540 



known as the "quadrantal" deviation, shows, that while in wood- 

 built steam-ships it seldom exceeds 1 or 1|, it rises in iron-built 

 ships from 1| to 6 and 7; the Liverpool Compass- Committee 

 recording even a point of the compass. 



The chief characteristics of the quadrantal deviation, as developed 

 in this investigation, are 



1. That it has invariably a positive sign, causing an easterly 

 deviation in the N.E. and S.W. quadrants ; and a westerly deviation 

 in the S.E. and N.W. quadrants. 



2. Its amount does not appear to depend on the size, or mass of 

 the vessel, or direction when building ; or on the existence of iron 

 beams. 



3. That a gradual decrease in amount has occurred, after the 

 lapse of a number of years, in nearly every vessel that has been 

 observed. 



4. That the value remains unchanged in sign and amount, on 

 changes of geographic position. 



5. That a value not exceeding 4, and ranging between that 

 amount and 2, may be assumed to represent the average or normal 

 amount in vessels of all sizes. 



Numerous examples are given in support of these propositions, as 

 also of the uniformity of the amount of quadrantal deviation when 

 determined in various parts of the ship ; and, assuming the normal 

 amount in iron steam-ships as from 2 to 4, an analysis is given by 

 which it is seen that 75 per cent, of the iron ships of the Royal Navy 

 are included in this condition. 



Two questions of importance here arise ; are the results of this ana- 

 lysis conclusive, and if so, under what conditions do large quadrantal 

 deviations occur? Reverting to the Astronomer Royal's early experi- 

 ments in 1838-39, in the iron ships ' Rainbow' and * Ironsides,' whose 

 values were very small, and presuming that those vessels were built 

 of good material from their then experimental character as also 

 that similar conditions of material of good quality exist in the iron 

 ships of the Royal Navy, it is assumed that the value (2 to 4) re- 

 presents the average condition of a ship built of the best or superior 

 iron. 



On the other hand, can the inference be drawn that large quadrantal 

 deviation in an iron ship implies that inferior material has been used 



