Data points and fitting functions

Niles

1. The problem statement, all variables and given/known data
I have made an experiment with a Hall-meter and I have got some data. I know have an expression on the form

[tex]
B(a) = \mu IN\frac{{a^2 }}{{2\left( {a^2 + z^2 } \right)^{\frac{3}{2}} }}
[/tex]

where z = 0.02 m and a is my only parameter varyring. I have plottet these data, and I want to fit a regression to these. But which regression should I choose? The dependency is on the form

[tex]
\frac{{a^2 }}{{\left( {a^2 + k^2 } \right)^{\frac{3}{2}} }}
[/tex]
where k is a constant. The data looks linear, but still - what regression should I choose here?