Considerations on the Prolilom of Latitvde Variation. 



Although systcinatically different values are obtained from 

 tlie different elongations on the different assumptions, the mean 

 value from hoth elongations on the first assumption is nearly equal 

 to that from both elongations on the second assumption. ï:^o tliese 

 mean values may be looked upon as giving approximately the real 

 value of the micrometer. 



An assumption, to hold good, must l)e of sut-h a nature as to 

 yield theoretically the same value from both elongations. From 

 this point of view both of the above assum])tions are to be rejected, 

 and Ave must find an intermediate one which will bring Ijoth elon- 

 gations into conformity. 



It is manifest that such a reduction, provided its validity be 

 detinitely granted, Avill need a diminishing factor less than unit}^ 

 In order to determine the diminishing factor I made a complete 

 reduction of my series of elongation observations, forming the 

 equations of condition in the following form, — 



A + Bt + CO = a + Db 



