i8 9 



serious, as some of our worst storms occasionally give 

 very slight indications of their proximity by instrumental 

 readings. 



I have represented on the screen a recent instance of 

 this, the storm of October 24th, 1882. You will see that 

 the map for 6 P.M. on the evening before showed, in its 

 barometrical indications and its wind reports, not a sign of 

 the disturbance which lay over the west of England at 8 

 o'clock in the morning. 



Similar instances might easily be cited, and others 

 arising from miscalculation of the path which the storm 

 was following, but the time would fail me were I to 

 continue the subject further. 



There is, however, one point to which I ought to allude 

 before I have done. It may seem strange that when a 

 storm made its appearance in the night time, like that of 

 October 24th just cited, I should have said nothing about 

 the possibility of receiving intimation of it by means of 

 later, say midnight reports, and distributing the news to 

 the coast stations at once. 



The explanation of not doing so is, that though we 

 might organise night reports, the only portion of the popu- 

 lation who could benefit by them would be the readers of 

 newspapers. We could send forecasts to the press at a 

 later hour than 9 P.M., but we could not send warnings to 

 the fishermen on the coast ; and it is for them in the first 

 instance that the warnings are meant. The local telegraph 

 stations close at 8 P.M. at latest, and so the messages could 

 not possibly reach the coasts before about 9 A.M. next day, 

 when, as will be seen by the map, they would have been 

 too late for the coasts of Devon and Cornwall. 



The development of our storm-warning system is regu- 

 lated mainly by the completeness of our telegraphic arrange- 



