( 451 ) 
dered, the process of scattering by the molecules can only interfere 
to a much smaller extent with the consequences of anomalous refraction. 
Several of the stronger lines of the chromospheric spectrum appear 
double. 1 ) This may be partly due to common reversal by absorption; 
but ray-curving no doubt contributes appreciably to the effect. Indeed, 
if we remember that, in the spectrum of the solar disk, the light 
of the nearest vicinity of the absorption lines appears weakened 
because a part of it is curved back towards the photosphere, and 
therefore fails to leave the sun 2 ), we easily see that the same thing 
must hold good for the spectrum of the lower chromosphere. The spec¬ 
trum of the limb passes quickly but gradually into that of the chro¬ 
mosphere. Those sorts of light, for which the refracting power 
(R m A m )i$ or {R ia A,„ jvr of the mixture of gases is uncommonly 
great, have a chance to return from the chromosphere to lower 
levels; so they are present in the chromospheric light with less 
intensity than kinds of light for which the refracting power is 
!) On the plates, obtained with the prismatic camera by the Dutch Eclipse- 
expedition of 1901, all chromospheric lines were double. After carefully dis¬ 
cussing and examining the circumstances with the other members of the 
expedition, we could not find an explanation of that phenomenon on the basis of 
errors in the instruments or in the way of using them; and because in the pre¬ 
ceding year I had concluded from my theory that many chromospheric lines must 
show a dark core (cf. the quotation on p. 447), I unfortunately gave way to the 
temptation of considering the photographed duplication to represent a real pheno¬ 
menon, and, therefore, a general property of chromospheric lines (Proc. Roy. 
Acad. Amst, Vol. IV, p. 195, 1901). On further developing the refraction theory 
of solar phenomena, however, I found it impossible to explain by it the enigmatical 
observation, that the distances between the components of those various lines 
should differ so little in magnitude, and should on an average be as great, as they 
appeared on the photographs. So I could no longer, in the above result of the 
eclipse observations, see any special proof of the validity of that refraction theory. 
This I wrote to Prof. Kayser (cf. Handbuch der Spectroscopie IV, S. 595); and 
it was also communicated in March 1907 by Prof. Nijland, in my name, to Prof. 
J Hartmann. Evidently, however, the latter did not receive that communication, 
for in April 1907 (Astron. Nachr. 174, S. 353) and again in July of the same 
year (Astron Nachr. 175. S. 352 — 853) he criticized the above paper on chromo¬ 
spheric double lines just as if 1 continued to consider that result of the Dutch 
expedition as a confirmation of my theory. This, however, not being the case, 
Hartmann’s criticism has in no way'rendered our hypothesis less probable, ac¬ 
cording to which the chromospheric light is mainly refracted light, coming from 
the photosphere. Hartmann’s explanation of the double lines appearing on the 
Plates of the Dutch expedition — viz. that they might be due to a slight astig¬ 
matism of the prisms — seems to me very probable. 
Proc. Roy. Acad. Amst. Vol. XII, p. 280, 1909. 
