As a part of an experiment in radioactivity, we placed an alpha particles emitting source (Am) and a sensor in a vacuum chamber, and measured the power (number of particles per time unit) for various angles of the sensor. Our goal is to fit a theoretical function to our measurements (power vs. angle) so we can integrate it and get the total power of the source. At first we've been told to fit a gaussian, but then we saw that it's peak is "cut". We found in some book that a fit to f(x)=e^-(x^4) will give better results, and that the reason to this have something to do with the fact that the source isn't a perfect source point, i.e. it has finite dimensions. I will be glad if someone could think of a reasonable explanation to this. Thanks a lot!