24 



Also much data have been recopied and put in a more compre- 

 hensible and practical manner for future reference. By special 

 request of the Weather Service at Washington a series of observa- 

 tions have been taken for ascertaining the temperature at which a 

 killing frost is possible, and the temperature at which a frost is 

 possible, as well as the relation between the temperature on the 

 tower and that of the shelter a few feet above the ground. These 

 temperatures are taken with the standard minimum thermometer, 

 exposed in the regulation thermometer shelter. Much study has 

 been made of the weather maps, two of which are received daily, 

 and the local and government forecasts compared with the actual 

 conditions of the weather at this station for the same period. In 

 fact, verifying daily forecasts has been quietly carried on at this 

 station for several years, and we believe no other station has ever 

 continued this work for so long a period. The local and Wash- 

 ington forecasts are carefully studied and marked according to a 

 certain standard which was decided upon at the outset. Although 

 the work has required a large amount of time and patient applica- 

 tion, yet the value of energy and time expended is small when 

 compared with the results obtained. The conclusion arrived at, 

 based upon actual observation is, that the efficiency of our weather 

 service, as shown in weather predictions, is certainly gratifying. 

 The average percentage of accuracy of the local service for the 

 whole period is ninety-one per cent and the Washington ninety 

 per cent. While the local service is slightly ahead of the depart- 

 ment at Washington, it is due the latter to say that the slight 

 difference in the averages in favor of the Boston office should not 

 alone be considered as indicating superior foresight of the officials 

 at the latter place, as they have a small area to consider in making 

 weather predictions for New England, while the officials at Wash- 

 ington make forecasts for the whole country. The predictions of 

 Foster in St. Louis, based upon electrical and periodical weather 

 changes, which predictions are made two weeks in advance, and 

 those of Clayton at Bk;e Hill, whose bulletins have been carefully 

 watched, show that these gentlemen* also have methods of fore- 

 casting the weather which give remarkably good results. 



All official telegrams received at the observatory are recorded, 

 the time they are received and the time the weather signals are 

 hoisted, it being thought proper to make a record of this for 

 future reference. 



While certain lines of investigation already begun have not 

 been fully completed, yet much careful and thorough work has 

 been accomplished during the year. 



