GALES OF WIND AND APPEARANCE OF THE AURORA. 461 



pare simultaneously the records of tlie weather from day to day tbrough- 

 out all the uortheru and southern temperate zones, and to ascertain 

 whether the maximum of these changes have any fixed relation in time 

 to the changes of the moon. The fact that the problem has not been 

 considered from this point of view, may account for the failure, in the 

 study of a series of records at a single place, to furnish evidence of the 

 action of the moon. 



The changes of the moon take place at a given moment on every part 

 of the earth; the greatest efitect of a lunar tide ought, tlierefore, to be felt 

 in succession entirely around the earth in the course of about twenty- 

 four and one-half hours. 



The i»roblem, however, has not been solved and cannot be determined 

 by such casual observations as those which you narrate. I have not 

 the least idea that the attraction of Venus produces any a])preciable 

 effect. It is too small to produce a result which would be indicated by 

 any of our meterological instruments. 



I am far from subscribing to the justice of your remarks in regard to 

 Mr. Espy, since I have a great respect for his scientific character, not- 

 withstanding his abberation, in a practical i)oint of view, as to the 

 economical production of rain. The fact has been abundantly proved 

 by observation that a larg'e fire sometimes produces an overturn in the 

 unstable equilibrium of the atmosphere and gives rise to the beginning 

 of a violent storm, but it was not wise in him to insist on the possibility 

 of turning this principle to an economical use. — [J. 11.] 



CONx\ECTIOX OF GAIES OF WIND AND APPEARANCE OF THE AURORA, 



By R. T. Knight, of Piuladklphia. 



"An officer of the British navy states that from eleven years' observa- 

 tion, six years in the Arctic regions and five years in the north of Scot- 

 land, he has ascertained that tremendous gales follow from twelve to 

 twenty -four hours after the appearance of the aurora borealis.'' I never 

 thought proper to call your attention to the above extract from the Phil- 

 adelphia Ledger of the 4th instant, because it agrees with wh;it I pub- 

 lished in 1804, and also in 18G5. 



Eemarks. — We have had frequent communications from observers 

 suggesting a connection in the time of the appearance of the aurora 

 borealis and the occurrence of storms of wind and other meteorological 

 phenomena; but on referring to our records we have never been able to 

 verify the existence of such connection. On the receipt of the foregoing 

 communication the records of the Institution were examined in relation 

 to this subject, with the following results : 



1. From the log-book of the brig Advance, Haven's Arctic expe- 

 dition, forty-six appearances ot the aurora were followed b}^ four storms. 



