52 
starlings, I have also found two young birds of the Molothrus 
nearly fully feathered in the nest of a starling ; but in this 
instance the young starlings had been ejected from the nest.” He 
then states that he had long kept in confinement a male and female 
of this species of Molothrus, which are now six years old, The 
hen began to lay at the age of two years, and has laid each time 
six eggs, which is the number laid by Icterus, a near ally of 
Molothrus, The dates on which the eggs were laid this year are 
as follows :—February 1, 6, 11, 16, 21, and 26; so that there 
was an interval of exactly four clear days between the laying of 
each egg, Later in the season she laid six additional eggs, but 
at much lenger intervals and irregularly, viz. cn March 8, 
April 6 and 13, May 1, 16, and 21. ‘These interesting facts, 
observed by Mr, Nation in relation to a bird so widely distinct 
from the cuckoo as is the Molothrus, strongly support the con- 
clusion that there is sone close connection between parasitism 
and the laying of eggs at considerable intervals of time. Mr. 
Nation adds that in the genus Molothrus, out of every three 
young birds he has invariably found two to be males; whereas 
with Sturnella, which lays only three eggs, two of the young 
birds are, without any exception, females, 
CHARLES DARWIN 
Down, Beckenham, Kent, November 7 
The Velocity of Light 
In reply to Mr. Macaulay (NATURE, vol. xxiv. p. 556) I 
will endeavour to explain more clearly the statements, made in 
my former communication on this subject (NATURE, vol. xxiv. 
p. 352). On one important point the explanation will include a 
correction. 
With reference to the group-velocity UV, we know from 
Fourier’s theorem that any disturbance travelling in one di- 
mension, can be regarded as resulting from the superposition of 
infinite trains of waves of the harmonic type, and of various 
amplitudes and wave-lengths. And we know that any one of 
these trains, of wave-length A, is propagated unchanged with 
a velocity V, which we regard as a known function of A, 
dependent upon the nature of the medium. 
Unless we can deal with phases, a simple train of waves 
presents no mark by which its parts can be identified. The 
introduction of such a mark necessarily involves a departure 
from the original simplicity of a single train, and we have to 
consider how in accordance with Fourier’s theorem the new 
state of things is to be represented. The only case in which 
we can expect a simple result is when the mark is of such a 
character that it leaves a considerable number of consecutive 
waves still sensibly of the given harmonic type, though the 
wave-length and amplitude may vary within moderate limits at 
points whose distance amounts to a very large multiple of A. 
We will therefore suppose that the complete expression by 
Fourier’s series involves only wave-lengths which differ but 
little from one another, and accordingly write it— 
a, cos|(z + 6m,)¢- (e + 5x,)x + ef 
+ a, cos} (7% + 6m,)t—(kK+OK,)xe +e f+ ... 
or in the equivalent form— 
cos (7¢—K x) Ba, cos (82, 2-8, x +e) 
—sin (zt—K m) Sa, sin (62, ¢-8 ,x+4,), 
where k=27/A, and x=xV. From this we see that, as in 
accordance with the suppositions already made, 
Om, 87%. _ _dn 
5 ky dk’ 
the deviation; from the simple harmonic type travel with velocity 
@n/d x, and not with velocity 2/«, that is with velocity d(« V)/d x, 
and not with velocity V. 
I now pass on to the theory of Foucault’s experiment. If D 
be the distance between the fixed and moving mirrors, @ the 
angular velocity of the latter, then the angle through which the 
mirror turns in the time occupied by the wave in making the 
double journey is 2D w/V, and the angular deflection @, which 
is the immediate subject of observation, is according to the u ual 
view— 
Oey 
@u4De 
NATURE 
| i | 
| Mov. 17, 1881 
Now it is here assumed that the deflection is due merely to the 
change of position of the mirror between the two reflections, and 
that the wave returns to the mirror with its front parallel to the 
position occupied immediately after the first reflection, as would 
be the case if the mirror were at rest. But if V be a function of 
A, this assumption is not true. Besides the deflection above con- 
sidered, there is another depending upon the fact that the wave 
front rotates in the air between the two reflections. The rota- 
tion is a consequence of the inclination to one another of succes- 
sive wave fronts, which involves a variation of wave-length and 
therefore of velocity at points situated on the same waye-front in 
a line perpendicular to the axis of rotation. Denoting distances 
measured along this line by x, we have for the angular velocity 
of the wave’s rotation— ‘ 
in which daA/dx, representing the angle between successive 
wave-fronts of similar phase, is equal to 2w A/V. Accordingly— 
‘as dlog V 
tae te AL 
and the actually observed rotation is— 
ea a - ek 
Va dloga 
The result of a calculation which leaves the aérial rotation out of 
account is therefore not V, but— 
sola 
1 — ZiogV 
d loga 
Now 
_ ake) = 5 d@ log V =P _ dilogV) , 
ire dk SIU SS = (: d loga i 
so that the result of the experiment is V*/U, and not as pre- 
viously stated the group velocity UVitself. The error arose from 
a mistake as to the direction of the effect of w’. 
The force of the arguments which I founded upon these con- 
siderations is increased rather than diminished by the correction, 
and with Mr. Michelson’s evidence on the same side of the ques- 
tion almost excludes any appreciable variation of VY. It 
should be noticed that by the combination of the two methods 
of the toothed wheel and of the revolving mirror we have the 
means of determining both V and JZ, and the results of Cornu 
and Michelson appear to prove, independently of astronomical 
observation, that there is no sensible difference between them. 
Indeed by a slightly varied arrangement it would seem possible 
to determine V directly from Foucault’s experiment. If a con- 
vex lens were so interposed at the distaut station that the fixed 
mirror occupied its focus, the sides of short and long wave-length 
would be in erchanged, and thus the rotation acquired during 
the outward journey would be neutralised during the return. 
RAYLEIGH 
The Struggle of Parts in the Organism 
I AM very glad to learn that Mr. Romanes fully accepts as 
‘well-known and unquestionable” the definition of the term 
Jaw of natuve which I propounded as expressing its true scientific 
sense ; but I would suggest to him, as to other writers who are 
accustomed.to speak of such laws as ‘‘ governing ” phenomena, 
whether the use of such ‘‘metaphorical” language is not objec- 
tionable, as tending to keep up in the wzscientific mind the 
notion of the “‘coercive” and “self-sufficient” agency of 
natural laws. Iam glad also to be able to express my entire 
accordance with Mr. Romanes in regard to the inferiority of the 
teleological argument based on spectal instances of adaptation of 
means to ends, to that which is based on the general order which 
we designate by the term law. For I maintained this view even 
in that remote pre Darwinian age in which my scientific life 
commenced, urging to the best of my young ability, forty-three 
years ago,” that the principles admirably laid down by Whewell 
in rezyard to physical inquiry, viz. that final causes should be 
excluded, because ‘we are not to assume that we know the 
objects of the Creator’s design, and put the assumed purpose in 
the place of a physical cause,” and that ‘‘the notion of design 
and end is transferred by the researches of science froia the 
© I continually meet with this phrase in the pages of NATURE. 
2 See British and Foreign Medical Review for April, 1838 : “‘ Physiology 
an Inductive Science.’” 
; 
