### Peccata

A Taniyama product for the Riemann zeta function. In: Exploring the Riemann Zeta Function: 190 Years from Riemann's Birth, H. Montgomery, A. Nikeghbali, and M. Rassias eds., Springer (2017).
I deeply apologize for being completely unaware of, and hence not acknowledging, the following prior literature:
[1] K. Joshi, R. Ragunathan, Infinite product identities for L-functions, Illinois J. Math. 49 (2005), 885-891.
[2] K. Joshi, C. S. Yogananda, A remark on products of Dirichlet L-functions, Acta Arith. 91 (1999), 325-327.
[3] J. Kaczorowski, A. Perelli, Some remarks on infinite products of L-functions, J. Math. Anal. and Appl. 406 (2013), 293-298.
In particular, Kaczorowski and Perelli consider product identities for a vastly more general class of L-functions, the identity for $1/\zeta(s)$ being merely a special case. I am grateful to Alberto Perelli for drawing my attention to the above literature and for showing me a proof of Theorem 2 valid for the subclass of Euler products in [3] which are of "polynomial type."

### Errata

Quaternionic Artin representations of Q , Math. Proc. Cambridge Phil. Soc. 163 (2017), 95-114.
In the first paragraph of the introduction, the asymptotic relation proved by Klueners is misstated: It should be $Z(k,G;x)\sim cx^a$ with a positive constant $c$.

Self-dual Artin representations. In: Automorphic Representation and L-functions, D. Prasad, C.S. Rajan, A. Sankaranarayanan, J. Sengupta, eds., Tata Institute of Fundamental Research Studies in Mathematics Vol. 22 (2013), 455 -- 499.
The following errors have been corrected in the above pdf file but disfigure the published version:
(i) The sentence following (4.1) on p. 465 should be corrected as follows: "Indeed let $m$ be a positive integer not divisible by 4 such that the greatest common divisor of $m$ and the discriminant of $F$ divides 2, and let ..."
(ii) In the second line from the bottom of p. 472, $\mathfrak q$ should be $q$.
(iii) The constant in the asymptotic relation (9.3) on p. 487 is incorrect, both because of double-counting and because the formula for $q(\rho)$ holds in general only if $q(\psi)$ is relatively prime to $d_K$. However, the bound $\vartheta^{\ab, K}_{\Q,2}(x) \ll x$ still holds, and if (9.3) is replaced by this bound then the rest of the argument goes through as before.

A deformation of the Tate module, J. of Algebra 229 (2000), 280 -- 313.
The formula for  $\theta(c)$  on p. 294 is in error: The coefficient of  $X^2$  should be  $(c^4-c^2)/12$,  not  $(c^4+2c^2-3c)/24$.  Although very embarrassing, the error is inconsequential in the sense that there was no need to display the coefficient of  $X^2$  in the first place.

Galois theory, elliptic curves, and root numbers, Compositio Math. (1996), 311 -- 349.
In the first line of the paragraph containing displayed formula (3.6) on p. 331, the equation  $K=F(\Delta^{1/e})$ should be  $K=F(\varpi^{1/e})$.