The author takes an array \(\{x_{n,k}\}\), \(n\geq1\), \(k\geq0\) such that for each \(k\geq0\), \(\gamma_k\) exists for which \(x_{n,k}=O(n^{-\gamma_k})\), \(x\to\infty\) (the reviewer could not find where the latter assumption is used or even mentioned again in the paper). And he takes a family of nonnegative functions \(\phi_{n,k}\in C^1(\mathbb R_+)\), \(n\geq1\), \(k\geq0\), satisfying \[ \sum_{k=0}^\infty\phi_{n,k}(x)=1,\quad\sum_{k=0}^\infty x_{n,k}\phi_{n,k}(x)=x,\quad x\geq0, \] and such that there exists a sequence of positive functions \(\psi_n\in C(\mathbb R_+)\), for which \[ \psi_n(x)\phi'_{n,k}(x)=(x_{n,k}-x)\phi_{n,k}(x). \] He defines the operators \[ L_n(f,x):=\sum_{k=0}^\infty\phi_{n,k}(x)f(x_{n,k}),\quad n\geq1,\quad x\geq0, \] for all functions \(f\) defined on \(\mathbb R_+\) for which the sums converge. Special cases are the Szász-Mirakyan and the Baskakov operators. (Note that in the paper there is a mistake in the definition of the Baskakov operators.) The author investigates the properties of the \(r\)th central moment of \(L_n\), namely, \[ \Lambda_{n,r}(x):=\sum_{k=0}^\infty\phi_{n,k}(x)(x-x_{n,k})^r, \] not noticing that nothing in his assumptions guarantees the convergence of the infinite sum for \(r\geq2\). Thus without additional assumptions Lemma 1 and hence Theorem 1, are invalid. Assuming that we add the appropriate assumptions above, the reviewer fails to see the proof of Lemma 2 since the author indicates nothing about the behavior of \(\Lambda'_{n,2m-1}(x)\) that is needed in his induction step. In addition, the reviewer wonders why Lemma 2 “might be of interest in its own right”. The methods of proof are completely standard, notwithstanding the above comments.