Errors in Arnott's Elements of Physics. 
323 
1829.] 
in.-Notice of Errors in Arnott’s Elements of Physics, 3rd Ed. — 1828. 
There cannot be a greater service conferred on the student or general reader, than 
the detection of errors in systems, or books of authority. Such works are almost 
always looked to as oracles, by those who are not masters of the subject they treat 
of ; and in proportion to the reputation of the author, or the celebr ity of the book, is 
the mischief which may be occasioned by their mistakes, their partial views, or false 
reasoning. In works on mathematical science there is less room for these correc- 
tions and criticisms, owing to the simplicity and small number of the general prin- 
ciples, which again, are such as are intuitively known to be true, and are always ap- 
pealed to with the greatest confidence. But when observation forms part of the 
ground work,— when we have to build on the testimony, rather than the reason of man, 
the ease is altered. In chemistry, in physics, in all experimental science, we have 
daily proof how little safeguard the greatest names afford against the intrusion of 
error. So, when we find the name of Davy sanctioning the opinion, that the air in 
a barometer proceeds not from the tube, but from the mercury ; when we find Danniel 
unwarily asserting, and afterwards endeavouring to maintain, that, in correcting for 
the expansion of mercury in the barometer, we should allow for the expansion in 
the glass tube ; when we find Thomson offering a theory of latent heat, the result 
of which would be, that to raise water to a higher temperature, we should lower the 
fire ; when we find Ure giving us as the measure of the heat contained in a body, 
the number of degrees which its temperature maybe above 32% as if there Were no 
heat at that point ; when we find these and similar errors, all propped up at one 
time or another, by the greatest and most illustrious names in science, we may well 
have distrust of humbler men, and satisfy ourselves of the danger of trusting too 
implicitly to these miscalled systems. We are convinced that a very useful book 
might be made, by collecting the mistakes and paralogisms which are found, not 
only in the writings of these system-mongers, but even in those of the first philoso- 
phers of the day. We do not speak of those grosser errors, in which eagerness to 
make a discovery, outruns discretion, and which form the great justification the un- 
initiated have, iii throwing ridicule on science and on scientific men. We do not 
mean such dogmas as “ all nature alive for these are ephemera] errors, which are 
horn one day and die the next. We allude rather to those mistakes in experiment, 
ing, arising either from defective observation, from pnrsiiinga wrong course, from the 
warping of the mind by preconceived notions, or from that mixture of speculation 
with observation so common with many men, which would always lead them rather 
to guess an answer, than interrogate nature for one. If to these sources of error, to 
which the greatest men are more or less prone, we add those arising from halt know- 
ledge, we shall be persuaded that there is much in our modern books deserving ot 
correction, and that a list of their errors would be longer than, perhaps, their authors 
These observations have been prompted by the following communication from one 
of our correspondents, noticing some errors in Arnolt’s Physics, a work which though 
but recently published, has already reached a third edition. It is highly spoken 
of, both by reviewers and readers ; our correspondent is, therefore, the more en- 
titled to their thanks in noticing them, as well as to those ot the author, who will, 
doubtless, take the opportunity of another edition being called for, to correct them. 
We must add, that we have not seen thework ourselves, andcannot judge, therefore, 
whether its general character is deserved. Our remarks have been perfectly general, 
and had no reference, whatever, to this work in particular. 
p 2 >. The author assumes as an established fact, that sulphuret of mercury con- 
sists of equal numbers of atoms of sulphur and of mercury ; without explaining the 
reasons which have led chemists to infer the relative numbers of atoms in any suh- 
St 33. e 'He countenances the abandoned scheme of strengthening wood, by conden- 
39 / He states that Caoutchouc becomes permanently elongated by too much 
stretching; unaware of the recently established fact, that its loss of contractility 
is occasioned by the loss of latent heat, and capable ol being restored by heating. 
51. States that a cannon bull, or a shrapnel, if burst into fragments during their 
flight, would still retain their previous velocity ; although it is clear, that it must he 
speedily diminished by the increased resistance of the air, to the increased surfaces. 
64. A new theory of the spinning top, according to which, its line of rotation 
ought always to be perpendicular to the surface on which the apex turns, whether 
that surface be horizontal or not ! 
