yuga, VarNa and colour

[B. Reusch]

At 22:29 +0100 5/17/97, Luis Gonzalez-Reimann wrote:
>
>The names of the dice throws, K.rta TretA, DvApara, and Kali, are associated
>with the descending numerical sequence 4-3-2-1 (like the Pythagorean
>tetraktys).  TretA comes from tri (3), and DvApara from dva (2).  They are
>obvious cognates of other IE names for the same numbers.  K.rta is the past
>passive participle of the root k.r, "to do" (Pisani has a different
>etymology for K.rta that connects it with 4, see Mayrhofer).

Mayrhofer says Pisani's reconstruction does not agree with known
phonological laws.
Yet doesn't some of the Skt. literature refer to the yuga sequence also in
a different way at times? I mean, instead of 4-3-2-1 for k.rta, etc. there
would be the first yuga (k.rta), the second (tretA), etc.  If I remember
correctly,  in one of your own previous posts there was a MBh quote that
inverted the colors of the second and third yuga. In this kind of account
of the yuga sequence, of course, the names tretA (etymologically connected
with 3) and dvApara (etymologically connected with 2) are in metathesis, so
to speak.
What's the logic behind one and the other accounts of the yuga sequence?

Regards,
Beatrice