A typical example of cyclogenesis in the western tropical Pacific Ocean is investigated. In the low troposphere, data do not support the concept of an equatorial front extending across the entire Pacific Ocean. They indicate the presence of several large clockwise-rotating eddies in the equatorial trough. Typhoon development begins in consequence of instability of the northern-hemisphere trades, not as a result of interaction between currents from northern and southern hemispheres. No evidence can be adduced in the low troposphere that suggests a reason for this instability. Large vortices are found at 200 mb instead of the easterlies characteristic of low levels. Their spacing and speed are studied, and it is found that they are systems of the order of magnitude of long waves in the westerlies. After computations on the structure of the basic zonal current in the trade-wind belt, relations between high-tropospheric vortices and waves in the easterlies are discussed. These two types of perturbations are not propagated at the same speed, but there is relative motion between them. Instability of the trades sets in just as a lower wave trough is bypassed by a broadscale ridge in the high troposphere. The pressure field superimposed on the low levels by the upper vortices appears to furnish an explanation for the instability of the trades. Finally the question is raised as to why some incipient storms attain typhoon intensity and others do not. It.is found that a storm will grow when it becomes located under a ridge aloft, so that upper outflow from the storm is facilitated by the general pressure distribution.