between the months contiguous to those for which the data 

 were fragmentary or missing. The means depend in part, there- 

 fore, upon graphic interpolations, using the monthly charts of 

 the Monthly Weather Review for that purpose. The year 1896 

 was rejected in taking the average. 



The figures are the hundredths of an inch and range from 

 29.54 to 30.12. The figures in italics in Table i imply that a 

 number of observations were missing. 



