192 Pedler and Crombie— On tlie Tornado which [No. 2, 



Dacca tornado under discussion, an ordinary nor'wesfcer was actually 

 in progress when the tornado suddenly appeared in close proximity to, 

 if not in actual connection with, the storm, but still quite distinct from 

 it, and moving in its own particular path. This would apparently point 

 to the fact that the general conditions which produce nor'westers, or 

 perhaps dust-storms, may with the addition of some cause or causes, 

 possibly almost accidentally present, also generate tornadoes of a violent 

 type. What such additional circumstances or causes are which deter- 

 mine the actual formation of the tornado, it is impossible at present to 

 say, but it is fortunate for India that the combination of circumstances 

 required is rare. 



The conditions as to great contrasts of temperature and humidity 

 previously referred to, and the passage of areas of barometric minima 

 may be looked upon as predisposing causes only, but they clearly cannot 

 be classed as proximate causes, for such predisposing conditions obtain 

 very frequently indeed in India, and it is rarely that they are followed 

 by tornadoes. 



Tornadoes are also found in America to be formed frequently 

 during still or almost calm weather, and there are accounts of observers 

 having actually seen the formation and growth of a tornado taking place 

 on an almost perfectly calm but sultry day.* This would perhaps point 

 to some such action as described before in Sir George Airy's words 

 being the possible explanation of their formation ; and, if such is the 

 case, any forecast of their probability from any regular or definite series 

 of meteorological conditions would become impracticable, for it is mani- 

 festly impossible to say when any sudden uprush of heated air may take 

 place over any large and highly heated area, or at what point such action 

 might occur. 



The above statements may perhaps be said to be confessions of an 

 ignorance more or less complete of the subject of tornado formation, but 

 such only are the facts at present known. 



Meteorological Conditions in Bengal previous to the Tornado. 

 — The meteorological conditions of the earlier parts of the year do 

 not call for any particular comment. In fact, the phenomena of the 

 formation and existence of a tornado are of such a brief and transitory 

 nature that it would be useless to seek for anything like proximate or 

 even predisposing causes in the meteorology of the previous months 

 of the year ; so that an extremely brief record will saffice. 



January and February were months of the normal cold weather 

 type in Bengal, and in March the usual rapid increase of temperature 



* H. S. Whitfield, Tornadoes in the Southern States, American Journal of 

 Science, 3rd Series, Vol. II, p. 96 and others. 



