Blair's recent theory of excitation is analysed with the following conclusions: 1. The theory is inapplicable to currents of long duration; i.e., slowly increasing currents and the opening excitation. 2. The theory is a modification of the condenser theory of excitation but the modification is to be rejected on three grounds: (a) The modification has no obvious physical significance. (b) It does not in fact remedy the divergence between calculation and observation. (c) It leads to certain conclusions of a surprising kind which are contrary to observed fact. 3. The qualitative value of the condenser theory is demonstrated by the fairly close agreement between calculation and observation over a considerable field of enquiry.