752 IOWA DEPARTMENT OF AGRICULTURE 



mum of December, 1877, was 63° or 10° above the mean of the maxima 

 for this month. The abnormal temperature of this particular month 

 was shown by the fact that only on thirteen nights did the thermometer 

 fall to the freezing point. 



If we take the month of January as observed at Iowa City beginning 

 with 1860, we find that the mean temperature of the month just passed 

 has been equaled or exceeded only for the corresponding months in 

 1880, 1891 and 1900, or but four times in fifty years have we had as 

 warm a January. The temperature of December, 1913, has been equaled 

 or exceeded only in the years 1877, 1891, 1892 and 1897. There have 

 been then but five sucli warm Decembers in fifty years. 



Since a warm month in any one year may or may not be preceded 

 or followed by another warm month, it does not follow that an extremely 

 warm December or January means a really mild winter. For example, 

 the January of 1880, as said above, was the warmest ever recorded at 

 this station; yet because the December preceding was decidedly below 

 normal, the result was not a higli average for the two months. 



When the mean then of December and January is considered, we find 

 that the two months just past have been equaled or exceeded only for 

 the corresponding months in 1877-78 and 1890-91, or as warm a December 

 and January together has occurred but three times in fifty years. 



There is but little doubt but that these months in 1857-58 were as 

 warm as the ones just past, yet the observations made at Muscatine can 

 not be compared rigorously with those at Iowa City. 



The average deviation from the mean may be taken as a measure of 

 variability. The average deviation, for December, from the mean of 

 23.6° is, when computed from fifty years' observations at this station, 

 found to be 4.7°; while this deviation in 1913 was 8.6°. The correspond- 

 ing value for January is 5.9° and for January, 1914, it was 8.0°. 



From the standpoint of probability we may expect a December as 

 warm as that of 1913, about once in ten years, while a January like 

 that just past may be expected perhaps once in thirteen years. A warm 

 December with a January following such that a mean temperature 

 equal to that of the past two months is observed, may probably occur 

 once in seventeen years. 



If the chance of a warm December and a warm January immediately 

 following were entirely independent of one another, we should then have 

 for the probability of such an occurrence the product of the two inde- 

 pendent probabilities, or in this case one-tenth times, one-thirteenth or 

 one divided by one hundred and thirty; while the combination seems 

 ■to occur with a probability of one-seventeenth; we then must think 

 that a warm December and January are not independent of each other. 

 The data are too limited to be dogmatic in our conclusions yet one seems 

 to be justified in assuming a strong correlation between two successive 

 months. 



