[en] Plasma ignition delays were observed in a ‘main’ inductively coupled plasma (ICP), in tandem with an ‘auxiliary’ ICP. The Faraday-shielded ICPs were separated by a grounded metal grid. Power (13.56 MHz) to the main ICP was pulsed with a frequency of 1 kHz, while the auxiliary ICP was operated in continuous wave (cw) mode. In chlorine plasmas, ignition delay was observed for duty cycles greater than 60% and, in contrast to expectation, the delay was longer with increasing duty cycle up to ∼99.5%. The ignition delay could be varied by changing the auxiliary and/or main ICP power. Langmuir probe measurements provided the temporal evolution of electron temperature, and electron and positive ion densities. These measurements revealed that the plasma was ignited shortly after the decaying positive ion density (n ₊), in the afterglow of the main ICP, reached the density (${n_{+}}_{,\text{aux}}$) prevailing when only the auxiliary ICP was powered. At that time, production of electrons began to dominate their loss in the main ICP, due to hot electron injection from the auxiliary ICP. As a result, $n_{\text{e}}$ increased from a value below $n_{\text{e},\text{aux}}$, improving inductive power coupling efficiency, further increasing plasma density leading to plasma ignition. Plasma ignition delay occurred when the afterglow of the pulsed plasma was not long enough for the ion density to reach ${n_{+}}_{,\text{aux}}$ during the afterglow. Besides Cl₂, plasma ignition delays were also observed in other electronegative gases (SF₆, CF₄/O₂ and O₂) but not in an electropositive gas (Ar). (paper)