[fitsbits] How is NAXISn supposed to work?

[RPEHLM]

Hello folks,

I'm a simple sort of chap, so I have a simple question! <g>

Reading the IAUFWG v3 (2008 July 10) documentation for the Primary Data 
Array (page 14, section 3, sub 3.3.2).  I'm not understanding what is 
meant by:

"Arrays of more than one dimension shall consist of a sequence such
that the index along axis 1 varies most rapidly, that along axis 2 next 
most rapidly, and
those along subsequent axes progressively less rapidly, with that along 
axis m, where
m is the value of NAXIS, varying least rapidly."

Does axis 1 equate to NAXIS1 and axis 2 to NAXIS2 etc.?  Should the 
value of NAXIS1 be greater than NAXIS2, which in turn should be greater 
than NAXIS3 etc.?
Or should the value of NAXIS1 be the smallest and NAXISn the greatest?  
Or something else entirely?  For example, take a 100 by 200 array, would 
the following be correct:

NAXIS = 2
NAXIS1 = 100
NAXIS2 = 200

or this:

NAXIS = 2
NAXIS1 = 200
NAXIS2 = 100

(Yes, I know the formatting is not exactly right <g>)

The documentation would seem to indicate the former, but I have seen 
plenty of commercial applications implementing the later.  Hence the 
confusion to my poor abused grey thing.

Regards
Robin
P.S. My first posting so please be gentle!!!