Several analyses of variance tests were performed on data obtained from a dust collection network. The results indicate that very careful attention to design of the system, strict control of modifying influences, and close adherence to rigid collection and analytical procedure are necessary for duplicable results to be recorded. Human activities, natural phenomena, and atmospheric differences encountered at moderate distances between sites all tend to introduce modifying variables.