Abstract
The nature of ground discharge has been investigated using the data of electric field measurement, thunder recording, flash photographing, etc. The height of charge center neutralized by a return stroke was found not to increase with the stroke order in contrast with the report given by Malan and Schonland, instead the stroke channel was often found to progress laterally successively. The existence of a continuing discharge process between and after return stroke has been verified in agreement with the report given by Malan, and Kitagawa et al. The charge neutralized by return strokes in a ground discharge is, on the average, smaller than that neutralized by a cloud discharge. However this does not always mean that the total charge neutralized by a ground discharge is smaller than that of a cloud discharge, because a ground discharge is often accompanied by a continuing discharge process neutralizing considerable amount of negative charge. A new leader stroke which is proposed by us has the structure of corona-like streamers on the top part of the stroke channel. If the electrical structure of a thundercloud in South Africa is the same as that in Japan, the charge height reported by Malan and Schonland seems to be too high and should be reduced to a smaller value.
