792 
others it is unmistakably different. There 
appears to be a second component whose 
spectrum contains the Hy line and the rather 
prominent iron lines. On the plates of 
September 16th, October 3d and October 
5th, these lines are shifted toward the violet 
with reference to the solar type spectrum ; 
and in the spectra of November 12th, and 
February 24th they are shifted toward 
the red. (Papers published in Astrophysical 
Journal. ) 
Kurr Laves: On the Determination of the 
Constant of Nutation from Heliometer 
Observations of Eros. 
The opposition of Eros at the end of 
next year will grant very valuable material 
for the determination of NV, the constant 
of nutation. In No. 3,156 of the A. N., 
I have pointed out that this constant 
could be well determined from heliometer 
observations of such small planets as 
come nearer to the earth than the astro- 
nomical unit. In the discussion of the 
heliometer observations of Victoria, made 
for the purpose of deriving a reliable value 
of the mean solar parallax, Dr. Gill has 
carried out this plan. He has found VN = 
9.2068 = 0.0034 (see Annals of the Cape 
of Good Hope, Vol. VI., part 6). This re- 
sult agrees excellently with the value of NV 
derived from direct observations. The 
reason for this agreement is due to Dr. 
Gill’s new value derived for L, the constant 
of lunar equation. Indeed, in a former at- 
tempt to obtain NV by this method, I had 
employed Leverrier’s value I = 6’.50, and 
I was led to a value of N=9’’.26. The 
classical work of Dr. Gill has shown, be- 
yond doubt, that the correct value of LZ is 
much smaller; Dr. Gill gives for it 6.414 + 
0.009. With this value he has obtained 
the result mentioned above. The quanti- 
ties on which the final determination of NV 
depends are: the constant of luni-solar pre- 
cession p’, the mean solar parallax 7°, the 
SCIENCE. 
[N.S. Vou. X. No. 257. 
constant of lunar equation Z, and the incli- 
nation of the moon’s orbit to the ecliptic c. 
We find 
dN = 0.45 dL — 0.34 dx° + 0.17 dp® 
+ 0.002 de. 
It is thus evident that the small probable 
error of Z in Dr. Gill’s value has mainly 
reduced the formerly large probable error 
of N. We may regard that the probable 
errors of L, x°, p°, ¢ are as follows: 
AL = +0".009; 
Ap’ = 0.004; 
4x = 0.005; 
4e= 1" 
4Z and 47° are the values obtained from 
the heliometer observations of Victoria. 
The smallest geocentric distance of this 
small planet was 0.82. The greatest geo- 
centric displacement observable for the pur- 
pose of determining LZ was 15” (this dis- 
placement is due to the semi-monthly 
translatory motion of the Earth’s center 
about the center of gravity of the System 
Earth-Moon). 
During the next opposition of Eros this 
planet will be ata mean geocentric distance 
of 0.34 for about eight weeks, the displace- 
ment witnessed will amount to about 37”. 
From this it is evident that a systematic 
series of heliometer observations of this in- 
teresting object, will both give us a much 
more accurate value of z° and grant an ex- 
cellent redetermination of LZ. It is found 
that the magnitude of Eros will vary be- 
tween 8.8 and 9.7 during this time. The 
most favorable opposition possible for Eros 
will reduce the geocentric distance to 0.15. 
Assuming the mean distance for two weeks 
to be 0.20 we shall have an angle of nearly 
64” available for heliometric measure- 
ments. 
Inner Potential Forces in Astronomy. 
In No. 445 of the Astronomical Journal an 
‘investigation was published concerning the | 
