The statement of the exercise is pretty simple to understand, but I'm having trouble to prove it. The exercise is the following:
(Exercise 3, from Introduction to functional analysis, by A. E. Taylor, page 190).: If $X$ is a complex normed linear space and $X_r$ is the associated real linear space, we write $x'(x)=x_1'(x)-ix_1'(ix)$, where $x_1'\in (X_r)'$ and $x'\in X'$. Show that $\|x_1'\|=\|x'\|$.
Here, $X'$ denotes the dual space of $X$, i.e. the space of the linear continuous transformations $x:X\to \mathbb{K}$, where $\mathbb{K}$ is the field of the scalars. And $X_r$ denotes the same vector space $X$, but considering the scalars to be only in $\Bbb{R}\subset \Bbb{C}$.
Okay, we have to show that $\|x_1'\|=\|x'\|$, i.e. $$\sup\{\|x'(v)\|\,:\, \|v\|\leq 1\}=\sup\{\|x_1'(v)\|\,:\, \|v\|\leq 1\}.$$
I've thought to show that $\|x_1'\|\leq \|x'\|$ and $\|x'\|\leq \|x_1'\|$. The first one is easy, because, since $x_1'(v)\in \Bbb{R},\forall v \in X_r\approx X$, we have $$\|x_1'(v)\|\leq \|x_1'(v)-ix_1'(v)\|=\|x'(v)\|.$$
I could not prove the second one. Then I've tried by contradiction: supposed that $\|x_1'\|<\|x'\|$, made some computations, but could not find a contradiction at all...