B vectors F-94 toward strongest of three returns.


F-94 gets airborne radar contact on moving target at range and bearing close to vector information, has to do hard starboard turn to keep onscope as target moves with acceleration across scope.


After 90 seconds pursuit, with no lock-on achieved, target moves offscope at high speed; Shiroi GCI tracks both unknown and F-94 into its ground clutter, where both are then lost in clutter.</td>


Shiroi releases F-94 from scramble-search


F-94 visually spots another C-54, over Johnson.


F-94 lands back at Johnson.


Thus the period 2330 on 8/5 through about 0018 on 8/6 is of present interest: Next, events in that period will be
examined in closer detail.


A. Initial attempts at radar detection from Shiroi GCI. When, at about 2335 or so, Haneda requested Shiroi to
search the area of the bay to the NE of Haneda (SSW from Shiroi, roughly), Lt. A, then duty controller at Shiroi,
found his CPS-4 giving too much ground clutter to be useful for the relatively low estimated heights Haneda had
suggested. Those heights are indicated as 1500–2000 ft in one portion of the case-file, though Airman A elsewhere
gave 5000 ft as his impression of the height. Clearly, lack of knowledge of size and slant ranges precluded any exact
estimates from Haneda, but they offered the above-indicated impressions.

Trying both low and high beams on the CPS-1 search radar, Lt. A did detect three or four blips “at a position 050
deg. bearing from Haneda, as reported by the tower, but no definite movement could be ascertained...” The report
gives no information on the range from Shiroi, nor inferred altitude of those several blips, only the first of a
substantial number of missing items of quite essential information that were not followed up in any Bluebook
inquiries, as far as the case-file shows. No indication of the spacing of the several targets is given either, so it is
difficult to decide whether to consider the above as an instance of “radar visual” concurrency or not. One summary
discussion in the Bluebook case-file so construes it: “The radar was directed onto the target by visual observations
from the tower. So it can safely be assumed that both visual and radar contacts involved the same object.” By
contrast, the Condon Report takes the position that there were no radar observations that ever matched the visual
observations. The latter view seems more justified than the former, although the issue is basically unresolvable. One
visual target won’t, in any event, match 3–4 radar targets, unless we invoke the point that later on the main radar
target split up into three separate radar targets, and assume that at 2335, 3–4 unknown objects were airborne and
motionless, with only one of these luminous and visually detectable from Haneda.


That is conceivable but involves too strained assumptions to take very seriously; so I conclude that, even in this opening radar search, there was not obvious correspondence between visual and radar unknowns. As we shall see, later on there was definitely not correspondence, and also the F-94 crew never spotted a visual target. Thus, Haneda cannot be viewed as a case involving the kind of “radar-visual” concurrency which does characterize many other important cases. Nonetheless, both the visual and the radar features, considered separately, are sufficiently unusual in the Haneda case to regard them as mutually supporting the view that inexplicable events were seen and tracked there that night.

One may ask why a radar-detected object was not seen visually, and why a luminous object was not detected on
search radar; and no fully satisfactory answer lies at hand for either question. It can only be noted that there are
many other such cases in Bluebook files and that these questions stand as part of the substantial scientific puzzle that
centers around the UFO phenomena. We know that light-sources can be turned off, and we do know that ECM
techniques can fool radars to a certain extent. Thus, we might do well to maintain open minds when we come to
these questions that are so numerous in UFO case analyses.

B. F-94 scramble. When Lt. B came on duty at 2345, he was soon able, according to Capt. Malven’s summary in
IR-35-52, “to make radar contact on the 50-mile high beam,” whereupon he and Lt. A contacted the ADCC flight
controller at Johnson AFB 35 miles to their west, requesting that an interceptor be scrambled to investigate the
source of the visual and the radar sightings.

An F-94B of the 339th Fighter-Interceptor Squadron, piloted by Lt. P, with Lt. R operating the APG-33 airintercept
radar, was scrambled, though a delay of over ten minutes intervened because of fuel-system difficulties during engine run-up. The records show the F-94 airborne at about 0003-04, and it then took about 10 minutes to reach the Tokyo Bay area. The APG-33 set was a 3-cm (X-band) set with 50 KW power, and lock-on range of about 2500 yards, according to my information. The system had a B-scope, i.e., it displayed target range vs. azimuth. The case-file notes that: “The APG-33 radar is checked before and after every mission and appeared to be working normally.”

At 0009, Shiroi picked up a moving target near Haneda and alerted the F-94 crew, who had no difficulty
identifying it visually as an Air Force C-54 in the Haneda pattern. The crew is quoted in the report as reporting
“exceptional visibility.” Shiroi instructed the F-94 to begin searching at 5000 ft altitude as it got out over the Bay.
But before proceeding with events of that search, a GCI detection of a moving target at about 0001 must be

C. First GCI detection of orbiting object. Just before the F-94 became airborne out of Johnson AFB, Lt. B picked
up the first definitely unusual moving target, at about 0000-01. His statement in the Bluebook case-file reads: “At
the time of the scramble, I had what was believed to be the object in radar contact. The radar sighting indicated the
object to be due south of this station over Tokyo Bay and approximately eight (8) miles northeast of Haneda. The
target was in a right orbit moving at varying speeds. It was impossible to estimate speed due to the short distance
and times involved.” That passage is quoted in the Condon Report, but not the next, which comes from Malven’s
summary and indicates that Lt. B only meant that it was impossible to estimate the target’s speed with much
accuracy. The omitted passage is interesting, for it is one of a number of indications that anomalous propagation
(which is the Condon Report’s explanation for the radar sightings) is scarcely creditable:
An F-94 was scrambled to investigate. The object at this time had left the ground clutter and could be
tracked (on the CPS-1) at varying speeds in a right orbit. Although impossible to accurately estimate
speed, Lt. B gave a rough estimate of 100–150 knots, stopping, and hovering occasionally, and a
maximum speed during the second orbit (just before F-94 was vectored in) of possibly 250–300 knots.

A map accompanying IR-35-52 shows the plotted orbiting path of the unknown target. The orbit radius is
approximately 4 miles, centered just off the coast from the city of Funabashi, east of Tokyo. The orbiting path is
about half over land, half over water. The map sketch, plus the file comments, imply that GCI had good contacts
with the target only while it was moving out over the Bay. The ground-clutter pattern of the CPS-1 is plotted on the
same map (and on other maps in the file), and it seems clear that the difficulty in tracking the target through the land
portion of the roughly circular orbit was that most of that portion lay within the clutter area. The presumption is
strong that this set did not have MTI, which is unfortunate.

The circumference of the orbit of about 4-mi radius would be about 25 miles. Taking Lt. B’s rough estimate of
100–150 knots in the first of the two circuits of this orbit (i.e., the one he detected at about 0001), a total circuit-time
of perhaps 12–13 minutes is indicated. Although the basis for this time-estimate is quite rough, it matches
reasonably well the fact that it was about 0012 when it had come around again, split up into three targets, and looped
onshore again with the F-94 in pursuit this time.
If the object executing the above orbits had been the luminous object being watched from Haneda, it would have
swung back and forth across their sky through an azimuth range of about 30 deg. Since no such motion seems to
have been noted by the Haneda observers, I believe it must be concluded that the source they watched was distinct
from the one radar-tracked in orbit.

D. Second orbit and F-94 intercept attempt. The times given in Lt. B’s account of this phase of the sighting do not
match those given by the pilot and radarman of the F-94 in their signed statements in the file. Other accounts in the
file match those of the aircrew, but not the times in Lt. B’s summary. This discrepancy (about 10–12 minutes) is
specifically noted in Capt. Malven’s IR-35-52 summary: “The ten minute difference in time between the statement
by Lt. B, 528th ACGW SQ, and that reported by other personnel concerned, is believed to be a typographical error,
since the statement agrees on every other portion of the sighting.” That Lt. B and the aircrew were describing one
and the same intercept seems beyond any doubt; and in view of Malven’s quoted comment, I here use the times
recorded by the aircrew and accepted as the correct times in other parts of the case-file. Further comment on this will
be given below.

After completing the first of the two orbits partially tracked by GCI Shiroi, the target came around again where it
was out of the CPS-1 ground- clutter pattern, and Lt. B regained contact. Malven’s summary comments on the next
developments as follows: “At 0012 the object reportedly broke into three smaller contacts, maintaining an interval of
about 1/4 miles, with one contact remaining somewhat brighter. The F-94 was vectored on this object, reporting
weak contact at 1500 and loss of contact at 0018. Within a few seconds, both the F-94 and the object entered the
ground clutter and were not seen again.”

The same portion of the incident is summarized in Lt. B’s account (with different times), with the F-94 referred
to by its code-name “Sun Dial 20.” Immediately following the part of his account referring to the first starboard
orbit in which he had plotted the target’s movements, at around 0001, comes the following section: “Sun Dial 20
was ordered to search the Tokyo Bay area keeping a sharp lookout for any unusual occurrences. The object was
again sighted by radar at 0017 on a starboard orbit in the same area as before. Sun Dial 20 was vectored to the target.
He reported contact at 0025 and reported losing contact at 0028. Sun Dial 20 followed the target into our radar
ground clutter area and we were unable to give Sun Dial 20 further assistance in re-establishing contact. Sun Dial 20
again resumed his visual search of the area until 0014, reporting negative visual sighting on this object at any time.”

If Malven’s suggestion of typographical error is correct, the in-contact times in the foregoing should read 0015 and
0018, and presumably 0017 should be 0012. But regardless of the precise times, the important point is that Lt. B
vectored the F-94 into the target, contact was thereby achieved, and Lt. B followed the target and pursuing F-94
northeastward into his ground clutter. I stress this because, in the Condon Report, the matter of the different times
quoted is offered as the sole basis of a conclusion that ground radar and airborne radar were never following the
same target. This is so clearly inconsistent with the actual contents of the case-file that it is difficult to understand
the Report rationale.

Even more certain indication that the GCI radar was tracking target and F-94 in this crucial phase is given in the
accounts prepared and signed by the pilot and his radarman. Here again we meet a code-designation, this time “Hi-
Jinx,” which was the designation for Shiroi GCI used in the air-to-ground radio transmissions that night and hence
employed in these next two accounts. The F-94 pilot, Lt. P states:

“The object was reported to be in the Tokyo Bay area in an orbit to the starboard at an estimated altitude of 5,000 feet. I observed nothing of an unusual nature in this area; however, at 0016 when vectored by Hi-Jinx on a heading of 320 degrees, and directed to look for a bogie at 1100 o’clock, 4 miles, Lt. R made radar contact at 10 degrees port, 6000 yards. The point moved rapidly from port to starboard and disappeared from the scope. I had no visual contact with the target.”

And the signed statement from the radarman, Lt. R, is equally definite about these events: “At 0015 Hi-Jinx gave
us a vector of 320 degrees. Hi-Jinx had a definite radar echo and gave us the vector to intercept the unidentified
target. Hi-Jinx estimated the target to be at 11 o’clock to us at a range of 4 miles. At 0016 I picked up the radar
contact at 10 degrees port, 10 degrees below at 6,000 yards. The target was rapidly moving from port to starboard
and a ‘lock on’ could not be accomplished. A turn to the starboard was instigated to intercept target which
disappeared on scope in approximately 90 seconds. No visual contact was made with the unidentified target. We
continued our search over Tokyo Bay under Hi-Jinx control. At 0033 Hi-Jinx released us from scrambled mission...”
Of particular importance is the very close agreement of the vectoring instructions given by Shiroi GCI to the
F-94 and the actual relative position at which they accomplished radar contact; GCI said 4 miles range at the
aircraft’s 11 o’clock position, and they actually got radar contact with the moving target at a 6000-yard range, 10
degrees to their port. Nearly exact agreement, and thus incontrovertibly demonstrating that ground-radar and
airborne radar were then looking at the same moving unknown target, despite the contrary suggestions made in the
Condon Report. Had the Condon Report presented all of the information in the case-file, it would have been difficult
to maintain the curious position that is maintained all of the way to the final conclusion about these radar events in
the Condon Report’s treatment of the Haneda case.

That the moving target, as seen by both ground and airborne radar was a distinct target, though exhibiting radar
cross-section somewhat smaller than that typical of most aircraft, is spelled out in Malven’s IR-35-52 summary:
“Lt.B, GCI Controller at the Shiroi GCI site, has had considerable experience under all conditions and thoroughly
understands the capabilities of the CPS-1 radar. His statement was that the object was a bonafide moving target,
though somewhat weaker than that normally obtained from a single jet fighter.” And, with reference to the airborne
radar contact, the same report states; “Lt. R, F-94 radar operator, has had about seven years’ experience with
airborne radar equipment. He states that the object was a bonafide target, and that to his knowledge, there was
nothing within an area of 15–20 miles that could give the radar echo.” It is exceedingly difficult to follow the
Condon Report in viewing such targets as due to anomalous propagation.

Not only were there no visual sightings of the orbiting target as viewed from the F-94, but neither were there any
from the Shiroi site, though Lt. B specifically sent men out to watch as these events transpired. Also, as mentioned
earlier, it seems out of the question to equate any of the Haneda visual observations to the phase of the incident just
discussed. Had there been a bright light on the unknown object during the time it was in starboard orbit, the Haneda
observers would almost certainly have reported those movements. To be sure, the case-file is incomplete in not
indicating how closely the Haneda observers were kept in touch as the GCI directed radar-intercept was being
carried out. But at least it is clear that the Haneda tower controllers did not describe motions of the intensely bright
light that would fit the roughly circular starboard orbits of radius near four miles. Thus, we seem forced to conclude
either that the target the F-94 pursued was a different one from that observed at Haneda (likely interpretation), or
that it was non-luminous during that intercept (unlikely alternative, since Haneda observations did not have so large
a period of non-visibility of the source they had under observation 2330–0020).

Condon Report Critique of the Radar Sightings
The Bluebook case-file contains essentially no discussion of the radar events, no suggestion of explanations in
terms of any electronic or propagational anomalies. The case was simply put in the “Unexplained” category back in
1952 and has remained in that category since then at Bluebook.

By contrast, the Condon Report regards the above radar events as attributable to anomalous propagation. Four
reasons are offered (p. 126) in support of that conclusion:

1. The tendency for targets to disappear and reappear;
2. The tendency for the target to break up into smaller targets;
3. The apparent lack of correlation between the targets seen on the GCI and airborne radars;
4. The radar invisibility of the target when visibility was “exceptionally good.”

Each of these four points will now be considered.

First, the “tendency for the targets to disappear and reappear” was primarily a matter of the orbiting target’s
moving into and out of the ground-clutter pattern of the CPS-1, as is clearly shown in the map that constitutes
Enclosure #5 in the IR-35-52 report, which was at the disposal of the Colorado staff concerned with this case.
Ground returns from AP (anomalous propagation) may fade in and out as ducting intensities vary, but here we have
the case of a moving target disappearing into and emerging from ground clutter, while executing a roughly circular
orbit some 4 miles in radius. I believe it is safe to assert that nothing in the annals of anomalous propagation
matches such behavior. Nor could the Borden-Vickers hypothesis of “reflections” off moving waves on inversions
fit this situation, since such waves would not propagate in orbits, but would, at best, advance with the direction and
speed of the mean wind at the inversion. Furthermore, the indicated target speed in the final phases of the attempted
intercept was greater than that of the F-94, i.e., over 400 knots, far above wind speeds prevailing that night, so this
could not in any event be squared with the (highly doubtful) Borden-Vickers hypothesis that was advanced years
ago to account for the 1952 Washington National Airport UFO incidents.

Second, the breakup of the orbiting target into three separate targets cannot fairly be referred to as a “tendency
for the target to break up into smaller targets.” That breakup event occurred in just one definite instance, and the
GCI controller chose to vector the F-94 onto the strongest of the resultant three targets. And when the F-94 initiated
radar search in the specific area (11 o’clock at 4 miles) where that target was then moving, it immediately achieved
radar contact. For the Condon Report to gloss over such definite features of the report and merely allude to all of this
in language faintly suggestive of AP seems objectionable.

Third, to build a claim that there was “apparent lack of correlation between the targets seen on the GCI and
airborne radars” on the sole basis of the mismatch of times listed by Lt. B on the one hand and by the aircrew on the
other hand, to ignore the specific statement by the intelligence officer filing IR-35-52 about this being a
typographical error on the part of Lt. B, and, above all, to ignore the obviously close correspondence between GCI
and airborne radar targeting that led to the successful radar-intercept, and finally to ignore Lt. B’s statement that the
F-94 “followed the target into our radar ground clutter,” all amount to a highly slanted assessment of case details,
details not openly set out for the reader of the Condon Report to evaluate for himself. I believe that all of the
material I have here extracted from the Haneda case file fully contradicts the third of the Condon Report four
reasons for attributing the radar events to AP. I would suggest that it is precisely the impressive correlation between
GCI and F-94 radar targeting on this non-visible, fast-moving object that constitutes the most important feature of
the whole case.

Fourth, it is suggested that AP is somehow suspected because of “the radar invisibility of the target when
visibility was ‘exceptionally good.’” This is simply unclear. The exceptional visibility of the atmosphere that night
is not physically related to “radar invisibility” in any way, and I suspect this was intended to read “the invisibility of
the radar target when visibility was exceptionally good.” As cited above, neither the Shiroi crew nor the F-94 crew
ever saw any visible object to match their respective radar targets. Under some circumstances, such a situation
would indeed be diagnostic of AP. But not here, where the radar target is moving at high speed around an orbit
many miles in diameter, occasionally hovering motionless (see Malven’s account cited earlier), and changing speed
from 100–150 knots up to 250–300 knots, and finally accelerating to well above an F-94’s 375-knot speed.

Thus, all four of the arguments offered in the Condon Report to support its claim that the Haneda radar events
were due to anomalous propagation must be rejected. Those arguments seem to me to be built up by a highly
selective extraction of details from the Bluebook case-file, by ignoring the limits of the kind of effects one can
expect from AP, and by using wording that so distorts key events in the incident as to give a vague impression
where the facts of the case are really quite specific.

It has, of course, taken more space to clarify this Haneda case than the case is given in the Condon Report itself.
Unfortunately, this would also prove true of the clarification of some fifteen to twenty other UFO cases whose
“explanation” in the Condon Report contains, in my opinion, equally objectionable features, equally casual glossing over
of physical principles, of important quantitative points. Equally serious omissions of basic case information
mark many of those case discussions in the Condon Report. Here I have used Haneda only as an illustration of those
points; but I stress that it is by no means unique. The Condon Report confronted a disappointingly small sample of
the old “classic” cases, the long-puzzling cases that have kept the UFO question alive over the years, and those few
that it did confront it explained away by argumentation as unconvincing as that which disposes of the Haneda AFB
events in terms of diffraction of Capella and anomalous propagation. Scientifically weak argumentation is found in a
large fraction of the case analyses of the Condon Report, and stands as the principal reason why its conclusions
ought to be rejected.

Here are some other examples of UFO cases considered explained in the Condon Report for which I would take
strong exception to the argumentation presented and would regard as both unexplained and of strong scientific
interest (page numbers in Condon Report are indicated):

  • Flagstaff, Ariz., 5/20/50 (p. 245)

  • Washington, D. C., 7/19/52 (p. 153)

  • Bellefontaine, O., 8/1/52 (p. 161)

  • Gulf of Mexico, 12/6/52 (p. 148)

  • Odessa, Wash., 12/10/52 (p. 140)

  • Continental Divide, N.M., 1/26/53 (p. 143)

  • Seven Isles, Quebec, 6/29/54 (p. 139)

  • Niagara Falls, N.Y., 7/25/57 (p. 145)

  • Kirtland AFB, N.M., 11/4/57 (p. 141)

  • Gulf of Mexico, 11/5/57 (p. 165)

  • Peru, 12/30/66 (p. 280
    Holloman AFB, 3/2/67 (p. 150)

  • Kincheloe AFB, 9/11/67 (p. 164)

  • Vandenberg AFB, 10/6/67 (p. 353)

Go Back