MARYLAND GEOLOGICAL SURVEY 149 



for St. Mary's County is that for Charlotte Hall in the northwest portion 

 of the county. Here Prof. J. Francis C'oad took meteorological obser- 

 vations from 1893 to 1904, but owing to frequent absence from station 

 the records are unfortunately not continuous, varying in length from 

 7 years for June to 11 years for January. Table XXIV, page 176, gives 

 the necessary information in regard to each station in the county. St. 

 Mary's City, St. Inigoes, Eidge, Cherryfields, and Porto Bello are located 

 not far apart in the southern portion of the county, near St. Mary's 

 Eiver; Leonardtown is at the head of Breton Bay; and the most northern 

 station, Charlotte Hall, lies near the boundary of Charles County. 



From the fact that records of temperature extending over many years, 

 such as the period of 200 years at Florence, 100 years at Paris and 88 

 years at Baltimore, show no change from the annual course of tem- 

 perature at present observed, in spite of the great variation in the 

 monthly means, it is concluded that every place has a normal march 

 of temperature which can be ascertained only by a series of observations 

 of considerable duration. The length of the period required to obtain 

 a correct normal depends largely upon the character of the climate. In 

 tropical regions where the seasonal variations are small, five years are 

 sufficient to give an annual mean temperature accurate to within 0.1°, 

 but such is by no means the case in a variable, continental climate like 

 that of the eastern United States. At Baltimore the probaljle error of 

 the means for 88 years is still greater than 0.1°. Means for less than 

 20 years at any station in Maryland are far from representing true 

 normal temperatures. At Baltimore the January mean temperature has 

 varied from 43.9° in 18.58 to 24.3° in 1893, a difference of 19.6°; at 

 Solomons the variation in the February means was from 40.8° to 26.2°, 

 a difference of 14.6° in 14 yeai's. Therefore, in comparing short term 

 records at several stations it is impossible to determine whether the 

 differences observed are really due to actual differences in climate, or 

 are due to the special character of the short period from which the 

 means were derived. This is well shown by comparing the excellent 

 record for 14 years at Solomons (1892 to 1905) with the 88 year means 

 at Baltimore, when it appears that Solomons is colder than Baltimore 



