48 KNUT LUNDMARK, GLOBULAR CLUSTERS AND SPIRAL NEBULiE. 



magnitude as those of the nebulse, and that, on an average, the radial motions of 

 nebulse are only 9 ti mes as great as the velocities that it has been possible to 

 examine independently of spectroscopical methods (e. g. rotations of planets, radial 

 velocities of planets and comets). 



Unfortunately, there exists no older extensive series of nebular positions than 

 the determination of the coordinates of 500 nebulse performed in 1863 — 74 by Schultz 181 

 at Upsala. These measurements have been reduced anew by Wirtz 227 , who in exten- 

 sive investigations has also discussed låter measurements and derived proper 

 motions for the nebulse and calculated the apex-value mentioned above. 



For the 9S small nebulse, probably spirals, investigated by Wirtz, we obtain 

 the following value for the mean proper motion: 



|A m = 0",036 



Curtis 4MT , who has made extensive investigations of the proper motions of 

 spiral nebulse by measuring Crossley plates with an average interval of 13,85 years, 

 has published a communication regarding his results and finds, for 66 great spirals: 



\ ] 'm — 0",033. 



A few other results are further to be found concerning the lateral displace- 

 ments of spiral nebulse. Barnard 17 in A. J. 715 has communicated the results of 

 his micrometrical measurements regarding the Andromeda nebula, and finds that 

 during the 20 years covered by his measurements no proper motion can be shown. 

 He gives the following positions for the nebula, referred to one of his comparison stars: 



Epoch 



Distance 



Angle of Position 



1836 



125".5 



261° ; ii 



1867 



124,75 



261,28 



1898 



125,42 



261,28 



1898,7 



124,67 



261 ; 28 



1909,9 



124,67 



261,31 



1916,3 



124,63 



261,20 



From this measures results ;j. <0%04. 



Lampland 112 has measured a series of photographs of the Andromeda nebula, and 

 obtains for the proper motion of the nucleus [i=0".oi9. Bohlin in his parallax-determ- 

 ination obtained \x a = 0".o9o ± O.oso, while van Maanen obtained |j. a = — 0".ooi3 ± 0".oon. 

 If we take this value into account, w T e obtain from the two parallax-determinations 

 as a mean jj- = 0".o4o. From a discussion of 11 meridian observations distributed 

 över the years 1800 — 1892, Bohlin obtains ijl = 0".oio. A simple mean from these 

 four determinations has the value: 



!J-Andr. = 0",018. 



