TRANSACTIONS OF THE SECTIONS. 43 



subsidiary communications to the central stations of Paris and London. Authority 

 being thus given to collect and communicate, by the telegraph, particular meteoro- 

 logical intelligence, a commencement may be made on the 1st of September, as the 

 plan proposed is simple, and the machinery is ready. Once a day, at about nine 

 a.m., barometer and thermometer heights, state of weather, and direction of wind 

 will be telegraphed to London, from the most distant ends of our longest wires, — 

 namely, Aberdeen, Berwick, Hull, Yarmouth, Dover, Portsmouth, Jersey, Plymouth, 

 Penzance, Cork, Galway, Londonderry, and Greenock. Facts sent thus from five of 

 these places, will be put into one telegram, and sent to Paris immediately, when a 

 corresponding communication will be made from the southward Atlantic coasts. 

 When threatening signs are not apparent, no further notice will be transmitted to 

 or from London on that day, respecting weather. But when indications are such 

 as to warrant some cautionary signal at a certain part of, or along all our coasts, 

 the words "Caution, — North" (or "South") will be sent to some of the thirteen 

 places specified, or to all of them, on the receipt of which a cone (or triangle) will 

 be hoisted at a staff (point up for north, down for south), indicating the side whence 

 wind may be expected. This signal will be repeated along part of the coast by the 

 Coast Guard, at such of their stations as may be authorized (at most of their stations 

 flagstaff's are visible to coasters). Danger will be implied by a drum (or square), a 

 cone, and perhaps, in addition, very great danger by a cone, a drum, and a second 

 cone. [The cones and drums may be made with hoops and black canvas, to collapse, 

 without top or bottom. They will be the same in shape from all points of view, 

 and unlike any other signal, such as a time-ball, used ordinarily.] As the Coast 

 Guard extends all along the frequented parts of our shores, and as the telegraph 

 companies are liberally willing to have instruments and signals placed at their ex- 

 treme stations, in charge of and used by their officials, only the necessary materials 

 and instructions will be required, all of which are ready or in progress. By vigilance 

 at the central station, and by taking great care to avoid signalling too frequently, 

 much may be done towards diminishing the losses of life on our increasingly crowded 

 coasts. Property alone may be duly insured, but every wise precaution for the safety 

 of life should, of course, be used. As an auxiliary measure, a concise Manual of 

 Instructions for the Barometer will be circulated among maritime communities; 

 who, though they may have frequent access to " weather-glasses " of various kinds, 

 do not generally know how to use them most advantageously. The following details 

 may be useful, as well as interesting, to those who wish to investigate these subjects 

 and examine the diagrams more critically : — The probable limits of error of the 

 barometric curves on the synoptic sheets, 21st of October— 2nd of November 1859. 

 The observations at the regular observatories, such as Greenwich, Oxford, Cam- 

 bridge, Highfield House, Kew, &c, have had all corrections applied, and have been 

 reduced to sea-level, and the temperature of 32°. The returns from members of the 

 British and Scottish Meteorological Societies (neatly ninety in number) have nearly 

 all been corrected for the exact height above sea-level, all within a few feet. The 

 corrections due to instrumental errors and reduction to 32° have (in most cases I 

 believe) been applied by the observers. The Continental observations have been 

 collected partly from the Dutch papers and partly from the ' Moniteur.' Those from 

 the former have been reduced to 32°, and, it may be presumed, have also been cor- 

 rected for instrumental errors. The heights of some stations are known ; the cor- 

 rections due to those heights have been applied, and others are known to be little, if 

 at all, above the sea-level. Any error in laying down a curve from such data can 

 scaicely exceed two or three hundredths of an inch. The observations obtained 

 from the ' Moniteur ' it is assumed are given duly corrected. The heights of the stations 

 of ordinary observers are known for the most part pretty nearly, and corrections for 

 such heights have been applied to the returns. Other corrections have only been 

 applied in a few cases — observations sometimes recorded only to the nearest tenth, not 

 being deemed worthy of any further correction. Those returns, however, of which 

 the barometrical observations are evidently erroneous (from comparison with other 

 more reliable neighbouring and contemporaneous obsevations), have been rejected 

 altogether. On the whole, we may safely assume that even these observations, 

 as laid down, are less than a tenth in error. The heights of the lantern above the 

 sea-level and of the tower, from the base to the vane, being known, the probable 

 height of the barometer can be ascertained. The proper correction for the height 



