76 On the Rates of Chronometers. [No. 1. 



where her course again lies more or less athwart it to England, like the 

 homeward bound vessels of which we have already spoken. 



If we allow any influence at all to terrestrial magnetism, the error of 

 thirty-three seconds which Captain Fitzroy describes does not at all 

 seem excessive or surprizing ; nor again, that while amongst our own, 

 or with our own and foreign navigators, many admirable coincidences in 

 Chronometric measurements are to be found, some hitherto unaccount- 

 able discrepancies, from which some discussion and ink-shed have 

 arisen, should also exist. 



It is clear, I think, that, wholly apart from the ship's local attraction, 

 and all the precautions which science can devise, the agreement or dis- 

 cordance of any two sets of Chronometric measurements, even by the 

 same Chronometers and observers, may depend upon the ship's track ; 

 upon the position of the "XII. VI. line (or other polar line) of the ba- 

 lances of the Chronometers in relation to the keel ; and all this again 

 upon the degree of polarization of the balances ! Here are surely the 

 elements of a great and delicate scientific investigation yet to be made ?* 



It would seem then to result from the foregoing facts and views, 

 though writing in Calcutta I have been unable to consult a host of au- 

 thorities to which I should have been desirous of referring, such as 

 Gauss, Sabine, Duperrey, Blosseville, &c. that temperature is by no 

 means " the principal cause of the variation of the rates of Chronome- 

 ters" and indeed we have of late years had some extensive experiments 

 made to prove that Chronometers may undergo great variations of tem- 

 perature without any considerable change of rate, though to these also 



* It should be made by a double Chronometric voyage ; one ship proceeding 

 East and another West. Both should rate their Chronometers, specially and 

 independently of all other rating, as near as may be to the Magnetic Nodes (say at 

 Bahia and Manila which are about 12h. apart), and while measuring their chain of 

 distances should particularly endeavour to ascertain, at various spots, the effect of 

 the placing of the XII. — VI. or polarized line of the balances coinciding with, or 

 athwart, and at various angles to the Magnetic meridian. Perhaps part of the 

 Spitzbergen variations recorded by Mr. Fisher, (page 62) may have been due also to 

 this cause, and if the Chronometers had been placed in the Magnetic meridian they 

 would have given different results. He evidently overlooks the terrestrial magnet- 

 ism and attributes the change of rate to the absence of the ship's local attraction 

 only. 



