188 EIGHTH REPORT — 1838. 



If we compare the mean value of u derived from the Irish 

 series, — 34° 6' (varying in the several partial results from 



- 31° 20' to -38° 00'), with its mean values in England and 

 Scotland — 50°, (the partial results varying from — 40° 38' to 



— 55° 46'), we find, notwithstanding the amount of the partial 

 differences, a general and consistent indication that the isody- 

 namic lines are less inclined to the meridian in Ireland than in 

 Great Britain, The two Irish series which give the least values 

 for this angle, are those which were the earliest obtained, — 

 which had consequently the disadvantages of less experience in 

 the observers, and less perfection in the instruments ; and of 

 combining in one series observations at different epochs, and 

 results by different observers, and with different instruments. 

 The two series of Captain Ross were, on the other hand, ob- 

 tained by one observer with the same instruments ; were well 

 distributed over the country ; and were made in immediate 

 and rapid succession. We may therefore safely infer, as Mr. 

 Lloyd has done (pages 184, 185), that the values of u derived 

 from Captain Ross's series are entitled to weight beyond the 

 proportion which the number of the stations wliich they repre- 

 sent bears to the number of stations in the other Irish series. 

 Still the difference in the angle with the meridian in Ireland and 

 in Great Britain cannot, in any consistency Avith the observa- 

 tions, be less than several degrees. I have employed — 35°, 

 the value deduced by Mr. Lloyd, pages 184 and 185, as the 

 general mean value of u in the Irish deductions. 



If we compare generally the mean results of the horizontal 

 with those of the statical series, we are not able to discover any 

 ap])arent systematic differences whatever in regard to the values 

 of u and r. The individual observations by the horizontal me- 

 thod do indeed exhibit much greater discordances with each 

 other than is the case in the statical method. This has been 

 already shown in detail in the analysis of the observations by 

 the two methods in Scotland, in pages 20, 21, of the Sixth 

 Report of the British Association : and Mr. Lloyd has else- 

 where pointed out the causes of the advantage in this re- 

 spect of the method for which we are indebted to him. Al- 

 though, therefore, the accordance of the two methods, when 

 the observations are grouped, is a satisfactory confirmation of 

 the conclusions which they unite in estabhshing, the horizontal 

 observations are less fitted than the statical to be employed in 

 a graphical representation of the particular nature adopted in 

 this report, in which the discordances of individual observa- 

 tions are brought strongly into notice, and if exceeding a cer- 

 tain limit might produce inconvenience, by in some degree 

 perplexing the judgment. In extreme cases they might entirely 

 mislead it; as, for example, if the point furnished by an obser- 

 vation for a particular line should fall nearer to an adjacent 



