In the past, most of the investigations of cosmic-ray showers by means of counters have been made by using a triple coincidence method. In the present experiments a quintuple coincidence method is used. It offers the following advantages. The casual coincidence rate is greatly reduced by adding two more counters, so that the correction for casual rate can be omitted altogether. If the five counters are given the pentagonal arrangement shown in figs. 1 and 3, then at least three rays are required to operate all the counters to cause a coincidence wherever the rays come from. Thus all showers counted contain at least three rays each. Although in Gilbert’s three-counter arrangement* the lead plate is limited so that a shower produced in the lead is only counted if it contains three or more particles, two rays are sufficient to cause a triple coincidence, if they come not from the lead but from the air or from the surrounding materials. In the quintuple coincidence pentagonal arrangement, there is no need to limit the size of the shower-producing material. Thus more material can be used and so the rate of the shower production can be raised. The sheets used to produce showers are bent into semi-cylindrical form (fig. 1). In this arrangement, all the shower particles which cause a coincidence have roughly an equal path in the shower-producing material.