1: \begin{definition}\label{def:MmuLine}

2:    \begin{enumerate}[label=(\roman{*})] 

3:     \item\label{it:defMv}

4:    Given a non-decreasing, continuous function $v:(0,m)\to\mathbb{R}$

5:    satisfying $\lim_{x\to0+}v(x)=-\infty, \lim_{x\to m-}v(x)=\infty$, we define its \textit{generalised inverse} $M_v:\mathbb{R}\to(0,m)$ via 

6:    \begin{align}\label{eq:MvLine}

7:      M_v(r) = \sup\big\{x\in(0,m):v(x)\le r\big\},\quad r\in\mathbb{R}.

8:    \end{align}

9:    \item\label{it:defmuv} It is elementary to see that $M_v$ in~\eqref{eq:MvLine} is increasing, right-continuous

10:    and satisfies 

11:    \begin{align*}

12:      \lim_{r\to-\infty} M_v(r)=0,\qquad \lim_{r\to\infty} M_v(r)=m.

13:    \end{align*}

14:    Hence, $M_v$ is the cumulative distribution function (cdf) of a measure $\mu_v\in \mathcal{M}^+_b(\mathbb{R})$ whose mass equals $m$ (see e.g.~\cite[Chapter~20.3]{royden2010real}). The measure $\mu_v$ is uniquely determined by 

15:    \begin{align*}

16:     \mu_v((-\infty,r])=M_v(r),\quad r\in \mathbb{R}.

17:    \end{align*} 

18:    \item 

19:    Given $u$ as in Theorem~\ref{thm:line} and $t\ge0$ we denote by $M(t,\cdot):\mathbb{R}\to(0,m)$

20:    the generalised inverse of $u(t,\cdot)$, i.e.

21:     $$ M(t,\cdot):=M_{u(t,\cdot)},$$

22:     where we used the notation~\eqref{eq:MvLine}. We further let $\mu(t)\in \mathcal{M}_b^+(\mathbb{R})$ denote the measure associated with the cdf $M(t,\cdot)$ as introduced in Definition~\ref{def:MmuLine}~\ref{it:defmuv}, i.e. $\mu(t)=\mu_{u(t,\cdot)}$.

23:    \end{enumerate}

24:  \end{definition}

25: