The experimental conditions of the Santa Barbara Project over the three years 1957 through 1959 were not uniform. In 1957, there was no seeding in the adjoining areas. In 1958, seeding went on in Ventura County at every opportunity. In 1959, the seeding operations in the two counties were factorially randomized. These changes created partial confounding of possible effects of seeding with possible effects of changes in weather pattern. Whatever the cause, some of the apparent effects were very large. In particular, the average apparent increase in the rain in Santa Barbara, obtained in the absence of seeding in Ventura, amounted to about 100 per cent. The last section of the paper is concerned with recommended changes in the design intended to increase the accuracy of the experiment.
1 Prepared with the partial support of the National Science Foundation, Grant G-8211.