this is an interesting question. we don`t know what materials u used to get the particles. i am a phd student , my research is about mass- weight dispersity . as polymers never be just one pure weight. when u depart depends on particle size, there are two posibility : firstly , if ur material used here have great separation depends on wolecualr weight , than the small size particle mighht form by small-weight componants and big particles formed by large-weight componants. secondly , what is more interesting, if there is no obviously weight separation in your systerm while making the partcles, maybe in both small size particle and large particle have a distribution of  molecular weight. ie , in the small partcles , it may have both small weight molecualrs and large weight molecualrs . so as to large particle. in this situation , even u depart the particles by size . maybe in thermodynamics , in the small particles u obtain , u will find it forms some large paticles again, as restructuring happends . aslo in the large particles u obtain , u will also get some small particles again .

so here , i am more interesting in which way u depart the particles by size ?

Dear Carl Chen,

We have determined particle size distribution using light scattering, viscosity as well as magnetization measurements. In all the cases we have assumed log-normal distribution. Here ,one has to determine two parameters, geometrical mean and standard deviation from the measurements. We have found that agreement between theoretical and calculated from measurements plots depends on truncation. For example, forward scattering method give better agreement at smaller size. Some details are discussed in our attached papers.

fitting to log-normal does not mean only truncation if I understand well. I think often you get larger and smaller particles as well, not present in dispersion or particle mix.

I agree.