196 Report of the Meteorological Committee . [Jan. 20, 



II. Weathee Telegeaphy. 



The first action prior to the resumption of the issue of storm-warnings 

 was the inspection of the stations. 



This has been annually carried out ever since. Much that was objec- 

 tionable has been gradually improved, and at present 16 out of the 29 

 stations are provided with the Stevenson's thermometer-screen. 



Observers have been changed from time to time, so that at present 

 only 12 are telegraph clerks, and all are distinctly and immediately 

 responsible to the Office. 



The Office may fairly claim for its service a higher degree of scientific 

 accuracy and completeness than exists in any at present in operation in 

 Europe. It must always be remembered that, as a rule, telegraphic 

 stations are not likely to be good stations for general climatology, the 

 conditions which determine the choice of locality being widely different 

 in the two cases. It seems therefore less incumbent on the Office to 

 publish means for temperature, &c. for these stations than would appear 

 to have been held by some authorities, to judge from the evidence on the 

 subject laid before the recent Science Commission. 



A difficulty in dealing with Weather Telegraphy is to be found in the 

 frequency of telegraphic errors, which renders any absolute dependence 

 on figures received by telegraph impossible. Some idea of the extent to 

 which this evil affects the scientific prosecution of weather study and 

 interferes with the formation of correct views of the essential facts with 

 the promptitude that is required for the practical application of the 

 deduced results — the issue of warnings to the coasts — may be gathered 

 from the fact that in the case of one single station, Oxford, 49 errors 

 were detected (on receipt of the original MS. messages) in the space of 

 18 months, which were all proved to be due to inaccurate transmission, 

 and were in addition to a considerable number which had been discovered 

 at first inspection of the telegrams (owing to the glaring discrepancy of 

 the reports from those of adjacent stations), and had been corrected by 

 repetition of the message. 



This number gives about 32 errors per annum ; so that on this hypo- 

 thesis there would be 32 x 29 or 928 errors from British stations coming 

 in every year, which it is apparently impossible to detect by simple 

 inspection of the telegrams. 



Of errors in barometrical and thermometrical readings as received by 

 telegraph and suspected from their discrepancy inter se, the Office dis- 

 covers more than 1000 every year, frequently eliciting a correction by 

 repetition of the telegram and correspondence with the observers. 



The following memorandum shows the present condition of the Weather 

 Service : — 



