In the magnetometer signal, we found long-term (~hours) glitches in the CS and the X-end just after the Earthquake.