1887.1 loO [Blasliis. 



of Delaware, South-eastern Pennsylvania and for New Jersey. The tor- 

 nadoes actually occurred in these States, about eight hours from the time 

 of prediction." 



Now if we ask how valuable to the owners of the houses damaged by 

 that tornado in Philadelphia was the Signal Service's prediction eight 

 hours before, that tornadoes would occur somewhere in " Delaware, 

 South-eastern Pennsylvania or New /ersey, " we cannot fail to see, I 

 think, the entire impracticability of the whole scheme. 



The Signal Service is perhaps obliged by circumstances to devote most 

 attention to those things which will show most apparent results to the 

 general public and to Congressmen who vote for the annual appropriation. 

 It is hampered too by its routine methods and its army rigidity of disci- 

 pline. It cannot however but be a cause of disappointment that whether 

 owing to these causes or others, it has added so little of scientific value to 

 the knowledge of meteorology during the twenty years of its existence. It 

 has been following the old methods in reference to which Sir William 

 Ilerschel says : " In endeavoring to interpret the weather, we are in the 

 position of a man who laears, at intervals, a few fragments of a long his- 

 torj^ related in a prosy, unmethodical manner ; a host of circumstances 

 omitted or forgotten, and tlie want of connection between the parts pre- 

 vents the hearer from obtaining possession of the entire story." And the 

 great Biot, after enumerating the efforts to advance this science, says : 

 " What has come of it? Nothing, and nothing will ever come of it. No 

 single branch of science has ever been fruitfully explored in this way." 



No, the methods followed have been wrong. Would astronomy be in 

 its present position if the great astronomers had been dependent on the 

 data furnished by ol)servations made according to arbitrary rules, and for 

 a minimum of time at one or two hours of the night, and for the most 

 part, too, by observers of meagre training and intelligence? And how 

 much would the great naturalists have learned, liad they been content to 

 send out into the fields three times a day for five minutes, and sit in their 

 closets to generalize upon the data thus obtained ? 



Tliirty five years ago, I urged that only by continuous observations could 

 we hope for anything beyond mere empirical knowledge ; that we must 

 gather up the "host of circumstances " now "omitted or forgotten," and 

 supply "the connection between tho parts."' I am not unmindful of the 

 practical difficulties that are in the way of a method of observation that 

 alone can give a continuous knowledge of a storm as it passes ; but it 

 seems to me that this may be obtained with the means now available, if 

 the meteorological organization would devote more attention to the dis- 

 covery of general laws than to the more sensational part of their duties — 

 the weather predictions, which the newspapers now make a matter of busi- 

 ness enterprise. If we know the laws, there will be little trouble about 

 the prediction. Each of us can do this for himself sufficiently well for all 

 practical purposes. 



