1094 A Seventh Memoir on the Law of Storms in India. [No. 131. 



be, I think, no stronger proof than this of the utility of every man's 

 notes and remarks, how little important soever he may imagine them 

 to be, and (a notion which I fear yet prevails sufficiently to deter 

 many from assisting us, though well disposed to do so,) the value of 

 every plain common sense account of the weather, whether within or 

 without the limits of a storm, I repeat here again, that all observa- 

 tions are of value to us; that scientific ones are of course the best, but 

 that the plain ones are often quite as useful at certain points, and 

 that as our science is as yet in its infancy, our main business is to 

 collect and register evidence. It will be observed, and it is a striking 

 proof of what I have just said, that two of the most important deduc- 

 tions we have obtained in this investigation, have been proved by three 

 simple unscientific statements of this kind. Mr. Martin of Sooraje- 

 gunge factory, and Mr. A. Pinard of Kunjirpore factory near Bhaugul- 

 pore, with the Newspaper notice from Jungypore, the author of which 

 I do not know. From these we have been enabled fully to establish 

 the highly curious fact, that a violent hurricane, with a strong mon- 

 soon setting in at about right angles to its course, sometimes breaks 

 up into several smaller storms, none of them equalling the main one 

 in fury, but all obeying the law of rotation. These reports also 

 explain a phenomenon, the cause of which we suspected before, but 

 of which we had no distinct proofs. I mean those instances in which 

 what is usually described both at sea and on shore as "wind con- 

 tinually veering" takes place, though we have full evidence, that the 

 places are at a distance from the centre of the storm, and this again 

 leads us to the strongest confirmation of Col. Reid's theory of the 

 variable winds. I trust it is unnecessary after such instances to repeat, 

 that all observations may be of use. 





