\section{Introduction} Sampling a reaction system at discrete times does not reveal the intervening reaction counts or holding times. Both can carry information about a parameter that multiplies every transition rate. We quantify how much of the complete-trajectory Fisher information about such a common clock survives when only the sampled states are observed, and we locate the first loss exactly. The inferred clock is a fixed parameter; no distribution of random external inputs is assumed, and the physical sampling interval $\Delta$ is held fixed under differentiation. With $t=\theta\Delta$ the dimensionless sampling time, the information fraction retained is a function of $t$, or of $\lambda=\Omega t$ for a population of size $\Omega$. \paragraph{Results.} \emph{Reaction-count aliases} (Section~\ref{sec:alias}). For a finite known mark law with nonzero integer marks, the \emph{alias depth} $h$ is the smallest value of $\max\{|z|,|z'|\}$ over pairs of reaction-count vectors $z,z'$ of different total length with the same increment. Theorem~\ref{thm:alias} gives the loss of the compound-Poisson count experiment as $1-F_p(\lambda)=C_p\lambda^{h-1}+O_p(\lambda^h)$ with an explicit positive \emph{alias coefficient} $C_p$, and $F_p\equiv1$ when $h=\infty$. Corollary~\ref{cor:alias-rw} writes $C_p$ in terms of the $h$-step mark random walk and evaluates it for nearest-neighbour and simplex mark laws. Two mark laws with identical Gaussian moments can have different exponents (Section~\ref{sec:gaussian-counterexample}). \emph{Holding-time loss} (Section~\ref{sec:holding}). For a fixed finite irreducible generator $Q$, Theorem~\ref{thm:classification} classifies the missing log-clock information $\mathcal M_Q(t)$ between the complete trajectory and its endpoints: $\mathcal M_Q(t)=A_2(Q)t^2+O_Q(t^3)$, and $\mathcal M_Q(t)=[A_3(Q)+H(Q)]t^3+O_Q(t^4)$ when $A_2(Q)=0$. Here $A_2,A_3$ are explicit path-alias coefficients and $H\ge0$ is the holding-time term, positive exactly when the escape rates are not all equal. Every reversible generator has $A_2>0$ (Remark~\ref{rem:birth-death}), so the $t^3$ regime requires a generator without two-cycles. \emph{The stationary Ehrenfest chain} (Sections~\ref{sec:protocol}--\ref{sec:window}) is the worked example, observed at its full endpoint pair for a known population size $\Omega$. Theorem~\ref{thm:finite-gaussian} gives, for every $\Omega$ and finite $t>0$, the finite Gaussian envelope $B_n(t)\le R_n(\lambda)