Sunday, April 20, 2014

The primary cilium cannot sense the moon's gravity (sorry, you'll have to read the intro to learn more)

A physiologist interested in the effects of gravity on biological process, such as I, must understand how living organisms transduce those effects. In my favorite organism, the human body, there are many avenues for such transduction. At the scale of a meter, the weight of the limbs on the joints is readily sensed in even the most casual of circumstances. The weight of the extended column of bodily fluids as it distends tissues and organs is only slightly subtler, but dramatically noticeable when one hangs upside down by the ankles. At the scale of centimeters, there are the organs of balance, which are collections of cells forming the vestibular system that is specifically sensitized to both static gravity and induced acceleration (which Einstein told us is indistinguishable from gravity). 

But some biologists are interested in the effects of gravity at the cellular level, and ask whether a single cell, with a dimension measured in millionths of a meter, can sense gravity independently. There are structures within the cell that might be sensitive enough, but in a normal living cell, the gravitational stimuli would be overwhelmed by non-gravitational stimuli such as thermal agitation.  Determining gravitational sensitivity in isolated single cells requires carefully controlled, well-designed experiments. Unfortunately, some biologists interested in this topic don’t understand the physical phenomena involved (just as some physicists don’t understand biological systems).

In 2009, I read an article in the journal Developmental Dynamics that claimed to demonstrate the gravitational sensitivity of single nerve cells from the developing spinal cord of the zebrafish. I cannot challenge the authors’ knowledge of neural cell physiology, but their interpretation of the physical phenomena involved was in error. Some colleagues (Maneesh Arya and Susan Steinberg) and I drafted a letter to the editor of that journal explaining the errors, but the journal does not print letters to the editor (1).  The editor recommended a full article, which we were not able to undertake.

So, there it ended, with a flawed experiment unchallenged in the published literature. In the interest of making this discussion slightly more public and discoverable on the Internet, I decided to put it here in my blog:

This is all you need to know about the primary cilium for this article.
To the Editor, Developmental Dynamics

I have pondered the article, “The primary cilium as a gravitational force transducer and a regulator of transcriptional noise” (Moorman and Shorr, Dev. Dyn. 2008, vol. 237, pp. 1955-9), for well over a year now, troubled that my understanding of several points of gravity and its influence did not seem to match that of the authors, as detailed on pages 1957-8.  

The authors seem to report that isolated cells in culture can detect changes in ambient earth-surface gravity environment due to the passage of the moon and the sun overhead.  They reported a clever and imaginative comparison of neurogenin-3.1:gfp (ngn3.1:gfp) expression in the Rohon-Beard neurons of the developing zebrafish spinal cord at the times of the local high and low tide as evidence of an influence of the moon’s gravity on ngn3.1:gfp expression, presumably mediated through the primary cilium (2). Their report, while recent, already seems to be becoming accepted by others who cite their findings without further question (3, 4).

Unfortunately, their assertions regarding lunar and solar gravity at earth’s surface are flawed (5).

The authors are correct that the effective gravity at earth’s surface is not quite constant.  In addition to extremely small regional and latitudinal variations, there is indeed the influence of the gravities of the sun and the moon.  

However, the authors mischaracterized the dominance of the moon in generating earth’s oceanic tides as being due to its gravitational pull on objects at earth’s surface. Instead, the tides are actually due to the greater gradient (e.g., decrease with distance) of the Moon’s gravitational force across the width of the earth than that of the sun.  

The tides in the Raritan River near the authors’ laboratory are not reliable indicators of the moon’s gravitational influence in New Brunswick, NJ.  High tides recur at 12.42-hr intervals, first when the moon is above the horizon, and its gravitational influence opposes earth’s downward pull—and again 12.42 hr later when the moon is on the other side of the earth, and its influence would augment earth’s downward pull.  Thus, any randomly selected high tide might be associated with either lunar gravity adding to or subtracting from earth’s surface gravity environment. Low tides occur at the halfway points between successive high tides, so the moon’s gravity influence should be at some intermediate value, neither greatest nor least.

In addition, local high and low tides may be offset from the overhead passage of the moon by many hours.  For example, in New Brunswick on Aug. 15, 2010, moon transit—when the moon was highest overhead and its gravitational influence should have been greatest—was at 17:55 EDT (6) while local high tide was 7.9 hr later (7).

A more appropriate indicator of lowest effective gravity level might be the time of the moon’s transit overhead, when its gravity acts opposite to earth’s; highest effective gravity level might then come 12.42 hours later when the moon is on the opposite side of the earth and its gravity sums with earth’s, which may be calculated as 9.81 m•s-2, defined here as 1.0 G.  In this highly-simplified example, the variation in effective gravity due to the moon’s influence would 0.000034 m•s-2, approximately 0.0000035 G (3.5 micro G), over a period of 12.42 hr, due to the earth’s apparent daily rotation under the moon.  Note that even in this simplified, optimized case, the Rohon-Beard neuron would be required to detect a variation in earth’s surface gravity of one part in 300,000 occurring gradually over more than 12 hours!

The sun’s influence on earth’s surface gravity environment is much greater.  In the simplified case of the Rohon-Beard neuron on earth’s surface directly on a line between earth’s center and the sun, the “weight” sensed by the primary cilium would be greatest when it is on the side of the earth directly opposite from the sun (e.g., at local “midnight”), and thus subject to their summed gravitational pulls, and it would be least when it is directly between the earth and the sun (e.g., “noon”).  At midnight, it would be subject to the sum of earth’s surface gravity plus the sun’s effective gravity at the distance of the earth from the sun, 0.0059 m•s-2, or approximately 0.0006 G (600 micro G); the sum is 1.0006 G.  At noon, it would be subject to the difference, or 0.9994 G.  In our simplified example, the neuron would be subject to this solar variation in effective surface gravity of 0.06% on either side of the average of 1.0 G repeatedly at 24-hr intervals.

Note that the moon’s gravitational influence would be 1/176th of the sun’s, and periodically in phase and out of phase due to its monthly motion relative to the earth.

Thus, Moorman’s and Shorr’s reliance on local tidal occurrence as indicators of greater or lesser lunar gravity at earth’s surface for comparisons of gene expression at high tide and low tide is flawed in at least four respects: 
  1. being timed to the influence of the far weaker of the two supposed gravitational influences (e.g., the moon instead of the sun)
  2. overestimating the supposed lesser lunar influence at low tide which would actually be of an intermediate value
  3. assuming that any particular high tide reflected the greatest lunar influence when it might very well have been the weakest, depending on whether the moon was above the horizon or below at the time of high tide, and 
  4. neglecting a multi-hour lag between the passage of the moon over the laboratory and the occurrence of high tide.  

The entire foregoing discussion assumes that the primary cilium of the Rohon-Beard neuron could even detect the gravitational influence of the moon or the sun.  However, the estimated variation of less than 6 percent of one percent is much less than other ambient influences acting upon the neuron, including thermal, vibrational, atmospheric pressure and others.  A 70-kg laboratory technician approaching to within 4 cm of the cell culture before receding again while periodically tending the culture would provide almost as much gravitational influence as the moon!  Modern gravimeters are more than capable of such sensitivities, but they are much more massive than the Rohon-Beard neuron and may require magnetic suspension and liquid helium cooling to eliminate extraneous environmental signals (8).

Even if possible, reliable detection of such a small signal-to-noise ratio would probably require prolonged integration of the signal.  Melvill Jones and Young (9) have proposed that the gravity sensors of the mammalian vestibular system do not signal detection of acceleration until sufficient acceleration has occurred to produce a threshold velocity of approximately 0.20 m•s-1 (recalling that velocity is the integral of acceleration).  Arbitrarily assuming a comparable velocity threshold in the primary cilium of the Rohon-Beard neuron exposed to the moon’s gravitational attraction (0.000034 m•s-2) on earth’s surface, the neuron would need nearly 5,900 seconds (over 1.6 hr) for detection, assuming the force is provided completely and continuously, instead of increasing and then decreasing over a period of many hours.

The authors propose a clever test of their observation of the influence of extraneous gravitational variations by use of a technique they refer to as “gravity clamping.”  Unfortunately, this technique is not further described, but must be assumed to involve centrifugation, as there are no other techniques to increase gravity at earth’s surface by 10% for an extended period of time.  However, centrifugation at 1.1 G would merely increase the background level against which the supposed variation would occur by 10%, since there is no such thing as a “gravity shield” to exclude the presumed influences of the moon and the sun in the authors’ laboratory.  Rotation of the cells during centrifugation would have continually rotated their orientation with respect to the external environment, which might have been responsible for the reported effects if those effects could have been demonstrated convincingly.  An additional control, perhaps using a laboratory test-tube rocker to randomize cell orientations without the putative gravity-clamping confounder, might have produced the same result. 

In short, the authors’ assumptions about the presence of a gravitational perturbation for the primary cilium to transduce appear to be unsupportable and their measurements not well controlled.

Then there is the role of the primary cilium of the Rohon-Beard neuron as a gravitational force transducer.  The authors calculated the approximate shear force that would be applied to the cell’s primary cilium by the assumed cyclic changes in earth’s gravitational field.  They calculated the mass of a 10-micro-m diameter cell by multiplying its spherical volume, 523.6 micro-m^3, by the density of water, which they did not state but which I provide as 10-12 g•micro-m^-3 (i.e., 1 g•cm^-3).  The cell’s mass is thus 5.24x10^-14 g, or 5.24x10^-17 kg—one-millionth of the value calculated by Moorman and Shorr.  Thus, earth’s gravitational force on a single cell is 5.24x10^-17 kg  X  9.81 m•s-2, or 5.14x10^-16 N, a factor of 10 less than the sensitivity of the primary cilium reported by Resnick and Hopfer (10) and cited by the authors.

But even if the sensitivity was appropriate, any such stimulation of the primary cilium would require the capacity for its relative displacement with respect to the cell body—such as embedding the free end of the cilium in a large mass other than its own cell body, so that the presumed gravitational variations could physical displace either the cell body or the cilium but not both.  The authors describe no such mechanism, but it is the subject of at least one recent report (11). 

Finally, the authors find evidence of a gravitational effect on cells in culture through, not their altered rates of gene expression, but their increased variability in gene expression at local high tide, a time of supposedly reduced gravity on earth due to the moon’s pull.  However, they did not hypothesize that such variability would result, and gave no reason why it was a logical or meaningful result from such exposure.

For these reasons, Moorman’s and Shorr’s results cannot be accepted as supporting any gravitational force transduction role of the primary cilium of the Rohon-Beard neuron of the developing zebrafish spinal cord, and, by extension, a direct gravitational effect on any single cell that depends on a mechanism ascribed to the Rohon-Beard neuron.  That is not to say that such an effect may not exist, nor that it could not explain many of the observed responses to weightlessness of biological specimens ranging from isolated cells in culture to intact higher mammals.  But, there is ample justification for most—if not all—of the observed effects of hypo- and hypergravity on complex biological systems to be derived from the influence of gravity on multicellular organ systems at a macroscopic level.  

Extraordinary claims require extraordinary evidence; unfortunately, the evidence provided by Moorman and Shorr does not suffice.

Key words: primary cilium, Rohon-Beard, zebrafish, neurogenin-3.1:gfp (ngn3.1:gfp), gravitational transducer, Moorman, Shorr, Developmental Dynamics, high tide, low tide, lunar tide, solar tide, Raritan River, New Brunswick, Melvill-Jones, gravity clamp

  1. See
  2. Primary Cilia: the sensory organs of our cells?, accessed Apr. 20, 2014.
  3. Satir, P., Pedersen, L., and Christensen,S.  The primary cilium at a glance.  J. Cell Science 123, 499-503 (2010)
  4. Alaiwi, W., Lo, S., and Nauli, S.  Primary cilia: highly sophisticated biological sensors. (Review.)  Sensors 9, 7003-7020 (2009).
  5. I have relied on NASA’s Cosmicopia webpage (, accessed on August 14, 2010) for most of the geophysical background information used in this letter.
  6. Source: U.S. Naval Observatory,, accessed Aug. 14, 2010.
  7. Source: Mobile Geographics,, accessed Aug. 14, 2010.
  8. Source: Gravimeter,, accessed Aug. 17, 2010.
  9. Melvill Jones, G., and Young, L. R.  Subjective Detection of Vertical Acceleration: A Velocity-Dependent Response?  Acta Oto-Laryngologica 85(1): 45-53, 6 January 1978.
  10. Resnick, A., and Hopfer, U.  Force-response considerations in ciliary mechanosensation.  Biophys. J. 93: 1380-1390, 2007.
  11. Blanco, C., Drayer, I., Kim, H., and Wilson, R.  Mathematically modeling chondrocyte orientation and division in relation to primary cilium, published online at, August 6, 2010.

Saturday, February 15, 2014

Tales of Immersion 2: Two Pictures, No Answers (But Many Inferences)

In October 2013, I purchased a pair of photographs on eBay which appear to illustrate an early episode in the history of neutral buoyancy by water immersion: Figure 1, labeled 69442B, and Figure 2, labeled 69471B. They were listed on eBay under the title, “2 ASTRONAUT [sic] UNDER WATER GENERAL DYNAMICS (181234004788#).” The eBay seller provided no provenance or additional information. All I know is what was stamped and written on the backs of the photographs: the photo numbers and the notations “unclassified,” “General Dynamics,” and “Convair Division.” Since then, I have been trying to uncover the story behind these pictures.

Figure 1. General Dynamics, Convair Division, photo 69442B.
Figure 2. General Dynamics, Convair Division, photo 69471B.

Nowadays everyone knows that astronauts train underwater for their spacewalks, or extravehicular activities (EVA). It was even featured prominently in the movie “Armageddon” (IMDb, 1998), probably the closest thing to “correct” in the whole movie.

The Convair Division of General Dynamics was in San Diego, California. A quick query to Francis French at the San Diego Air & Space Museum, repository for many of the company’s documents and records, did not yield any insights. It became clear that if I was to learn more about this particular activity, it would come from an analysis of the photos themselves.

Luckily, this pair of images seems as if it were composed specifically to tutor me on a step-wise analysis, starting with the human figure, working outward to the setting and ending (I hope) with a date and a context for the activity in the pictures. The steps of the analysis practically suggested themselves:
  • Person(s)
    • Central person(s), typically the test subject or trainee
    • Supporting person(s)
  • Body-worn equipment
    • Garments
    • Real or simulated life support equipment
  • Tools
  • Activities
  • Test fixture
  • Setting
    • Water tank
    • Location
  • Time factors
    • Approximate calendar date
    • Temporal relationship of images
I already knew that GD was an early developer of neutral buoyancy techniques using water immersion for astronaut training and procedures development (Mattingly and Charles, 2013). The company was certainly no stranger to neutral buoyancy, but at a different scale than a single astronaut: it has been building submarines for the U.S. Navy since 1900 (General Dynamics, 1998)

GD came close to becoming the first contractor to provide such training to NASA’s Gemini astronauts. In July 1966, NASA astronaut Scott Carpenter visited GD’s underwater facility in San Diego and was impressed with their capabilities. He was familiar with Convair from previous visits, starting in 1959 to monitor the development and construction of the Atlas rockets which boosted the Mercury capsules into low Earth orbit (Voas, 1960).

Carpenter was on his way back to Houston intending to recommend that the company train the Gemini astronauts. But he was diverted[1] to Baltimore, Maryland, apparently with no more information than to visit a group of contractors at the McDonogh School in nearby Owings Mills. He arrived unannounced at the school’s pool where he witnessed the neutral buoyancy capabilities developed by Environmental Research Associates (ERA). What he saw at McDonogh changed his mind, and soon thereafter Gemini astronauts arrived at the McDonogh School for EVA training (Mattingly and Charles, 2013).

I don’t know what he had seen in San Diego, but whatever it was is important for an understanding of NASA’s decision to go with ERA.

Persons. There was one central person in both images. Figure 2 also included a second person, probably a safety diver or in-water observer: note the second person’s right hand and forearm in a wetsuit visible behind the central person’s right elbow, and his right leg visible through the circular opening to the left of the central person’s right leg.

Body-worn equipment. The central person was wearing a full-pressure suit, apparently an Arrowhead version of the B.F. Goodrich Mark 4 garment (Young, 2009,  pp. 16-23), judging from its distinctive features including its unique zipper configuration (Figure 3). The suits shown in Figures 1 and 2 may have been different suits based on some visible structural details, although differences in lighting may explain the variation. The garments were apparently pressurized with water, not air, as indicated by the absence of external ballast on the torso and limbs and by the scuba mask visible inside the closed helmet. It appears that the transparent helmet visor has been removed to accommodate the scuba mask, but the visor frame assembly appears intact.  The helmet was prominently (and conveniently) labeled “General Dynamics” and “Convair Division.”

Figure 3. Arrowhead variant of B.F. Goodrich Mark 4 suit. Identification based on specific features in both Figures 1 and 2 and in image of Mark 4 from National Air and Space Museum: red, lower left leg fitting; green, upper left leg fitting; blue, right thorax fitting; yellow, right forearm corrugation unique to Arrowhead variants (see also corrugated sections visible above and below knees, and below elbows in all three images; corrugations around waist not visible in General Dynamics suit, and above elbow only in Figure 1). (Photo credits: General Dynamics, left and center; National Air and Space Museum, right.)
Life support for the central figure appears to have been provided by a large diameter pressurized hose extending horizontally underwater to the central person from out of frame at the left. It seems to have connected to the upper back of the test subject, between the helmet and the middle of his back. A thinner black cable, possibly a telephone line, was bundled with the pressure hose. In Figure 2, the pressurized hose was secured to the test fixture, possibly to provide stress relief and facilitate mobility by the test subject.

The central person was also wearing what appears to be a white mockup of a Garrett AiResearch EVA Life Support System (ELSS) (Figure 4) on his chest in Figure 2 only. This identification is based on the location and size of the unit, and its distinctive top panel (Thomas and McMann, 2006, pp. 66-7).

Figure 4. Astronaut Edwin Aldrin in 30-foot vacuum chamber of the McDonnell Aircraft Co., St. Louis, Mo., on Aug. 15, 1966, wearing Garrett AiResearch EVA Life Support System (ELSS) and Ling-Temco-Vought Co. (LTV) Astronaut Maneuvering Unit (AMU) for U.S. Air Force experiment D012. ELSS is on his chest, and AMU is his large backpack. (Photo NASA S66-51073, 1966.) 
Tools. In both pictures, the central person appeared to have his left leg inside a loose strap attached to what might have been a simplified wooden mockup of the Astronaut Maneuvering Unit (AMU) (Figure 4). The AMU was developed by the Ling-Temco-Vought Co. for the U.S. Air Force experiment D012 (Shayler, 2004, p. 57). Its tentative identification derives from features suggesting over-the-shoulder thruster assemblies and side-mounted hand controllers, and from the curved cut-out at its top to accommodate the wearer’s helmet, all present in the AMU (Figure 5). However, its front-to-back depth and the detail of the over-the-shoulder and side-mounted features merely suggest the structure of the corresponding features of the AMU; perhaps the GD mockup represented only the aspect of the AMU which would have interfaced with the astronaut’s suit and body. The visible straps around the subject’s left leg differ between the photographs: in Figure 1, there was a section of black fabric, presumably Velcro, while in Figure 2 the corresponding area appears white.

Figure 5. General Dynamics’ possible mockup (left image) of LTV AMU (right image). Detail from Figure 1 shows mockup elements corresponding to features on AMU: semi-circular cut-out for wearer’s helmet (bracketed at top of both images); over-shoulder protuberances in mockup corresponding to reverse thrusters (middle two-headed arrow); lower protuberances on mockup correspond to hand-controller mounts on AMU (bottom two-headed arrow). Scale model at right illustrates astronaut position during AMU operations. (Photo credits: General Dynamics, ca. 1964, left image; John Charles at National Museum of U.S. Air Force, 2011, right image).
A thin curved pipe or tube, apparently metallic and connected to the top of the AMU mockup (seen more clearly in Figure 1), was used by the central person as a hand-hold, but was not present on the LTV AMU.

The presumed AMU is seen in both photographs to be in proximity to the large test fixture, orthogonally aligned in Figure 1 but slightly askew in Figure 2, indicating that may have been loosely affixed to the test fixture.

Activities. In Figure 2, the central person’s umbilical appeared to originate out of frame to the left, and was routed through the open top side of the cylindrical structure, or tunnel, before exiting through the open end of the tunnel and then connecting to the figure. Apparently the central person had translated out of the tunnel through its open end before arriving at his work site, suggesting egress from an airlock for the EVA.

The central person interacted with the faux AMU in an unlikely manner. His left leg, but not his right leg, was behind a loose strap attached to the AMU. This is true in both photographs, suggesting it was a common practice and therefore intentional. However, it is not consistent with astronaut interaction with the AMU in other testing or spaceflight settings. There is no sign of any mobility aids or restraints as were provided for in-flight AMU activities (see Figure 6). In fact, there are no hand holds or mobility aids visible anywhere on structure.

Figure 6. Astronaut Eugene Cernan in weightlessness training aboard KC-135 wearing ELSS and AMU. Note handholds and foot restraints as flown on Gemini-9, which were notably absent in Figures 1 and 2. (Photo credit: NASA, 1966.)
Test fixture. The dominant feature of both photographs was a wooden framework about 10 feet (3.0 meters) tall, about 6 feet (1.8 meters) wide at its widest point, and about 12 feet (3.7 meters) in length. It was apparently painted, and what would have been contiguous surfaces were substituted by a metal mesh. The largest portion was laterally dissimilar: a large half-cylinder was mated to a rectangular structure. There was a triangular brace in the top half of main cylinder. The smaller portion was the tubular tunnel, about 3 feet (1 meter) diameter and 8 feet (2.4 meters) long, attached to the lower center of main section. A wire mesh disc at the unattached end of the tunnel may have simulated a hatch cover, shown swung open toward the right (Figure 2). Both the hatch end of the tunnel and the larger section appeared to be weighted down by ballast in fabric bags about 6 inches across.

Given the aerospace nature of work done by GD and the astronautical appearance of the central person, the test fixture may have been a representation of a then-current spacecraft design. The pairing of a tunnel and a large broad-faced bulkhead suggest either the Gemini-B/Manned Orbiting Laboratory (MOL) program of the Air Force or NASA’s Apollo Lunar Module (LM) docked to its Command Module.

The Gemini-B/MOL interpretation is unlikely for several geometrical reasons: the tunnel appears straight, whereas the tunnel from the MOL to the Gemini-B heat shield was angled and bent; the tunnel terminated too close to the periphery of the presumed heat shield, and at its bottom instead of at its top; and the diameter of the large element appears greater than the Gemini heat shield diameter (Figure 7).

Figure 7. Gemini-B/MOL tunnel configuration in artist’s concept from McDonnell Aircraft Company in 1967 compared with configuration of General-Dynamics mockup in Figures 2. Large bulkhead (bracketed in upper image) in mockup has different features and size than Gemini-B heat shield (bracketed in lower image). Tunnels in mockup and concept (indicated by two-headed arrows) have different diameter and shape. (Photo credit: McDonnell Aircraft Co., ca. 1967.)
Figure 8. Two-headed arrows indicate points of correspondence supporting the tentative identification of the General Dynamics test fixture as a version of the Lunar Module (LM) TM-1 mockup configuration, showing ascent stage (upper portion) with newly-added triangular windows for standing (not seated) astronauts but retaining round forward hatch originally intended for docking with Command Module. TM-1 was current between March 1964 and January 1965. (Photo credit: Grumman, 1964.)
Instead, the large element had some features which may correspond to the structure of the ascent stage and crew compartment of an early version of the LM (Figure 8). Of prime significance is a recessed triangular frame with a wire mesh interior, which appears to represent the forward facing window in the left half of the LM crew cabin. The triangular windows are unique to the LM, so their inclusion is strong evidence that this is a LM mockup. This is further supported by the combination of the rectangular central portion, from the bottom of which the tunnel exits, and the recessed rounded portion in the right half of the picture, which represented the left half of the cylindrical crew cabin. The round opening at the bottom of the rectangular section could have been the circular hatch once planned for the LM, which by 1964 was simply a vestige of its discarded role as a docking port (Aviation Week & Space Technology, 1964); by January 1965 it had been replaced by the familiar square hatch. Together they resemble most of the LM crew cabin structure, minus the corresponding rounded portion on the photo-left of the rectangular structure, which would have represented the right half of the cylindrical crew cabin. These features correspond to the stage of LM evolution represented by the TM-1 mockup, which was current between March 1964 and January 1965 (Godwin, 2007, pp. 72-3).

The tubular tunnel terminating in the circular LM hatch might have simulated the tunnel of another spacecraft which would have been docked to the front port of the LM. However, there is no external structure around the tunnel, and as described above such a docking is inconsistent with the inferred stage of LM development. Therefore, the tunnel attached to it does not reflect a contemporary aspect of the Apollo LM or its counterpart the Command Module.

Perhaps it represents an airlock attached to the front port of the LM (Figure 9). This might have been part of a proposal to use the LM in an Earth-orbiting research or operations role. Such an application is not shown among the options for LM-derived vehicles advertised by Grumman (Grumman Aircraft Engineering Corp., 1970), but it may have been considered briefly nonetheless.

Figure 9. Author’s concept of an airlock or other tubular structure attached to LM forward hatch. The right-hand panel in this image illustrates a hypothetical explanation of the mockup shown in the left-hand panel, but no such spacecraft concept is known from the available literature. Or it might just be a product of my pareidolia. (Photo credits: left, General Dynamics; right, John Charles, 2014.) 
There was a proposal to use the LM for long-duration lunar surface habitation (NASA, 1967, p. 29) which incorporated an airlock to minimize the loss of cabin atmosphere during repeated EVAs (Figure 10). However, that was a 1967 concept for use in a gravitational environment (albeit 1/6-g), not a ca. 1964 concept for weightlessness, and that airlock was noticeably different from the structure in Figures 1 and 2.

Figure 10. Lunar Module Shelter concept with airlock (circled in red). LM Shelter was proposed to support multi-week lunar surface missions in the Apollo Applications Program. This is the closest match to the concept illustrated in Fig. 9. (Photo credit: NASA, 1967.)
In light of this uncertainty, it is possible that the test fixture in Figures 1 and 2 may simply represent a generic LM-derived spacecraft concept for the neutral buoyancy activities being undertaken, and its specific features were not intended to correspond to actual or planned spacecraft.

The test fixture was apparently repositioned in the interval between Figures 1 and 2. In Figure 1, the rectangular component appeared nearly aligned with the three vertical lights between two windows. In Figure 2, the rectangular component was more aligned with the middle of the observation window to the left of the three lights. The ballast bags appear not to have been rearranged, indicating that the fixture was moved but not emersed and resubmerged.

Setting. Both photographs were made in the same facility, a water tank about 10 feet deep. If the large element of the test fixture represents the Lunar Module ascent stage (Grumman Aircraft Engineering Corp., 1964), then the depth of water required to submerge it completely (as demonstrated in Figure 2) was 9.4 feet (2.9 meters). The scenes in both images were well-lighted for photography purposes: from above water, perhaps using natural sunlight, suggested by the refraction patterns of surface ripples on the floor in Figure 1, and also apparently from a subsurface light source, out of frame left, judging from the shadows in Figure 2. The facility had at least three large distinctive observation windows and arrangements for subsurface lighting, suggesting a public facility for underwater shows or exhibitions. Thus, the setting was presumably an existing underwater exhibition facility in the San Diego area, in proximity to Convair.

A Google search of public aquariums near San Diego produced two possible venues. The Aquarium of the Pacific[2] in Long Beach, California, was established in 1998, according to the response to an email query, and was thus not a candidate. The Birch Aquarium at the Scripps Institute of Oceanography[3] was opened in 1992, and was likewise not a candidate. However, it was preceded by the T. Wayland Vaughn Scripps Aquarium-Museum in La Jolla, California, which existed from 1951 to 1992. A photograph (Figure 11) from the 1967 annual report (Scripps Institute of Oceanography, 1967, p. 35) shows what may be an interior view of one of its distinctive windows; it appears similar to the windows in Figures 1 and 2. Thus the Vaughn Aquarium may have been the setting for the neutral buoyancy activities in the photographs. However, some questions remain to be answered, such as how a tank which was presumably a habitat for its marine occupants came to be emptied, cleaned and made available for GD’s use.

Figure 11. Fred Spiess (right), Director of Scripps Institute of Oceanography, speaking with Donald Wilkie (left), Curator of the Vaughn Aquarium-Museum ca 1966. Note similarity of observation windows to windows in Figures 1 and 2. (Photo credit: Scripps Institute of Oceanography, 1967.)
Time. GD’s use of neutral buoyancy techniques was limited to the early to mid 1960s based on other evidence (Mattingly and Charles, 2013). The Vaughn Aquarium was in existence from 1951 to 1992. The Mark IV-type full pressure suits were introduced into service by the U.S. Navy in the 1950s (Young, 2009, pp. 19, 22) and decommissioned suits were available for widespread ground testing by 1964 (Mattingly and Charles, 2013). The presence of the mockup ELSS in Figure 2 indicates a date between January 1964 when the ELSS design was finalized (Thomas and McMann, 2006 p. 66) and November 1966 when the end of the Gemini program also ended use of the ELSS and presumably its usefulness in neutral buoyancy simulations (a different system was already in development for the upcoming Apollo flights). The faux AMU placed the date between May 1964, when LTV won the contract to build three flight units for U.S. Air Force experiment D012 (Shayler, 2004, p. 57) and September 1966 (Hacker and Grimwood, 1977), when the AMU was definitively deleted from the Gemini flight program (although the Air Force remained interested in using it for some period of time thereafter (Wade, undated)). The LM configuration TM-1, with its round forward hatch, was current between March 1964 and January 1965 (Godwin, 2007,  pp. 72-3). Thus I estimate the approximate calendar date for the two images as no earlier than May 1964, no later than September 1966, and probably not later than early 1965.

The scenarios in the two photographs are superficially similar but contain differences indicating they are from two different sessions. Presumably 69442B preceded 69471B, and the differences between them represent improvements or intentional modifications in the test scenario. These differences have been discussed above and are summarized in Table 1:

Table 1. Similar and different elements of Figures 1 and 2.

Common to both Figure 1 and 2
69442B (Figure 1) only
69471B (Figure 2) only
Test subject
Support diver
Body worn equipment
Water-pressurized Arrowhead Mk. 4 suit
Forearm bellows section visible
Forearm bellows section absent or covered
Faux AMU
Mounted orthogonal to test fixture
Black Velcro strip
Mounted rotated leftward re: test fixture
White Velcro strip
Left leg behind strap
Test fixture
LM mockup
Airlock mockup
Ballast bags apparently not re-stacked between photographs
Near 3 vertically-aligned lights
Aligned with window to left of 3 vertically-aligned lights
Vaughn Scripps Aquarium
Lighted from above, possibly natural light
Lighted from left of scene
Time factors
Ca. 1964
Not same session as 69471B
Not same session as 69442B


The available evidence indicates that the spacecraft and tool mockups in the photographs were likely used between mid-to-late 1964 and early 1965, although they might have continued to be used after those dates for general studies in neutral buoyancy.

The similarities in arrangement of the central person, the test fixture, the background and the perspective of the photographs suggest that they were made within a single session, perhaps showing a sequence of activities. On closer inspection, it becomes apparent that there were differences in lighting, body-worn equipment, features of the tools, possibly the space suit and even the location of the test fixture (see Table 1). These differences indicate that they were made during separate sessions, perhaps even on different days.

It is of interest that, despite their differences, the two images show the central person in nearly the same position and orientation with respect to the test fixture. He was above the top of the faux AMU, with his left leg inserted behind the same strap. This suggests that the action was intentional and significant. However, this does not represent an effective interface with the backpack-style AMU. Assuming a leg strap was part of the AMU restraint system, it should not have been necessary to enter it from the top of the unit. In addition, by early 1966, when Gemini astronauts were training to use the AMU in flight, there was a significant set of fixtures associated with donning it: handholds, footholds, straps and hoses (Figure 6). Despite those aids, Gemini 9 astronaut Eugene Cernan, the first person who attempted the task in spaceflight, abandoned the task it because the required physical effort overwhelmed his suit cooling system (Evans, 2013).

Perhaps the simplistic approach to AMU donning shown in these images was an early stage in the progressive development of the more robust (but still inadequate) donning aids used on Gemini 9, but any role of GD in preparations for AMU flights, including donning studies, has yet to be uncovered. If, however, the tool I have identified as a faux AMU was not, in fact, a mock AMU, then this analysis is moot, and the device being simulated remains to be identified.

Besides the faux AMU, there is the ambiguous test fixture and the mission it supported. Its resemblance to an early design for the Grumman’s Lunar Module suggests that GD was evaluating extravehicular activities associated with the Apollo program or perhaps the follow-on Apollo Applications Program (AAP) (Wikipedia, 2014; Shayler, 2002), which would have utilized surplus Apollo vehicles for scientific missions in Earth orbit as well as on the Moon. Judging from the course of the life support umbilical, the test subject had translated from the open end of the tubular tunnel, presumably representing an airlock, to the faux AMU which was affixed to the lower left front panel of the LM mockup, below its left forward window. While the AMU or similar maneuvering units were considered for AAP use (Figure 12), I have not seen any plans to use them in conjunction with a LM-derived vehicle.

Figure 12. Illustration of AMU application during an AAP mission outfitting a Saturn IVB upper stage for future in-orbit use. (Photo credit: Douglas Missile and Space Systems, 1965.)
What does all this mean? A great amount of detail can been inferred, but how close have my inferences come to the reality of GD’s role in spaceflight neutral buoyancy techniques? Were these photos really made in 1964? Where? What were the test fixtures intended to simulate? Does 69442B predate 69471B? Why was the scene duplicated so closely on a second occasion, and why were the obvious changes introduced? Why have no other photos surfaced?

If in fact these images date from 1964, it was two years before Carpenter almost recommended GD for the NASA contract. During those two years they must have made as much progress as ERA. My colleagues Francis French and Alan Renga at the San Diego Air & Space Museum, continued searching their General Dynamics records and found a document dated July 1968 (Braxell and Thomson, 1968) which reported on a different type of underwater work by subjects wearing the Mark 4 suit as well as in a Gemini-type suit; however, it is silent on the neutral buoyancy work illustrated in Figures 1 and 2.

I have taken my forensic analysis of these two photos as far as I can, without new information. Additional questions include: where are the photos and reports from 1965 and 1966 which document the capabilities that impressed Carpenter? Who funded that work? How much longer did they pursue neutral buoyancy work after it became obvious that NASA would not support it?

Maybe I can learn more when I am in San Diego in May 2014 for the 85th Annual Scientific Meeting of the Aerospace Medical Association. I will visit the San Diego Air & Space Museum, and possibly the Scripps Institute of Oceanography. Perhaps some of the answers are awaiting me there.

[Edited for clarity, 17 Feb. 2014]

My thanks to Addie Eure, Marketing Coordinator, Birch Aquarium at Scripps, the anonymous Guest Support Specialist, Aquarium of the Pacific, and especially to Francis French and Alan Renga, San Diego Air & Space Museum.


Aviation Week & Space Technology. 1964. Aviation Week & Space Technology. February 10, 1964.

Braxell, R. R. and Thomson, W. G. July 1968. Selected Experiments in the Assembly of Structures in Space. Life Sciences Department, Convair Division. San Diego : General Dynamics, July 1968. Sponsored by Aero Propulsion Laboratories, Wright-Patterson Air Force Base, Ohio. Contract F33-615-67-C-1302, "Assembly and Maintenance of Lightweight Metallic Structures in Space".

Cunningham, Walter. 1977. The All-American Boys. New York : Macmillan Publishing Co., Inc., 1977. ISBN 0-02-529240-4.

Evans, Ben. 2013. Date With An Alligator: The Trials of Gemini-IXA. Space Safety Magazine. [Online] June 7, 2013. [Cited: Dec. 27, 2013.]

General Dynamics. 1998. General Dynamics Electric Boat. [Online] 1998. [Cited: Jan. 1, 2014.]

Godwin, Robert. 2007. The Lunar Exploration Scrapbook / A Pictorial History of Lunar Vehicles. Burlington : Apogee Books, 2007. 978-1-894959-69-8.

Grumman Aircraft Engineering Corp. 1970. Apollo Program: Grumman's "Lunar Module Derivatives for Future Space Missions," Circa 1970... Heritage Auctions. [Online] 1970. [Cited: Jan. 19, 2014.]

Grumman Aircraft Engineering Corp. 1964. Inboard Profile of the Lunar Excursion Module, 1964. National Archives, Online Public Access (OPA). [Online] March 10, 1964. [Cited: December 4, 2013.] Entry 55, Box 3. 2657372.

Hacker, Barton C. and Grimwood, James M. 1977. On the Shoulders of Titans. Washington : National Aeronautics and Space Administration, 1977. p. 371.

IMDb. 1998. Armageddon (1998) Trivia. [Online] 1998. [Cited: Jan. 24, 2014.]

Mattingly, G. Sam and Charles, John B. 2013. A personal history of underwater neutral buoyancy simulation. The Space Review. [Online] February 4, 2013. [Cited: December 4, 2013.]

NASA. 1967. Lunar Mission Planning Data Book. Lunar Missions Office, NASA Manned Spacecraft Center. Houston : NASA Manned Spacecraft Center, 1967. p. 142. Library, Center for Advanced Space Studies, Houston TX 77058. TL 789.8 .U6 M282 1967.

NASA. 2008. NASA - T-38A Talon Performance and Specifications. Dryden Flight Research Center. [Online] Sep. 22, 2008. [Cited: Feb. 15, 2014.]

Scripps Institute of Oceanography. 1967. Annual Report for the Year Ending June 30, 1967. Scripps Institute of Oceanography. La Jolla, Calif. : s.n., 1967. p. 35.

Shayler, David J. 2002. Apollo: The Lost and Forgotten Missions. 1st. Chichester : Springer, 2002. p. 344. ISBN 978-1-85233-575-5.

Shayler, David J. 2004. Walking in Space. Chichester : Praxis Publishing Ltd., 2004. 1-85233-710-9.

Thomas, Kenneth S. and McMann, Harold J. 2006. US Spacesuits. Chichester, UK : Praxis Publishing, Ltd., 2006.

Voas, Robert B. 1960. Project Mercury Astronaut Training Program. NASA Space Task Group, Langley Field, Va. 1960.

Wade, Mark. undated. 1966.11.18 - NASA plans to include the DOD's astronaut maneuvering unit "back pack" aboard AAP flights. Encyclopedia Astronautica. [Online] undated. [Cited: Dec. 16, 2013.]

Wikipedia. 2014. Apollo Applications Program. Wikipedia. [Online] Jan. 10, 2014. [Cited: Jan. 24, 2014.]

Young, Amanda. 2009. Spacesuits / The Smithsonian National Air and Space Museum Collection. 1st. Brooklyn : powerHouse Books, 2009. 978-1-57687-498-1.

[1] If Carpenter was piloting a NASA T-38, this diversion was not trivial. Google Earth gives the straight-line distance from San Diego to Baltimore as about 2,300 miles. The T-38 is usually refueled every 700 miles (Cunningham, 1977 pp. 70-75), so a direct flight from San Diego to Baltimore would have needed three intermediate stops. A NASA website (NASA, 2008) gives the range of the T-38 as about 1,140 miles, although practical considerations such as weather and safety margins required more stops. Or, he might have flown commercial.

[2] Aquarium of the Pacific, 100 Aquarium Way, Long Beach, CA 90802,; email response received 25 Oct. 2013.

[3] Birch Aquarium at Scripps, 2300 Expedition Way, Scripps Institute, La Jolla, San Diego, CA 92037,; email response received 2 Dec. 2013.