SPECTROSCOPIC OBSERVATIONS OF THE ROTATION OF THE SUN. 91. 
seem to me so remarkable and unexpected, that I trust this early publication may 
induce other observatories to take part in the work, at this opportune time of the 
beginning of what seems to be an energetic sunspot cycle. 
As I intended to make the following observations directly comparable to those of 
Professor Dunsr, I have used the same group of lines. The wave-lengths of these are 
given on page 55 of his treatise, ‘Recherches sur la rotation du Soleil.” * The 
measurements therefore refer to the same photospheric level. Fig. 1 represents this 
group of lines as seen in the viewing telescope. The displacements of the solar lines 
shown in the figure correspond approximately to the shift at the solar equator. 
With regard to the instrument, some deviations from Dun&r’s arrangement have 
been suggested by the apparatus at my disposal. Since they increase the stability of 
the instrumental plant, these alterations may be considered as essential improvements. 
The most important of them was attained by the use of a siderostat. Dun&r’s spectro- 
scope was mounted on the great refractor of the Lund Observatory. The extraordinary 
dimensions of the apparatus gave rise to flexures which were bound to lessen the 
accuracy of the observations. The influence of such flexures is, of course, avoided if 
the whole spectral apparatus, including the front telescope, is mounted on fixed and 
insulated tables, and if the solar light is thrown upon the object-glass of the telescope 
by means of a siderostat. ‘The considerable advantage of such an arrangement is indeed 
shown by the fact that the probable error of a single observation appears to be only 
half the probable error of one of Professor Dunir’s measurements. This increase of 
accuracy has to be ascribed chiefly to the greater stability of the instrument, and 
perhaps also to the easy and comfortable position of the observer during the 
observations. 
A second essential departure from the design of the apparatus used by Professor 
Duner is to be found in the arrangement by which the focal images of the two opposite 
limbs of the sun are thrown upon the slit of the spectroscope. Duntr employed a 
system of right-angled prisms arranged according to a device previously suggested by 
Lanciey. By successive inner reflections from the hypotenuse surfaces of these prisms 
the light of the solar limbs can be thrown upon neighbouring points near the centre of 
the slit, which lies in the optical axis of the telescope. In a much simpler way, 
however, this same purpose can be attained by using as a front telescope a heliometer 
of sutticient optical power. By separating the halves of the object-glass, we can at 
once bring opposite points of the solar limb into contact, and these images may be 
thrown upon the centre of the slit without any further auxiliary apparatus. Besides, 
by altering the position angle of the heliometer, all the opposite points of the solar disc 
ean be successively brought into contact. Thus we are enabled to determine the 
rotational velocity for any desired heliographic latitude simply by turning the 
heliometer into a position which corresponds to that latitude. 
Tt seemed advantageous to throw the two solar images at each observation into such 
a position that the line joining their centres coincided with the slit. This could be 
* Nova Acta Regiae Societatis Scientiarwm Upsaliensis, 3rd series, vol, xiv. fase. ii., 1891. 
