88 
lengths are as follows, in inches: .98, .97, .95, .95, .97, .95, .95, .96. The 
mean of these is .960, corresponding to a vibration frequency of 6,820. 
The following table gives the series of readings made and the values of 
m deduced, all lengths being in feet: 
SERIES. r. R. | Temp m 
Bere ee he Mees rate 1260 1660 80° F .0010 
gee a, Soe cl ord abe ieee 1333 1699 74 0012 
eee ciobrasie SE SUHAN tb actin,» «1/2 2078 2473 60 | 0013 
OE idee Ck Day | SEN ane Sane | 1502 1834 | 70 0015 
I) A es Bes Ai ah rl Re 1959 2335 68 .0013 
; 
To enable us to appreciate the meaning of these figures, it may be 
noted that, since the eight whistles gave a sound of four times as great 
an intrinsic intensity as the mean pair of whistles, this sound should, if 
no cause except spherical spreading affected the intensity, be audible 
just twice as far, while, as a matter of fact, owing to the other causes 
enumerated, it was usually audible only about one and one-quarter times 
as far. The value of m may be expressed by saying that, excluding the 
effect of spherical spreading, the intensity of the sound died off about 
one-fourth of 1 per cent. for each foot of advance. 
For such rough determinations of a quantity very difficult to determine 
at all, the agreement between the values of m, as shown by the last 
column, seems very satisfactory. 
If now we return to the theoretical investigation it can be shown that 
the value of m consists of the sum of three parts, due, respectively, to 
viscosity, conduction and radiation. Since the constants of viscosity and 
conduction are known, while the full value of m has been found by ex- 
peiment, it can be calculated that the effect of viscosity is to produce a 
diminution of intensity amounting to one-fortieth of 1 per cent. per foot 
of advance; of conduction, one-seventieth of 1 per cent., while the re- 
mainder, amounting to one-fifth of 1 per cent., is due to radiation. 
It may be questioned whether all of the part of the effect thus attrib- 
uted to radiation is really due to that cause and whether part of it may 
not be due to refraction and heterogeneity of atmosphere, as stated earlier. 
That it is not due to refraction is pretty certain, for refraction tends to 
produce actual sound shadows of the surface of the earth or water at 
