( 298 ) 



Through these values we have drawn a curve which from the 

 equator towards the southern declinations ascends steeper and steeper 

 and which gives the values of the last column. According to this 

 curve we have applied the following corrections, for 



zone 1° 2°-5° 6°-8° 9°-10' 11° 12° 13° 14" 15° 16° 17° 18' South 

 neg. corr. 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1,1 1,3 



We may assume that by these corrections the variation of colour 

 due to atmospheric absorption has at least for the greater part been 

 eliminated. 



§ 4. After the two corrections (^ 2 and § 3) had been applied we 

 could determine for all spectral classes the mean values of the colour; 

 they are given in the following table. Class XV was again subdivided 

 into 3 classes according as the spectrum agreed with a Bootis (A) 

 or with a Cassiopeiae (C) or was not accompanied by any such 

 remark; the result shows indeed that here class C is considerably 

 redder than class A while the B's lie between the two. 



The deviations of the separate values from these mean values 

 give, as a measure for the accuracy of the results, for the mean 

 error of a colour-number, [/ 0,20 = 0,45 ; the real accuracy will be 

 greater, however, and the mean error smaller because in these values 

 are also included the errors of the adopted corrections for brightness 

 and declination, the errors which may have been made by Maury while 

 classifying each star in a delinite class, and also the real deviations 

 of the single stars from their class-means. 



