J216 



increases as the Milkj^ Way is approached, and since all the plates 

 of group III, with a single exception, have their centres close to the 

 Milky Way, it is not impossible that the difference between the 

 curves I and III originates in the systematic behaviour of the star- 

 colours. Since the colours of the majority of the fainter stars are 

 still unknown, the influence, mentioned above, cannot be examined 

 numericall}'. 



Another cause of the difiference between the two curves might be 

 found in the influence of galactic condensation. Each |)late of the 

 Harvard Map of the sky covers about 900 square degi-ees of the sky, 

 and it is obvious that such a large area will show the influence of 

 galactic condensation and this in a higher degree where the centre 

 of the plate is lying nearer to the galactic circle. In the couise of 

 my first investigation, Dr. P. J. van Rhun of Groningen drew my 

 attention to this fact; my objection then was that of the eight plates 

 witii centres near the galactic circle only three belong to group III. 

 This, however, does not settle the question because the remaining 

 five plates have not been included in group III for the only reason 

 that the}' showed a meau density less than 35 stars per square 

 degree. 1, therefore, submitted the 11 plates of group III to a further 

 investigation in order to get the distance-correction after eUniination 

 of a possible influence of the galactic condensation. For the 100 

 fields which have been counted on each plate we not only do know 

 the star-density — the number of stars pei- square degree — , but 

 also the galactic latitude ; and this coordinate can easily be computed 

 for the centre of the plate as well. F'rom table V of Grou. Publ. 

 N". 27 ^) I have, for every degree between 6 = 0° and h = 50°, 

 derived the value of log. \^ for the visual magnitude 1 1.0 (this 

 being the mean limiting magnitude of the Harvard plates) by graphical 

 interpolation. With these logarithms the density of all the fields on 

 a plate could easily be reduced to the galactic latitude of the centre. 

 If, e.g. the density of a field is 20.7, its galactic latitude 14° and 

 the galactic latitude of the centre 3°, then : 



log 20.7 -f \_logN\o — [log N\,o 



represents the logarithm of the i-educed density of that field. After 

 the densities for each of the II plates had been reduced in this 

 way to the galactic latitude of their centres, the distance-correction 

 has been re-determined in the way described above. The result 



1) Dr. P. J. VAN Rhun. On the Number of Stars of each Photographic Magni- 

 tude in Different Galactic Latitudes. Gron. Publ. W. 27, p. 63. 



