METEOROLOGY. 



581 



converter in use is only 1-4 metre high and one 

 metre in diameter. The holes from the tuyeres 

 cover an area equal to that of a circle 200 mil- 

 limetres in diameter. There are ninety holes, 

 from about 3 to 3 millimetres in diameter, 

 and the area is so calculated that after the 

 blast has passed through the metal there is 

 little or no free oxygen left in the escaping 

 gases. The charge weighs from 170 to 765 

 kilogrammes, and only eight tenths per cent, of 

 70 per cent, ferro-manganese is used to recar- 

 bonize. The pressure of the blast is 1'04 kilo- 

 grammes to the square centimetre. The bottom 

 of the apparatus is forced into place by a screw- 

 arrangement, and can be detached and replaced. 

 Magnesite, a native carbonate of magnesia, 

 is being introduced into England as an ex- 

 tremely refractory substance for lining fur- 

 naces. It comes from Styria, Austria, where 

 large deposits have recently been discovered. 

 It is chiefly employed, after being dead- 

 burned, in the form of magnesia-stone for fur- 

 nace-bottoms. Furnaces lined with tiles of 

 this stone can be worked three turns in twenty- 

 four hours, and continuously for six or seven 

 months; and so powerful is the magnesite 

 that iron containing as much as 3 per cent, of 

 phosphorus can be effectually treated in such 

 furnaces, and made to turn out a material 

 equal to the best Swedish steel. 



METEOROLOGY. Under a system of observa- 

 tions organized by Lieut. John P. Finley, of the 

 United States Signal Service, a great advance . 

 has been made in our knowledge of the char- 

 acter and movements of tornadoes, and it has 

 become possible to predict with some approach 

 to accuracy, a few hours beforehand, when 

 any region in the settled parts of the coun- 

 try is about to be visited by a storm of this 

 character. Lieut. Finley adopted the practice, 

 whenever a tornado occurred, of sending to 

 postmasters or other officers, to be given to 

 competent persons who would supply the in- 

 formation requested, blanks, with questions re- 

 specting the prominent features of the phe- 

 nomena, and the more important circumstances 

 attending their appearance. The Signal-Serv- 

 ice Office has now upon its books the names 

 of more than 1,500 observers east of the Rocky 



; mountains, the most of them in the States 

 which are most frequently visited by destruc- 

 tive tornadoes. Lieut. Finley has divided the 

 country east of the Rocky mountains into 



> 18 districts, for which he makes predictions 

 with a degree of accuracy which has steadily 

 increased since they were begun. During 

 1884, of 3,228 predictions unfavorable to tor- 

 nadoes, which were made as assurances of 



, safety to the people of the several districts, 

 3.201 received positive verification. A more 

 trying test is that of actual success in predict- 



f ing tornadoes, and in this certainty is lessened 

 by the fact that with the present limited re- 

 sources of the Signal Service the reports are 



:. defective or wholly wanting from sparsely set- 



. tied regions. Lieut. Finley found that of 38 



predictions that tornadoes would occur, made 

 in April and June, 1884, 18 were verified, and 

 of 19 made in Juno and July, 1885, 15 were 

 generally verified. In all cases there were 

 violent storms, either tornadoes, hurricanes, or 

 hail. Many predictions may seem to fail, though 

 they are really fulfilled, through the limited 

 extent of the phenomenon, which in a treeless, 

 thinly settled prairie might escape observation. 

 The conditions favorable to the development 

 of a tornado are positive and noticeable, and 

 consist, according to Lieut. Finley, generally 

 in a departure of from 15 to 50 from nor- 

 mal conditions of temperature, accompanied 

 by abnormal conditions of humidity, pressure, 

 wind-direction, and cloud formation and move- 

 ment. A remarkable and interesting feature 

 in the case is that the tornadoes nearly al- 

 ways form southeast of a moving center of 

 low pressure, and their tracks, conform close- 

 ly to the progressive direction of the main 

 storm. The average number of tornadoes oc- 

 curring yearly in the region east of the Rocky 

 mountains is more than a hundred. The 

 number in 1884 was 172. Of 600 tornadoes 

 which are recorded as having occurred be- 

 tween 1794 and 1881, 304 moved from south- 

 west to northeast, and the remainder, with 

 very few exceptions, kept close to that direc- 

 tion. So comparatively certain has this move- 

 ment been found to be, that the tornado-track 

 can be escaped by running southeast from its 

 center. It is not quite so safe to run north- 

 west, on account of the northward motion of 

 many of the storms. The average width of 

 the storm- path is 1,085 feet, and the velocity 

 of progression is 30 miles an hour. The form 

 of the cloud is nearly always funnel-shaped, 

 varied at times with the forms of the hour- 

 glass, cone, and inverted funnel, modifications 

 which are caused by different altitudes and ve- 

 locities of air-currents. It is estimated that 

 in the center of the funnel the air sometimes 

 attains a speed of 2,000 miles an hour. The 

 whirling movement is almost invariably in an 

 opposite direction from that taken by the 

 hands of a clock. Though the tornadoes pur- 

 sue a course parallel to that of the storm- 

 center, by the side of which they are formed, 

 they are developed at a distance of several 

 hundred miles from it, and keep at that dis- 

 tance. They are generally formed under con- 

 ditions of partial sunshine and cloudiness and 

 high humidity. 



The "sun-glows," or "red sunsets," which 

 were so conspicuous all over the world dur- 

 ing the winter of 1883-'84, have recurred 

 from time to time since then, with various 

 degrees of intensity and persistence. Prof. 

 Kiessling, of Hamburg, has made a series of 

 experiments in illustration of the phenomena 

 by the artificial formation of all manner of sun- 

 set effects, from which he has deduced the 

 law that " an intense purple glow, visible over 

 a considerable area, may occur when, in close 

 proximity beneath a lofty and highly attenu- 



