The example of husky/non-husky and black/buff on the last page leads us neatly into situations where two independent genes come into play. As a second gene is introduced, our punnet diagram can have sixteen possibilities instead of just four.

This diagram shows a cross between two rats carrying husky and buff:

Obviously, any chance you get to simplify this kind of diagram should be taken. In the case of our surprise on the the last page, the sire was true-breeding self, QQ (so there's no q's), and the dams were huskies, qq (so there's no Q's), then our example simplifies to:

If you think you've got it, what would you expect from a cross between Forth (qqRr) and the buff, husky-carrier from Skye's litter (Qqrr)? If you are uncertain, start with the full grid at the top of the page. Work out which columns are needed for Forth (clue: they're the same as above as it's the same doe!). Now work out which rows are needed for the sire... (scroll down or click here)

So we would expect a quarter of black selves (carrying both), buff selves (carrying husky), black huskies and buff huskies.

So far, we have discussed the effects of genes either one at a time, or as a combination of several genes acting in this way. However, a lot of the colours in rats are brought about by a number of genes working together.

There are two main families of colours which are determined by the agouti gene, A. If the dominant agouti gene is present (AA or Aa), the agouti colours will result. The recessive characteristic is black so all the non-agouti colours are based on black (aa).

The first two tables shows the effect of how agouti and black are modified by the dilute gene, D, and the mink gene, M.

The next two tables show how colours can be further modified by the brown gene, B. When the recessive chocolate characteristic is present, the eight colours in the tables above become: