V O L U M E 9 , D O C U M E N T 2 1 7 a 2 3 3
Zu Ihrer Arbeit noch folgendes. Die Parallaxenbestimmung schein mir doch auf
Spiegelfechterei zu beruhen, da in die Grösse der Gravitationseffekt gar
nicht einzugehen scheint, sondern einzig und allein die durch Bedeckungen be-
stimmten Radien. ist eben ein Volumen, hat also mit dem Gravitations-Effekt
so gut wie nichts zu
thun.[7]
Ferner argwöhne ich, dass Ihr angemnommener Wert
Sonnendichte eine inkonsequente Annahme involviert; kann diese „übliche“
Annahme unabhängig von den „üblichen“ Meinungen über die Massen der B-Ster-
ne aufrecht erhalten werden? Sitze ich nicht auf dem Ast, den ich selber abgesägt
habe?[8] Hierüber müssen Sie mich unbedingt aufklären.[9]
Beste Grüsse von Ihrem
A. Einstein.
ALS (NNPM, MA 4725 (14)). [11 218].
[1]Dated on the assumption that the paper discussed in it is Freundlich 1919b, published in the 15
December 1919 issue of the Physikalische Zeitschrift.
[2]Earlier in the year, Einstein had responded enthusiastically (Einstein to Erwin Freundlich, 29
November 1919 [Vol. 9, Doc. 15]) to Freundlich’s report on his research on a redshift in fixed stars
(Erwin Freundlich to Einstein, 27 March 1919 [Vol. 9, Doc. 14]), which was subsequently published
as Freundlich 1919b.
[3]The most reliable estimates of the size of distant stars were obtained from observations of light
curves in binary star systems. As the stars eclipsed each other, the light from the binary system
decreased and the duration of the eclipse gave an estimate of the physical size of the objects. See, e.g.,
Russell 1912.
[4]Einstein’s reference to the Orion nebula is evidence that this letter refers to Freundlich 1919b
(see note 1 above), in which considerable use is made of redshift data from stars in the nebula (see
also Erwin Freundlich to Einstein, 27 March 1919 [Vol. 9, Doc. 14]). The “parallax” of the Orion neb-
ula is not measured directly but through comparison of luminosity curves for stars in the nebula
(assumed to be a group) with luminosity curves for nearer groups whose parallax is directly measured
(see, e.g., Kapteyn 1918).
[5]If a star’s parallax, or distance from the Earth, is known, then its absolute magnitude can be com-
puted from its apparent magnitude. If one determines the surface temperature spectroscopically, one
can then estimate the surface area, and thus the size, of the star.
[6]Einstein proposes that, if the physical size of stars in the Orion nebula is known, the measured
redshift of these stars, assumed to be purely gravitational in origin, be used to determine their masses.
Then the theoretical prediction of the gravitational redshift could be tested by comparing the masses
so derived with the measured masses of stars of similar types in binary systems.
[7]In a 1916 letter to the astronomer Ejnar Hertzsprung, Einstein discusses the particular way in
which the gravitational redshift depends on the mass and physical size of a star (see Vol. 8, Doc. 282a,
in this volume).
[8]Since Freundlich assumes a “typical” density for giant stars, Einstein worries that this value is
actually based upon the binary system measurements of the masses of stars of different spectral types.
In this case, the same assumption underlies both Freundlich’s calculations based upon the gravita-
tional redshift prediction, and the values against which those results are compared.
M
d
---- -
2 /3
M
d
---- -
1
10
----- -