216 



average spring mean. Tlie above averages are all based upon 14 

 years' observations. 



Thus looked at in every way we see how abnormally low the 

 mean temperature was from November 1878 to May 1879, both 

 these months being included. Let us turn now to the second 

 table showing the extreme temperatures for the same period. 



In this table the mean of all the daily maxima is shown to vary 

 much in the same Avay as the monthly mean. The difference 

 between it and the average (14 years as before) being greatest in 

 January, least in March, running up again in April, and still con- 

 siderable in May as compared with March and February. The 

 mean of the daily minima varies not very differently from that of 

 the daily maxima, except that in May it is not so much below the 

 average as the latter. It will also be seen that, considered as a 

 whole (for December and March are exceptional months in this 

 respect), the mejin maxima are more depressed hclow the average than 

 the mean minima. 



With regard to the alsohte maximum and minimum of each 

 month (not given in the table) the maximum here too is remark- 

 able for its greater depression below the average than the minimum. 

 Looking to back registers I find that, taldng the same average of 

 fourteen years, the absolute maximum in each of the above months, 

 November — May, in 1878-0, was beloic the average except March, 

 when the maximum rose to 63°, being above the average by 4°-l. 

 This high temj^erature was attained on the 19th, there being no 

 other day in that month on which the thermometer rose even to 

 60°. That it was an exceptional day is shown by there being 

 very little difference between the mean of the daily maxima for 

 March 1879 and the average as shown in the second of the tables 

 given above. 



The difference between the absolute maximum and the average 

 in the other months ranged from 1°8 in December and February 

 to 10° -9 in April, and 5° -2 in May. These two last months appear 

 to have been decidedlv those in which the maximum extreme was 



