ENGHLMANN MEAN A: EXTREME TEMPERAT E IN ST. LOUIS. 499 



slighter one in the two weeks succeeding them. After that the 

 temperature sinks rapidly to about Dec. loth, from which time 

 till the beginning of February the changes are not very marked. 



The mean temperature of April 17th to 19th and from October 

 1 2th to 17th correspond with the mean of the year. 



The tables, and still more distinctly the diagram, show us also 

 that the extreme highest and lowest temperatures diverge most in 

 winter and least in summer, and that their values are much more 

 variable in the former than in the latter season. The possibilities 

 of range from the middle of December to the middle of March are 

 80 to 95 degrees, while in June and July they amount only to 40 

 or 45 degrees. 



The same law is found when we compare the actually ob- 

 served lowest minimum and highest maximuni of every month ; 

 their divergence is greatest in January, and least, not much more 

 than half, in July. 



.January . 

 February 

 March . . 



April 



May 



June 



Nearly the reverse is the case — i.e. the range in winter is much 

 smaller than that in summer — if we compare the difference of the 

 average daily minima and maxima for each month : 



.January.. 13.27 April 18.29 l July 18.24 October .. 18.00 



February. 14.72 I May 18.77 August... 17.75 November 14.06 



March ]U. 40 June 18.14 | September 19.05 December. 11.97 



The range, it will be seen, is, on the whole, least in the cooler 

 and greatest in the warmer months of the year ; but this differ- 

 ence is not due to the lower or higher temperatures of those 

 months, for it will be seen that in December the range is the 

 smallest (smaller than in January) and in September greatest 

 (greater than in July). This variation in the range of maxima 

 and minima is undoubtedly owing to the condition of the sky in 

 the different months. Gloomy weather prevails in the beginning 

 of winter, and a clear sky with abundant evaporation, and thus 

 a reduction of night temperature, in the autumn. The little table 

 can give us an indication of the prevailing weather in the differ- 

 ent months. Thus the difference, and its cause, the clearness of 



