, b o THE POPULAR SCIENCE MONTHLY. 



halos are white, we may conclude that the particles are frozen, and ex- 

 pect cold weather. Crossed halos, mock moons, or highly-developed 

 phenomena, indicate larger crystals of ice, and probably frost, hail, 

 snow, or heavy rain, after three or four days, according to the season 

 of the year. Similarly the laws of reflection of light indicate that 

 the cause of a deep-purple morning or evening sky is the large 

 amount of moisture present in the atmosphere. Another effect of the 

 moon, when at the full, is to clear the sky of cloud, traceable, says Sir 

 John Ilerschel, to a distinct physical cause, the warmth radiated from 

 its highly-heated surface ; though, why the effect should not continue 

 for several nights after the full, remains, in the opinion of the same 

 accurate observer, problematic. Other lunar prognostics, founded on 

 arbitrary rules, as to the time of the day or night at which the changes 

 or quarterings take place, are worse than useless, for they are calcu- 

 lated to mislead, and are generally included in almanacs or note-books 

 intended for sale only, being in some cases attributed to an eminent 

 meteorologist or astronomer Sir W. Herschel or others. 



It is of course far from our purpose to enter here into a disquisition 

 on the theory of the trade and anti-trade winds, and their barometri- 

 cal indications subjects that can be usefully discussed only in a trea- 

 tise on meteorology : we limit ourselves to the present position of 

 weather prognostics, although it must be admitted that any advance 

 yet made or likely to be made in prognosticating the weather arises 

 from the study of such recurrent phenomena, the investigation being 

 much aided by the highly-developed character of the laws of the ex- 

 pansion of gases, upon which laws the theory of the wind is founded. 

 Thus we know that a rise in the barometer, together with a fall in tem- 

 perature, as shown by the thermometer, indicates the approach of a 

 cold, northerly current of air ; while a fall of the mercury in the ba- 

 rometer, with a rise of that in the thermometer, indicates that a south- 

 erly or warm air-current is on its way. Northerly currents may in- 

 clude winds from the northwest and northeast, as well as from the 

 north ; similarly, southerly currents may include winds from the south, 

 southeast, or southwest. When the barometer rises while a northeast 

 wind is blowing, with prevalent hail, rain, or snow, there may be no 

 change. Of barometrical indications alone, it is generally known that 

 a rapid rise portends changeable weather; a slow rise, the contrary ; 

 a rapid fall, heavy wind, rain, and snow ; while a fluctuating height of 

 the column of mercury indicates unsteady weather. With a heavy gale 

 of wind in the east or southeast, changing south, the barometrical col- 

 umn may fall until the wind shifts its quarter. Upon such observa- 

 tions did Admiral Fitzroy base his code of instructions, now to be 

 found by the side of every barometer, his forecasts depending on the 

 indications of the barometer and thermometer, with observations as to 

 the direction and force of the wind with regard to time and place, and 

 its previous course taken altogether. These indications are thus not 



