tions in fact in any region of similar 1931-32 to date. These diagrams seasonal rainfall characteristics. Such should be issued with the weather map diagrams should be available for gen- at least once each week, thereby keep- eral distribution. ing the public informed at a glance A sample diagram of the character not alone of the current 's rain- referred to is herewith presented. fall in relation to the normal, but also On this sample diagram there is of the seasonal regional rain history also shown the mass curve of the to the date of issue. rainfall at San Francisco, California On the sample herewith submitted in 1861-1862, this being California's the rain at San Francisco to March great year with 223 per cent of 26th for the seasonal year 1931-32 ap- normal rain at San Francisco. The pears at 90.6 per cent of the annual record of daily rainfall at San Fran- normal, 22.0 inches, or 19.93 inches. cisco was used in preparing this spe- The normal curve at the same date cial curve. shows 85.0 per cent, or 18.70 inches. The annual normals of all points of Every Weather Bureau station importance to which a diagram ap- should prepare such a diagram for plies should, for convenience, be noted office use and for the convenience of on the diagram. The final diagram those who desire to make compari- issued at the end of the seasonal year sons. covering the twelve months from At stations with long time records, September 1st to August 31st would frequent revision of such a diagram be in a convenient form for perman- would not be necessary even though ent preservation. the station's annual normal may pass The value of a record of this type through a considerable range. in graphic form is obvious. It would As each rain progresses the be particularly valuable to have such rainfall should be platted (in percent- diagrammatic records, based upon the age of the annual normal) on such a daily rainfall, when the influence of diagram in the form of a mass curve a rain storm or a sequence of storms —substantially as shown on the upon maximum stream flow is under sample diagram for San Francisco study.

JUNE TEMPERATURE INDICATES CORN MATURITY IN IOWA* By CHARLES D. REED Weather Bureau, Des Moines, Iowa Whether or not the average tem- September has much more to do with perature of June is above or below the amount of corn that escapes frost 67 ° F. largely determines the extent of damage than does the earliness or late- frost damage to corn in Iowa. ness of the first killing frost in au- In former papers, it was shown1 tumn. Not until recently was it real- that June temperature is a good indi- ized that the average temperature cator of the temperature of the next of June is a better early indicator of three months; and 2 that the average corn maturity than any other of the temperature of June, July, August and several weather elements that might be thought important. This is due to * Atlantic City Meeting, Am. Met'l. Soc., Dec. 27-28, 1932. the good start given to the corn plants 1 Monthly Weather Review, June, 1925, 53: 249-251. by a warm June and to the favorable 2 Monthly Weather Review, November, 1927, 55: 485-488. weather for maturity during following warm Junes. However, a average temperature of the month of high yield is not always associated June, July, August and September with high maturity, for a warm June shows a close relation to corn matur- is frequently followed by deficient pre- ity, the complete data are not available cipitation and too high temperature long enough in advance to be superior during July and early August. in a practical way to a direct statistical Besides June temperature, efforts inquiry by crop reporters on October 1, have been made to correlate corn ma- as to the advancement of the crop. turity with monthly and seasonal rain- Dropping September to make the data fall, temperature of other months and available a month earlier, and using seasons, snowfall, March snow- the mean temperature of the months fall and April snowfall, monthly and June, July and August does not im- seasonal sunshine, and fall prove the prediction curve over that frost dates, and monthly and seasonal based on June alone. sunspot numbers, with the result that A dot chart (Fig. 1) made by plot- June temperature is far in the lead, ting June mean temperatures for Iowa with the per cent of possible sunshine as abscissas and the per cent of the in June next in importance. While the corn that matured without frost dam-

Fig. I. June Temperature Indicates Corn Maturity. Broken lines show average June temperature (69.4°) for Iowa in 43 , 1890-1932; and average per cent of corn (87.3%) that matured without frost damage in the same period. The curve, drawn by eye, serves as a ready means of estimating in June the probable maturity of the corn crop the fol- lowing autumn.