In Hardy and Wright, sixth edition we have theorem 28 which states that if $\dfrac{h}{k}$ and $\dfrac{h'}{k'}$ are two consecutive terms in a Farey sequence: $$kh'-hk'=1$$ and theorem 29 states that, if $\dfrac{h}{k}$, $\dfrac{h'}{k'}$ and $\dfrac{h''}{k''}$ are three consecutive terms in a Farey sequence: $$\dfrac{h''}{k''}=\dfrac{h+h'}{k+k'}$$ Later, they prove that those two theorems are equivalent. First they prove that theorem 28 implies theorem 29 as follows:
The thing I don't understand is why $kh''-hk''=1$ since theorem 28 only states that $kh'-hk'=1$. Also I'm not entirely sure about how they would have solved those equations to get to the part where they say: $$h''(kh'-hk')=h+h',\ k''(kh'-hk')=k+k'$$
You have the misstated the hypothesis of Theorem 29. The terms are in a different order than you write.
Theorem 28 is a statement about adjacent terms in the sequence. Under the hypothesis of Theorem 29, $h/k$ and $h''/k''$ are adjacent terms, and $h''/k''$ and $h'/k'$ are adjacent terms, and Theorem 28 can be applied to either pair.
Further information added after OP’s comment:
I’m not sure what you changed. Your question still says “if $\dfrac{h}{k}, \dfrac{h'}{k'},$ and $\dfrac{h''}{k''}$ are three consecutive terms” which is not the correct hypothesis. Under the correct hypothesis ($\dfrac{h''}{k''}$ is the middle term), the equations follow from Theorem 28 applied to consecutive terms.
Regarding the second part of your question, assume that $kh''-hk''=1$ and $k''h'-h''k'=1$. Solve the first equation for $h''$ to obtain
$$h''=\frac{1+hk''}k,$$ then substitute this into the second equation.
$$k''h'-\frac{1+hk''}k\cdot k'=1$$
$$k''h'k-(1+hk'')k'=k$$
$$k''(h'k-hk')=k+k'$$ as claimed. The second equation can be obtained by solving for $k''$ first.