KUNGL. SV. VET. AKADEMIENS HANDLINGAR. BAND 60. N:0 8. 



35 



During the last few years considerable progress has been made as regards the 

 study of the nebulae, and although it cannot be said, that any definitive fact has 

 yet been established as to their cosmogony, we can, however, say that almost all 

 observational facts tend to show that spiral nebulse must be considered as very 

 remote objects. The reasons for this statement are principally the following. 



1. Direct 'parallax measure?nents. 



From what has now been published concerning attempts to measure directly 

 parallaxes for spiral nebulae, we have compiled the following 



Tabh VIT. 



Object 



- 



Observer 





' 0,"o 



Hall 73 





-0,32 ±0,"l2 1 

 -0,11 ±0,14 I 



Franz 68 





» 



M 32 



0,0 

 + 0,171±0,051 1 

 + 0,132 J 



Barnard 72 



BOHLIN 23 





» 





0,0 



Barnard 17 





0,004 ±0,005 



van Maanen 1 ' 4 



M 31 



j 0,073±0,055 1 

 l 0,090±0,060j 



Strömberg 220 



M 51 



0",005±0'',008 



VAN MaANEN 120 



The first three values of the parallax for M 32 are relative parallaxes for the 

 nova that appeared in the said nebula in 1885. The measurements are uncertain 

 and do not extend över more than two and six months resp., and no great im- 

 portance will therefore be ascribed to them. The different parallax-measurements 

 for M 32 show so little mutual concordance, that forming a mean out of all the 

 values is peihaps not to be justified. However, a mean value of all the determina- 

 tions would give as a final result a very small parallax, as consideration must be taken 

 also to negative parallax values, as pointed out by Kapteyn 104 . Especially puzzling 

 is Bohlin's great positive value for the parallax of the Andromeda nebula, but as 

 van Maanen has not been able to confirm this result, and the great focal-length of 

 the Mount Wilson reflector and the appropriateness of this instrument for parallax- 

 measures show that great importance is to be assigned to van Maanen's parallax- 

 value, it seems not to be improbable, that Bohlin's parallax must be due to some 

 other phenomena with a yearly period than the parallactic displacement of the object. 

 As Strömberg's measurements of the parallax for M 31 has been made with the use 

 of the same plates and probably also the same method of reduction, as Bohlin has 

 used, the same remark will also ref er to the result found by him. 



