222 TRANSACTIONS OF THE AMERICAN INSTITUTE. 



stood in order to make the barometer a useful instrument, and yet they 

 should be explained in order to be understood by men who have not made 

 the subject a study, as I have ; and that explanation I may wish to make, 

 reviewing my rules in every article. Rule 1. Tlaere is no point at which 

 the barometer must stand to indicate rain or wind. 



This is simply to counteract an impression in the public mind, that if 

 the barometer stands at a given point it will rain or be clear, as the case 

 may be. In England, where the climate is not as variable as in this country, 

 they use on their barometers the words fair at 30 inches, change and rain 

 29| inches; now, if you mark your scale at these points with the words 

 above indicated, at the sea level, and then remove your barometer to a 

 locality say one thousand feet higher than the sea level, your barometer 

 will stand in clear weather at the point marked rain, so that if for no other 

 reason this alone is sufficient for having no marks on your barometer. 



Rule 2d. The judgment must be governed by the rising or falling of the 

 barometer. 



The falling of the barometer indicates the approach of a storm, the ex- 

 tent of which will be proportionate to the amount and rapidity of the fall. 

 The ordinar}^ variation is about one inch, the extreme variation is about 

 two inches; the extremes are seldom reached — only three or four instances 

 that I have noticed for thirty years. My rules divide storms into three 

 classes, but strictly speaking there are but two classes, as showers are 

 but the reaction or the passing off of a storm, and that only in the season 

 of showers, say June, July and August. 



Northeasterly storms. In this class of storms the barometer usually 

 falls from a high point, and when only wind blows from the N. E. the 

 barometer seldom falls much, and sometimes even rises with a N. E. wind, 

 but when much rain falls in a N. E. storm, the barometer falls from 5 to | 

 of an inch; when the storm comes to a crisis, or when the wind changes to 

 N. W., the barometer begins to rise, and more or less rain falls from the 

 N. W. and with a rising barometer. (Rain only falls with a rising barom- 

 eter when a storm is passing off.) 



Southerl}^ storms. My rules say that in a southerly storm the barometer 

 falls from one to four-tenths of an inch, and varying in time from six to twelve 

 hours; this variation, in time, is too limited; instead of twelve hours, it 

 often happens that the south wind blows two, three and sometimes four days 

 after the barometer has fallen, and during this time the barometer does not 

 vary much from the point to which it fell at first. Rain seldom falls with 

 a south wind; but on a change of wind to the southwest or west, from which 

 points an abundance of rain usually falls, the action of the barometer 

 during this class of storms is vei'y slight after the first fall. About the 

 time of change of the wind to southwest or west, there will or may be a 

 slight fall of tlie barometer; but if there is no change in the barometer at 

 this stage of a southerly storm, the first fall of the instrument is to be relied 

 upon, and after it has rained freely, if the index does not rise, depend upon 

 it there is more to come. 



In the season of showers, southerly storms usually pass off in a series of 

 showers, instead of passing off in a steady rain. There will be showers 

 for several days in succession, each of which will be indicated by a fall 



