Thayer Watkins Silicon Valley & Tornado Alley USA 

Planetary Satellite Systems 
There are a number of examples of resonance phenomena in the solar system that suggest that the regularity and magnitude of the parameters of TitiusBode type of rules for satellite systems should be in terms of the periods of the orbits rather than the average radii of the orbits. For example, among the asteroid there are none having a period which is one third, one half or two fifths of the period of Jupiter. On the other hand there are 19 which have periods which are close to two thirds that of Jupiter. And there are the Trojan and Apollo asteroids which have the same period as that of Jupiter but move in orbits such that they along with the sun and Jupiter form equilateral triangles. Another resonance phenomenon is the Cassini gap in the rings of Saturn. The Cassini gap is a 1700 mile wide gap that corresponds to a region in which particle would have a period of revolution near to one half the period of Saturn's innermost moon, one third the period of the second moon or one fourth the period of the third moon.
The distances and the length of the years of the various satellites of Jupiter are as follows:
Discovery Order  (000s km) 
(Earth days) 


Amalthea  V  181  0.498 
Io  I  422  1.770 
Europa  II  671  3.554 
Ganymede  III  1071  7.166 
Callisto  IV  1883  16.754 
VI  11470  266.00  
VII  11740  276.67  
X  11850  253  
XII  21200  631  
XI  22560  692  
VIII  23500  735  
IX  23700  758 
There appears to be three groups of satellites.
This information indicates that the second and third groups are the results of the capture of asteroids sometime after the formation of the innermost satellites.
For the inner group if one hypothesizes a missing satellite between Amalthea and Io then the plot of the period of revolution versus the order number of the satellites is as shown below:
This suggests an exponential relation of the form
The plot of log_{10}(T) versus order number:
The regression line is
For Saturn the data are:
Discovery Order  (000s km) 
(Earth days) 


Janus  X  157  0.749 
Mimas  I  184  0.9424 
Enceladus  II  237  1.370 
Tethys  III  293  1.882 
Dione  IV  374  2.737 
Rhea  V  523  4.517 
Titan  VII  1213  15.945 
Hyperion  X  1470  21.276 
Iapetus  VIII  3536  79.331 
Phoebe  IX  12864  550.417 
The first observation is that compared with Jupiter's satellite system Saturn really has them packed in. The inner eight seem part of a system and they all have inclinations of their orbits near zero. Definitely the outer moon, Phoebe, is of a different nature. Its distance is of another order of magnitude and it has retrograde motion. Its angle of inclination is 30° with retrograde motion or 150° if its motion is not consider retrograde. The next to the outer moon, Iapetus, has an angle of inclination of 14.7° and its distance make is uncertain whether it is of the same family as the inner group.
First a plot of the basic data:
The data points for the inner six moons do approximate a straight line. The data for the outer two moons were made to fit by hypothesizing two missing moons in the system. This is of dubious justification but not totally devoid of validity in that the last moon in the set more or less fits the line without any further manipulation.
The regression equation for the first six moons is:
The slope of the regression line for Saturn's inner moons is about half of that for Jupiter's inner moons. For the inner moons of Saturn there appears to be two systems interleaved. If Janus, Enceladus, Tethys and Rhea (System I) are taken as one system and Mimas, Tethys and Titan (System II) taken as another then the data looks as follows. It is hypothesized that there is a missing fifth moon in System I and two missing moons (third and fourth) in System II.
The regression equations for the two systems are:
Again the magnitude of the slope is in the range of 0.25 to 0.30.
Nepture has only two moons and the largest of them has retrograde motion. There is not a system of satellite in the nature of what Jupiter, Saturn and Uranus have.
The graph of log_{10}(Period) where Period is the Period of Revolution relative to that of Earth versus the order number:
The regression equation is
The magnitude of the slope is on the same order as the 0.3 found for Jupiter's satellite system.
The estimates of the regression slope coefficients for comparison are:
System  Slope Estimate 

Solar  0.3493 
Jupiter  0.3026 
Saturn I  0.2883 
Saturn II  0.3076 
Uranus  0.2510 
The magnitudes are correlated with the mass of the primary bodies in the systems. The plot of the slope coefficients versus the logarithm of the mass of the primary is shown below. (The masses are in units of Earth masses.)
On the basis of the preceding analysis there seems to be a systematic relationship between the periods of planets/satellites revolving around a primary body. The relationship may arise from resonance phenomena that prevents material from achieving a stable orbit if its period is one half, one third or two fifths of the period of an outer body in the system. The resonance could also work in the other direction preventing an outer body forming in orbits with periods equal to two or three times the period of the inner body. Of course if material is prevented from achieving stable orbits in some range that material gets concentrated in orbits of other ranges. Suppose resonance precludes orbits with periods of two fifths or one half of the period of the outer body and so the material gets concentrated an orbit with a period of 0.45 of that of the outer body. The reciprocal of 0.45 is 2.22.
Consider a period/order relationship of the form
where n is the order number of the planet or satellite. If b=2.22 (the reciprocal of 0.45) then log_{10}(b)=0.3468, essentially the value found for the planets of the solar system. The value of b is marginally different for the Jovian planets' satellite systems and may be weakly a function of the mass of the primary body of the system.
Because of Kepler's Law, R^{3}=αT^{2}, the relationships concerning periods can be converted to relationships involving the radii of orbits; i.e.,
If b=1/0.45 then b^{2/3}=1.703.
(To be continued.)
HOME PAGE OF Thayer Watkins 