A survey of the data available indicates that lightning flashes in all thunderstorms have similar characteristics as regards the times involved and the number of strokes in a flash, and it is shown that Norinder's divergent values are due to misinter-pretation of his data.It is suggested that while the charge neutralized in a lightning flash is probably about 50% greater than has hitherto been believed, the cloud potential is only a few per cent of the accepted value, and the energy of a flash about 250 kWh, as compared with Wilson's value of about 2800 kWh. A theory is advanced bearing on the mechanism of the leader and return strokes, which enables the shape of the lightning current wave to be calculated; in addition, a form is derived for the whole current wave in a multiple flash. A synthesis made from the results of many investigations leads to a picture of the mechanism of the return stroke and to the calculation of the various field changes produced, which agree well with those observed both in shape and in amplitude. New light is thrown on the analysis of the electrostatic field change caused by a stroke. The importance of space charge is emphasized in several connections, including the neutralization of cloud charge thereby, the frequency of “air discharges” and the peculiarity of flashes to the Empire State Building, the currents to which have been re-analysed.The effect of direct strokes to transmission lines is investigated, and it is shown that successive strokes which do not cause flashover are unlikely. The polarity of flashes to earth is shown to differ in tropical and temperate regions, and it is concluded that the polarity of strokes to phase wires cannot be deduced from that of strokes to towers. Suggestions are made as to further investigations.