however, without this division of the temporal stream
　　   into  the  three  moments  of gone-to, etc.  In other
　　   words, a necessary  condition of our recognizing  the
　　   commencement  of going is our being in a position  to
　　   speak of a time where  going has ceased, a time where
　　   going is presently  taking  place, etc.  If we are to
　　   succeed in designating the commencement  of going, it
　　   must take place  in one of these  three moments--and,
　　   of course, the not-yet-gone-to  may  be exluded  from
　　   our  considerations   as  a  possible  locus  of  the
　　   commencement  of going, since by definition  no going
　　   may take place in it.  Thus the commencement of going
　　   must  take  place  either   in  the  gone-to   or  in
　　   present-being-

　　　　　　　　　　　　  p.296

　　   gone-to.  This is impossible.  however. since neither
　　   of.  these  moments  may be designated  prior  to the
　　   commencement of going.
　　　　  Candrakiirti's  commentary  on II:13  appears  to
　　   support   this   interpretation:  "If  Devadatta   is
　　   standing, having  stopped  here.  then  he  does  not
　　   commence  going.  Of him  prior  to the beginning  of
　　   going  there is no present-being-gone-to  having  its
　　   origin  in time.  nor is there a gone-to  where going
　　   should be begun.  Therefore from the non-existence of
　　   gone-to   and  present-being-gone-to,  there   is  no
　　   beginning of going."(20)
　　　　  A moment's  reflection  will show, however.  that
　　   this  interpretation  is not substantially  different
　　   from   the  "mathematical"   interpretation   of  the
　　   argument, particularly the second version, which made
　　   use of infinitesimal  increments of duration.  Indeed
　　   on this interpretation  the argument  seems  specious
　　   unless  we make  the additional  assumption  that its
　　   target  includes  a ''knife-edge''  picture  of time.
　　   Thus  if one  assumes  that  time  is continuous  and
　　   infinitely  divisible, then at the instant  (that is,
　　   time-point)  at which going actually commences, there
　　   is in fact no real  motion, since  this  is just  the
　　   dimensionless  dividing-line  between  the period  of
　　   rest and the period of motion. And no matter how many
　　   infinitesimal  increments  one adds to the period  of
　　   rest  after  it has  supposedly  terminated, the same
　　   situation  will prevail.  Moreover, as long as one is