40 BappickKER—On the Influence of Magnetism on the Rate of a Chronometer. 
festera; mais si ce pole se trouve en dessous, c’est-’-dire vers la partie concave de 
la section du spiral, la force magnétique tendra au contraire & rapprocher cette 
section de la ligne droite: allongement dans la projection horizontale du ressort et 
retard dans la marche; mais l’action magnétique continuant, le spiral dépassera 
cette position normale et se déformera en sens contraire, et alors l’accélération se 
manifestera.” 
The above passage is not very intelligible, since the basis of Ansart-Deusy’s 
deductions is not in accordance with the facts. For firstly, he obviously thinks 
throughout of a flat spiral only, while all the box, and almost all the pocket 
chronometers have, on account of the easier adjustment for isochronism, cylindri- 
cal spiral springs. And further, his considerations apply only to the ideal vertical 
section of the spiral, by which the single parts of the spring, though in reality 
curved in the plane of the spiral, appear as straight lines; and, naturally, the 
results obtained for the sections (or projection) do not necessarily hold good for 
the spring itself. And the action which he assumes to take place would only be 
possible if both ends of the sections of the spring (- ) were fixed, which is 
certainly not the case. But even if all these objections did not exist, a very strong 
magnet indeed would be required to produce any deformation of a spring, even when 
acting in the direction of the greatest elasticity, ¢.e. vertically to the flat side; and 
it is not probable that the rigidity of the spring would admit of any deformation 
when the magnet acted on its edge. We see, therefore, that the whole deduction 
of Ansart-Deusy is totally erroneous, and, since it forms the point of issue for all 
his speculations, it will hardly be necessary to reproduce them. Suffice it to say, 
that all cases of observed acceleration serve him as proofs of his theory, and that 
also the accidental variations, such as sudden changes of the chronometer’s correc- 
tion without a change of rate, are according to him due to magnetic action. Such 
irregularities, which generally will find their explanation in a fault of the mecha- 
nism,* as perhaps in a weakness of the escapement spring, by which more teeth of 
the escapement wheel than one slip by at a time (‘‘ tripping”), are, for instance, 
observed by Mouchez (No. 24), and, significantly enough, in chronometers of 
one maker (Winnerln) only. And since Ansart-Deusy imagines by Mouchez’s 
Tables, that these chronometers are less subject to the influence of temperature, 
and the reverse in the other instruments, he is led to the strange remark :— 
‘‘ Est ce que dans certains instruments le compensateur agirait 4 bord comme com- 
pensateur magnétique et que dans d’autres chronométres il agirait comme compen- 
sateur thermométrique seulement ?” which is hardly compatible with a thorough 
knowledge of a chronometer’s mechanism. And further on, though he knows and 
criticises Fisher’s (No. 5) and Barlow’s (No. 6) Papers, still their results do not 
affect his theory at all—The experiment which he communicates illustrates this 
* Similar remark by A. Ledieu (No. 80, on p. 70). 
