OF ARTS AND SCIENCES. 345 



on newspaper notices, on circulars to members of our society, and to 

 co-operating observers, and in some cases on circulars to postmasters ; 

 but in this day of many circulars, they give an uncertain dependence. 

 Members in Classes B and C are enlisted directly or by promotion from 

 A, and include many careful observers, whose interest in the society 

 is a great gain in its work. Indeed, our relation to a body of intelli- 

 gent correspondents, thus enrolled in all parts of New England, is a 

 most valuable assistance in a variety of ways. We can at any time 

 address them on special questions, and feel fairly sure of interested 

 attention to our requests : for example, in a recent attempt to trace 

 out the limits of an earthquake in southern New Hampshire, I have 

 had much assistance from our volunteer observers of thunder-storms. 

 At the same time, we may feel that the attention thus awakened to 

 matters of a scientific nature constitutes one of the results that we 

 desire to reach ; for in the constitution of our society its object is 

 announced to be " the cultivation of meteorological science in New 

 England." 



Discussion of Observations. — Until the systematic observations were 

 begun in France, most investigations of thunder-storms were devoted 

 to studying their electrical action, to the neglect of their mechanism. 

 This was much as if a physicist should carefully examine the pecu- 

 liarities of electric sparks, but neglect to study the construction of the 

 machine that produced them. There were, of course, exceptions to 

 this practice, but they were distinctly exceptions. Since the estab- 

 lishment of systematic observations, it has generally been customary to 

 present their results in the form of averages of occurrence according 

 to months, hours, and places ; also in averages of direction and velocity 

 of motion, and attitude with respect to the neighboring and controlling 

 centre of low pressure ; and again in maps that represent the path 

 of the storm across the country. But while these statements are of 

 much value and interest, they generally consider the storm only as a 

 whole. More detailed study is required to determine the relation of 

 the different parts of the storm, the mechanism of its winds, the dis- 

 tribution of its temperature, pressures, clouds, and electrical phenom- 

 ena. Some may, however, say, on first approaching this question. Do 

 thunder-storms possess sufficiently persistent features to admit of a 

 regular classification of their parts? The results of European obser- 

 vations — and especially of those discussed by Dr. Ferrari of Rome — 

 answer this question very definitely in the affirmative. We know 

 that cyclonic storms have been found susceptible of detailed dissec- 

 tion, and that their several parts appear with much ri'gularity in 



