We have the following probabilites:

Following the definition of conditional probability:

In this case, we want to calculate the conditional probability that a person has a daughter given that he/she already has a son. Then, the probability is:

therefore, the conditional probability that a person who has a son also has a daughter is 34%