A number of different avenues of study have shed light on the problem of long‐range forecasting. If we may judge from the results achieved thus far in a practical way, the most profitable one is to associate weather in one locality with antecedent conditions in some other distant region, as was done by Sir Gilbert Walker in the monsoon‐forecasts for India. Unfortunately little work of this character has been done in the United States and it was with the idea of laying a foundation for studying conditions in this country that the work which will be described was undertaken. With a medium so labile and so subject to change as our atmosphere, it would be idle to suppose that some extremely simple relationship connected the weather in a nearby region with antecedent weather in some remote region. It was realized, therefore, that chances were extremely remote that the results of this investigation would expose any such simple relationship. Nor was it with the expectation of solving finally the problem of long‐range forecasting in this country that the study was undertaken but rather with the conviction that by such studies our knowledge would be enhanced, with the likelihood of developing suggestions that would lead by successive stages to a solution of the problem in the f