Since the appearance of Rogowski's investigation of the spark discharge by a cathode ray oscillograph as well as those of Dunnington and others using a Kerr cell, the physical investigations on the spark formation have undergone almost a complete transformation. The previous theory of the ionization by collision of positive ions was modified to the theory of the electron avalanche proposed by Frank, v. Hippel, and others. In this stage of the investigation it is most important to get thorough knowledge on the preliminary stages of spark formation immediately preceding the main spark. for this purpose the authors have tried to study the production and distribution of ions associated with the initial stages of spark formation by using the Wilson chamber, and have confirmed the possibility of its application. Almost simultaneously Snoddy and Bradley, and Kroemer published the results of their experiments which were quite similar to those in our investigation. In our previous experiments an influence machine was used as the high voltage source and a pendulum for the synchronization of the expansion of the chamber and the application of the voltage impure. It was difficult, however, by these methods to keep the voltage constant and to avoid undesirable distributed inductance of the circuit. In the new series of experiments the influence machine was replaced by a system of transformer, kenotron, and condenser, and an apparatus of special design was used for the purpose of the synchronization in place of the pendulum. With this apparatus thus developed cloud photographs were taken of sparks in various gases and air mixed with small quantities of some organic vapours, the results of which will be described in this paper. Raether's recent paper deals with similar experiments in nitrogen, carbonic oxide, and hydrogen. some of his photographs are similar to ours, but most of them are quite different, as will be described below.