It is often quoted that most of the solar neutrinos passing through earth do just that – pass through – without interacting.
What's generally not said is what fraction of them do interact, and I've just noticed I never questioned this.
Wikipedia tells me that the characteristic electroweak cross-section $\sigma$ of electron neutrinos is 3.2 nanobarn. From this I would have estimated the differential loss fraction as $$ \frac{1}n \frac{\partial n}{\partial s} = \frac{\sigma\times\sum\text{atomic number}}{(\text{lattice constant})^3} \equiv \frac{\sigma\cdot k}{a^3} $$ Well, if I put in some rough numbers corresponding to $\mathrm{SiO_2}$, WolframAlpha gives me a characteristic length of $10^7\ \mathrm{m}$.
Not that much, is it? It's certainly not compatible with the statement that the vast majority of neutrinos pass through earth unhindered.
What did I do wrong in the calculation? What fraction of neutrinos actually interacts with earth, or what's the interaction half-life of a neutrino passing through rock?