According to Velikovsky , the sun-earth-moon system has undergone two periods of extreme disturbance which have changed the parameters of the system by large amounts within the past few thousand years. In the first period, the earth had two near-collisions with Venus, the first being in the time of the Hebrew exodus from Egypt and the second being on the day during Joshua's campaigns when the sun and moon stood still; see the discussion of the “record” of the eclipse of −1130 September 30 in Section 11.2.
These near-collisions were about 50 years apart, and they took place sometime around −1400, according to Velikovsky. After the second one, the sun-earth-moon system was left with a year of 360 days and a month of 30 days. Velikovsky [pp. 334ft] emphasizes that 30 days is not a mere approximation to 29.53 days, and that the month was almost exactly 30 days at that time. Venus was left in a highly elliptic orbit.
Velikovsky also emphasizes that a calendar year as long as 365 days was unknown until sometime after −700, and that all known calendars in the period from −1400 to −700 had only 360 days. This, as I understand what he writes, is his basis for saying that the astronomical year was only 360 days during this period. On his page 340, in reference to the Egyptian calendar, he explicitly writes: “In the eighth or seventh century (before the common era) 1 the five epagomena days were added to the year under conditions which caused them to be regarded as unpropitious”. By the five epagomena days (usually called the epagomenal days) he means the days that were added to a period containing 12 months of 30 days each (see APO, Section 11.3), thus bringing the Egyptian calendar year to 365 days. Parker  explicitly contradicts Velikovsky's statement. According to Parker, the Egyptians introduced a year of 365 days early in the third millenium before the common era. This is a time between −3000 and −2500. Yet according to Velikovsky [p. 333] the year at this time was even shorter than 360 days.
During the second period, the earth had two or more near-collisions with Mars, and Velikovsky [p. 218 and p. 241] gives exact dates for the first and last of them. The first was on −746 February 26 and the last was on −686 March 23 2. Since −686 March 23, the parameters of the sun, moon, and planets have been undisturbed. For at least part of the time between −746 and −686, the month had 36 days while the year continued to have 360 days, according to Velikovsky; thus he explains the fact [Velikovsky, pp. 347ff] that the oldest known Roman calendar (see Appendix I) had only 10 named months.
At some time during the series of near-collisions with Mars, the earth's spin axis shifted with respect to the crust [Velikovsky, pp. 316ff] by 20− or so. The longitude of the older North Pole was about 90-W in terms of present maps. That is, the North Pole was once near the west end of Baffin Island. Also during this time, Venus and Mars had one or more near-collisions with each other. These changed the eccentricity of Venus's orbit from a large value to its present small value of only about 0.007. At the last of these near-collisions, Mars must have had an orbit that came close to the present orbit of Venus and that also extended as far out as the present distance of Mars from the sun. I did not find any place in which Velikovsky explained how the orbit of Mars assumed its present shape.
Data already presented in this chapter allow us to prove rigorously that Velikovsky's conclusions are wrong. Before I present this proof, however, I wish to discuss some aspects of Velikovsky's work and the nature of his arguments. Let us start with his “demonstration” that the last collision with Venus occurred on the day that the sun and moon stood still so that Joshua could finish his battle.
As Velikovsky [p. 58] points out, the Biblical story of Joshua recounts a shower of large stones (Joshua 10:11) that occurred just before the day when the sun and moon stood still, and the stones killed more of Israel's enemies than Israel's soldiers killed in combat. Velikovsky also quotes from a non-scriptural poem which says that earthquakes and whirlwinds occurred at the same time, and he writes [p. 59]: “A torrent of large stones coming from the sky, an earthquake, a whirlwind, a disturbance in the movement of the earth — these four phenomena belong together 3. It appears that a large comet must have passed very near to our planet and disrupted its movement; a part of the stones dispersed in the neck and tail of the comet smote the surface of our earth a shattering blow.” He later develops the thesis that the “comet” was the planet Venus, and hence that comets are massive dense bodies and not light tenuous ones.
This shows vividly the danger in taking a passage out of context, or of considering only one passage without examining the implications of other similar passages. For example, as we saw in Section II.3, in a year near +30, there was darkness over the earth for three hours, the curtain of the temple (in Jerusalem) was rent in two, the earth shook, and rocks were split and tombs torn open. In AAO [p. 78], I discussed a passage from the Song of Roland about what happened at the death of Roland. There was immeasurable rain and hail 4. There was an earthquaka so severe that “there was not a house whose walls did not burst” over a large area of France, and there was also a great darkness.
These passages provide almost exact parallels to the passage from Joshua. In particular, in all three passages we have an event that is impossible according to orthodox astronomy. The main difference is that the event is the stopping of the sun and moon in Joshua while it is an impossible eclipse in the other passages 5. Thus, if we accept Velikovsky's argument that there must have been a near-collision of the earth with a massive body in the time of Joshua, there must have been other near-collisions around +30 and around 780, when Roland died. However, we have ample evidence that no such thing happened in or near 30 and 780. Therefore there is no basis for assuming such an event in the time of Joshua.
Velikovsky frequently misunderstands the nature of the evidence that he cites in support of his conclusions, and I shall give two examples. On pages 251ff of his work, he argues that the Trojan War took place sometime around −800 or −750, and part of his argument reads as follows: “The tradition about Aeneas who, saved when Troy was captured, went to Carthage (a city built in the ninth century) 6 and from there to Italy, where he founded Rome (a city first built in the middle of the eighth century) 6, implies that Troy was destroyed in the eighth or late in the ninth century”.
According to ancient Roman tradition, Rome was founded in the middle of the eighth century before the common era, and −752 is the year most commonly given. We may join Velikovsky in assuming that this year is reasonably accurate. However, the founding of the city did not mean the first building there. Or the contrary, according to Italian practice of the time, it meant the formal organization of a well-inhabited area into the legal entity known as a city. This Implies, among other things, that there must have been a fairly large population already living there, and hence that Rome had probably first been built at an earlier period.
The second example concerns the date −746 February 26 that Velikovsky takes as the date of the first near-collision between the earth and Mars. Velikovsky [p. 218] introduces his discussion of this date by writing: “In −747 7 a new calendar was Introduced in the Middle East, and that year is known as ‘the beginning of the era of Nabonassar’. It is asserted that some astronomical event gave birth to this new calendar, …” Velikovsky does not say who made this assertion.
The date in question is indeed the era of Nabonassar, which is discussed in connection with the Egyptian calendar in APO [Section 11.3]. However, there is no evidence that it ever formed the basis of a new calendar, there is no evidence that it was ever used by anyone until it was introduced into astronomy about 900 years later, and there is no evidence known to me that it was chosen because of some singular event. Ptolemy needed an epoch from which he could count years and days, and there is no reason to believe that he chose the era of Nabonassar on any basis other than convenience. The most probable explanation is that the oldest astronomical records known to him were made in the reign of Nabonassar. If so, it would be natural to take the beginning of Nabonassar's first year as the basic epoch for astronomical calculations.
In sum, there is no reason to assume that the epoch of Nabonassar was chosen because of a catastrophic event. It was merely the first “New Year's Day” after the accession of Nabonassar.
Velikovsky also fails to take certain physical facts into account. For example, as we have already noted, Velikovsky [pp. 316ff] concludes that the North Pole was once near the point on the earth's crust that is now at latitude 70-N and longitude 90-W. As a result of a near-collision with Mars around the year −700, the axis suddenly shifted to its present position.
At the present time, the earth is almost exactly in hydrostatic equilibrium under the joint influence of its self-gravitation and its rotation at its present rate and about its present axis. If the earth has come to equilibrium in 2600 years, as Velikovsky's conclusions demand, we can safely assume that it was in equilibrium before −700 as well. Let us see what would happen to an earth in equilibrium if its axis suddenly shifted, ignoring the problem of finding a force system that could cause the shift.
One consequence is that Siberia shifted from a temperate part of the earth to a frigid one. In regard to this point, Velikovsky [p. 330] writes: “The sudden extermination of mammoths was caused by a catastrophe and probably resulted from asphyxiation or electrocution. The immediately subsequent movement of the Siberian continent (sic) into the polar region is probably responsible for the preservation of the corpses”. Later on the same page he concludes that they died from a “lack of oxygen caused by fires raging high in the atmosphere”. He does not explain why the lack of oxygen did not affect other areas. If it killed mammoths in Siberia, it should have devastated animal and human life over the entire earth, but Velikovsky implicitly assumes that this did not happen. Velikovsky does not explain what there is in the upper atmosphere that could support a raging fire.
If this shift of the axis occurred, I agree with Velikovsky that the mammoths would have died from asphyxiation, but not for the reason he gives. Siberia and China would have increased their latitude by 20− as a result of the shift, and this shift in latitude changes the equilibrium radius of the earth. The equilibrium radius at the new latitude is less than at the old latitude by about 7 kilometers for any point in Siberia or China. In other words, Siberia and China suddenly found themselves, presumably in the period of about a day, at an altitude of at least 7 kilometers.
The rigid massive crust and mantle could not adjust themselves to this change as rapidly as the light fluid atmosphere. Hence all the inhabitants of Siberia and China suddenly found themselves in the atmospheric conditions appropriate to an altitude of 7000 meters. At such an extreme altitude, it is doubtful that many of the Chinese people would have survived. However, there is no evidence in the records that there was any large discontinuity in the Chinese population or in its conditions of living.
Further, the ocean would also adjust more rapidly than the earth. Hence all the islands just east of mainland Asia would have become joined to the mainland by dry land. On the other hand, the temperate parts of North America would have found themselves at a depth of 7000 meters below sea level and hence they would have been immediately inundated by the oceans. Even the peaks of the Rocky Mountains would have been under water. We may safely say that no such thing happened.
From these and many other examples, we see that Velikovsky has not produced any important evidence to support his conclusions. From one point of view, this should be sufficient; speculation that is not supported by any significant evidence has no place in the scholarly literature. However, I prefer not to rely upon the negative proposition that Velikovsky has failed to support his conclusions. Instead, I wish to present positive proof that a major part of his conclusions is wrong. We can do this by using one of the sources that Velikovsky himself relies upon.
I mentioned above that Velikovsky takes −686 March 23 8 as the exact date of the last collision between the earth and Mars, and that the sun-earth-moon system has been undisturbed since then. He takes this date from the Annals of Lu [Confucius, −479], using the following passage [p. 241] to do so: “The year 687 B.C., in the summer, in the fourth moon, in the day gin mao (23rd of March) during the night, the fixed stars did not appear, though the night was clear [cloudless]. In the middle of the night stars fell like a rain”. Velikovsky says that this translation is taken from a work by Edouard Biot that I have not consulted. The parenthesis, the brackets, and the Italics all appear in Velikovsky's text; I do not know whether they appear in Blot's text or not.
The term sin mao designates a day whose cycle number (Appendix II) is 28. The cycle number for −686 March 23 is indeed 28. If the year should be changed to −685, the day would have to be changed to March 18, whose cycle number in −685 was also 28.
Velikovsky concludes that this meteor shower was occasioned by the earth's near-collision with Mars, and hence he is able to assign an exact date to the event.
As it happens, the Annals of Lu contains many records of astronomical importance besides the record of the meteor shower. In particular, it contains the records of 34 solar eclipses that are listed in Table V.2; these are the records up to the one designated −480 April 19 C. Of these records, as I have noted in the table, the reading of the date that I have listed as −644 August 28 is questionable; this leaves 33 records. Calculation with provisional values of the accelerations shows that all the eclipses took place in accordance with the records with only one exception. The eclipse of −552 August 31 was not visible in China, so that the record must be the result of a scribal error of some sort, unless it represents an early attempt to predict an eclipse. This leaves us with 32 valid records.
Twenty-nine of these eclipses occurred after −686 March 23. Since Velikovsky concludes that there have been no significant perturbations since that date, these eclipses have no direct bearing upon his conclusions. However, the agreement of the calculations and the records for these 29 eclipses shows that the chronology of the Annals of Lu is correct for years after −686. Further, the internal chronology of the annals is certain. Thus, if the chronology is correct after −686, it must also be correct for years before −686.
Table V.2 shows us that 3 eclipses took place before −686, and the preceding paragraph shows us that the recorded dates of these eclipses are correct. Further, calculation with the provisional accelerations shows agreement with tho records. The record of −708 July 17 is particularly valuable because the record asserts that the eclipse was central, and calculation shows that the eclipse must have been quite large. I estimate that the agreement for −708 July 17 could not occur if the calculated time of the eclipse were in error by more than about an hour.
Between −708 July 17 and −686 March 23 there are about 265 months. During this interval, the accumulated error in the ephemeris of the moon 9 cannot amount to more than about an hour, or 3600 seconds. Thus the length of the month cannot have changed by more than about 14 seconds in −686. Yet Velikovsky concludes that it changed by more than 6 days, from 36 days to 29½ days. The change that he claims is not consistent with any of the records before −686.
Thus we have proved, by using a source that Velikovsky himself relies upon, that the “catastrophe” of −686 did not occur. Further, it proves that there could have been no such catastrophe as late as −719 February 22, the date of the earliest eclipse in Table V.2. Still further, Velikovsky's arguments connecting the “catastrophe” of −746 with that of −686, and in fact with the entire set of “near-collisions” with Mars, are tightly connected, and if we have proved that the near-collision of −686 did not occur, we have proved that the entire set did not occur.
I do not see any way to extend the argument based upon Table V.2 to the alleged near-collisions with Venus around −1400. However, since the same kinds of argument are used for both the Venus and Mars events, the case for the near-collisions with Venus is considerably weakened from its already frail condition.