SERIES A.— LATITUDES AND LONGITUDES. 25 



termine the mean annual direction of the wind, are useful in obtaining monthly 

 results, and hence the annual curve. To determine the latter, with the same 

 accuracy that we do the mean direction for the year, we need much more exten- 

 sive data; and these monthly collections serve to swell the list, and increase the 

 number of months on which the average is based. More complete series, and also 

 collections of observatioiis from additional places,^ might, no doubt, have been 

 obtained in many cases by applying directly to the observers ; but I had already 

 taxed my friends so far that I felt unwilHng to put them to anymore trouble- 

 especially as a long time must necessarily intervene between furnishin"- the data 

 and seeing any fruit of their labor. 



' See Appendix A. 



