Experiments on the Velocity of Sound in Air. 447 
sary, such that a very small deformation corresponds to a very 
great expenditure of energy. 
(14) To deal with communications of energy from bodies 
outside the system, all we need do is to include them in the 
system. Treat as before the whole system thus increased, and 
then consider only the motion of the original parts of the 
system. This method applies with perfect generality whether 
the external masses be themselves free, constrained, or resisted. 
(15) Another method of applying the same principles is 
then given. Starting from the definition dK=Xm$pclp, the 
kinematical properties of A are developed. Then, by the help 
of § 2, these are exhibited in their physical translations. 
(16) The paper concludes with a brief comparison of the 
fundamental principles of the science as they have been intro- 
duced by Newton, Lagrange, Hamilton, Peirce, Kirchhoff, 
and Clerk-Maxwell respectively, and also as they appear in 
the unique Vortex-system of Thomson. 
LXV. Experiments on the Velocity of Sound in Air. 
By J). J. Blaikley*. 
THE method of experiment which I venture to bring before 
you this afternoon is the outcome of various attempts 
made by me to determine with greater accuracy than had 
hitherto been done the velocity of sound in small tubes, such 
as are used for musical instruments ; in addition to which 
practical purpose the idea presented itself to my mind that, if 
a series of tubes were taken, having their diameters in a 
definite ratio, the observed results might by calculation be 
extended so as to include a value for a tube of infinite dia- 
meter, that is for free air. Some of the results were brought 
before the Musical Association in June last ; but the experi- 
ments have since been repeated with greater accuracy. 
It will not be necessary to take up your time by referring 
to more than a few of the many determinations that have been 
made. A useful summary was given by Mr. Bosanquet in the 
' Philosophical Magazine ' for April 1877; but we may note 
that only those observations in which corrections both for 
temperature and for moisture have been made can be con- 
sidered at all accurate. 
Such corrections being made, there remained at the time of 
Ptegnault's great series of experimentsf considerable differences 
in the results arrived at by different observers, partly due, 
* Communicated by the Physical Society. Read November 10, 1883. 
t Memoires de V Academie des Sciences, tome xxxvii. 
