Blasius.] ibU pjay 6, 



rigid rules. The duties assigned to them individually were, however, 

 and still are of a very simple nature. The "observers" at the various 

 stations, at certain hours of the day, record in tabular columns the readings 

 of their meteorological instruments, and telegraph these to Washington ; 

 and since my work on " Storms"* was published in 1875, in which I called 

 especial attention to the Bureau's neglect of cloud forms, some simple 

 observations upon the clouds are added to the telegraphic reports. Upon 

 receipt of these reports at Washington, they are written down upon blank 

 maps, on which the respective stations are marked. The points of lowest 

 equal barometric pressure are then united by a line called the Isobar 

 which usually form an ellipse. In the same way the stations of five or 

 ten points of higher pressure are joined by a line. The field enclosed by 

 these somewhat concentric Isobars is the "area of low pressure," or the 

 "cyclonic storm" so called. From the results of the next reading the 

 next position of the area of low pressure is ascertained in the same way, 

 and by comparing the distance traveled with the time occupied, the 

 probable position for some hours ahead is calculated, and predictions are 

 issued. The labor is thus of a routine character 



"Areas of high barometer" are likewise noted, but these are thought to 

 bring fair weather. The barometer therefore is still the chief reliance in 

 the prediction of storms, and those storms which are distinguished as 

 "areas of low pressure" are practically all that the Signal Service is able 

 to predict. 



Now, in the report of the Chief Signal Officer for 1884, there are noted 

 as having occurred during the year 152 of these areas of low pressure, 172 

 tornadoes, 947 hailstorms, and 1745 thunder-storms, so that if every one of 

 the 152 "cyclonic storms" — the "areas of low pressure" — are correctly 

 predicted, we have some 2864 storms, of which the Bureau knew nothing 

 until after they had occurred. Relying on the barometer and on machine 

 methods, it could not be otherwise. 



Lieut. J. P. Finley, Chief of the Tornado Division of the Bureau, in 

 "Signal Service Notes, No. XII," says: "Probably if a barometer were 

 placed in the immediate track of a tornado cloud, it would not with any 

 certainty indicate the presence of the storm until the crushing winds had 

 fallen on the instrument." Indeed, although the Bureau appears to pro- 

 ceed on the old rule that has obtained predominance since the daj'sof Otto 

 von Guericke, that a falling barometer denotes an approaching storm, it 

 has long been well known that the most destructive storms often arrive 

 with a rising barometer, and this for reasons that I explained as long ago 

 as 1852. One of these storms took place at Colon, or Aspinwall, Panama, 

 December 2, 1885. The New York Herald of December 18 says : "This 

 storm was not preceded by any distinct precursory signs. The barometer 

 on the Isthmus apparently remained stationary or slowly rose during the 

 progress of the tempest. Much property, many vessels and lives were 

 lost." 



* Storms : Their Nature, Classification and La-\ys. Porter & Coates, Philadelphia, 1875. 



