Abstract Quantitative description of connections between neurons are scarce in neuroanatomy and this lack severely restricts theoretical developments. An “anatomical transfer function” is proposed that provides a consistent and simple notation to express a variety of numerical relationships: convergence, divergence, net divergence, numbers of input and output neurons, total number of connections, and degree is possible to include statistical information, synaptic strenghts, and of course dynamics. A few examples from mammalian cerebellum are worked out to illustrate these concepts of connectivity.