From calib/partial0.fit and calib/partial1.fit, I estimate the radius
of the sun to be 402 pixels. 

2.5*402 = 1005
SQRT((1005**2)/2) ~ 711

From images 0, 3, and 6 (1s exposures through 3875, 4050 and 4410
filters respectively), I estimate the center of the sun to be located
at -37,17 in image pixel coordinates. So 674,728 is a point NE of the
sun at 2.5Rsun. I've plotted cross sections through one of the three
second images from each polarizer to give a feel for how much data
we have. 

I've made bias-subtracted copies of all the polarization images (they
are in the redbias/ directory), averaged the 3s exposures through each
polarizer, then calculated the polarization angle using the imexpr
IRAF task and the expression

atan( ( (2*b) - a - c )/( a - c ) ) / 2 

where a is the 0d, b the 45d, and c the 90d averaged 3s image.

[from Billings, D.: A Guide to the Solar Corona, New york, 1966.]