> But the Mars effect doesn't simply depend on something as vague as
> "born in the fall."  Of course, any decent study would've used time
> of year of birth as an independent variable, along with position of
> Mars at time of birth;  I would imagine that Gauquelin did this.
> Can anybody out there tell us?

The Mars effect has been conjectured to be a consequence of 
a kind of interference effect of demographical and astronomic effects.

(1)  there is an annual rhythm in births; this annual rhythm 
     has changed over time, and used to be very strong two centuries
     ago.
(2)  around sunrise more children are born than at other times of the
     day; for first-borns there is an extra delay of a few hours.
(3)  as Mars moves very slowly in conjunction with the sun compared
     to opposition (because the distance in conjunction with the sun is
     5 times as large as in opposition, the apparent speed is slower
     too by the same factor), a Mars rise will disproportionately
     nearly conincide with a sun-rise.

All this was known to Gauquelin, and he corrected (mostly for (2) and
(3)) for it.  The infamous Zelen test tried to establish whether that
theoretical correction was computed right; as Gauquelin had already 
support from data of about 24000 non-sports people, this Zelen test
was more or less superfluous.  I'll spare you the details of the bickering
about it.

It has been conjectured by my friends that these birth rhythms (1-3)
might be different for future athletes; for people of a very healthy
constitution; more specifically, it has been conjectured that in 
summer births are early in the morning and in winter later in the morning,
at least for the population of athletes.  Such a fact would go in the
direction of at least giving a natural explanation for the Mars effect.
It would be strange though, because a natural birth is a drawn out 
process; it is not exactly known what triggers it.  But if summer
births would be early and winter births late, the trigger can't be very
well have anything to do with the onset of darkness.

Actually my friends gave up trying to pursue this explanation.
The reason is (among others) that the Mars effect is just one of
a family of claims of planetary effects; only the Mars effect is the
largest one, and the only that has been the subject of independent 
tests. 

They (and I) are thinking about ways in which a small error in 
data collecting methods could lead to this result.  People don't
realize that the computation of a person's Mars sector isn't that
easy.  One obstacle is that one has to convert the local time 
(as given by town hall records) into universal time. So knowledge
of time zones (also during the German occupation), summer time and so
on back to 1870 is necessary.  I don't know how carefully the geographical
location was taken into account. 

One source of bias could be that the researcher recalculated the Mars
sectors of famous champions when they happened to fall just outside
a favorite sector; if he was just a little bit less enthousiastic about
doing that work for famous athletes just inside a sector, a bias might
result.

Tracking down a subtle experimental bias can give just as much
insight as finding a remarkable birth rhythm pattern.  There are
zillions of scientists who could use a clear example of what to
avoid in applying statistics. The birth rhythm pattern of pre-WWII
born athletes on the contrary is about as interesting as the mating
ritual of the dodo.

JWN