All of the streams of Utah that flow from mountain slopes are subject to great fluctuations. They derive a large share of their water from the melting of snow, and not only does the melting vary as to its rapidity and season, but the quantity of snow to be melted varies greatly from year to year. A single measurement standing alone is quite inadequate to determine the capacity of a stream for irrigation, and as it was rarely practicable to visit a stream more than once, an endeavor was made to supplement the single determination by collating the judgments of residents as to the relative flow of the several creeks and rivers at other seasons and in other years. In districts where the water is nearly all used and its division and distribution are supervised by “watermasters”, those functionaries are able to afford information of a tolerably definite character, but in other districts it was necessary to make great allowance for errors of judgment. Certainly, that element of my estimates which is based on inquiries cannot claim so small a probability of error as the element based on measurements. Streams that are formed in high mountains reach their highest stage in June, and their lowest in September or October. Streams from low mountains attain their maxima in April or May, and reach their low stages by August or September. In the low valleys the irrigation of wheat and other small grains begins about the first of June, and continues until the latter part of July. The irrigation of corn and potatoes begins in the early part of July, and continues until the middle of August. In the middle of July all of the land calls for water, and if the supply is sufficient at that time, it is sure to meet all demands at other times. It will be convenient to call that time _the critical season_. In the higher agricultural valleys corn and potatoes are not grown, but the irrigation of small grains and hay is carried on from the middle of June to the middle or latter part of August. Through all this time the volume of the streams is diminishing, and if they fail at all it is at the end of the season. The critical season for the higher valleys is about the middle of August. In order to estimate properly the agricultural capability of a stream, it is necessary to ascertain its volume at its critical season. In the investigations of the past summer, this was accomplished by direct measurement in but a limited district. For the remainder of my field of operations I was compelled to depend on the estimates of others as to the relation between the volumes of streams at the time of measurement and at the critical season.