EXPERIMENT STATION REPORT. 547 



while the third in rank. 1897, has its July with the record of 11.42 

 inches against it, the largest for any month for the past sixteen 

 years. 



The dry and rainy seasons are somewhat in groups, that is, the 

 first two taken together are decidedly wet ones followed by six 

 that are all above the average for dryness while the last four sea- 

 sons average high for rainfall. The rank which each season takes 

 in wetness and dryness is shown in the lower lines of the table. 



Temperature: In a table on page 550 corresponding to the 

 one of rainfall it is seen that there has been some variation in 

 temperature during the past sixteen growing seasons. In order 

 to determine the relative temperature of a month it may be com- 

 pared with the average for that month in the column at the ex- 

 treme right. The warmest summer was in 1900 and the coldest in 

 1903, in fact the last three summers include the three coldest dur- 

 ing the sixteen years while with one exception (1900) the three 

 warmest came in succession in 1894-96. 



By comparing the two tables, namely, that of rainfall and tem- 

 perature, it is seen that the warmest summer was next to the low- 

 est in rank of rainfall and the three successive seasons which were 

 the coolest ranked "5th, 7th and 4th respectively in precipitation. 

 That there is not a closer relation between warmth and dryness or 

 coolness and moisture is due, for one thing, to heavy downpours 

 of rain that may cover only short periods of time. For example 

 the last three months of the season of 1889 were all above the 

 average for rainfall and they were all below in temperature. The 

 same thing is true of the corresponding months in the next 

 (1890) season. In 1900 there was a shortage of rain in all 

 months excepting one (May) and in that the amount above the 

 average was slight. In this season the temperature was above the 

 average for all the months excepting May, which was exactly the 

 average. July of 1897, previously mentioned as leading all other 

 months in rainfall, had a temperature a little above the average 

 and is remarkable in that the showers were heavy and the 

 warmth kept close to the normal for the month. 



Sunshine: A third table (page 551) gives the sunshine for 

 each month, the percentages being obtained by dividing the total 

 of clear and partly cloudy days by the total number of days in the 

 month. It is seen from the column of averages that August is the 

 brightest month while April and May are the most overcast. 

 Upon the whole summers are practically three- fourths (74.6 per 

 cent.) fair weather. Some years have been more bright than 

 others, as seen bythe average for 1897 (80 per cent.) as compared 



