I've found the complete question online.
The exercise ends with the following question: "How many genes resulting in short tails are represented among these four colonies?"
With the whole context of the exercise, it can be concluded that at least four different genes are represented among these four colonies for the same phenotype characteristic - short tail. If all of the colonies are true-breeding for the recessive allele conferring the short tail characteristic, it would be expected that independently of the colonies, if the mice were crossed between them, the expected result would always be short-tailed mice. Because this is not what is verified, it means that all of the mice were only true-breeding for the recessive allele of the gene conferring the short tail characteristic when bred within the same colony, and that from colony to colony the gene involved in such characteristic is different. Therefore, interbreeding colonies would result in long-tailed mice for true-breeding was not a guarantee.
