ALLISON — OX THE METEOROLOGY OF HALIFAX. 185 



ed in the daily papers. It will be desirable hereafter to put up 

 large maps, with variable symbols, which may indicate in a visible 

 manner the progressive changes. In addition to receiving telegraph- 

 ic intelligence from the continent at large, it would be desirable 

 for each large centre of commerce, such as Ilalifiix, St. John, &c., 

 to be surrounded, as it were, by a cordon of ports at a moderate 

 distance, from which telegrams need not be sent regularly, but only 

 when they are asked for. This supplementary arrangement had 

 better be postponed till the larger system is in operation ; but its 

 establishment should be kept in view as an important adjunct." 



Mr. Kingston gives further valuable information regarding ex- 

 pense ; but as that is, and will be, subject to modifications of 

 different kinds, I add nothing on that head just now. 



From the Meteorological office at London, G. B., I have lately 

 received the followincr useful suo:2:estions. Mr. Robert II. Scott 

 the Direct(3r there, writes to me : — 



*' I should say that it would be best to place all the stations in 

 connection with the Central office." 



" The various nations here are establishing centres of their own. 

 These exchange reports dailv, and send extra telegrams to each 

 other whenever a storm is reported. Each office then decides for 

 itself whether or not it will warn its own coast. This is the plan 

 we have introduced, and it is adopted by Holland and Norway ; 

 Sweden and Denmark are also about to take it up, and Russia will 

 probably follow suit. The French system of extensive generaliza- 

 tion gives a magnificent view of the general conditions, but does 

 not enable you to draw conclusions for local storms and weather. 

 It is also perfectly impossible for a distant central station to keep 

 its observers in check. Telegraphic errors are our bete noir, and 

 it is expensive work asking for repetitions over long lines ; conse- 

 quently the French reports are seldom if ever corrected before 

 publication, and errors of an inch sometimes appear in their charts, 

 and are never subsequently corrected. I should, therefore, say 

 that Halifax ought to warn its own coast ; speaking purely from 

 a geographical point of view. A central station in Canada can- 

 not get news of sudden changes in time. We send by post 



